Most importantly, the core areas of later development—France, Italy, England, the Low Countries, western Germany—were particularly well sheltered from the steppe. This was also the zone in which post-Roman state deformation progressed the farthest. What little impetus clashes with organized herders gave to state-(re)building—the German victory over the Magyars in 955 temporarily strengthened the ruler’s hand—soon fizzled in the absence of sustained challenges. Just as pressures from the steppe encouraged state-strengthening and empire-building in China and elsewhere in the Old World, their absence favored the opposite outcome in Latin Europe.79
From this perspective, it is telling that precocious state formation in early medieval England can be linked to meta-ethnic conflict with invaders from Scandinavia that prompted Wessex’s expansion from the 860s to the 920s and concurrent domestic reforms. This process was but a pale shadow of what nomadic aggression might have accomplished, both in terms of conquest and in terms of the agriculturalists’ response. No other European frontier generated similarly severe friction: conflict between Germans and Slavs could be managed on a more local scale, and other credible challenges were missing.
Remoteness from the bulk of the Eurasian steppe was a constant, invariant across European history. Just as it did not matter if Latin Europe’s states were weak, it also did not matter if a large empire was in place. Unlike Chinese dynasties, the Roman empire did not bring forth a nomadic “shadow empire”: there was no ecological potential for it. The Pontic steppe, where Sarmatian tribes might have coalesced in response to the inducements of Roman wealth, was too detached from the Roman heartlands that lay behind the Carpathians, the Alps, and the Adriatic. To the west of the plains of Eastern Europe, both components of the “steppe effect” were conspicuous by their absence: and so—at least after Rome—was empire-building on a large scale.80
Old World Parallels
Conditions in other parts of the Old World mirror the stark contrast between persistent imperiogenesis along the steppe frontier in East and eastern Central Asia and persistent political polycentrism in post-ancient Latin Europe: while those in South Asia and the Middle East bear some resemblance to those farther east, Southeast Asian outcomes were more similar to those in Europe. In all these cases, the relative strength of the “steppe effect” can be shown to have been of considerable importance.
SOUTH ASIA
India is situated at a considerable remove from the Eurasian steppe. However, the vegetated river plains of the Amur-Darja and Syr-Darja and zones of open shrubland form conduits for the southward movement of horses between them. In antiquity, the west of northern India used to be dominated by pastoralists. Over time, flows of invaders and the horse trade from or through that frontier region exerted growing influence on state formation on the subcontinent. Just as in early China, the first steps toward empire were not yet heavily affected by this factor: the Maurya military of the fourth through second centuries BCE mostly relied on infantry and elephants.81
For the next several centuries, conquerors from the Central Asian steppe took their place, first the Sakas and then the Kushanas, who boasted large numbers of mounted archers. Both the Kushanas and the homegrown Gupta empire in the fourth and fifth centuries CE faced challenges from Hunnic horse warriors. These conflicts led to the adoption of heavy cavalries of archers and swordsmen. Campaigns were waged across the Indus valley against Huns and Sasanian Iranians, and also in order to procure horses from beyond; the composite bow was introduced from the steppe. A massive Hun incursion in the late fifth century did not last long, hamstrung perhaps by the lack of sufficiently large pastures: in the longer term, sedentarization, as practiced by Kushanas and later Turks, was the only viable option, just as it was for steppe invaders of the Hungarian plain.
In keeping with Barfield’s and Turchin’s theoretical predictions, the end of the Gupta empire coincided with the cessation of nomad invasion and ushered in a period of prolonged polycentrism. Inasmuch as it did occur, sporadic empire-building relied on horses: in the first half of the seventh century, Harsha used a large cavalry force—established to defeat the Huns—to impose short-lived imperial rule. By contrast, in deference to ecological conditions, successive smaller regional states such as the Pala empire and south Indian polities deployed large herds of war elephants sourced from Bengal, Assam, and Orissa alongside infantry troops. At that time, in the late first millennium, only the Gurjara-Pratiharas of Kanauji marshaled a cavalry strike force to fight the Arabs in Sind.
