• 400
  • 500
  • 600
  • 700
  • 800
  • 900

The Uyghurs were a people of the Tiele, the Nine Oguz of the inscriptions, and for a century they were the power of the eastern steppe. In 744 their leader Kutlug Bilge Kül (Guli Peiluo), with the Basmïl and the Karluk, overthrew the last khan of the Second Göktürk Khaganate. The coalition did not outlast its victory. The Uyghurs and Karluk turned on the Basmïl and subjected them. The Uyghurs then brought the eastern Karluk under their control and stood alone. From the Orkhon, from a walled capital on the site the later sources call Karabalghasun, they ruled until 840. In that year the Kirghiz came down from the Yenisei with a hundred thousand horse, killed the khan and burnt the court.[1]

The inscription of their founder repeats the Göktürk creed of sky and earth. It names Ötüken as the centre of his realm. The thirteenth-century historian Juvaini records a Uyghur tale of origin. In it two trees stand at the meeting of the Tola and the Selenga, there is a mound, and of five children Bugu Tegin, “the stag prince”, is chosen khan. A silver plaque from the Noin Ula graves shows a broad-antlered animal between two slender trees on a mound. It has been understood as an early form of this tale.[2]

Sources

The connected accounts are Chinese. Chapter 195 of the Jiu Tang Shu is a chronological account of Uyghur political history under the Tang. Chapter 217 of the Xin Tang Shu is in two parts. The first takes the history to 814. The second completes it and adds notes on the tribes of the empire, on peoples related to the Uyghurs, and on the Kirghiz who drove them from Karabalghasun. The older history is dated much more precisely, often to the month or day. The newer rarely gives more than the year. It is briefer, but it has material the older lacks. Both rest almost entirely on the Veritable Records of four reigns, those of Suzong, Daizong, Dezong and Muzong. The Chinese accounts are weighted accordingly. About ninety-five per cent of the older history’s account of the empire concerns the years 756 to 795 and 821 to 822. Together these are less than half the century the empire lasted. The Chinese wrote about the Uyghurs in China, not about the Uyghurs themselves. No Chinese source mentions the khan’s conversion to Manichaeism, the most consequential event in the empire’s inner history. The name Uyghur itself changed its range. From the middle of the eighth to the middle of the ninth century it named the Turkic empire of high Mongolia that followed the Göktürks. It then became the name of the kingdom of the Turfan region.[3]

Two stones and a traveller supply what the Chinese left out. The Shine-usu inscription, found in northern Mongolia in 1909, gives the wars of the second khan in the first person. The trilingual inscription of Karabalghasun was found in 1889. Its Chinese text is the best preserved. The inscription is the only early source for the introduction of Manichaeism and for the restoration of the empire under its second dynasty. Among Muslim authors, everything known of the Uyghurs before 840 goes back to Tamim ibn Bahr. He saw Karabalghasun with his own eyes. His account is lost and survives in quotation.

Further texts of the Uyghur age survive in pieces or with disputed readings. The stele of Tes is broken, the painted runiform texts of Hoyto Tamir are known from copies, and the gravestone of Ulaangom is cut in Uyghur letters, though its place in the khaganate is doubted.

History (744–840)

The Göktürk monuments name the Uyghurs only once. From this it has been concluded that the Uyghurs, unlike other Turkic peoples, obeyed the Göktürk khan, paid tribute and lived quietly. Bilge Khan marched against the Uyghurs of the Selenga while he was busy with the Oguz. Their elteber fled east with about a hundred men. The captured herds may have ended the famine that the loss of animals and the war with the Oguz had brought to the Göktürk camp, so that the campaign had an economic purpose.[4] The Old Tang History names the earlier leaders. When the chief Shijian died, the tribe chose his son Pusa, who was brave and good at planning. In the first years of Zhenguan he and the Xueyantuo raided the northern border of the Göktürks. Yukuk Shad, a son of Ilig Khan, rode against him with 100,000 horsemen, and Pusa beat him at Mount Maliao with 5,000 and pursued him as far as the Tianshan.[1][5]

Kutlug Bilge Kül (Guli Peiluo) founded the state. His son Bayanchur (Moyanchuo), who reigned from 747 to 759, made it. By one count, the Uyghurs, Karluk and Basmïl drove the last important Göktürk ruler from the steppe in 742.[6] Kutlug Bilge Kül (Guli Peiluo) defeated the Basmïl and took the title of khan in 744.[7] According to the Shine-usu inscription, he tightened the Uyghur hold on the restive Basmïl and Karluk. He also built the capital and the palace, and a second city on the Selenga. He sent his eldest son to serve the Tang against the rebellion of An Lushan, which lasted from 755 to 763.[7] In 757 the Uyghur troops helped to retake Chang’an and Luoyang, and they plundered Luoyang for three days. The elders of the city paid them ten thousand rolls of silk to make them stop.[8] The Uyghurs also asked for settlements of their own in the two cities. These were avoided only after very large gifts had been offered.[9] His second son Bögü (Mouyu) reigned from 759 to 779. He brought the empire to the height of its power, and to its turning point. In 762 and 763 he was at Luoyang, where he had come to help the Tang against the rebellion of An Lushan. There he met followers of Mani and was converted. Shortly afterwards he imposed the faith on his people.[10] He met opposition at his own court. He found it necessary to organise his subjects in groups of ten, with one man in each answerable for the souls of the other nine. Manichaeism was a Sogdian religion. Its adoption raised the Sogdians at the Uyghur court to something like a state within the state. They were merchants, builders and advisers, and their marks are on the ruins of Uyghur buildings.

