by John Keay
Meanwhile, far to the west, Ma himself began grappling with the proto-Tibetan Qiang people. Between AD 35 and 39 he subdued the region from modern Xining to Kokonor and settled vast numbers of Qiang along and within this sector of the frontier. It was hoped that they would become taxable subjects as they forsook pastoralism in favour of sedentary farming; and for twenty years peace did indeed return to the area. But here, as along the northern frontier, a generation of instability had already prompted an exodus of the existing population. The ingress of non-Chinese only slewed the population balance still further, so that even within striking distance of Chang’an concentrations of Qiang and Xiongnu numerically challenged those indigenous descendants of the erstwhile Xia and Qin who may now be called ‘Han Chinese’. This would have unexpected consequences; the Qiang problem, unlike the Qiang themselves, was anything but settled.
General Ma’s next assignment could hardly have been farther away. In AD 40 revolt broke out in the south of Nanyue, or what is now northern Vietnam. For Ma it meant crossing the whole of China, exchanging near-tundra for tropics, raising a new army, and adapting the tactics of frontier patrolling to jungle warfare. The trouble was again ethnic, although it would acquire strong nationalist overtones. The indigenous Yue had risen against Han rule and more especially against Han immigration into the Red River basin. Whereas in the north of China Han settlers were retreating, in the south they were encroaching; a great population drift that over the next several centuries would change the whole pattern of Chinese demography was gathering pace.
Untroubled by such trends, Ma marched into Vietnam with overwhelming force, his supplies followed by sea from Guangdong, and the revolt was all over by the end of AD 43. But the written character for ‘Yue’ being read as ‘Viet’ in Vietnamese, and leadership of the Yue having fallen to two wildly courageous sisters (called Zheng in Chinese, Trung in Vietnamese), it was inevitable that later Vietnamese patriots would hail the revolt as the first ‘national uprising’ of an all too often oppressed people. Heavily romanticised, the story of Viet resistance to General Ma became a national epic rich in detail. It tells how one of the Boadicea-like Trung sisters rode into battle on elephant-back with her breasts flung over her shoulders like saddlebags – an arrangement that could be editorially glossed as the earliest authentic reference to the halter-neck top, or perhaps the brassiere. Chinese sources credit General Ma with extending to the Yue/Viet the opportunity of peaceful assimilation to Chinese ways and an end to their incorrigible tribalism. Certainly northern Vietnam underwent intensive sinicisation; and despite constant troubles, the next ‘national uprising’ would be a long time coming. In fact the Red River valley would remain under some form of Chinese rule for all of the ensuing millennium. But the Vietnamese prefer to forget this hiatus in their national struggle, and not surprisingly, temples dedicated to Ma Yuan are notably absent in Hanoi.
No sooner had General Ma reported back to Emperor Han Guang Wudi in Luoyang than he received another troubleshooting assignment, this time to the northern steppe-land on the Shanxi–Inner Mongolia frontier. There and throughout the ‘Great Wall’ borderlands the Xiongnu had taken advantage of the recent civil strife to stage a remarkable comeback. Partly through their own efforts, partly thanks to the erratic support of renegade Chinese warlords and of the nomadic Wuhuan and Xianbei peoples of the north-east, the Xiongnu had recovered practically all the territories once ruled by the great shanyu Maodun. Raiding parties were reaching what remained of the Former Han’s ancestral tombs near Chang’an and, far to the west, were threatening the ever loyal Wusun. The Ordos was back in Xiongnu control; so, to the west of it, were many of the oasis-cities of northern Xinjiang and, to the east of it, even the arable parts of Shanxi and Hebei.
The last of these was General Ma’s new theatre of operations. In AD 45 he set up headquarters west of where Beijing now stands. From there he marched north to Mongolia. But the Xiongnu evaded him, and in what was supposed to be a surprise raid on their Wuhuan allies, Ma himself was surprised. He lost a thousand cavalry, and though he remained in the region for another year, he did not again take the offensive. This could be because he did not need to. For in AD 46 the shanyu died, the succession was disputed, and the Xiongnu became the prey of their Wuhuan allies. ‘From the Chinese point of view’, writes Rafe de Crespigny, an authority on the northern frontier, ‘there was hardly any purpose in an aggressive policy while their enemies were carving one another up in such a satisfactory fashion.’16 Far from disgraced, General Ma moved on again. He would live to fight and die while suppressing yet another insurrection, his fourth, in northern Hunan.