Whereas the caliphate had been kept at bay, later Turkic conquest regimes made deeper inroads. In the tenth century, Ghaznavid cavalry proved superior to the Rajput armies of northwestern India. In the early eleventh century, the Ghaznavids conducted raids far into India to fund wars against Central Asian nomads who threatened their sultanate in Afghanistan, and recruited horse archers from Transoxania into their service. Once again, South Asia was drawn back into the orbit of belligerent steppe formations.82
The next frontier regime to take the Ghaznavids’ place, the Ghurids, allegedly fielded a cavalry numbering 120,000—that is, “a lot” of cavalry—when they trounced the Rajputs in the northern Ganges basin in 1192, and came to rule not only Afghanistan but also the Indus basin and northern India. Their appearance marked a fateful shift from raiding to conquest, a project made possible by their access to horses from Central Asia and even Arabia and Iran.83
The next iteration, the Sultanate of Delhi, was credited with even vaster cavalry forces. Having subdued northern India, it managed to contain even the Mongols and briefly expanded into the Deccan. Horsemen were the backbone of their rule: in yet another example of the tax-and-empire nexus typical of steppe conquest regimes, much of the revenue from land taxes was used to remunerate them in cash, and the authorities maintained control over the supply of war horses. As their polity fractured, Timur invaded with a large cavalry army, sacking Delhi in 1398.84
The dramatic success of cavalry warfare and its crucial role in propping up large empires impelled imitation even in parts of the subcontinent that lacked native sources of horses. In the Deccan plateau, the Vijayanagara empire arose in direct response to Muslim aggression spearheaded by a superior cavalry: in the fourteenth century, the ruler of the Bahmani Sultanate in the northern Deccan plateau (a successor state to the Sultanate of Delhi) was known as Ashavapati or “Lord of Horses.” Only once Vijayanagara began to recruit Muslim soldiers and built up sizable cavalry forces by importing horses from the Middle East was it finally able to compete, helped along by Bahmani’s eventual fissioning. Vijayanagara became the most militarized of the non-Muslim southern polities: the semiarid environment of western India permitted the diffusion of high-quality horses—local breeds were inferior—and horse-riding equipment from the northwest, such as the stirrup. This enabled locals to catch up with established steppe-sourced cavalry regimes. As a result, cavalry assumed a decisive role in combat even this far south.85
When Vijayanagara was defeated by Muslim sultanates in 1565, it was cut off from the horse trade, retreated, and soon collapsed. A generation earlier, Babur’s invasion of northern India had succeeded thanks to the deployment of a combination of horse archers and firearms. His Mughal empire continued to privilege cavalry over local elephant forces. Once again, an ambitious tax system supported a huge and expensive cavalry of maybe 100,000 to 200,000 men that built the largest and most centralized empire India had seen thus far.86
Even though the influence of steppe groups on earlier Indian state formation had already been considerable, the appearance from the late twelfth century onward of conquest regimes that drew on steppe-sourced horses and sometimes personnel was a major turning point. It reversed a trend toward regionalization that had commenced after the Gupta collapse and might have steered Indian state formation onto a rather different trajectory.
Between the sixth and twelfth centuries, South Asia was divided among regional states that were quite populous in absolute terms but not relativ
e to its total population. Their intermediate size and fairly symmetric interstate competition fostered enhanced military and fiscal capabilities and extended their longevity beyond that of the earlier hegemonic empires. In Lieberman’s view, their “increasingly coherent personalities suggest that South Asia may have been headed toward a permanent competitive multistate system not unlike that of Europe.”87
Even so, exposure to Inner Asian pressures made this outcome less likely in the long run. Similar to the “marginal zone” between northern China and the great steppe, Afghanistan served as a conduit for exchanges with Central Asia and as a launchpad for invasions of northern India. In a further analogy, south Indian militaries were generally deficient in cavalries (but strong in naval assets), just as south Chinese ones were. South Asian ecology placed serious constraints on riders and herders: the lack of extensive grasslands limited their numbers, and high levels of humidity damaged their bows.
