The An Lushan Rebellion
For about a century before the winter of 755 the Tang dynasty had been the largest, wealthiest and most cosmopolitan state on earth. Its capital, Chang’an, was a planned city of broad avenues, monasteries and merchant quarters; Japanese, Korean, Sogdian, Turkic and Persian residents lived alongside Han Chinese populations, and the city’s population at the height of Xuanzong’s reign has been estimated close to two million. The Tang court had been a centre of Buddhist patronage and Confucian scholarship alike; Tang clothing and Tang poetry had become the prestige register of elite culture across East Asia; the Tang Empire’s tax revenues, flowing through a household registration system that the dynasty had inherited from the Sui, financed the largest professional army seen anywhere on the Eurasian continent since Rome. The first century of Tang rule is sometimes labelled a golden age by later Chinese historians because there was a single government reaching from the Tarim Basin to the Pacific, a single legal code, a single imperial cult of the Khan-of-Heaven and a single aristocratic culture shared across continents. The second century of Tang rule would end with the dynasty’s collapse at the hands of its own military governors in 907, and the dividing line between the two centuries is the long civil war that began on the 16th of December 755 at Fanyang.
The war itself is the An Lushan rebellion, also known by the Chinese historiographical name An Shi zhi luan, the rebellion of An and Shi, that is, of the two non-Chinese generals who led the Yan state against the Tang during the eight years that followed. The rebellion is among the most studied events in Chinese history precisely because of its consequences. It cut deep and irreversible rifts in the Tang state, ended its westward reach, shattered its household registration and taxation system, and began a long farewell between the central government at Chang’an and the powerful regional military governors, the jiedushi, who would hold substantial parts of the empire in their own hands for the remainder of the dynasty. The political history of the Tang state in 756 is genuinely different from the political history of the Tang state in 754. A period of consolidated imperial monarchy, regional coherence and centralised tax-and-service obligations gives way, after the rebellion, to a century of partial compensation, permanent military decentralisation, eunuch-dominated palace politics and agrarian trouble. Reading the Tang as two halves is an old habit among Chinese historians, and the rebellion of An Lushan is the place the halves meet.
The Tang state Xuanzong inherited
Emperor Xuanzong, the second Tang emperor to take this posthumous title, came to the throne in 712 after a contested regency period that had ended with his aunt, Princess Taiping, falling to a coup arranged by his father Ruizong. The young emperor inherited an empire whose administrative and fiscal armature had been carefully constructed over the previous century. The Tang model of government rested on three pillars. First, a household registration system maintained by local officials, in which every household had its assigned tax and service obligations. Second, a system of equal-field land grants that allocated to each farming household a fixed amount of state land and obliged it to return grain, cloth and corvée labour in return. Third, the fubing, a militia system inherited from the Western Wei and Northern Zhou, in which farming households owed a certain number of months of garrison duty each year in return for their land grants. The result was a state whose military manpower was drawn directly from the farming population of the north China plain and whose tax revenues flowed through the same household channels.
The fubing system had begun to drift into difficulty from the seventh century onwards. As lands got locked up in private estates, as aristocratic families grew richer and more politically powerful than the smallholder class the fubing had been designed to draw from, and as foreign wars against Tibet and the steppe increased the scale of frontier deployment, the fubing obligations became increasingly onerous. The Tang had therefore begun a slow shift to professional soldiers who were paid and garrisoned for longer terms. By Xuanzong’s reign the new pattern had produced a class of full-time soldiers, known by some sources as changzong suwei in their palace-guard form and as changzheng jian’er in their frontier form, who served their state year-round and required cash salaries. To command this professional force Xuanzong formalised the position of jiedushi, military commissioners who held combined military, civil and fiscal authority over large circuits consisting of multiple prefectures. By the late 740s ten jiedushi had been appointed to the northern and northwestern frontiers. Their regional commands made strategic sense in the short term; the jiedushi were closer to the threats they faced than the imperial centre could be, and they had the resources to recruit and pay professional troops whom the imperial centre could not easily supervise. In the longer term, of course, those same authorities became the bones of the rebellion.