The Sogdians brought about Bögü’s end. In 779 they advised him to use the mourning for a dead Chinese emperor to invade China. His chief minister and first cousin Tun Bagha (Dun Mohe) opposed the plan. Tun Bagha saw the tide turn against him. He murdered the khan together with many of the khan’s Sogdian advisers and took the throne.[10] The Chinese court recognised him in 780, and he reigned until 789.[10] A period of rule against the Sogdians followed. There was probably also a turn away from Manichaeism to the old cults of the steppe. After the deaths of Tun Bagha (Dun Mohe) in 789 and of his successor in 790, the empire declined. At court the authority of the ruling Yaoluoge clan passed to a general. In the west the same general failed to hold Beiting against the Tibetans. He acted as regent for a khan still in his teens. In 790 he twice failed to relieve Beiting.[11]

In 795 the throne passed to another clan, the Xiedie. Its khan, whom the Chinese call Huaixin, was probably the general who had lost Beiting. A Chinese source says that he came of the Ediz, whom the Chinese write Xiedie. According to the same source, he took the surname of the Yaglakar when he became khan. He also sent the sons, grandsons and young relatives of the khans since Tun Bagha (Dun Mohe) to the Tang court.[12] Under him the empire recovered. Beiting was retaken. The reach of the Uyghurs extended to Ferghana. The Sogdians and the Manichaeans returned to influence. Beiting had been retaken from the Tibetans in 791 or 792, while he was still a minister. By contrast, most accounts have the Tibetans holding it much longer.[13] The Chinese text of the Karabalghasun stele, written after 808, tells of a khan’s restoration of the empire. He may be this founder of the second line. Another dating puts him somewhat later.[13] His successor, called Baoyi by the Chinese, reigned from 808 to 821. The state was still strong in the west but already declining. Its court was given to luxury, and its Tibetan neighbours were growing confident. Most accounts give 805 for the death of the founder of this line and so for the accession of Baoyi. A study of the succession of the khans argues for 808, and later work follows it.[13] A Manichaean hymn-book names dignitaries of Beiting, Kocho, Kucha, Aksu, Kashgar and Karashar. This points to strong Uyghur influence in those places, though not necessarily to outright rule. Some date the text to the reign of Baoyi. Others date it to that of Hesa (824 to 832).[14]

After 821 the empire came apart. There was war with the Kirghiz to the north, intrigue at court and rebellion at home. In 839 a bad season and a hard winter killed much of the livestock on which everything rested. Both Chinese histories record the end. They do not agree. In the older history, a minister who kept soldiers outside killed the khan in 839, out of resentment at the execution of two conspirators. He set up another khan.[1] A general who hated the minister fled to the Kirghiz.[1] At the head of a hundred thousand Kirghiz cavalry he attacked the Uyghur fortresses. He killed the new khan, beheaded the minister, and burnt and all but destroyed the court. The Uyghurs scattered across the barbarian lands.[1][15] In the newer history, the minister led some Shatuo against the khan. The khan took his own life, and the people set up a prince in his place.[16] That year there was famine, pestilence and heavy snow, and many of the sheep and horses died.[16] The Tang envoy who came to announce the accession of a new emperor in 840 found the state in disorder.[16] Then the general and the Kirghiz, a hundred thousand horse together, attacked the fortresses. They killed the khan, executed the minister and set fire to the royal camp. All the tribes were scattered.[16] The survivors went in several directions. A group of fifteen tribes is said to have gone to the Karluks in September 840 under their chief and his five nephews, and a branch of it to have moved on to the Tibetans and to Anxi. Another group took refuge with the Shiwei, whom the Kirghiz drove out in 848. Thirteen tribes went to the Yellow River and were scattered in 847 by Kirghiz and Chinese. Small groups joined the Khitan, the Mongols and the Tungus.[17] The same two main movements are given in a Turkish history of the Oguz. Thirteen Uyghur tribes fled to the Chinese frontier, and fifteen went west, some of them to Tibet and then to Gansu, some into the Tianshan.[18] The ruler of one western group had taken the title of khagan by 848, and the Tang court recognised him in 856.[19] Some scholars call the Kirghiz the last of the Turkic peoples to rule in Mongolia.[20] The compiler of the newer history closes with a sentence that sounds like satisfaction. It says that their state in the long run did not rise again.[16]

Peoples and court

At the foundation, the core of the population was a union of nine Turkic tribes, the Toquz Oghuz or Nine Surnames. They were the Uyghurs, the Bugu, the Hun, the Bayirku and the Tongra, and the Sijie, Qibi, Abusi and Gulunwugu.[21] The Chinese count them all among the Tiele. The last two were put on a level with the others only about 742.[21] An embassy of 788 was led by an official of the Ediz, who therefore belonged to the union by then. The Chinese text of the Karabalghasun stele, written after 808, still calls the khan “khagan of the Nine Surnames”.[21] By then the division had become political rather than ethnic.[21] The ruling tribe, the Uyghurs themselves, was split into ten clans, the On Uyghur. The Yaglakar was the leading one, and until 795 every khan came from it.[21]

The Basmïl and the eastern Karluk were subjects and counted for less. In battle the Uyghurs put them in the forward guard.[22] Others hold that in law they stood level with the Uyghurs themselves.[23] After 774 the Uyghur khan took the title of the Basmïl ruler, iduq qut.[24] The loss of Beiting to the Tibetans in 790 cost the Uyghurs their allegiance until they retook the western regions not long after.[22] Among the minorities the Chinese and the Sogdians were the most important. A Chinese of the Lu family had gone to Karabalghasun and become the khan’s adopted son. In 792 the Tang honoured him with the surname Yaglakar.[22] To the east the Khitan acknowledged the Uyghurs as overlords until 842, when they went over to the Tang.[25] Following the old Turkic custom, the Uyghurs also put the Karluk, Telengut, Tatar, Khitan, Kirghiz and others in a position of vassalage and tribute. They sent special officials to their nomad vassals to see that the yearly tribute was collected.[23] A line of the Šine-Usu stone names the Ten Uyghurs and the Nine Oguz one after the other as the people that remained. This is taken to show that the Nine Oguz were a people distinct from the Uyghurs.[26][27] Some scholars hold that the Nine Oguz stood in law almost where the Göktürk people stood. They received no ruler of their own or from the dynasty, and the khans governed them directly.[28] Some hold the Shatuo to be the part of the Toquz Oghuz that moved east into the Tarim lands before 840 and held Beshbalik early in the ninth century.[29]

A Chinese record gives the reach of the khaganate at the time of Kutlug Bilge Kül (Guli Peiluo). The Shiwei marked the eastern end, the Altai the west and the Gobi the south. The khaganate therefore held the whole of the old Xiongnu land.[16] No northern limit is given. The khan probably thought his possessions ran at least as far as Lake Baikal.[14] After the loss of Beiting the Karluk took the valley of Futu from the Uyghurs. The valley lay probably north-west of Ötüken, dangerously close to Karabalghasun. The Uyghurs, the Chinese say, trembled with fear and moved the north-western tribes to the south of the royal camp.[14]