Yet these obstacles could be overcome by learning from local practices and, above all, by taxing large subordinate sedentary populations to sustain large cavalry forces that were replenished by ongoing imports of horses from farther north and west. High-endurance turki-horses were Central Asia’s main export item to Mughal India: as many as 100,000 of them are said to have arrived there each year. Medieval Latin Europe, by contrast, lacked not only sufficient grasslands but also the fiscal infrastructure that might have made comparable forms of Magyar or Mongol rule viable.88
For close to 2,000 years, from the arrival of the Sakas in the first century BCE to the decline of the Mughal empire in the early eighteenth century, empire-building in South Asia was increasingly fueled by inputs from the steppe. As foreign conquest regimes relied on cavalry, indigenous polities sought to adapt and imitate accordingly. In this respect, the Indian and the Chinese experience had much in common.89
IRAN
Beginning in the early first millennium BCE, similar influences shaped Iranian state formation. Iran is connected to the Inner Asian steppe through rivers east of the Caspian Sea, which traversed terrain that was probably less arid than it is now. Farther west, the Caucasus, a formidable geological barrier yet endowed with meadows that nourished passing horsemen, could not prevent complementary infiltration.
While there is no solid evidence for nomadic pastoralism in prehistoric and Bronze Age Iran, the early Iron Age state of Urartu in the northwestern Iranian highlands was already associated with horse breeding and cavalry warfare. Cimmerian and Scythian raiders from the Pontic steppe assaulted Asia Minor and Mesopotamia. These advances coincided with the establishment of a culture of “equestrianism” in the high plains east of the Zagros Mountains that affected a whole range of traits from military tactics to costume, most notably the adoption of trousers that became a hallmark of Iranian culture.
The emerging hegemony of the Medes was the most visible political manifestation of these developments. Horse tributes from western Iran had greatly contributed to the apogee of Assyrian power from the mid-eighth century BCE onward. For all we can tell, the Medes’ ascent was rooted in the fusion of cavalry forces into organized armies that managed to cut off the Assyrian empire from these vital supplies and subsequently wipe it out near the end of the seventh century BCE. It appears that the Persian Achaemenid dynasty that took over the Median domain likewise heavily relied on mounted troops when it assembled its enormous empire in the second half of the sixth century BCE. Horses and horsemanship played a prominent role in royal self-representation.90
Through co-optation and conflict alike, steppe power was crucial to Iranian empire-building. Various Iranian-speaking groups from the northeastern Iranian steppe frontier—Daians, Mardians, and Sagartians—served in the Achaemenid military. In the third century BCE, the Parthian empire of the Arsacid dynasty originated from a Daian confederation east of the Caspian Sea, in what is now Turkmenistan. Folkloric traditions that the Achaemenid and Arsacid dynastic founders Cyrus II and Arsaces I both originated from among these mobile herders reflect the latter’s contribution to state formation.
At the same time, confrontation was endemic. In the sixth century BCE, Cyrus II died fighting the Massagetians, a grouping of nomads south of the Aral Sea, and Dareios I campaigned against the Scythians in the Pontic steppe. In the late second and early first centuries BCE, several Parthian kings battled Scythians, two of them losing their lives in the process. Not much later, a coalition of Daians and Sakas attacked the Parthian empire, and in 73 and 135 CE Alans from the Pontic steppe invaded by crossing the Caucasus.91
These conflicts continued under the Persian Sasanian dynasty. From the early fourth century CE onward, a succession of warlike steppe federations—Huns, Hephthalites, and Turks—posed a severe threat to the Sasanian empire, pushing back its garrisons across its northeastern frontier region. During a peak of hostilities in the late fifth century, King Peroz I was slain by the Hephthalites (or “White Huns”) and his forces were annihilated. His successor, Kavadh I, deposed by his own nobles, fled to the Hephthalites and depended on their support to regain power later in exchange for accepting tributary obligations.92
These are not mere anecdotes. The largest Iranian empires of antiquity, those of the Achaemenids and Sasanians, were established and maintained in a symbiotic relationship with the equestrian societies of the Medes, Parthians, Huns, and Turks: both accommodation and conflict provided vital inputs. Interstate competition with Assyrians, Babylonians, and Romans, which is better documented in the surviving sources, tends to overshadow these dynamics, which were as foundational as they were persistent. Only the most dramatic incidents surface in the record: it is telling that whereas three Parthian and Sasanian kings died fighting steppe enemies, none of their peers succumbed to Macedonian or Roman opponents. Nor was the heartland of the Parthians ever threatened by Romans the way it was by challengers from the steppe. Cumulatively, these experiences were among the most potent drivers of state formation. It was not by accident that “Iran” came to be defined in contradistinction to “Turan,” the term applied to the steppe regions beyond the northeastern reaches of settled Iran. Antagonism ran deep.93
Inputs from arid frontier zones further intensified beginning in the seventh century CE, as military groups first from the Arabian peninsula and then from Inner Asia took over Iran. Their cavalry forces played a central role in creating imperial polities and in reconfiguring domestic power relations in their own favor. Their regimes dominated the region for more than a millennium, from the early seventh through the late eighteenth century.
