The emperor had been raised a devout Buddhist. ‘With the armed might of a Cakravartin [a world-ruling ‘turner-of-the-wheel’ in Indian Buddhism], We spread the ideals of the Ultimately Enlightened One,’ declared Sui Wendi’s first edict. ‘With a hundred victories in a hundred battles We promote the practice of the Ten Virtues. The weapons of war We regard as incense and flowers offered to the Buddha . . .’7
Nearly four thousand temples, pagodas, nunneries and monasteries were founded during Sui Wendi’s reign, many of them in his rebuilt Chang’an, where one pagoda reportedly measured 100 metres (109 yards) in height and 120 metres (131 yards) in girth. The number of new Buddha images erected ran to over a hundred thousand, while those repaired following the Northern Zhou’s iconoclasm exceeded 1.5 million.8 In China there was no nobler exemplar of the Buddhist sovereign as maha-danapati (supreme donor) thanSui Wendi. Towards the end of his reign he organised a ceremonial distribution of Buddha relics to thirty specially built reliquary sites dotted throughout the land. Packed and sealed in jars by the emperor himself, the relics were distributed by an army of monks and were all enshrined on the same day. Offices everywhere were closed for a week. It was an act of empire-wide dedication consciously modelled on a distribution of the Buddha’s remains by the Indian emperor Ashoka, Buddhism’s first imperial patron. Ashoka’s action had been intended not just as a demonstration of piety but as a celebration of sovereign power over a sprawling empire. So it was with Sui Wendi. The distribution twice repeated, the number of Sui-sponsored reliquary shrines, which also served as places of congregation and pilgrimage, rose to 114.
But if the emperor’s actions stemmed from a concern to expiate his sins and improve his karma (rather than from a complete misunderstanding of the Buddha’s non-violent doctrines), his munificence needed to be on the grand scale. All too easily roused,SuiWendi sent many to the executioner, readily awarded ‘the privilege of suicide’ to others, penalised even his sons, and himself caned to death those whom he considered not chastised vigorously enough by his officials. His justice was stern, his energy prodigious. Yet, for an emperor, his tastes were modest and his personal life exemplary. With his non-Han wife, he shared the burdens of state in what was tantamount to a joint sovereignty. She bore him seven children, and by mutual consent he had no others. The imperial couple were ‘inseparable and . . . in the palace they were referred to as “the two sages”’.9
Something similar might well have been said of Sui Yangdi (r. 604–18), Wendi’s son and successor. Also a Buddhist benefactor, he too combined personal brutality with operational clemency, while his marriage to a southern princess seems to have been equally harmonious. But the Standard History of the Sui was written under the direction of their nemesis, the Tang; and while Sui Wendi, as the reunifier, could scarcely be denied his share of praise, Sui Yangdi, as the last of his line, could conveniently be credited with more than his share of opprobium. Sui Yangdi brought the empire to the brink of ruin; therefore rumours of his having murdered his father (the most heinous crime in an ultra-filial society) and having ravaged his father’s consorts (which was deemed incest) must be true. V. C. Xiong, his recent biographer, observes that even the name by which he is remembered, Yang (plus the imperial signifier‘-di’), was in fact a post-humous pejorative reserved exclusively for those who ‘lust after beautiful women . . . abandon ritual . . . defy Heaven and abuse the people’. Summarising Sui Yangdi’s many-sided persona as revealed in the traditional histories, Xiong notes ‘a hedonistic philanderer, a prodigal spend-thrift, an oppressive ruler, a cold-blooded murderer, an impulsive aggressor, a hater of remonstrance, a lover of sycophancy, and above all, a tyrant . . .’10 Yet Yangdi’s achievements were far from mean and his crimes not dramatically worse than his father’s – or than those of his immediate Tang successors. As commander-in-chief of his father’s southern invasion and then a sympathetic governor of the south, he had played a leading role in his father’s reunification; and when the apprentice became the autocrat, he pursued almost identical policies with equal, if eventually disastrous, zest.
Sui Wendi’s declared preference for the Han dynasty and the Cao family’s Wei dynasty had been intended to signify dynastic ambition and imperial resurgence. The Cao family’s Wei dynasty provided an acceptable blueprint for a legitimised usurpation and the Han an unassailable example of Confucian rectitude. Wendi’s new Chang’an was called Daxiangcheng, meaning ‘Great Revival City’, and of the Five Phases or Elements, he chose fire and its colour red, they being the same as those adopted by Han Wudi. No scholar himself, Sui Wendi had little interest in the niceties of Confucian morality but saw its value as a basis for social order and bureaucratic centralisation.