Who An Lushan was
By 755 the most powerful of the frontier jiedushi was An Lushan, who held three commands in the northeast, Pinglu, Fanyang and Hedong, and commanded more than 150,000 professional troops. His biography is one of the most unusual in Tang history. An Lushan was of mixed ancestry: his mother was a Göktürk of the Ashide clan, reported in Tang sources to have been a shaman; his stepfather, An Yanyan, was a Turkic general who fled the Eastern Turks during the early Kaiyuan era. As a young man An Lushan lived in Yingzhou, a frontier settlement in present-day Liaoning which was a mixed community of Kumo Xi, Khitans and other non-Chinese populations. The Sinicised surname An, which he took from his stepfather, is thought by several modern scholars to derive from the Chinese transliteration of Anxi, the name given to Sogdiana in Tang diplomatic vocabulary, indicating that the family had arrived in the Tang frontier through the Sogdian merchant network that connected Bukhara to northern China.
His given name Lushan is plausibly a Sinicised form of a Sogdian or Iranian name meaning bright or light, the same Indo-European word as the Persian roxshanae that gives us the name Roxana. Tang historical sources report that An Lushan could speak at least six non-Chinese languages and that he served in his early career as an interpreter at one of the frontier markets where Tang silk was exchanged for Central Asian horses. Such markets were regular features of the Tang frontier economy and they functioned as both economic and informal recruiting grounds for frontier troops. By 736 An Lushan had joined the Tang army in some capacity and by 744 he had been made, on the recommendation of the chancellor Li Linfu, military governor of Fanyang. Li Linfu is widely thought to have favoured the appointment of non-Chinese generals precisely because they were less likely to undermine his own grip on the chancellery; he wanted generals he could rely on to be obedient and not to develop their own political bases inside the capital.
From 744 to 755 An Lushan combined the duties of a frontier governor with the cultivation of an extraordinary relationship with Xuanzong. The emperor, who had been a brilliant and busy monarch in the first half of his reign, had grown tired and indulgent in the 740s. He doted on his favourite consort Yang Guifei, a lady of the Yang clan who had previously been the wife of one of his own sons. The young consort and her cousin Yang Guozhong, who was later to inherit the chancellorship, were important figures at court. The Tang sources are unkind to both Yangs, blaming them for the emperor’s loss of interest in state business and for the political chaos that followed, but there is no serious historiographical doubt that the late Xuanzong period was marked by indulgence, factional struggle and a steady detachment of the imperial centre from the actual business of government. An Lushan cultivated Yang Guifei with elaborate deference. Court sources say that he visited Chang’an several times, that he made much of his physical bulk and his talent for low comedy in the consort’s presence, that he was adopted by her in a ritual of make-believe sonship, and that he used the resulting access to the imperial household to gain steadily more authority.
By the late 740s An Lushan had accumulated three jiedushi commissions and the loyalty of an army that was, in absolute size, comparable to a quarter of the formal Tang household population. He had been granted an iron certificate promising exemption from execution except for treason, which is an extraordinary immunity for a non-imperial-clan official to receive. He was, in effect, the most powerful single territorial figure the Tang state had produced in over a century. The question was whether the central government would tolerate that position or whether Yang Guozhong, who became chancellor in 752 after the death of Li Linfu, would attempt to bring An Lushan into line. The result, plainly, was the rebellion.
The march from Fanyang
On the 16th of December 755 An Lushan mobilised his army and marched out of Fanyang, near modern Beijing. He declared, apparently falsely, that he had received a secret imperial command to march on the capital to remove Yang Guozhong. His first major stroke was not to advance directly on Chang’an but to seize the eastern capital Luoyang by way of the Grand Canal corridor through Chenliu and Yingyang. The Tang generals Feng Changqing and Gao Xianzhi attempted to oppose him at the Hulao Pass region, but their hastily levied troops were no match for An Lushan’s frontier veterans. Luoyang fell in January 756. On the first day of the lunar new year, the fifth of February 756, An Lushan declared himself emperor of a new dynasty he called Yan, taking the title Great Emperor of Yan and formally inaugurating what would become an eight-year-long civil war.
The fall of Luoyang was only the opening stage of the campaign. An Lushan’s next strategic target was Tong Pass, the strategic chokepoint that guarded the western approach to the capital from the eastern plains. The defence of Tong Pass in 756 was given to General Geshu Han, a Tujue general whose earlier career had included a costly and unsuccessful campaign against the Western Turks. Geshu Han’s defence was tactically competent but politically constrained; Yang Guozhong’s interference at the capital prevented him from withdrawing to a more defensible position, and on the seventh of July 756 Tong Pass fell to An Lushan’s assault. The road to Chang’an now lay open.
