The Han empire had reached its greatest extent around 90 BC. Nanyue (including all of northern Vietnam), Hainan Island (though this was abandoned in 46 BC), parts of Yunnan, all of Gansu, the trans-Xinjiang ‘seaways’ and many of that province’s oasis cities could now be counted as Han territory. In addition, new commanderies had been established in the extreme north-east to include much of Manchuria and north Korea; and farther afield, feudatory or tributary relations had been established with the Wusun and many of the central Asian states. No significant additions would be made by the Han after Wudi’s 87 BC death, nor by any subsequent dynasty until some eight centuries later.
The Xiongnu had been reduced from a threat to a nuisance. Their confederacy splintered; a new treaty was signed with one of its segments in 51 BC, and large numbers of these subdued Xiongnu settled on marginal lands within the frontier. Border alarms now as often involved Qiang from the Tibetan plateau, or a people called the Wuhuan from the Manchurian steppe-forest. Instead of expansion, Chang’an pursued a policy of consolidation through colonisation. Under Wudi’s successors, it would increasingly become one of retention, then retraction, as intrigue overwhelmed the empire from within. The dynastic cycle had reached its zenith.
Elsewhere it was empires which rose and fell – Persian, Greek, Indian and Roman. In China it was dynasties. Empire remained a constant. Heaven’s Mandate might be transferable but its terms were fixed; there could be but one legitimate emperor, one ‘Son of Heaven’. Though variable in extent and often flatly contradicted by political realities, imperial authority was becoming as inseparable from the notion of Chinese identity as was its self-consciously literate culture. Courtesy of Han, of its centralising and staying power, and of its statesmen, generals and scholars, ‘zhongguo’ was being mentally and physically reconfigured as the emphatically singular ‘Middle Kingdom’ or ‘Central State’.
Rome’s near-contemporary empire was comparable in size and would last several centuries. But China’s would weather the millennia. Many explanations – geographical, psychological, even genetic – have been offered for the contrasting fortunes of the two empires and for the different political orders to which they would give rise – fragmentary and increasingly consensual in the case of Europe’s nation-states, unitary and increasingly authoritarian in the case of China’s empire. Explanations that emphasise imperial China’s more effective administrative structure and its more flexible ideology are certainly not the most exciting. Yet they carry conviction; and both – the administrative machinery and its ideological lubricant – were honed under the Han.
The Han bureaucracy has been called by one of its recent exponents ‘the most impressive form of government that as yet had been devised in the world’.15 It was both pervasive and intrusive, insistent on participation, not unresponsive, but oblivious of representation. The entire population, 57.6 million as deduced from the first surviving count of AD 1–2, was registered among some 12 million family households. These were grouped into mutually responsible units of five or ten as in Qin times. The family groups were organised into hamlets, each with a headman, and the hamlets into communes, each with a chief. Neighbouring communes constituted a district or county, neighbouring districts a prefecture, and neighbouring prefectures a commandery (or one of the now few and much-reduced subsidiary kingdoms).
All the last three administrative units were manned by salaried officials who had been chosen on merit and posted from outside to avoid any local conflict of interest. Their functions were not simply extractive. Besides being responsible for registration and judicial duties, they organised relief and public works (roads, bridges, dams, granaries), regulated local markets and manufactories, and oversaw public order and security. Of the revenue they collected through taxation, whether in coin or kind, only part was remitted annually to the central government, the rest being retained for local expenditure. The same was true of the labour levies for civil and military service. Conscription (typically two years’ service for every adult male) and corvée labour (one month per year) did not normally mean transportation to the northern frontier or labouring on one of Chang’an’s imperial extravaganzas; terms were mostly served within the district or prefecture and to some mutually beneficial purpose, such as suppressing banditry or improving irrigation.
In 5 BC the total number of bureaucrats serving in both central and local government was estimated at over 130,000. Office was open to those with the requisite education or attributes, and as is the way with civil services, all posts were graded into a hierarchical pecking order that cut across departmental divisions. Pay scales conformed to this grading, although the values given to each rank as expressed in ‘bushels’ were survivals from a coinless past and had little to do with current pay equivalents.
Not even the emperor’s several thousand concubines and handmaidens were exempt from the system. By the beginning of the first century AD they too had been bureaucratised with ranks equivalent to those in government. Thus a ‘Brilliant Companion’ enjoyed the same status as the Chancellor, who was one of the highest officials in the land with a ‘10,000 bushels’ ranking; a ‘Beautiful Lady’ was ranked at 2,000 bushels, a ‘Compliant Lady’ at 1,000 and a ‘Maid for All Purposes’ at 300. Mere ‘Night Attendants’ and ‘Soothing Maids’, at 100 bushels, held roughly the same rock-bottom rank as ‘Accessory Clerks’.