The offices were in the main those of the Göktürks. In the 640s, when the Tang first organised the Uyghurs, their chief had already styled himself khagan. He had also set up official names after the manner of the Göktürks. Nothing shows that the duties changed afterwards.[1] The tribal leaders bore the title tutuk, from the Chinese dudu, “governor-general”. There were eleven of them: one for each of the nine tribes, and one each for the Karluk and the Basmïl. The tutuk of the Uyghurs was normally a close relative of the khan but not the khan himself.[30] They lived among their own people. They kept a garrison at court and had influence there. Nothing shows that the khans controlled what they did in their tribes.[31]

Succession ran from father to son among the Yaglakar. Bögü (Mouyu) came to the throne in 759 only because his elder brother was dead. This suggests that the Uyghurs followed primogeniture. Tun Bagha (Dun Mohe) was the first to set the rule aside, by the coup of 779.[32] When he died in 789 his son succeeded him. The next year the new khan was killed by his younger brother. An immediate rebellion set the murdered khan’s son, hardly more than a child, on the throne. It was probably the strength of feeling for father-to-son succession that brought the rebellion.[32] Power was not confined to the royal house. The founder of the second line had been a minister under his predecessor. The chief minister of Tun Bagha (Dun Mohe) and of later khans appears to have been unrelated to them. The Sogdians, most of them Manichaeans, held power at court from the conversion of Bögü (Mouyu) on.[32] Of the thirteen khans, at least five were assassinated. A sixth was murdered or took his own life after a minister’s rebellion succeeded.[33]

The principal wife of at least seven of the thirteen khans was Chinese.[34] Of the more than twenty women the Tang gave as brides to foreign rulers, only three were daughters of emperors: the princesses of Ningguo, Xian’an and Taihe.[35] When Bayanchur (Moyanchuo) died in 759, his widow the princess of Ningguo, a daughter of the emperor Suzong, was asked to be buried with him. She refused. She calmed the Uyghur court by slashing her face, an ordinary sign of grief.[36][37] Marriage was also trade. After the Tang envoy had invested Bayanchur, the khan presented 500 horses, sable furs and white felt, and he shared the gifts he had received among his officers and chieftains.[8][38] In 768 the khatun of Bögü died. The next year the Tang gave him a younger daughter of Pugu Huai’en as his second wife, and the envoy Li Han invested her. The gift was 20,000 rolls of coloured silk. The court was short of money and levied mules and camels from its officials to equip the mission.[16][38] In 788 the khan sent his chancellors with more than a thousand men, and his sister with the wives of great chiefs, to fetch Princess Xian’an. The two histories count the wives at fifty and at fifty-six. He paid the bride price in 2,000 horses.[8][38]

Seven of the thirteen full titles contain one of the phrases teŋridä, ay teŋridä or kün teŋridä, meaning “from Heaven”, “from the Moon God” and “from the Sun God”. They show that the khans held their power to come from above.[39] The Sogdian text of the Karabalghasun stele calls a khan of the second line the great Turkic ruler of the world who has received his splendour from Heaven.[40] The Chinese histories record the “appointment” of each khan as though his right to rule were a favour of the emperor. The Uyghurs probably took the ceremony as a formal act of recognition and did not regard themselves as under Chinese guardianship.[40] A leaf of a Manichaean book from Qocho bears a khan’s title in Old Uyghur, with the name torn away. Its first editor set it beside four titles of khans that open with the words ay teŋridä qut bulmïš.[41] The same ruin yielded a leaf with a list that its finders described as the court of the Uyghur khagan, in Manichaean script and Middle Persian.[42]

The writ of the khans ran short in the distant parts. Beiting and the other western cities lay within the Uyghur domain, yet until at least 790 the Chinese kept governors there as well and counted them as protectorates.[43] Tun Bagha (Dun Mohe) charged a toll on those who came from the city through his territory. The Buddhist pilgrim Wukong passed through the Uyghur country in 789. He left his Sanskrit books in Beiting for safety, because he knew that the khan was not a Buddhist.[43]

Manichaeism after 763

Bögü (Mouyu) stayed in the region of Luoyang from November 762 until about February 763.[44] His soldiers plundered the neighbourhood. He himself visited a group of Sogdian Manichaeans and took four of them home. The chief of them, called Ruixi in Chinese, is praised in the Chinese text of the Karabalghasun stele as learned in the Doctrine of Light and as eloquent as a cascade.[44] At Karabalghasun a court faction led by a senior official opposed the faith bitterly. Ruixi’s party, however, persuaded the khan to override it. He decreed that his subjects embrace the religion.[44] The sources do not say why he chose it. Four explanations are offered. One is that the faith joined sophistication to the discipline a warlike people demanded.[45] A second is that he wished to show his independence of the Tang. The Tang disliked the faith, had condemned it in 732 and allowed only non-Chinese to practise it.[45] A third is that he wished to ally himself with the wealth and the trade of the Sogdians.[45] A fourth is that he was sincerely moved.[45] In 768 the Tang ordered Manichaean temples to be built, probably in Chang’an and Luoyang. Three years later the Tang let the Uyghurs build one each at Jingzhou, Nanchang, Shaoxing and Yangzhou.[46] The khan may have wanted them to supervise the Uyghurs living in China.[46] A Turkish handbook adds that the Manichaean priests were Iranians and that a temple was built at Karabalghasun.[47] The Uyghurs are the first Turkic people known to have taken up, as a people, a religion with a moral code in place of shamanism.[48]

The faction hostile to the faith gained ground under Tun Bagha (Dun Mohe). A Latin record has a “Turkic king” ask the Nestorian patriarch Timothy I in the early 780s for an archbishop. That see, however, lay on the Syr Darya, a long way west of Karabalghasun. The king therefore cannot have been Tun Bagha (Dun Mohe).[49] There is better evidence for his bond with the old religion. In 765 the Uyghur generals consulted their magicians before a battle against the Tibetans and had them call up wind and snow. When the weather turned to the Uyghurs’ advantage, the magicians were given the credit.[1] After his death Manichaeism came back and stayed strong at court. The titles of every khan of the second line but the last contain the Moon God or the Sun God. These phrases do not occur in a khan’s title before 789. They are understood as symbols of the darkness and light of the faith.[50] In 807, after a break of more than thirty years, Uyghur envoys again asked the Tang for leave to build temples. From then on, Manichaeans came with the Uyghur embassies.[51]