Though by no means the end of the Xiongnu people as a frontier presence, the ructions over the shanyu’s succession in AD 46–49 sounded the death-knell of the Xiongnu state. Repeating events of exactly a hundred years earlier, in AD 51 one shanyu claimant finally turned to Luoyang for support and tendered his allegiance, plus a modest tribute. In return he received Han recognition plus decidedly less modest gifts – 10,000 bales of silk fabric, 2.5 tonnes of raw silk, 500,000 kilos of rice and 36,000 head of cattle. Tributaries never came cheap; in fact the shanyu of these ‘Southern Xiongnu’, in trading kowtows for commodities, and sovereignty for security, may have struck a better bargain than his ancestors under the old ‘peace-through-kinship’ treaties. Moreover, the gifts were just the first instalment of what was in effect an annual subsidy. By AD 91 the Southern Xiongnu were estimated to be costing the Han exchequer 100 million cash per year.
Nor were they the only beneficiaries of the Later Han’s policy of accommodation on the frontier. The Wuhuan and, on occasion, the Xianbei were also handsomely paid for assisting the Han, whether against the still-hostile ‘Northern Xiongnu’ or against one another. Perhaps as many as 3 million Wuhuan were eventually settled within the frontier. The Han army welcomed the addition of their fearsome cavalry, while the Wuhuan welcomed a subsidised existence like that of the Southern Xiongnu. It was different with the Xianbei (Xianbi, Sarbi). Having assisted the Han in driving off the Northern Xiongnu in AD 91, they proceeded to fill the Mongolian void left by this Xiongnu exodus. By the second half of the second century AD the Xianbei headed a new nomadic confederacy that would pose an even greater threat to the Later Han than had the once united Xiongnu.
All these frontier peoples merit more consideration than the mostly Chinese sources permit. Tomb paintings discovered in Inner Mongolia portray them as shaven-headed, sometimes with a Mohican tuft, and shabbily dressed. The Wuhuan and Xianbei seem to have originated in eastern Mongolia or western Manchuria, though their ethnic identification as ‘proto-Mongol’ is more tentative than that of the Qiang as ‘proto-Tibetan’. They lived in tents, herded livestock and were good on horseback. More critically, as a result of Luoyang’s accommodating policies, they (plus the Qiang) came to occupy vast areas within the north, north-west and north-east frontiers of the Later Han empire.
For the Han they were a mixed blessing. Tributaries had been won, and the northern frontier had been returned to where it had stood under the Early Han. But the cost was great. It was argued that subsidies were cheaper than the military campaigns and garrisons that would otherwise be necessary. Yet incorporating alien populations that resented discrimination from their Han neighbours (though in many areas they outnumbered them) would have devastating consequences. Though initially illiterate and alien to all that constituted Chinese civilisation, the Qiang, Xiongnu, Wuhuan and Xianbei would come to play a decisive role in the post-Han period. In that age of warring kingdoms, they, just like the men of Chu and Qin in the earlier ‘Warring States’ period, would be invited into the central plain as mercenaries and then power-brokers. Insinuating themselves among the peoples who today consider themselves Han Chinese, the alien ‘tributaries’ would infuse the indigenous mainstream. Thereafter the flow of China’s history would be as much theirs as that of their erstwhile Han enemies.
As for the ‘Western Regions’, that appendage of island-o
ases and intersecting silk-ways beyond the Gansu corridor in Xinjiang, the Later Han were ambivalent. For nearly half a century (AD 25–73) no attempt was made to reclaim the area. Suche (Yarkand), the principal kingdom in the extreme south-west of Xinjiang, attained a brief hegemony over the other oasis-states, but the Northern Xiongnu were soon raiding and trading at will throughout the region. Colonies established by the Early Han remained in a state of siege – if they remained at all; and the ‘Jade Gateway’ to the Silk Road near Dunhuang was more often barred than open. None of this, however, prompted speedy action. For in retrospect those long-distance exchanges of the previous century, which the Chinese had habitually characterised as tribute and the tributaries as trade, looked less worthwhile. At one point the two-way traffic along the silk trails had grown so heavy as to be almost continuous. Yet its prestige value to the Han court had declined. The system, it seems, was being too much abused.