The flight to Sichuan and Mawei
What follows in the Tang annals is one of the most affecting episodes in Chinese imperial history. The emperor, his palace, the imperial clan, the major officials and the Tang imperial household abandoned Chang’an in panic and retreated toward Sichuan. The route through the Qinling mountains was the traditional safety corridor, and Xuanzong hoped to gather support in his ancestral base near Chengdu. At Mawei, a station on the road between Chang’an and Sichuan in what is now Xingping in Shaanxi, the imperial escort mutinied. The soldiers were exhausted and hungry. They blamed Yang Guozhong for the disaster: he had brought the war upon them by his vendetta against An Lushan and his arrogant mismanagement of Tong Pass. When the soldiers demanded his head, Xuanzong could not refuse. Yang Guozhong was executed. When the soldiers then demanded the death of Yang Guifei herself, the emperor hesitated. The story goes that he tried to spare her life, that she was his favourite, that the troops were not appeased by partial concessions. Finally he agreed; Yang Guifei was forced to hang herself at Mawei.
Yang Guifei’s death has been commemorated in Chinese poetry, drama, opera and popular narrative for fourteen centuries. The Qing-period play Changshengdian, the Hall of Everlasting Life, by the playwright Hong Sheng, is the most famous literary treatment, but the story recurs in Tang verse after Tang verse; it was understood by the Tang dynasty itself to be the central human drama of the rebellion. The emperor’s reluctance to kill her, his eventual surrender to the troops’ demand, his long and melancholy retirement into the imperial household at Chengdu afterwards, are all part of a set of stories that structured Chinese upper-class grief for the rest of the dynasty. The cultural afterlife of Mawei is partly a tribute to the political and personal importance of the moment.
At Mawei the emperor also lost practical control of his dynasty. The crown prince Li Heng did not retreat to Sichuan with his father; he slipped away separately to Lingzhou in the northwest, in modern Ningxia, where he was proclaimed emperor by a coalition of frontier generals and provincial officials who feared the worst. Xuanzong, still in flight, eventually recognised the fait accompli and abdicated in favour of Li Heng, taking the retired title of Taishang Huang, the retired emperor. The reign name under Li Heng was Suzong. Three emperors and a similar number of Yan claimants would hold the imperial seal during the next eight years; An Lushan’s death in 757, his son An Qingxu’s seizure of power and its overthrow by Shi Siming, Shi Siming’s death and his son Shi Chaoyi’s brief succession, the brief Tibetan seizure of Chang’an in 763, all crowd into a historical sequence that the Tang dynastic histories recount with an almost desperate narrative pace.
The long war and the Uyghur alliance
The first year of Suzong’s reign saw the Tang and Uyghur courts come to terms. The Tang had been unable to defeat Yan in the field; the Uyghurs, on the other hand, had the horse archers who could match the steppe cavalry tradition that An Lushan had recruited from. Suzong accepted that the Uyghurs would be paid in return for intervention: by long and detested tradition, payment would come out of the silk revenues of the eastern territories, and the Tang dynasty would later come to regret this contractual relationship because the Uyghur cavalry under Bögü Khan took the chance to sack Luoyang when their Tang allies proved slow in 757. The first year of the alliance was nevertheless decisive: in late 757 the joint Tang and Uyghur forces retook both Chang’an and Luoyang. The Tang administrative apparatus crawled back to the rebuilt city in the autumn of 757 with the help of the great Tang general Guo Ziyi, who had been operating in the northwest and who now took operational charge of the campaign.
The capture of Chang’an did not end the war. An Lushan had been assassinated in the previous January by his son An Qingxu and a eunuch named Li Zhu’er, in part because An Lushan’s court at Luoyang had descended into paranoia and cruelty; but his death did not end Yan. In 759 An Qingxu was killed by Shi Siming, a former Yan general, who declared himself emperor of a successor Yan. Shi Siming was killed by his own son Shi Chaoyi in 761, and Shi Chaoyi continued the war until 763, when the Tang with their Uyghur allies retook Luoyang for the second time. Shi Chaoyi fled north and committed suicide in February 763. The Yan dynasty collapsed, but at enormous cost.