The faith did not take hold everywhere. On one assessment it is unlikely that it spread beyond the khan, his retinue and a small educated group, since its asceticism did not suit a nomad life.[52] Tamim ibn Bahr says that Manichaeism was only one of two religions practised in the towns outside the capital. He says that it prevailed in Karabalghasun itself, although the edict that ordered every person to embrace it was then nearly sixty years old.[43] He adds that the king of the Toghuzghuz possessed pebbles with which rain, snow and cold could be brought down at will.[53] Uyghur Manichaean texts speak of the sacred forest of Ötüken with reverence.[53] In the west a dissident group existed. Sogdian letters to a Manichaean dignitary in Kocho complain of monks who washed in running water, felled trees themselves and dug in the garden.[54]

Steppe and city

The older history says of the Uyghurs before 744 that they had no fixed dwellings. They roamed in search of water and pasture and were excellent riders and archers.[1] Within a few years of the empire’s founding they had a capital. Work on Bay Balïq began in 757 on the khan’s order. Karabalghasun was built about the same time. Both were finished under Bayanchur (Moyanchuo).[55] Tamim ibn Bahr reports a royal castle and twelve iron gates of huge size. He also reports a populous town with markets and trades, villages lying close together and cultivated land around it. Outside the city, he is said to have found the khan’s army camp as a ring of troops with four gates, in which all the army’s beasts were kept.[56] Archaeology finds irrigation, millstones and pestles, the traces of smiths, potters, weavers and jewellers, mining of iron, copper, gold and silver, and grain buried with the dead. The tent stayed the commonest dwelling even in Karabalghasun, where ministers and commoners lived in tents. The golden tent that Tamim saw above the palace could hold a hundred people.[57] In this century the Uyghurs passed part of the way from the wholly nomadic life of their ancestors to one in which cities, commerce and agriculture counted. The passage was only partial. Stock-raising remained the base of the economy, and far from the capital the old ways went on unchanged.[58] The khan’s main camp lay on the Orkhon, about where the Mongol Karakorum later stood. To judge by the ruins, the city that grew up there was larger than the later Mongol one.[20] Other scholars hold that the Uyghurs already had real towns and were developing agriculture in the river valleys. It is not clear who worked the land.[59] The same account sees the beginnings of social stratification in Uyghur society as the work of outside influences and outside income.[59] A Song writer looked back on the early Uyghurs as simple and nearly equal. He says that the khan Tengri was the first to build palaces, and that the silks and finery paid by the Tang broke the old customs.[59]

The Sogdians are the other half of that story. The literature on Sogdian trade looks at the conversion from their side. The trilingual inscription of Karabalghasun relates the khan’s conversion to Manichaeism at Luoyang in 762 and 763. The Tang had just turned on the hu of northern China after the revolt of An Lushan. Manichaean Sogdians won the protection of their coreligionists and compatriots from the one military power the dynasty could not refuse. As a result, the Uyghurs became the protectors of the hu in China until 840, when the persecutions resumed. The same literature explains the khan’s agreement by what the Uyghurs needed. The Uyghurs had been masters of the steppe only since 744. They had to find a model on which to organise a new power. The Chinese model was too burdened with subservience. The only other was the Sogdian, anchored at the old Göktürk court and along the whole frontier. They took the religion, the script, much of the vocabulary and the urban pattern. The Sogdian influence on the Uyghurs was greater than it had been on the Göktürks.[60] The Shine-usu inscription records Bayanchur’s order to the Sogdians and the Chinese to build Bay Balïq on the Selenga.[61][62] Some scholars see in it three things that run through Uyghur history. They are close relations with China, close relations with the Sogdians, and an interest in towns.[61] In their view, the coup of 779 is the reverse of the fusion.[1] A party favourable to Göktürk tradition and hostile to foreign importations killed the khan, and with him the hu with nine names, two thousand people by the Jiu Tangshu’s count. It resembled the party at the court of the second Göktürk khaganate.[1] After that the Sogdians and their religion recovered and shared in the horse-for-silk trade to the end.[1]

The horse trade and the Sogdians

The Sogdians came with the Uyghurs into China. The Xin Tangshu records that whenever the Uyghurs came to the Chinese court they brought Sogdians with them. The Sogdians often stayed in the capital. In time there were nearly a thousand, and their property grew into great capital.[63] A caravan of Uyghur officials entered China in 780 with several thousand camels and horses, Uyghurs, Sogdians and their Chinese concubines, and more than a hundred thousand rolls of silk.[63] A Chinese official proposed to kill them. “The Uyghurs are not strong in themselves,” he said, “but they are helped by the Sogdians.”[63] Men of high rank and envoys of the khaganate bore Sogdian names. From 807 Manichaeans served as envoys and traded at Chang’an.[64] The Sogdians held in the Uyghur state the place they had held in the first Göktürk khaganate.[64]

The silk came through horses. The Tang bought horses from the Uyghurs at a high price: 38 to 40 pieces of silk a horse, against 25 before. They bought under the threat that the Uyghurs would change sides.[64] On average the Tang bought 7,500 horses a year for 300,000 rolls of raw silk.[65] The envoys took tea home as well. The Tang official Feng Yan writes that the Uyghurs drove fine horses to court and went home with tea.[66] The emperors tried to cut the purchases, most often in vain. The Chinese sources are full of complaint about the trade.[65] In the judgement of some historians, China became a huge hunting ground for the Uyghurs. It endured pillage and paid heavy contributions that the court called payments for services rendered, and it was freed from the tribute only after the khaganate had fallen.[59] At the accession of the khan in 821 an embassy of two thousand tribesmen brought twenty thousand horses and a thousand camels to the Tang court.[16][67] The Old Tang History puts the complaint in other terms. After the Qianyuan era (758 to 760) the Uyghurs came every year. They traded one horse for forty pieces of silk and brought horses by the tens of thousands.[1] The Chinese records of the trade are detailed for the 820s and break off in 830.[68] The Uyghurs also made money in China itself. After the rebellion of An Lushan some of them settled there, married, owned property and lent money at interest. The Tang tried with little success to check them.[69] Sogdian influence at the Uyghur court suffered a check between 780 and 790.[65] In the following decade the Uyghurs and the Tibetans fought for the oases of the northern Tarim basin, above all around Turfan and Kucha. The struggle seems to have ended in favour of the Uyghurs.[65] A peace of 822 or 823 between the Tibetans, the Uyghurs and the Tang ended the long fighting for the oases. Some scholars say it cost the Uyghurs direct control of much of that country.[70]