The point was well made in a report on relations with Kashmir. Here was a country, south of Xinjiang and on the other side of two of the world’s highest mountain chains, that was clearly beyond the reach of Han arms. It was almost beyond the reach of Han emissaries. Supplies were unobtainable on the mountain trails, a military escort was essential to discourage bandits, and such was the effect of the extreme altitudes that the Kun Lun and Karakoram mountains were known to the Chinese as ‘the ranges of the Greater and Lesser Headache’. For long stretches the trail – it probably followed the Hunza, Gilgit and Indus rivers – narrowed to ledges less than 45 centimetres (18 inches) wide. Travellers had to rope themselves together. ‘The danger of the precipices beggars description,’ said the report. Yet for the best part of a century substantial missions had been scrambling back and forth, often annually, to pledge fealty on behalf of the Kashmiris and to confer titles on behalf of the emperor. In both directions they also carried merchandise – woollens, embroidery and Indian produce from Kashmir, large consignments of silk from China. Secure in their Himalayan fastness, the Kashmiris brooked no interference in their internal affairs and massacred one Han mission that tried. But apologies had followed, the protestations of loyalty continued, and the exchanges had resumed. Inveterate traders to this day, the Kashmiris were clearly using the Han tributary system and the protection that it afforded to conduct purely commercial activities. As the report puts it, ‘envoys sent out on missions to carry the commands of the emperor’ were being diverted ‘to escort the merchants of the barbarians’.17
It was not tribute or trade which tempted the Han back into Xinjiang but the Xiongnu. In AD 73, 74 and 77 military expeditions launched from Gansu against the Northern Xiongnu pursued them deep into Mongolia and reoccupied the northern Xinjiang oases of Turfan and Hami. Ban Chao, the brother of the Hanshu’s authors, took part in these forays, and it was he who in the AD 90s famously led a series of further expeditions that re-established Han supremacy throughout Xinjiang. As Protector-General of the Western Regions, an office defunct since the conquests of the Early Han, Ban Chao supervised the resettlement of Han colonies along the silk routes and renewed contacts with the states of central Asia. In AD 97 a mission under his subordinate, Kan Ying, was dispatched to ‘Da Qin’, a distant realm with an apparently insatiable appetite for Chinese silk. It was probably the Roman empire, albeit its eastern provinces. Sadly the encounter that might have resulted was pre-empted by the mission’s detention in Parthia, whose merchants, like those of Sogdiana (Samarkand), had a vested interest in excluding competitors from the overland trade.
Han horizons, which had been at their most expansive in the early first century BC under the Former Han, seemed to be reopening in the late first century AD under the Later Han. On the other side of China in North Korea, twentieth-century discoveries of richly endowed Han tombs have substantiated a patchy textual record of conquests and commanderies in the Korean peninsula which date from Han Wudi’s time. But with no external threat from this quarter, Han pretensions ebbed and flowed without provoking much comment. Their effect was more notable in terms of acculturation. Confucian values and useful Chinese achievements, such as literacy and paper-making (developed in the first and second centuries AD), passed down the Korean peninsula and across the Tsushima Strait. In AD 57, and again in 107, Luoyang entertained its first recorded visitors from Kyushu in the Japanese archipelago. At almost exactly the same time, Han texts make their first mention of a revered teacher who was called ‘Fo’ in China but ‘The Buddha’ in India. From the Sea of Japan to the Ganges and the Tiber, Han China’s horizons now spanned the Eurasian landmass.