The middle years of the rebellion saw a great deal of internal Yan distress: factional fighting between the An and the Shi camps, defections to the Tang in the rear areas, a partial resurgence of central authority in the northeast. The Tang, for their part, were fighting three different wars at the same time: against Yan in the central plains, against the Tibetan Empire in the northwest (where Tibet had taken advantage of the Tang distraction to occupy border territory and briefly seize Chang’an in late 763, just as the Yan war ended), and against local uprisings in the south and in the lower Yangtze. The Chinese historian felt the cumulative weight of these conflicts in the demographic and administrative records of the period, and the modern debate over the demographic cost of the An Lushan rebellion partly reflects the difficulty of distinguishing demographic collapse from administrative collapse in those same records.
The demographic disaster
The demographic evidence for the An Lushan rebellion is striking because it is built on direct comparison of two Tang censuses. The last census undertaken before the rebellion, in 755, recorded 52,919,309 persons in 8,914,709 households. The first census undertaken after the rebellion, in 764, recorded only about 16,900,000 persons in 2,900,000 households. The drop is roughly two-thirds. The simple interpretation is straightforward: the rebellion killed tens of millions of people, by direct combat, by famine, by plague, by displacement and by the loss of entire rural administrative structures in the north. The cautious interpretation is more sophisticated: the same period also saw the collapse of the central government’s ability to conduct a census, the disappearance of large classes of people from the tax rolls (Buddhist monks, Daoist priests, urban merchants, foreign residents), and the partial secession of the north-eastern circuits from the Tang fiscal system altogether.
Modern Chinese historians, working on the demographic evidence of the Tang census and on a careful comparison with surviving local gazetteer and tomb-epitaph evidence, have tended to settle on a cautious reading: the literal death toll of the rebellion was substantial, perhaps 13 million and possibly as many as 36 million, but the steep decline in the registered household count was probably exaggerated by a near-total administrative collapse of the north China census apparatus. The result is that historians who take the maximalist view of the death toll treat the An Lushan rebellion as one of the most lethal single events in human history; those who take the minimalist view treat it as one of the great administrative catastrophes of the medieval world. Both views agree on the basic picture of an old agricultural and administrative order brought low.
The demographic disaster was compounded by population displacement. Southern China, which had been a relatively poor and lightly populated region in the early Tang, gained an enormous influx of displaced farmers and craftsmen from the north. The population of the lower Yangtze region doubled in the decades after 763. The administrative and economic gravity of the empire shifted permanently southward. The Tang’s political capital might remain in Chang’an, but its demographic centre, its fiscal strength and its cultural weight moved increasingly into the Yangtze basin.
The institutional transformation
The An Lushan rebellion did not by itself produce the institutional changes that defined late Tang; it did, however, render them irreversible. Three structural changes deserve particular attention. The first is the survival of the jiedushi in hereditary command. Many of the rebels who surrendered at the end of the rebellion in 763 were pardoned but allowed to retain their regional commands. The Tang court could not displace them without restarting the war. Generals who had been loyal to the Tang during the war were rewarded with new jiedushi commissions. Over the next half-century the title became effectively heritable, and from the late ninth century onward the same families of frontier military strongmen would pass the title from father to son or to a designated heir, with the emperor rubber-stamping the succession. The jiedushi had been Tang officials in 750; they were in effect independent regional warlords by 800.
The second structural change is the reorganisation of the fiscal system. The old fubing-and-household-registration system was wholly destroyed by the rebellion; peasants had been displaced by the millions; the population registered on the household rolls of north China had shrunk to a fraction of its pre-war level; and the equal-field system on which rural taxation depended had effectively collapsed. From the late 750s onwards the Tang court had to look elsewhere for revenue: salt monopolies became one of the largest single sources of tax income; the price of salt rose; transactions became more monetised; and ultimately, in 780, the chancellor Yang Yan introduced a new twice-tax system, the liangshuifa, that taxed each household twice a year on the basis of its actual landed income. The twice-tax system was, in its intentions, a more flexible and more accurate system than the poll-and-corvée system it replaced. In its actual operation it empowered whoever controlled local collection, and that was the jiedushi.