Reach to the west

In the first quarter of the ninth century the khaganate reached into western Central Asia. The Arab envoy Tamim ibn Bahr travelled to the Uyghur capital probably in 821. He found the whole steppe north of the Tianshan as far as Chach under Uyghur control, with relays and a postal service.[71] In the same year a Uyghur army stood in Ustrushana, a few days from Samarkand.[72] Al-Tabari calls it the army of the Toghuz-Oghuz. These may be the Oghuz of the Aral Sea. Others hold that they were the Uyghurs, who were then expanding fast.[71]

The Uyghurs had come west before. In 806 the great revolt of Rafi ibn Layth broke out at Samarkand and set the whole of Transoxiana ablaze. The Uyghurs were among the allies who sent Rafi troops.[71] The campaign to the Syr Darya that the Chinese text of the Karabalghasun inscription records belongs to the years of that revolt.[71] At the same time al-Ma’mun, in revolt against his brother, thought of taking refuge with the khan, whom al-Tabari calls king of the Turks. The Uyghurs were then subjugating the Karluks north of Transoxiana.[73]

List of khans

The standard histories give the following sequence. The titles in inverted commas are those conferred by the Chinese court. Where a Turkic form is given, it is the literature’s.

  • 744 — Kutlug Bilge Kül (Guli Peiluo), ‘Huairen’ Khaghan
  • 747 — Bayanchur (Moyanchuo), Gele Khaghan
  • 759 — Idijian, Bögü (Mouyu) Khaghan, Tengli Khaghan
  • 779 — Tun Bagha (Dun Mohe), Alp Kutluk Bilge Khaghan, ‘Tianqin’ Khaghan
  • 789 — Duoluosi, ‘Zhongzhen’ Khaghan
  • 790 — an unnamed usurper, younger brother of the last
  • 790 — Achuo, ‘Fengcheng’ Khaghan
  • 795 — Gudulu, that is Kutluk, ‘Huaixin’ Khaghan, first of the Xiedie line
  • 808 — ‘Baoyi’ Khaghan
  • 821 — ‘Chongde’ Khaghan
  • 824 — Prince Hesa
  • 832 — Prince Hu, ‘Zhangxin’ Khaghan
  • 839 — Prince Hesa

A caution attaches to every date of death in this list. Where the Chinese give a khan’s death to the month or the day, the day is usually that of the court mourning at Chang’an, not of the death on the Orkhon. The records of the mourning show this. The precise dates should therefore be understood as dates of Chinese events.[58]

The drought of 783 to 850

The weather of the empire’s last decades is known from measurements as well as from the written sources. A tree-ring record from the Orkhon country shows moderate to wet years when the empire arose. From 743 to 782 only fifteen years in forty were dry.[74] Then a drought set in. It lasted, with few breaks, sixty-eight years, from 783 to 850, and covered fifty-seven of the empire’s ninety-six years.[74] Its worst stretch began in 805 and lasted thirteen years. In the whole record of 1,773 years, only a drought of seventeen years, from 1005, was longer and harder.[74] Another of ten years began in 827, and one of five in 840.[74] A second tree-ring record from Khorgo, some 200 kilometres to the north-west, follows the same course. It ends the drought earlier, in 836. A lake record from the eastern Pamir puts one of the driest periods of two thousand years in the early 800s. A drought atlas of Asia shows drought from 783 to 844 over the land from western Tibet to northern China, the Silk Road with it.[75] A drought reconstruction over two thousand years marks a severe drought beginning in 804. The famine, pestilence and snow of 839 in the Xin Tangshu are the written end of the same sequence. Turkic Atlas sets the drought beside the Kirghiz under climate and the steppe empires.

The study that applies these measurements to history is careful with them. The Chinese and Uyghur sources record none of what such a drought is expected to bring. They record no sharp rise in the deaths of animals, no loss of people, no migration and no armed pressure on the neighbours.[76] The empire did not collapse. The capital did not move. Military activity against the neighbours did not rise.[77]

The economy may explain how the empire lived through such years. The Uyghurs gave the Tang horses and military service and received silk. Silk served as money along the Silk Road. The Sogdian merchant families at court controlled much of its trade.[78] The silk could have paid for the food and other supplies that the capital, the court and the army on the Orkhon drew from farming lands, and so softened the drought.[79] This explanation rests on circumstantial evidence. The parallel offered for it is the Mongol capital Karakorum, a short distance away, which was supplied from outside.[70]

The trade in horses shows the strain. It stopped between 795 and 805, for reasons the sources do not give.[80] The Mu Us desert, on the road to Chang’an, had been drying out since about 800. This probably made the road harder.[80] According to observers of the time, among them the poet Bai Juyi, the Uyghur horses left not a blade of grass on the road along the Yinshan. Sixty or seventy in a hundred died or were hurt on the way to China.[80] The trade was largest in the 820s. The Tang paid 50,000 pieces of silk for horses in February 822 and 70,000 more in March, 200,000 pieces in 827 and 230,000 in the first month of 829.[1][81] The last exchanges recorded are 5,750 horses in 829 and 10,000 in 830.[68] At forty pieces a horse, the 230,000 pieces of 829 match the 5,750 horses.[1] After 830, while the drought continued, no silk seems to have come from China.[82] The long decline of the grass may have cut the supply of horses for war and for trade at once. How far it did so is not known.[82]