It was the same throughout the emperor’s household and those of the empress and the heir apparent. The ‘Prefect Grand Butcher’, his assistants and apprentices, forty-two in all, were also ranked, although their bloodied subordinates who did the actual cutting and dicing were not; in 70 BC, by way of an economy measure, the latter’s number was reduced to 272.16
Neither the butchers and other provisioners nor the women of the bedchamber were employed merely to gratify a gargantuan imperial appetite. Both served a dynastic function. The ladies carried a heavy responsibility for the future of the imperial lineage, while the provisioners supplied the needs of its past, their prime carcases and first pickings being destined as sacrificial offerings to the ancestors. Conspicuous extravagance was an accepted measure of majesty, for the dead as for the living. Ancestral shrines had their own staffs for attendance on the deceased and for the maintenance of the tombs and their surroundings. Whole tomb towns, endowed and populated for the purpose, sprang up in their vicinity. According to the Hanshu, by the time of Han Yuandi (49–33 BC), there were over 200 ancestral shrines tended by some 57,000 officials, at which, in the course of a year, 24,455 meals were ritually served.17
Faithful service or imperial favour might lead to an advance in rank, but rankings were not a sure guide to professional competence or influence, only to social status. For the able, the ambitious or the well endowed, the system itself provided a genuine chance of recognition. Patronage and nepotism, though normal, rarely resulted in a post becoming hereditary. Although Sima Qian, for instance, had succeeded his father as Grand Astrologer (the Han equivalent of Grand Diviner), this was thanks not to inheritance but to a paternal apprenticeship for this highly specialised function, ‘the most versatile and technically trained . . . in the entire central government’.18 (At the time, there was no such office as ‘Grand Historian’; the term, though adopted by the author of the Shiji, became a title with a ranked office only when Ban Gu was officially encouraged to undertake the authorship of the Han dynastic history (Hanshu) in the late first century ad.)
The hierarchy of rankings provided a bureaucratic framework, but individual titles, ranks and responsibilities, both in central government and local government, changed with bewildering frequency. Neat flow-charts showing the most senior ‘Three Excellencies’ (typically Chancellor, Commander-in-Chief and Grand Minister for Public Works) atop the ‘Nine Ministries’ (Master of Ceremonies, Superintendent of the Household, Commandant of Justice, various treasurers, etc.), with each ministry then branching into its various subordinate bureaux, can be misleading. The system was far from static. And responsibility was as often divided as delegated: for all officials were subject to vigilant scrutiny from a separate censorate, registry, auditor or inspector. Direction might come from above, but correction came from beside or even below. ‘Parallel administration’ was an accepted feature of government long before its reinvention by the Chinese Communist Party.
Sometimes a post would be shared by two officials; sometimes posts were paired, with an official ‘of the Left’ being matched by one ‘of the Right’. More typically whole hierarchies were duplicated. The ‘Prefect Grand Physician’ and his staff of assistants and pharmacists were observed by the ‘Inspector of the Grand Physician’ with his own staff of assisting, attending and apprentice physicians. In the commanderies and kingdoms, teams of inspectors, sometimes called ‘shepherds’, roamed their flocks with bated writing-brush as they scrutinised the local administrations in search of improprieties. Yet merit, no less than misdemeanours, had to be reported; failure to do so could result in the censors themselves being censured. It was a system of checks and balances, worthy in every way of a civilisation that would invent the world’s first mechanical timepiece.
The emperor himself was not exempt from correction. Confucius had been adamant that senior officials, possessed by definition of the highest moral character, were duty bound to admonish the emperor as well as advise him. They did so obliquely, often fearlessly, and sometimes to effect. For example, while Heaven might sound an ominous warning with flood and famine, it was up to local officials to report such occurrences, ministries to mitigate them, astrologers and others to interpret them, and senior counsellors to bring them – or not bring them – to the emperor’s attention and urge unpalatable correctives. Autocracy was thus checked by bureaucracy. Ideally it was balanced by it; occasionally it was overbalanced by it.
Much depended on the character of the emperor. Wuwei rulers such as the possibly aloof Wudi and his pacific successors Han Zhaodi (r. 87–74 BC, Han Xuandi (r. 74–49 BC) and Han Yuandi (r. 49–33 BC) were well suited to the system. All the latter (the Han emperors ‘Z’, ‘X’ and ‘Y’ for those allergic to assonance) won warm praise in the pages of the Hanshu. Xuandi so epitomised the ideal of the responsive ruler that his reign ‘could be called the renaissance [of the dynasty]’, while Yuandi was ‘broad-minded and had his inferiors express themselves completely’.19 These ‘Z-X-Y’ decades of the mid-first century BC were marked by retrenchment after the extravagant expenditure of Wudi’s long reign. ‘All within the four seas were exhausted,’ says the Hanshu, ‘the population having been reduced by half’ as a result of Wudi’s foreign adventures. But thanks to corvée exemptions, tax reductions and various amnesties and economies, the people now rediscovered ‘rest and repose’ and ‘became opulent’. Harvests were good, and Heaven held its devastating hand to smile on all that lay under it.
But a dictator like the Qin First Emperor (he of the ‘terracotta army’ and tantalising tomb) had obviously not welcomed officials who were given to ‘expressing themselves completely’. His book-burning minister Li Si had habitually tendered advice only after ascertaining that it was likely to prove acceptable or ensuring that it first came from someone else. Emperors of an independent character were a liability. Bureaucrats and palace officials must either conspire to corrupt them with dissolute distractions or simply bide their time. Worse still, though, was no emperor at all, or one too young or incompetent to appreciate his role. Regents and empresses might then fill the vacuum, fighting among themselves, and placing the entire bureaucracy under enormous strain as they vied for its support or strove to short-circuit it.
