by John Keay
Whether or not the political fragmentation of the tenth century encouraged these developments, the wars and disturbances reported by Ouyang Xiu prompted a new wave of migration to the south. Yet there is no sign of an overall decline in population like that in the ‘Period of Disunion’. Quite the opposite. In the mid-eighth century, the total population is thought to have been around 60 million, of whom 60 per cent were living in the Yellow River basin. By the end of the tenth century, it was nearing 100 million, of whom less than 40 per cent resided in the Yellow River basin; over 60 per cent now lived in the Yangzi basin and the south. Accelerated by territorial losses in the north, the long-drawn-out shift to the more productive south had finally tipped the demographic scales. As of the Five Dynasties/Ten Kingdoms period, more Chinese grew rice than millet and wheat; their winters were warmer and their summers wetter; their acquaintance with the Inner Asian steppes was slight; and they lived not under the turbulent Five Dynasties but among the opulent but historically neglected Ten Kingdoms.
SONG AND LIAO
An insistence on ‘dynasties’ in the north but mere ‘kingdoms’ or ‘states’ elsewhere betrays the need, born of the historians’ insistence on the preeminence of empire, for an unbroken pedigree of imperial validation, or ‘legitimacy’. This was usually expressed in terms of the Mandate, which, descending from dynasty to dynasty, could neither be shared nor suspended. Though many might claim it, only one at a time could actually enjoy it. But a difficulty lay in deciding which one. In the case of the third-century division of the Han empire into ‘Three Kingdoms’ (Cao Wei, Han-Shu and Sun Wu), the problem was never satisfactorily resolved. But when in the eleventh century historians such as Ouyang Xiu and Sima Guang addressed the same problem in respect of the Five Dynasties/Ten Kingdoms, they had an advantage. The longevity and distinction of the Song dynasty under which they lived made it indisputably a ‘legitimate’ holder of the Mandate, and this in turn provided a reassuring perspective from which to assess the status of its predecessors.
Like the long narrative scroll-paintings that became fashionable under the Song and were so rolled that, on unfurling them, the concluding scene came into view first, the historians spooled backwards. The Song had superseded the Later Zhou, who must therefore have been legitimate; the Later Zhou had succeeded the Later Han, who must be likewise; the Later Han the Later Jin, ditto; and so on back to the Later Liang, who had overthrown the indisputably legitimate Tang. The legitimate line must follow this northern succession of ‘Five Dynasties’. It mattered not that more Chinese in fact lived under the other, stabler dispensations of the Ten Kingdoms. Empire had traditionally emanated from the north, successful dynasties had almost invariably hailed from there, and the record allowed no other interpretation. Not until the Ming (1368–1644) would the entire empire be governed from the south, and then only briefly.
For Ouyang Xiu, a minor problem remained: the scheme meant recognising as legitimate the detestable rebel regime of the Later Liang. Besides tattooing its troops, a stigma as demeaning as torture in that it would stay to disfigure them into the afterlife, the Later Liang had murdered two Tang emperors and in general ‘represented an horrific evil that deserves eradication from history’. Earlier accounts of the period had done just that – eradicate them – hence in effect reducing the Five Dynasties to four. Ouyang Xiu considered doing the same but was dissuaded by an estimable regard for the facts, backed by Confucius’s insistence on rectifying names – or in this case calling a spade a spade. ‘Acknowledging vice’ would serve as a warning to posterity, he thought; and it could only enhance the general credibility of his account.21
The other four of the Five Dynasties posed no problem. Three were non-Han, but that was nothing unusual; and though all were usurpers, so had been the founders of most previous dynasties. All being descended from military governors of the Tang, it was their mutual jealousies which made for instability rather than their ethnicity. To gain power, any dynastic contender needed the support of his peers, which meant placating other governor-commanders with promises of titles and territories. But to rule he needed to rein in these same magnates and reduce their forces – which could all too easily provoke another challenge. The cycle was hard to break; and the troops themselves had a habit of propelling their commanders towards the throne. What mattered was that a new dynasty provide tangible evidence of its having inherited the Mandate, and this, in tumultuous times, meant enlarging its empire.