Causes of the collapse

Al-Jāḥiẓ, who died in 868 or 869, held that the conversion to Manichaeism cost the Uyghurs their edge. The Toghuzghuz, he wrote, used to excel the Karluk even if these were twice as many. They began to suffer defeats after they took up the faith.[83] Some scholars find this reasonable, since the faith turned minds away from war. They add the strain between the city-dwellers, drawn into foreign ways, and the nomads, who held to the old ones. Two societies grew apart within one empire, and the court’s hold outside the capital weakened.[84] Others call the claim exaggerated, because the Uyghur khan still defended his co-religionists in China and under the Samanids by force.[85] The khans commanded a very large army, much of it under the tutuk of the tribes outside the royal house. An idle soldiery of uncertain loyalty was a danger once the court split into factions.[84] Pressure from outside grew as well. The Tibetans came close to attacking Karabalghasun in 816. War with the Kirghiz broke out about 820. It went on intermittently until a general of the khan invited them in.[86] The famine of 839 may have been too much for an economy that had been failing for years.[86] A third explanation puts the weather first. Rebellions and quarrels at court began about 832, while the drought was still unbroken.[82] Then came the winter of 839 to 840. The Xin Tangshu records famine, pestilence and heavy snow in 839, and the death of many sheep and horses.[16] That winter has been described as a dzud. A dzud is the frost and snow of the Mongolian steppe that empty the herds, and it is the harder the drier the years before it. It was then the last blow to a pastoral economy the drought had already worn down. The tree rings record only the summers and cannot confirm it.[87] This account also gives the settled habits of the Uyghur elite and their weak leadership as large a part in the splits at court and in the defeat as the loss of pasture.[88] It does not claim that the drought by itself brought the khaganate down.[89]

Assessment

The Uyghurs were the theoretical allies of the Tang through almost the whole of the century. Under Bayanchur (Moyanchuo) and still more under Bögü (Mouyu), they used the Chinese dependence on them without scruple. Under Tun Bagha (Dun Mohe) the tone from Karabalghasun grew friendlier. After a tense period under Baoyi, the relation warmed again from about 820. The Chinese attitude did not change. They used their neighbours when they needed them, extended diplomatic friendship, and feared and distrusted them as barbarians throughout. What outlasted the empire was cultural. It grew from the two things the Chinese sources ignore: the adoption of Manichaeism and the absorption of the Sogdians. These are the roots of the later Uyghur script and art. The people of that name who settled the Tarim after 840 belong to another story. The empire period ended with the Kirghiz.

© Emre-E Yavuz 2026 The Uyghur Khaganate, 744–848, in eight time slices: from the founding on the Orkhon to the fall of Ordu-Balık in 840 and the dispersal of the Uyghurs. Modern map on a base of Natural Earth, after the Tang annals and the inscriptions of Tariat and Karabalghasun. A soft band marks a border that no source describes. Uyghursunder the rule of Kutlug Bilge Kül KhanKirghiz TangTang (Anxi, Beiting)KarlukShiweiKhitanToquz Tatar Basmïl chief to Beiting, 744 Ötüken Beiting Chang'an Uyghursunder the rule of Bayanchur KhanKirghizBasmïl and Karluk TangTang (Anxi, Beiting)ShiweiKhitanToquz Tatar Against the Chik, 750To the Bolchu, 751Chang'an and Luoyang, 757Against the Kirghiz, 758 Bay Balık Tariat inscription Shine-Usu inscription Chang'an Luoyang Uyghursunder the rule of Bögü KhanKirghizTibetan EmpireKarluk TangTang (Anxi, Beiting)ShiweiKhitanAbbasid Caliphate Luoyang, 762Karluk to Suyab, 766 Ötüken Ordu-Balık Taiyuan Luoyang Chang'an Liangzhou Suyab Talas Uyghursunder the rule of Tun Bagha Tarkan and his successorsKirghizShatuo and KarlukTibetan EmpireKarluk TangTang (Anxi, Beiting)ShiweiKhitanAbbasid Caliphate Envoys through Uyghur land, 781Beiting, 790–791 Ordu-Balık Beiting Kocho (Xizhou) Hami (Yizhou) Dunhuang Chang'an Uyghursunder the rule of Huaixin Khan and Baoyi KhanKirghizTarim oasis statesKarlukTibetan Empire TangFerghanaAbbasid CaliphateShiweiKhitan Kucha relievedTo the Pearl River and FerghanaLiangzhou, 808 Ordu-Balık Beiting Kocho (Xizhou) Karashahr Kucha Suyab Liangzhou Dunhuang Uyghursunder the rule of Chongde Khan and his successorsKirghizTarim oasis statesKarlukTibetan Empire TangShatuoAbbasid CaliphateShiweiKhitan Ordu-Balık Beiting Kocho (Xizhou) Kucha Chang'an KirghizTarim oasis statesKarlukTibetan Empire TangShiweiKhitanShatuo The Kirghiz, 840To the KarlukTo AnxiTo the TibetansWujie to the south Ordu-Balık Beiting Kocho (Xizhou) Kucha Tiande Ganzhou Uyghursunder the rule of Wujie KhanUyghursunder the rule of Pang TeginKirghizKarlukTibetan Empire TangShiweiKhitanXiShatuo Wujie driven north-east, 843The Kirghiz, 848 Ordu-Balık Tiande Zhenwu Youzhou Ganzhou Kucha Beiting
  1. 744–745 The founding

    In 744 the Uyghurs, the Basmïl and the Karluk overthrew the last Göktürk khans; the Uyghurs then broke the Basmïl, whose chief fled to Beiting. Kutlug Bilge Kül was raised as khan and moved his court from the Selenga south to the Ötüken and the Orkhon. For 745 the Xin Tang Shu gives the realm its only measure: east to the Shiwei, west to the Altai, south across the Gobi. No source describes its northern and eastern edges.

  2. 747–758 Bayanchur Khan

    Bayanchur, khan from 747, names his own land in the Tariat inscription: the Ötüken, the Selenga, the Orkhon and the Tola. He fought the Tatars, the Chik on the Kem and, across the Altai on the Bolchu, the Karluk, who withdrew west to the lands of the Ten Arrows. In 757 his horsemen helped the Tang retake Chang'an and Luoyang from the rebels of An Lushan. He had Sogdians and Chinese build Bay Balık on the Selenga, and in 758 the Uyghurs defeated the Kirghiz, whose ruler then took a title from the Uyghur khan.

  3. 759–779 Bögü Khan

    In 762 Bögü Khan led five thousand horsemen south by way of Taiyuan, and his brother defeated the rebel Shi Chaoyi at Mount Beimang by Luoyang. From Luoyang the khan brought Manichaean teachers into his land. Meanwhile the Tibetans took the Hexi corridor from the Tang, which left the Tang garrisons of Beiting and Anxi cut off, and in 766 the Karluk took Suyab and Talas. In 779 Tun Bagha Tarkan killed Bögü and seized the throne.