The third structural change is the rise of the imperial eunuchs. After the An Lushan rebellion and particularly after the brief Tibetan occupation of Chang’an in 763, the Tang emperors and chancellors became increasingly suspicious of the regular army and increasingly reliant on the palace eunuchs and the units they commanded directly. The Shence Army, the principal palace-guard force, came under eunuch command. The mechanism was a long-running feature of late Tang: whenever a chancellery reform threatened the palace’s grip on military power the emperor fell back on the eunuchs, and whenever the eunuchs became too powerful a chancellery reform threatened them. This dance of palace and bureaucracy would be a central feature of Tang politics for the next century-and-a-half. Its roots, however, lie in the An Lushan moment.
Why the rebellion mattered
Six consequences stand out. The Tang Empire lost its grip on the Western Regions. The An Lushan rebellion destroyed the long-standing Tang military presence in the Tarim Basin, the Hexi corridor and the Central Asian protectorates, and the Tibetan Empire moved into the vacuum. From 763 onwards the Tarim Basin and the Hexi corridor would be contested by Tibetan and Uyghur powers, neither of which was answerable to Chang’an. The Tang’s prestige in the wider East Asian world also declined. The Japanese diplomatic missions that had been a regular feature of the early Tang became rarer; the Korean kingdom of Silla, in receipt of Tang aid during its wars with Paekche, found the relationship harder to maintain; and the Vietnamese polity of Jiaozhi at the southern border became, by the 860s, effectively autonomous. The Tang state of the seventh and eighth centuries had been the centre of an international order; the Tang state of the ninth century was a regional power holding a rump of its original domain.
Second, the regional warlordism that characterised the late Tang originated here. The jiedushi of Hebei, Shandong and parts of Henan were granted in the 760s and afterwards and became hereditary within a few generations. The Hebei jiedushi in particular were quasi-independent for most of the second half of Tang, paying nominal tribute to Chang’an, refusing to transfer their revenues, and often treating the emperor’s appointees as figureheads. The Tang system had become a system in which the emperor’s writ ran more strongly in the south than in the north.
Third, the fiscal reorganisation that began in the late eighth century as a response to the lost household registration affected every subsequent Chinese state. The twice-tax system was, in essence, the first Chinese general-purpose income tax in modern form; it survived into the Song dynasty and would remain the basic Chinese fiscal arrangement for centuries. The economic monetisation that accompanied it produced a far more thorough commercial economy by the tenth century. The An Lushan rebellion, in this sense, is the seedbed of Song prosperity.
Fourth, the demographic shift south was permanent. The Tang census of 755 had 75 per cent of the registered population in the north and 25 per cent in the south; by the eve of the Song dynasty the same numbers were close to 50-50. The economic centre of Chinese civilisation, having been in the Yellow River basin for at least two thousand years, moved to the Yangtze and never moved back. Modern China is a Yangtze-centred civilisation; the move happened because of the demographic collapse of the An Lushan period.
Fifth, the cultural memory of the rebellion structured Chinese attitudes toward military power for the rest of the imperial period. The Yang Guifei episode became a school of personal tragic drama; the An Lushan period was remembered as a warning of the perils of frontier generals, of consort politics, of inefficient chancellery management. Confucian scholars of the Song dynasty would return repeatedly to the An Lushan period as a case study in dynastic decline.
Sixth, the rebellion institutionalised a pattern of palace-versus-bureaucracy tension that ran to the end of the dynasty. The emperor had come to depend on the palace eunuchs and on the Shence Army; the chancellery responded by trying to curb both; the emperor then relied on the eunuchs again, and so on. The pattern was set in the An Lushan moment and it recurred, with variations, until the dynasty’s final collapse in 907. The An Lushan rebellion, in this sense, did not just cause a particular catastrophe; it locked in a political style that would last for centuries.
The long aftermath
By 800, a full half-century after An Lushan’s death, the Tang dynasty was still functioning but was recognisably different from the Tang of 700. The jiedushi were entrenched in their hereditary commands. The eunuchs held the Shence Army and operated the imperial household. The fiscal bureaucracy relied on twice-tax assessments. The central state had lost the Western Regions and was holding the central and southern circuits reasonably tight. The state still produced Tang literature, Tang painting, Tang Buddhist scholarship; the dynasty had not ceased to be culturally civilised. But it had ceased to be an empire.