In alliance with the resurgent Khitan, the Later Tang (923–36), second of the Five Dynasties, had overcome the unspeakable Later Liang, so doubling the size of their territories. They had also briefly grabbed the state of Shu (Sichuan). The Later Jin (936–46), another Turk family, had fared less well. Puppets of the Khitan from the start, they were obliged to cede to them sixteen crucial prefectures lying in an arc from what is now Beijing to Datong in northern Shanxi. They also had to recognise the Khitan leader as their ‘father’ and overlord; thus ‘for the first time a Chinese regime openly acknowledged the suzerainty of an alien dynasty’.22 Heaven duly showed its disgust with a spate of natural disasters. In 946–47 the Khitan actually occupied Kaifeng. This ended the Later Jin’s wretched decade; assuming they had in fact possessed the Mandate, they clearly did so no longer. The Later Han (947–50), who were also Turks, reclaimed Kaifeng and might perhaps have regained the prefectures lost to the Khitan. But within a matter of months they were ousted by the Later Zhou (951–60), the fifth of the Five Dynasties.
The founder of the Later Zhou was ethnically Han Chinese, well educated and in every respect a worthy precursor of the mighty Song; so was his adopted son and heir. As is the way in the Standard Histories, their attributes foreshadow their success. Most of China north of the Huai River (but excluding the north-east and the sixteen ceded prefectures) was consolidated; and under the founder’s son (Later Zhou Shizong, r. 954–59) expeditions regained two of the lost prefectures and probed south as far as the Yangzi. There one of the largest of the Ten Kingdoms was forced to acknowledge Later Zhou sovereignty. In 955 Later Zhou Shizong felt strong enough to emulate Tang Wuzong and pillage the resurgent Buddhist establishment. ‘Zhongguo [i.e. the Middle Kingdom of the Five Dynasties] faced a shortage of cash at the time,’ explains Ouyang Xiu, ‘so the confiscation of all Buddhist bronze statuary was mandated for recasting as coin.’ In the process, some ‘3,336 monasteries were eliminated’. Yet Shizong himself made light of the affair. Since the Buddha, as he understood it, had attached no importance to physical existence and had resumed his bodily form only so that mankind might benefit from his teaching, ‘how’, enquired the emperor, ‘could he possibly begrudge us a bunch of his bronze statues?’23 Muffled by deference and often marred by translation, an imperial sense of humour can never be discounted.
The Later Zhou’s one great misfortune was the premature death of this Shizong, aged thirty-eight, in 959. His seven-year-old son succeeded him, but within months the dynasty was challenged by its most trusted general. A certain Zhao Kuangyin, fearless in battle and consummate in statecraft, this was the man who, as the next dynastic founder, would be posthumously known as Song Taizu (r. 960–76). At the time he appeared just another usurper. His background was identical to other contenders’. His prospects seemed no better than those of the five ephemeral dynasties to which he was about to add a sixth. Sure enough, move for move, his usurpation mirrored that of the Later Zhou. Other potential contenders were won over, while in public he proclaimed himself utterly unworthy of the throne; but his troops and their commanders clamoured for his elevation; the portents were adamant; and the matter was settled when, acknowledging his superior merit, the mother of the last of the Later Zhou insisted on her son abdicating in his favour.
The whole procedure was suspiciously predictable, yet the results were far from expected. Defying precedent, Song Taizu, and then his brother and successor, Song Taizong (r. 976–97), not only quelled all opposition within their northern empire but accelerated the progr
amme of southern conquest begun by the Later Zhou. One by one, most of the Ten Kingdoms were brought to heel so that within thirty years the empire was substantially reunited. To their still greater credit, the Song achieved this with a minimum of violence and some conspicuous gestures of magnaminity. The new dynasty had a measured feel to it, evincing intent without provoking antagonism. Song rule was revealed as more formal and much more restrained than that of the Tang. Executions and floggings would be comparatively rare, vendettas and purges comparatively few. Though the founding brothers campaigned vigorously, and though the forces under their direct control were massively augmented to nearly a million men, both they and their successors emphasised civilian rule. Instead of declaring a grandiloquent new beginning, they discreetly manipulated or bypassed existing offices to create a centralised and responsive system of control that was largely proof against the breakaway forces that had shattered the Tang empire. They cultivated the arts of peace, surrounded themselves with scholars, and encouraged a lofty imperial mystique that proved more effective in deterring internal challengers than the threat of punitive expeditions.