They soon retracted. The Han administrations in Korea were forced back into Liaodong province in the early second century ad, and all the Western Regions were abandoned – again – soon after. The main reason in the case of the latter was the cost of maintaining so many tributary statelets and far-flung colonies. But renewed trouble with the Qiang in the Gansu corridor threatened access to the region and was a contributory factor. The Qiang revolts came thick and fast. By AD 168 the Han had been forced to withdraw from three of Gansu’s commanderies, so narrowing its famous ‘corridor’ to little more than a crack. Proposals to withdraw completely from the whole of eastern Gansu were canvassed but met with strong resistance, most notably from Gansu itself. When in AD 184 the whole region erupted again in rebellion, its Han population would make common cause with their Qiang and Xiongnu neighbours. One rebel group even adopted the title ‘Ping Han’, meaning ‘Pacifying the Han’. The dynasty itself now became the target, and under Dong Zhuo, one of Gansu’s king-making leaders, this Gansu revolt would ‘play a key role in the decline and fall of the Han empire’.18
DECLINE AND FALL
As with the later Roman emperors, so with the Later Han emperors: few merit mention for what they did and not many for what was done in their name. After AD 88 the succession passed to a motley collection of infants, invalids, weaklings and imbeciles; though castigated in the Standard Histories, they seem more deserving of pity than censure. Their dates help to break up a substantial passage of time, but seldom can a change of emperor be taken to indicate a transfer of power or an adjustment in policy. Power changed hands between factions, not emperors; policies, where discernible, were dictated by events.
Much as the founding of the Han dynasty (in 206 BC) had been made to predate the installation of its first emperor (in 202 BC), so the formal demise of the dynasty (in AD 220) substantially postdates the empire’s actual disintegration (in AD 184–89, if not earlier). Dynasties and their empires would quite often not be coterminous. The Standard Histories, in emphasising a dynasty’s maximum span, sought to obscure the usually shorter duration of its effective empire. Adding a bit at either end – one might call it dynastic elastication – gave the desired impression: integrated empire was meant to seem the long-lasting rule, division the short-lived exception. Commissioned by the court and written by scholars from the central bureaucracy, the histories betray a vested interest in magnifying and prolonging central authority, plus a natural bias towards events and intrigues at court. A Luoyang perspective on the rest of the empire is standard.
But as seen from, say, Sichuan, Shandong or the Yangzi, let alone the border regions, things looked different. Never were the Later Han untroubled by insurrection, and only in the three decades after Han Guang Wudi (so AD 58–88) was the dynasty spared major regional or ‘religious’ revolts led by would-be emperors. In one year, AD 145, no less than three rebel emperors took the field; four others followed their example over the next decade. There were possibly as many competing emperors during the two centuries of the Later Han’s unitary empire as during any two centuries of the ensuing ‘Period of Disunion’.
The internecine intrigues and bloodlettings at the Later Han court reflected this strife in the wider empire, as well as encouraging it. At first all had gone well. After Han Guang Wudi’s death, his son and then grandson continued to manage the competing factions at court with some success. Han Mingdi (r. AD 58–75) chose a daughter
of General Ma Yuan as his empress, so appeasing the powerful Ma clan of the north-west while confounding its rivals, some of whose members were executed or exiled. On the other hand, Han Zhangdi (r. AD 75–88) chose a consort from the rival Dou grouping from Henan, which led to the downfall of the Ma clan, some of whose protégés were jailed and/or committed suicide.
The casualty rate was not exceptional; imperial favour was ever fickle, and dismissal was commonly attended by death. Moreover, both emperors managed the fraught issue of the succession rather well considering that neither of their empresses actually bore them sons. Normally this would have left an empress vulnerable to being replaced by some ‘graceful lady’ who did give the emperor a son; this might in turn bring the disgrace of the incumbent empress plus the downfall – or worse – of her clan. But such an eventuality was pre-empted in both these cases. Han Mingdi’s empress wisely co-opted into her own faction an imperially favoured mother-to-be and promoted the suitability of the latter’s son as if he were her own. Han Zhangdi’s empress went one better by championing in similar fashion two sisters, the Ladies Liang, one of whom duly gave birth to the future Han Hedi (r. AD 88–106). There was, however, risk in such an arrangement: the mother of the heir apparent might renege on it, turn the emperor against his empress and supplant her as consort and her clan as power-brokers. This is precisely what the Ladies Liang would eventually do. Better by far, then, indeed the ideal arrangement, was for the emperor to choose in the first place an empress whose attributes included a delectable sister. That way the same family got two bites at the cherry of providing a male heir and so of retaining its influence into the next generation. As an insurance policy, the empress-plus-sister combination would later become something of a commonplace.