This was what happened as the first century BC ground towards its end. It can hardly have been a surprise. Succession crises had dogged the Han ever since the death of their ‘Great Progenitor’ (Han Gaozu). The feisty Dowager Empress Lü had then manipulated the succession with a view to replacing the house of Han with her own Lü clan; and this seems to have set a precedent for similar manoeuvres by virtually every subsequent empress and her supporters.
Since an empress was by definition either the would-be mother, or the mother already, of a potential heir apparent, such jockeying should not have threatened the succession. But as in the bureacracy, so in the palace – duplication complicated matters. Rarely could it be assumed that an empress was for life; they came and went with dismal frequency. They might fail to produce a male heir, be eliminated by a rival, implicated in a plot, find another’s son preferred to their own, or simply fall from imperial favour. Delectable replacements from the ranks of all those ‘Brilliant Companions’ and ‘Compliant Ladies’ could be insistent and hard for an emperor to resist, especially if they had a promising heir apparent already in tow. Dowager empresses further complicated matters. As well as being experienced in palace intrigue and less vulnerable to imperial fancy, the dowagers retained great influence as residuary legatees of their husbands’ heavenly authority and as the mothers or grandmothers (in which case they were grand dowager empresses) of incumbent emperors.
Another succession-related crisis had darkened the last days of Han Wudi’s long reign. In 91 BC his empress, née Wei, had been implicated in a case of witchcraft. The origins of this affair are unfathomable, though its ramifications were horrific. Her son, the heir apparent, took up arms against his father, the ageing Wudi; Chang’an was plunged into civil war; and eventually almost the entire Wei clan was eliminated, including a chancellor, an inordinate number of generals, the heir apparent and the Empress Wei herself. But the proposed substitution of another favoured consort, plus son, met with much the same outcome; another chancellor was toppled and more generals disgraced, Li Guangli, the conqueror of Ferghana, among them. The Wei purge was then reviewed, mistakes were acknowledged, and the few surviving members of the clan reinstated. They included a toddler, the grandson of the rebellious heir apparent, who would one day reign as Han Xuandi, plus a stern and rather intimidating figure, the nephew of the deceased Empress Wei, called Huo Guang.
As Wudi’s health declined, Huo Guang emerged as a capable minister and stabilising influence. A new heir apparent, Wudi’s son by yet another consort, was soon in place, although when the emperor died in 87 BC, this Zhaodi was still only seven. Fortunately Wudi had made his wishes clear. Bypassing the ‘Three Excellencies’ (Chancellor, Commander-in-Chief and Grand Minister of Public Works), he had appointed what would in effect be a regency triumvirate headed by Huo Guang. Lest his meaning was still unclear, he had also presented Huo Guang with a specially commissioned painting that depicted the grand old Duke of Zhou, every Confucian’s hero, offering kindly guidance to the young King Cheng back in the early days of the Western (or Former) Zhou around 1043 BC.
To this challenge of emulating a revered figure from nearly a millennium earlier Huo Guang rose with almost indecent confidence. He exercised complete authority, buttressed it by supplying young Zhaodi with one of his granddaughters by way of an empress, and then when Zhaodi, still heirless and barely twenty, mysteriously died, shrugged off the inevitable suspicions and set about finding another compliant emperor. In this tortuous endeavour he was much assisted by the authority that reposed in the now dowager empress, who was otherwise his fifteen-year-old granddaughter.
As a first move, Wudi’s only remaining son was cold-shouldered, probably because, being more than twice the age of Zhaodi, he would have had little need for a regent. Instead, one of Wudi’s grandsons was summoned to Chang’an. According to the Hanshu, this young man responded to the call too promptly and then embraced his heavenly perks much too enthusiastically. The palaces rang with song and ribaldry when they should have been in mourning; Zhaodi’s womenfolk were liberally favoured while their former lord yet lay in state; and more credibly, the new emperor showed vindictive tendencies towards any who criticised his conduct. Sensing imminent danger as much as insupportable disrepect, Huo Guang moved quickly. Within a month the young man had been deposed by an edict issued on the authority of the little dowager empress. Another directive summoned Bingyi, the same great-grandson of Wudi and his Empress Wei who as a toddler had escaped the witchcraft purge of seventeen years earlier and was now (74 BC) installed as Han Xuandi.
For Huo Guang it was a case of ‘third emperor lucky’. The grateful Xuandi proved a model ruler and showered him with precious gifts, including the revenue and service obligations from an unprecedented 17,000 households. He also ensured that Huo Guang retained an unassailable position until, after a short illness, in 68 BC the great facilitator finally died of natural causes – itself a privilege seldom enjoyed by the most senior officials, especially those who dabbled in the succession. Huo Guang’s funeral was the grandest ever witnessed for someone of less than imperial rank. Xuandi himself attended; and, says the Hanshu, among the emperor’s parting gifts was the investiture, redolent of royalty, of a sea-grey suit of tailored jade.