Over the previous centuries the proliferation of provincial units and semi-autonomous fiefs had run unchecked. It has been calculated that, since the Later Han, the number of commanderies had increased by six and a half times and the number of prefectures by twenty-two times. There were so many mini-entities, so many parallel administrations, so many tiers of government and so many salaried officials that they absorbed most of the revenue; according to one adviser, it was like having nine shepherds for every ten sheep. Sui Wendi abolished nearly all of them and substituted ministerial offices, departmental hierarchies, graded ranks and parallel inspectorates as favoured by the Han. Grist to the mill of China’s bureaucratically inclined historians, the restored titles, portfolios and pecking orders had a certain rationality but would be subject to constant change. Adjustments by Yangdi further centralised, streamlined and, where convenient, sidelined the system, which would then be thoroughly overhauled by the Tang.
Much the same could be said of the legal codes introduced by Sui Wendi and Sui Yangdi and famously reworked for export throughout east Asia by the Tang. Both of the Sui codes professed superior logic and greater leniency; yet they were regularly flouted by emperors exasperated as much by the corruptibility of the judicial process as by the growing prevalence of crime. Wendi’s ‘Ten Abominations’ – crimes unpardonable even by imperial amnesties – were reduced by Yangdi to eight, ‘incest’ and ‘discord’ (or ‘plotting to kill or sell relatives who are of the fifth or closer degree of mourning’) being omitted, presumably because they were misdemeanours of which the emperor himself might be guilty. But given that for petty crimes like the theft of a copper cash, the purloining of a roof joist or even the picking of a melon offenders were routinely executed, penal severity seems not to have been significantly restrained by legal codification. Suffice it to say that in opting for Han precedent in matters of ritual, administration and justice, the Sui launched China’s ‘Second Empire’ as a fair approximation to the highly regulated, bureaucratised and draconian despotism that had characterised its First – and which would remain until its last.
Ferocious punishment, although probably no more revolting or, per head of population, more commonly administered in seventh-century China than in seventh-century anywhere, served as a deterrent in a society riven by conflicting loyalties. These were not simply the ethnic, political and regional residue of four centuries of strife. A downside of the Confucian emphasis on family ties was that it encouraged office-holders, for instance, to aim at perpetuating a monopoly of office within the circle of their kinsmen and dependants. Nepotism being respectable, corruption flourished, competence declined and the gap between statutory intent and actual practice widened. If the frequency and tenor of imperial appeals are anything to go by, discovering candidates for office whose ability was uncompromised by hereditary loyalties had challenged every emperor since the First. An efficient bureaucracy, and indeed the whole legitimacy of the regime, depended on its ability to attract men of calibre; yet this had to be done without antagonising existing magnates and, in the case of empire-builders like the Sui, while enlisting the loyalties of powerful new constituents in the south, the east and along the northern frontiers.
The Sui solution to expanding the base of civil service recruitment was to set up a Board of Civil Office to centralise all appointments and scrutinise the selection process. They also revived and developed the Han system whereby each commandery had to identify and recommend a certain number of outstanding candidates for future office; the chosen few were then dispatched to the capital, assigned to one of three academies and there tutored and examined. Though still rudimentary, the examination system was refined by the introduction of degrees in different subjects. Among them was one with a literary bias, known as jinshi, that would eventually become the acme of social and scholastic achievement. Though it would be left to the Tang, and more especially the Song, to elevate the examination system into one of imperial China’s most distinctive features, the Sui may be credited with having promoted the idea, and laid down the framework, for a genuinely meritocratic civil service.
As with the penal system, it is not clear how closely practice conformed to principle. But if one may judge by documentation relating to the early Tang period and recovered from Dunhuang and Turfan in the twentieth century, the imperial writ carried far and wide. Registration for the purposes of taxation and labour service is seen to have been remarkably thorough. Each household submitted detailed returns of its family members, servants, livestock, cultivable land (and whether hereditary or not), other property and crops. These were then checked, entered, assessed, and the assessments levied. On such meticulous record-keeping depended the revenue and manpower at the Tang’s disposal; and it is reasonable to suppose that the system worked just as smoothly under the Sui since their programme of public works, military expansion and ostentatious expenditure was second to none.