But what made all this possible, and what in posterity’s eyes seriously tarnished the early Song achievement, was a painful accommodation with the Khitan in the north-east. The abject capitulation of the Later Jin (third of the Five Dynasties) to the Khitan would not be repeated; ‘the suzerainty of an alien dynasty’ was repudiated and the son-to-father relationship with the Khitan qaghan was firmly terminated – but only to be replaced by that of brother to brother. The Khitan qaghan’s claim to be an emperor and an equal of the Song would be grudgingly recognised. The honours paid by him were reciprocated by those paid to him; and the subsidies disbursed to him were acknowledged as tribute in all but name. Thus China under the early Song presented an anomaly: though most of what was left of the Tang empire was reunited, imperial sovereignty remained divided.
Like a Damoclean sword, the Khitan menace had dangled over the Five Dynasties throughout their five fraught decades. Those who defied it, it felled; those who embraced it, it scarred. Either way, it took a hefty cut of state revenues in the form of indemnities and tribute. None of the Five Dynasties had found a satisfactory way of dealing with the Khitan presence in Hebei and it was a contributory factor to their instability. Meanwhile, under Abaoji (r. 907–47), a charismatic figure whose career has been likened to that of Chinggis Khan, the Khitan had spread their wings in other directions, overrunning much of Mongolia and various northern peoples beyond, dispossessing a sinicised kingdom called Bohai (Parhae) to the east in north Korea, subduing most of Manchuria, and so creating an empire that in its territorial reach dwarfed all others in east Asia, including the zhongguo of the Five Dynasties.
This Khitan empire is known as Liao after the dynastic name adopted by Abaoji’s successor, itself culled from a state of the ‘Warring States’ period that roughly corresponded to the Khitan heartland and which originally took its name from the Manchurian river that watered it. About Chinese culture in general, the Khitan of Liao were ambivalent. They considered themselves both superior to it and participants in it. Instead of extending steppe rule to the plains, let alone plains rule to the steppes, Abaoji introduced the novelty of a two-government system. A northern chancellery administered the clan-based society of the Liao’s nomadic subjects in accordance with traditional steppe norms, while a southern chancellery administered the empire’s settled subjects, principally Han, in accordance with Chinese bureaucratic practice. The former, based at a supreme capital in what is now eastern Inner Mongolia, was organised on military lines for tribute collection and offensive operations; the latter, based at what is now Beijing, was run by Han civilians in order to maximise tax and labour receipts. Over both, in the dual role of qaghan in the north and emperor in the south, presided a Liao dynast who, at moments of crisis, was not infrequently a woman.
Women in migratory societies probably shouldered more responsibility than their settled sisters. Raiding and herding kept the men from their hearths for long periods, while betrothals were more informal, and strict seclusion was precluded by the exigencies of steppe domesticity. But the Khitan, unlike many of their subject peoples, were at most only semi-nomadic. Their mainly Manchurian homeland was not open steppe but a configuration of forested hills and arable flood plains that limited pastoralism and localised transhumance. Fearless horsemen, they neither lived in the saddle nor slept in encampments. As among their Tangut neighbours in central Shaanxi, the tendency towards matriarchal rule seems to have been less a steppe legacy and more the result of the ruling clan attempting to impose a regular system of succession. Instead of the tribal councils and the trials of strength that traditionally decided the succession in steppe confederations, the Liao sought to impose a father-to-son system of primogeniture more in keeping with their imperial status and the concept of the heritable Mandate. This was frequently contested by their nomadic subjects, especially when, Liao longevity being far from guaranteed, the throne passed to a minor. But such minorities also afforded an opportunity for empresses and dowagers to exercise a de facto sovereignty that might continue after the emperor came of age.