  4. 780–792 The war over Beiting

    From 781 the envoys of the cut-off Tang garrisons in Beiting and Anxi travelled to Chang'an through Uyghur land. By then the Tibetans also held Dunhuang and Hami. In 790 they took Beiting; the Shatuo and the Karluk around the city, normally Uyghur subjects, went over to them, and the Uyghur army sent to relieve it was beaten. In 791 the Uyghurs defeated the Tibetans and the Karluk at Beiting and sent the captives to the Tang court. The Chinese histories date the fall of Beiting to 787, 790 or 791.

  5. 795–821 The height of the realm

    Under the Ädiz khans from 795 the Uyghur realm reached its greatest extent. The Karabalghasun inscription tells how the khan retook Beiting, relieved Kucha from a Tibetan siege and destroyed the Tibetans at Yushu west of Karashahr, and how Uyghur armies pursued their enemies to the Pearl River and west to Ferghana. The khan set a yabgu over the Karluk. A Manichaean hymn-book names Beiting, Kocho, Kucha and Karashahr among the Uyghur dependencies, and in 808 the Uyghurs took Liangzhou from the Tibetans.

  6. 821–839 Twenty years of war with the Kirghiz

    In 821 the Tang princess Taihe was married to the Uyghur khan. As Uyghur power weakened, the ruler of the Kirghiz called himself khan; the Uyghurs sent a minister against him without success, and the Xin Tang Shu counts twenty years of war. The Kirghiz kept ties with the Karluk, the Tibetans and the Arabs. In 839 a minister rose against the khan with the help of the Shatuo, and the khan died; famine, pestilence and heavy snow killed the herds.

  7. 840 The fall of Ordu-Balık

    In 840 a Uyghur general went over to the Kirghiz and led a hundred thousand horsemen against the Uyghur court. They killed the khan, burned Ordu-Balık and the golden tent, and took Princess Taihe. The Uyghurs scattered: fifteen tribes under the minister Sazhi and Pang Tegin went west to the Karluk, one branch to the Tibetans and one to Anxi, and thirteen tribes under Wujie Tegin crossed the Gobi south to the Tang frontier. The Kirghiz moved their own court south of the Lao mountain, fifteen days' ride from the old Uyghur court; the place is not identified.

  8. 841–848 After the fall

    Wujie Khan camped on the Tang frontier between Tiande and Zhenwu and raided the prefectures of Yun and Shuo. In 843 a Tang army surprised his camp near Zhenwu, freed Princess Taihe and drove him north-east, where he was killed. In the west Pang Tegin called himself khan and held the cities west of the desert; the Xin Tang Shu places him in Ganzhou, the Jiu Tang Shu in Anxi. In 848 a Kirghiz army of seventy thousand came south to the Tiande frontier and carried the last Uyghurs of the east back north of the Gobi.

  • Place named in a source (position as commonly identified)
  • Position disputed or approximate
  • Zone drawn after the Tang annals and the inscriptions
  • People under Uyghur overlordship
  • Oasis states and peoples subject to Uyghur levies
  • Border that no source describes (drawn uncertain)
  • Movement named in a source (route drawn)
The Uyghur Khaganate, 744–848, in eight time slices: from the founding on the Orkhon to the fall of Ordu-Balık in 840 and the dispersal of the Uyghurs. Modern map on a base of Natural Earth, after the Tang annals and the inscriptions of Tariat and Karabalghasun. A soft band marks a border that no source describes.

Rulers

  1. Kutlug Bilge Kül Khan744–747
  2. Bayanchur Khan747–759
  3. Bögü Khan759–779
  4. Tun Bagha Tarkan779–789

Relations with other states

Red is war, blue alliance, gold trade. Where several held at once, the line shows what prevailed; alliance and trade can run side by side. Each line rests on an article. Point at a state to see its weightiest relation, select it to see its events. The slider shows what stood in a given year, and Play runs through the years.