Not content with his father’s replanning of Chang’an, Sui Yangdi founded an alternative capital, more centrally located for a reunited empire, by rebuilding Luoyang. It was not as extensive as Chang’an/Daxiangcheng, being a mere 47 square kilometres (18 square miles) instead of the latter’s 80 square kilometres (30 square miles); and as usual, building materials from the previous city were recycled. Timbers could be rehung and roof tiles relaid; bricks baked easily in the hot dry summers and hangtu foundations required only the tireless tamping of an abundant labour force. The almost stone-free nature of most Chinese architecture goes a long way towards explaining the rapidity with which city after city rose and fell. Completed in 606, Sui Yangdi’s Luoyang had taken little over a year. Yet by all accounts the result was something special.
It was unrivalled in its extraordinary splendour [says the Standard History of the Sui]. Since the emperor, as imperial prince, had pacified the south in person, he assimilated the curvilinear and angular [forms] of Liang and Chen [architecture] . . . Its walls rose higher than the Mang mountains. Floating bridges spanned the Luo River. Above the golden gate and ivory watchtower were erected winged belvederes. Precipices were collapsed and rivers cut off to make way for pillars [or perhaps mounds] shaped like multi-coloured clouds . . .11
The main palace hall stood on a 2-metre-high (6.5-feet) pedestal. It was about the length of a football pitch and as tall as a stadium. Its roofline commanded the city; and between its pillars probably hung the mother-of-pearl blinds that had adorned its predecessor and which ‘at sunset flashed with a dazzling radiance’.12 Yuwen Kai, the architect who laid out the city as well as designing the main buildings, seems to have been preserved for just such work, he being one of the few members of the Northern Zhou’s Yuwen clan not to have perished in Sui Wendi’s purge. His Luoyang creation delighted Yangdi, and his abilities as a civil engineer would be further tested. Among several other palaces commissioned by Yangdi in the course of his extensive travels was one on wheels. In 607 it accompanied the imperial entourage on a grand tour of the northern frontier. Carried there in kit form by cart – a palace-on-wheels on wheels – it was assembled in situ and then trundled into an auspicious alignment. It was used for the reception of tribute missions and, in addition to the dignitaries, could apparently accommodate ‘several hundred’ imperial guardsmen.
Sui Yangdi’s 607 northern excursion took him beyond the Ordos, where, like his father, he sought to emulate the Qin and Han in respect of frontier management and wall-building. The strategic value of a walled or fortified northern frontier had been somewhat discredited over the preceding four centuries. In fact, as wave after wave of steppe-people poured into China, it was hard to think of any incursion that had been inconvenienced by it. Sui Wendi, and then Yangdi, ascribed this to the lamentable state of the surviving earthworks; and while Wendi had built westward, Yangdi set a workforce doubtfully put at 1 million, of whom half were said never to have returned, to construct a new wall east from Yulin to the Taihang mountains. But the new walls proved just as provocative, ineffective and ruinously expensive as their predecessors. Disowning Sui precedent for once, the Tang would make little attempt to maintain them. Across the Gobi desert in Inner Mongolia those parts of Sui Yangdi’s wall that survive may therefore be genuine examples of seventh-century workmanship. A recent visitor noted ‘a massy stretch of earthen rampart, perhaps 2.5 metres [8 feet] tall, running between solid towers rising twice as high out of archetypal Gobi vistas . . . it looks more like a termite-infested bank than a man-made defence’.13
Decidedly less redundant are the still-churning thoroughfares of what was Sui Yangdi’s, and indeed one of imperial China’s, most ambitious and rewarding creations. This was a canal – or rather it was the many canals and hydraulic features that, connecting numerous rivers, lakes and pre-existing conduits over a total distance of nearly 2,500 kilometres (1,550 miles), came to be known as the Grand Canal. Excavated between 605 and 611, the Grand Canal ran north-west from Hangzhou (south of the Yangzi delta) to the Yellow River near Luoyang, with a long extension from there north-east to where Beijing now stands. In effect it linked north and south, east and centre. It was an axial artery for a reunited empire. It was also ‘without doubt, the grandest navigation system ever undertaken by a single sovereign in pre-modern history’.14
Sui Yangdi followed the work closely, and by way of inauguration, in 611, made a stately progress up the length of it; his flotilla of extravagantly dressed craft was said to have stretched for over 100 kilometres (62 miles). But China’s Grand Canal, unlike its Venetian namesake, was rarely a stylish processional path, more a gigantic transport corridor. Much rerouted and often widened and dredged, in 1793 it would wring grudging admiration from the first British mission to imperial China; and some sections of the original alignment must still today be among the busiest waterways in the world. Past Suzhou (between Shanghai and Nanjing) tug-towed strings of wallowing barges snake continuously, head to tail, round the clock, in both directions, at a bank-sloshing pace;gondolas would be swamped, vaporettos vaporised. Nor is the canal’s utility exhausted. Twenty-first-century plans envisage its conversion into a giant aqueduct to divert the Yangzi’s flood waters to the now parched and polluted aquifers of Beijing and the north.