Arabs held on to power under the Umayyad, Abbasid, and Samanid dynasties, the last of these rooted in Khorasan and Transoxania right at the northeastern steppe frontier. The Umayyad caliphate inherited endemic conflict with the Central Asian Turks from its Sasanian predecessors. Not long after the ninth-century Abbasids had turned to Turkic mercenaries to maintain control, Turkic military regimes, sourced from the steppe, took over. They spawned a whole series of tribal-origin dynasties: Ghaznavids, Seljuqs, Khwarazm Shahs, Aq-qoyunlu, Qizilbash, and Qajars, while on occasion yielding to the Mongol Il-Khans and the Mongol-Turkic Timurids. Even in the northwest, Iranian Safavids mobilized nomadic Turkic tribes to establish their empire.94
THE LEVANT AND NORTH AFRICA
To the west of Iran, the Levant was similarly affected by the Arab conquests, the Turkic Seljuq takeover, the Mongol irruption, and the expansion of the Ottoman empire. By comparison, the contribution of pastoralists to regional state formation in the more distant past remains much less clear. The Amorites, who came to dominate the southern Mesopotamian core after the end of the third millennium BCE, cannot readily be identified as external herders. The same is true of the Hurrians who set up the Mitanni state in northern Mesopotamia and northern Syria soon after domesticated horses, war chariots, and composite bows had appeared in that region.95
Elsewhere, these new technologies as well as migrant groups associated with them, can nevertheless he shown to have been of some conse
quence. In Egypt, occupation by the Hyksos—most likely pastoralists from southwestern Asia—gave lasting impulses to later imperialism. They introduced horses as well as the composite bow and probably also chariots to the Nile valley. Their exclusive habit of burying horses in elite tombs called attention to their self-image as martial horsemen. Founded after the Hyksos’ expulsion, the New Kingdom state accorded unprecedented prominence to military affairs: it oversaw the creation of a professional army that campaigned widely across the Levant, whose terrain offered improved access to horses and allowed chariot warfare to be massively scaled-up. The resultant empire was larger than anything Egypt had previously sustained.96
In other cases, the appearance of pastoralists did not turn out to be particularly conducive to polity growth: thus, the movement of the Aramaeans into the Syrian-Mesopotamian core, completed by the tenth century BCE, did not generate large empires. For most of the region, the explosive expansion of Iranian power in the seventh and sixth centuries BCE, outlined above, marked the true turning point. The most economical explanation for this transition might be that it took the more sustained engagement with the vast “reservoir” of the Eurasian steppe to make a real difference, whereas other, smaller marginal zones had, at least at the time, been too small or underdeveloped to exert a comparable influence.97
In this respect, the Maghreb occupied an intermediate position. Sizable empires appear to have been difficult to maintain. Both the lower organizational capacity of North African pastoralist tribes and the lack of a deep agricultural hinterland have been held responsible: however, we have to allow for the possibility that the former may have been influenced by the latter.98
Escape From Rome Page 31