Such was the situation in the critical years at the turn of the tenth to eleventh centuries when the Song, having mopped up the Ten Kingdoms of the south, switched their attention to the Khitan Liao in a bid to reclaim the north-east and so complete their reunification of the empire. Fresh to the throne, in 979 Song Taizong overran a part of Shanxi to which the Later Han (fourth of the Five Dynasties) had retired, and in the process inflicted a heavy defeat on a Khitan Liao army sent to assist this time-warped enclave. Success then encouraged Song Taizong to defy all advice to the contrary and invade Liao itself. His exhausted troops reached the Liao capital at what is now Beijing ill prepared and short of supplies. They were routed. Vast quantities of booty and weaponry were lost. Taizong himself was reduced to fleeing the field of battle in a cart, ever the most ignominious of imperial fates.
Two years later the Liao emperor died in a hunting accident and was succeeded by the eleven-year-old Liao Shenzong (r. 982–1031). Sensing an opportunity for revenge, Song Taizong resumed the offensive. But he reckoned without the military skills and formidable resolve of a Khitan dowager empress. In 986 Dowager Empress Chengtian took the field in person at the head of the Khitan forces and thrice defeated the Song armies. A ten-year lull, broken only by border clashes, ensued, at the end of which Song Taizong died. His successor, Song Zhenzong (r. 997–1022), was handicapped less by youth than by a superstitious disposition that bordered on the timid. It was the turn of the Liao to take the offensive.
As of 999 Khitan armies poured into the Song territories north of the Yellow River annually. Though Liao Shenzong had now come of age, his mother continued to direct policy and in 1004 again commanded one of the Khitan armies that, in a full-scale invasion, marched on the Song capital of Kaifeng. They struck the Yellow River below Kaifeng at a place called Shanyuan. A decisive armageddon seemed inevitable. But behind the Khitan’s line of advance lay still-strong and uncaptured cities, while before them massed an enormous Song army. Both sides had reason to avoid a battle whose outcome was far from certain; and both sides enjoyed the services of Han intermediaries whom they trusted. The outcome, all too rare in Han-to-non-Han relations, was a frankly negotiated treaty, unmarred by trickery, that would stabilise Song–Liao relations for a century.
The Liao had demanded the concession of some strategic territories, while the Song sought recognition of their overall suzerainty. Both had to compromise. Instead of territory, the Liao were bought off with cash – a yearly payment of 200,000 lengths of silk and 2.8 million grams (100,000 ounces) of silver, plus gifts of comparable value on special occasions such as New Year and the Liao emperor’s birthday. The Song would interpret this as a triumph, arguing that the cost was trifling in terms of their empire’s revenue, far less than they feared, and certainly less than further military operations would have occas
ioned. Critics were not so sure, frequently citing the payouts to the Khitan Liao, and similar subsidies to the Tangut kingdom in the north-west, as an unacceptable drain on the Song economy. Latterly it has been suggested that the value of Song–Liao trade as regularised by the treaty far exceeded the cost of the subsidies and that, since the Khitan bonanza was used to purchase Song exports, most of it found its way back into the Song economy.
But while ceding no territory, the Song did suffer a major loss of face. In the double-speak beloved of diplomacy, they dismissed the payments to the Khitan as ‘contributions to military expenses’; the Khitan nevertheless regarded them as tribute, and so – since they were unreciprocated, ongoing and in 1042 substantially increased – they surely were. Using the kinship terminology familiar to both sides, the Liao were recognised as brothers of the Song, albeit ‘younger brothers’; thus Dowager Empress Chengtian was to be addressed by the Song emperor as ‘aunt’. Perhaps in analogy with the Liao system of dual government, it also became conventional to refer to the ‘Great Khitan state’ as the ‘Northern Court’, and the ‘Northern Song’ as the ‘Southern Court’, of an otherwise entirely fictitious imperial entity. That the object was peaceful partition rather than contentious aggregation is clear from other provisions. Both sides bound themselves to avoid further hostilities and to respect and demarcate their mutual frontier. The demarcation was conducted throughout the frontier’s 600-kilometre (370-mile) length from the westernmost reaches of the Yellow Sea near modern Tianjin to the northernmost reaches of the Yellow River west of Datong. ‘Constituting a genuine international frontier in the modern sense’ – and so quite unlike the often provocative and ineffectual fortifications known as the Great Wall – it was ‘something unprecedented in Chinese history’.24