To the pre-motorised economy, bulk transport had posed a logistical problem that was nigh insuperable. It frustrated trade in all but high-value commodities, inhibited urbanisation, impeded the development of industry and undermined efforts at major state-formation. This was especially so in Sui China, where roads were subject to annual inundation and coastal shipping was deemed precarious. Yet coinage being scarce and paper money still to be invented, the entire tax yield came in kind – usually grain and silk – and so in bulk. Moreover, the movement and storage of this agricultural surplus was the only insurance against the ever-present threat of famine and was a strategic necessity for provisioning frontier garrisons and supporting military ventures beyond.
The Grand Canal, linking the Yangzi region with its rice surplus to the heavily populated and famine-prone northern plains, thus had a similar effect to the first transcontinental railroads in North America. It made China’s economic integration feasible. Disparities of climate, terrain, produce and demographic distribution were suddenly converted into assets. Granaries – which were less mud-built silos than vast installations, walled and guarded, like oil-storage depots – were strategically located along the canal. Big government-owned grain barges, hauled by manpower wherever sluice and current required, constituted the bulk of the water traffic; a burgeoning private trade in salt, fish, vegetables and manufactured goods made up the rest. Along the route, irrigation schemes fed off the canal to increase crop yields, so boosting population figures, tax yields and corvée numbers; and below the Yangzi in ‘Nanjiang’ (all ‘south of the Yangzi River’) more land was brought into cultivation courtesy of the canal, so accelerating the pace of the population drift from north to south. Great cities – the future Hangzhou, Suzhou, Zhenjiang, Yangzhou and Kaifeng – grew up along the towpath. Existing metropolises, such as Chang’an, to which Sui Wendi had constructed a short canal from the Yellow River, and Luoyang, which Sui Yangdi similarly integrated into his system, could now outgrow the alimentary limitations of their immediate hinterlands.
But the cost was colossal and the human suffering incalculable as corvée demands took their toll of agriculture. Millions, or rather ‘tens of hundreds of thousands’, are said to have dug the channels and distributed the spoil of Sui Yangdi’s waterways, though whether such figures refer to the total labour force involved or the total number of corvée periods worked is unclear. They used spades and picks, plus wicker baskets balanced at either end of a pole to maximise carrying capacity and absorb jolts. Wheeled transport was provided more by barrows than ox-carts. The man-drawn wheelbarrow, a Chinese invention, had made its first appearance some time in the previous three centuries. By transferring the weight from the human frame to an axle, it more than quadrupled average loads. Men – and then women when the corvée pool began to dry up – graduated from being beasts of burden to serving as draught animals.
Ultimately it was all too much. The canal system could distribute only what the farmer could produce. But serious flooding of the Yellow River in 610/11 had reduced yields, while military requirements drained the labour pool and emptied the granaries. As the demand for manpower for both the army and public works raced ahead of supply, truancy increased and soon turned to popular revolt. Critics at court were ruthlessly silenced: powerful challengers began to mobilise within the provinces and the army. The final straw came when the last of three disastrous expeditions against Koguryo, a reluctant tributary state occupying much of Manchuria and northern Korea, was in 614 recalled in the face of minimal gains and escalating mutinies. Leaving the north in turmoil, in 616 Yangdi retired to the south. It had earlier been his adoptive homeland; now it became his last refuge. With Yangdi isolated from events as much by his tremulous courtiers as by distance, his megalomania subsided into melancholia. In 618, at his southern capital in what is now Yangzhou, he was murdered. The assassin was the son of one of his generals and a member of the once-purged Yuwen clan; but it could have been anyone.
According to the traditional histories, his fate was no less than he deserved. The sufferings Sui Yangdi had inflicted on his people had been intolerable and the strains he had imposed on the just-reunited empire unforgivable. Seldom had the Mandate been so obviously forfeited. Yet the products of so much distress would be acknowledged even by critics as an inestimable boon, ‘enormous indeed’ and ‘monumental’. Better still, according to a writer of the later Tang period, they came at no cost to posterity. The canal, for instance, ‘did not require a single [Tang] labourer to carry a basket of soil nor a single [Tang] soldier to hack through an obstruction’. ‘Is it not true’, concluded this observer with a smug flourish, ‘that Heaven has greatly benefited us with the help of the despotic Sui?’15 Such was the function, and such the fate, of an intercalary dynasty.