By 850 the dynasty was visibly declining. Peasant revolts, religious movements, the weariness of the bureaucracy and the recalcitrance of the northern jiedushi combined to push the state into a slow decline. By 880 the Huang Chao rebellion, a major agrarian rebellion led by a salt trader turned millenarian, would sack Chang’an, and the emperor would flee once again to Sichuan. By 907 one of the central jiedushi, Zhu Wen, would depose the last Tang emperor and begin the Five Dynasties and Ten Kingdoms period. Almost every feature of that later collapse is traceable to the An Lushan moment: the regional commanders, the soldiers of fortune, the chronic fiscal strain, the loss of the Western Regions. The An Lushan rebellion, in this sense, was the early sign of a deeper exhaustion rather than the random catastrophe it sometimes looks like in the Tang annals themselves.
What the rebellion looked like to its contemporaries
The Tang court chroniclers of the early post-rebellion period wrote the rebellion as a story of personal and dynastic tragedy. The Old Book of Tang and the New Book of Tang, both produced within a century of the events, present An Lushan as a traitor who succumbed to the temptations of court favour and personal vanity; they centre the Yang Guifei episode as the moral core of the rebellion’s tragic dimension; they treat the Muslim Uyghur ally as a saving but uncomfortable foreign power. The official histories stress the Tang emperor’s suffering, the final military triumph and the lessons to be drawn from inadequate vigilance. Modern historians have tended to deconstruct this narrative: the rebellion was systemic, not merely personal; Yang Guifei was a symbol rather than a cause; the social and economic stresses that produced the rebellion were visible from the 730s onward; and the empire’s institutional resilience in the second half of its history owed as much to its inheritance as to its innovation.
Reading the rebellion today
Three contemporary angles deserve particular attention. The first is the comparison with other continental collapses. The demographic and administrative catastrophes of the second half of the eighth century have their analogues in the contemporary Mediterranean collapse of the fifth and sixth centuries and in the central Islamic collapse of the ninth and tenth centuries. Treating the An Lushan period as one of a small number of large Eurasian cases of late-antique administrative collapse has been a productive comparative project in recent decades. The second is the question of regional militarism. The jiedushi of the late Tang and the daimyo of late Tokugawa Japan were not the same institution, but the late-Tang jiedushi and the European and Japanese analogues of regional military fiefdoms are parts of a recognisable late-imperial pattern. The third is the question of cultural memory. The An Lushan rebellion continues to be remembered in Chinese literature, drama, opera and popular novels; it is one of the most cited rebellions in Chinese historical writing. The cultural afterlife of the rebellion is itself a major topic in late imperial Chinese cultural history.
For the reader who comes to the topic through a Chinese-language source, the rebellion is a tragedy whose inner moral arc is the emperor’s loss of self-control and the consort’s role in producing that loss. For the reader who comes to it through modern scholarship, the rebellion is a structural episode in which institutional change and elite behaviour combined to produce a more regional, more militarised and ultimately less centralised late imperial Chinese order. Both readings matter; both are present in Tang and Song Chinese historiography itself. The intellectual challenge is to hold them together while recognising that neither alone captures the eighth-century Chinese state.
References
- An Lushan rebellion — Grokipedia, accessed 15 October 2026.
- An Lushan — Grokipedia, accessed 15 October 2026.
- Tang dynasty — Grokipedia, accessed 15 October 2026.
- “An Lushan” entry (Encyclopaedia Britannica) – Encyclopaedia Britannica; biography noting his Iranian-Turkic descent and the 8th-century Tang frontier context.
- “Political History of the Tang Period” (ChinaKnowledge.de) – Ulrich Theobald, ChinaKnowledge.de; detail on jiedushi, the financial reforms after the rebellion and the broader political history of the dynasty.
- World History Encyclopedia overview of the Tang dynasty (An Lushan section) – World History Encyclopedia; English-translation French summary of the rebellion’s human cost.
- “Consequences of the An Lushan Rebellion” (Warexplorer.com) – Warexplorer.com; summaries the demographic argument based on Tang census figures.
- Worldcat catalogue subject-heading search for An Lushan – Worldcat union catalogue.
- Worldcat catalogue search for Denis Twitchett (ed.), The Cambridge History of China, vol. 3, Sui and T’ang China, Cambridge: Cambridge University Press, 1979 – Worldcat record for the standard scholarly reference on the Tang.
- Worldcat catalogue search for Hugh R. Clark, Community, Trade, and Networks: Southern Fujian from the Tang to the Ming, and other Hugh Clark titles on the middle Tang period – Worldcat union catalogue.