  1. Second Göktürk Khaganate744

  2. Tang780 to 840

  3. Yenisei Kirghiz Khaganate820 to 840

References

  1. ^ a b c d e f g h i j k l m n Jiu Tangshu (Old Book of Tang), juan 195
  2. ^ Rásonyi 1993, p. 45
  3. ^ Pelliot 1930, p. 351
  4. ^ Sümer 1972, p. 18–19
  5. ^ İzgi 1989, p. 14–16
  6. ^ İzgi 1989, p. 16–17
  7. ^ a b Sinor 1990, p. 317
  8. ^ a b c Jiu Tangshu (Old Book of Tang), juan 195; Xin Tangshu (New Book of Tang), juan 217
  9. ^ Kwanten 1979, p. 51
  10. ^ a b c Kwanten 1979, p. 52
  11. ^ Sinor 1990, p. 318
  12. ^ Sinor 1990, p. 318–319
  13. ^ a b c Sinor 1990, p. 319
  14. ^ a b c Sinor 1990, p. 322
  15. ^ İzgi 1989, p. 25
  16. ^ a b c d e f g h i Xin Tangshu (New Book of Tang), juan 217
  17. ^ İzgi 1989, p. 25–26
  18. ^ Sümer 1972, p. 22
  19. ^ Sümer 1972, p. 22–23
  20. ^ a b Barthold 2004, p. 45
  21. ^ a b c d e Sinor 1990, p. 320
  22. ^ a b c Sinor 1990, p. 321
  23. ^ a b Khazanov 1994, p. 257
  24. ^ Kwanten 1979, p. 309
  25. ^ Sinor 1990, p. 403
  26. ^ Sümer 1972, p. 21–22
  27. ^ Tekin 2003, p. 227
  28. ^ Sümer 1972, p. 6–7, 22
  29. ^ Barthold 2004, p. 46
  30. ^ Sinor 1990, p. 323
  31. ^ Sinor 1990, p. 323,328
  32. ^ a b c Sinor 1990, p. 324
  33. ^ Sinor 1990, p. 327
  34. ^ Sinor 1990, p. 325
  35. ^ Sinor 1990, p. 325–326
  36. ^ Sinor 1990, p. 326–327
  37. ^ Jagchid 1989, p. 158–159
  38. ^ a b c Jagchid 1989, p. 160
  39. ^ Sinor 1990, p. 326; 35; Ahmetbeyoğlu 2020, p. 326; 35
  40. ^ a b Sinor 1990, p. 326
  41. ^ von Le Coq 1913, pl. 5
  42. ^ von Le Coq 1913, p. 9
  43. ^ a b c d Sinor 1990, p. 328
  44. ^ a b c Sinor 1990, p. 330
  45. ^ a b c d Sinor 1990, p. 330–332
  46. ^ a b Sinor 1990, p. 332
  47. ^ Ögel 1984, p. 349–350
  48. ^ Barthold 2004, p. 47–48
  49. ^ a b Sinor 1990, p. 333
  50. ^ Sinor 1990, p. 333; 85; 64; Wilkens 2024, p. 333; 85; 64; Şen 2018, p. 333; 85; 64
  51. ^ Sinor 1990, p. 325,333
  52. ^ Ögel 1984, p. 350
  53. ^ a b Sinor 1990, p. 335
  54. ^ Sinor 1990, p. 334
  55. ^ Sinor 1990, p. 337
  56. ^ Ögel 2000, p. 342
  57. ^ Sinor 1990, p. 337,339
  58. ^ a b Mackerras 1972
  59. ^ a b c d Khazanov 1994, p. 258
  60. ^ Özertural 2013, p. 94
  61. ^ a b Golden 1992
  62. ^ Ögel 2000, p. 185
  63. ^ a b c Xin Tangshu (New Book of Tang), juan 217; de la Vaissière 2005a, p. 308
  64. ^ a b c de la Vaissière 2005a, p. 308
  65. ^ a b c d de la Vaissière 2005a, p. 309
  66. ^ Fengshi Wenjian Ji, juan 6
  67. ^ İzgi 1989, p. 35–36
  68. ^ a b Cosmo 2018, p. 457
  69. ^ Sinor 1990, p. 338–339
  70. ^ a b Cosmo 2018, p. 456
  71. ^ a b c d de la Vaissière 2005a, p. 311
  72. ^ Tabari; de la Vaissière 2005a, p. 311
  73. ^ de la Vaissière 2005a, p. 312
  74. ^ a b c d Cosmo 2018, p. 448
  75. ^ Cosmo 2018, p. 450–451
  76. ^ Cosmo 2018, p. 443
  77. ^ Cosmo 2018, p. 451
  78. ^ Cosmo 2018, p. 455,459–460
  79. ^ Cosmo 2018, p. 455,460
  80. ^ a b c Cosmo 2018, p. 454
  81. ^ İzgi 1989, p. 36
  82. ^ a b c Cosmo 2018, p. 460
  83. ^ Sinor 1990, p. 340–341
  84. ^ a b Sinor 1990, p. 341
  85. ^ Barthold 2004, p. 50–52
  86. ^ a b Sinor 1990, p. 342
  87. ^ Cosmo 2018, p. 457,461–462
  88. ^ Cosmo 2018, p. 461
  89. ^ Cosmo 2018, p. 443,462
  90. ^ Sinor 1990, p. 330–331

Sources

Literature

  • Colin Mackerras, The Uighur Empire According to the T'ang Dynastic Histories (1972)
  • Emre-E. Yavuz, The Gokturk Empire: 2nd Edition (2025)
  • Étienne de la Vaissière, Sogdian Traders: A History (2005)
  • Peter B. Golden, An Introduction to the History of the Turkic Peoples (1992)
  • Nicola Di Cosmo; Amy Hessl; Caroline Leland; Oyunsanaa Byambasuren; Hanqin Tian; Baatarbileg Nachin; Neil Pederson; Laia Andreu-Hayles; Edward R. Cook, Environmental Stress and Steppe Nomads: Rethinking the History of the Uyghur Empire (744–840) with Paleoclimate Data (2018)
  • Amy E. Hessl; Kevin J. Anchukaitis; Casey Jelsee; Benjamin Cook; Oyunsanaa Byambasuren; Caroline Leland; Baatarbileg Nachin; Neil Pederson; Hanqin Tian; Laia Andreu Hayles, Past and future drought in Mongolia (2018)
  • Julian Struck; Marcel Bliedtner; Paul Strobel; William Taylor et al., Central Mongolian lake sediments reveal new insights on climate change and equestrian empires in the Eastern Steppes (2022)
  • Denis Sinor (ed.), The Cambridge History of Early Inner Asia (1990)
  • Zekine Özertural, Mitteliranische Einflüsse auf die manichäisch-türkische Wortbildung (2013)
  • Jens Wilkens (trans. Celalettin Bulut Bostan), Uygur Budizmi ve Manihaizm ile Yerel Dinlerin Etkisi: Dinî Terminoloji Örneği (2024)
  • Ali Ahmetbeyoğlu, Hüseyin Şen, Eski Türk Dininin Kozmolojisi ve Kozmogonisi (2018)
  • Ali Ahmetbeyoğlu, Eski Türklerde Kut ve Töre Bağlamında Hükümranlığın Hududları (2020)
  • Albert von Le Coq, Chotscho. Facsimile-Wiedergaben der wichtigeren Funde der ersten Königlich Preussischen Expedition nach Turfan in Ost-Turkistan (1913)
  • Paul Pelliot, Sur la légende d'Uɣuz-khan en écriture ouigoure (1930)
  • Luc Kwanten, Imperial Nomads. A History of Central Asia, 500–1500 (1979)
  • Özkan İzgi, Çin Elçisi Wang Yen-te'nin Uygur Seyahatnamesi (1989)
  • Faruk Sümer, Oğuzlar (Türkmenler). Tarihleri, boy teşkilâtı, destanları (1972)
  • V. V. Barthold, Orta Asya Türk Tarihi Dersleri (Orta Asya Türk Tarihi Hakkında Dersler) (2004)
  • Anatoly M. Khazanov, Nomads and the Outside World (1994)
  • Sechin Jagchid; Van Jay Symons, Peace, War, and Trade Along the Great Wall. Nomadic-Chinese Interaction through Two Millennia (1989)
  • Bahaeddin Ögel, İslâmiyetten Önce Türk Kültür Tarihi. Orta Asya Kaynak ve Buluntularına Göre (1984)
  • Bahaeddin Ögel, Türk Kültür Tarihine Giriş VII: Türklerde ordu, ordugah ve otağ (2000)
  • Bahaeddin Ögel, Türk Kültür Tarihine Giriş I: Türklerde köy ve şehir hayatı (2000)
  • László Rásonyi, Tarihte Türklük (1993)
  • Talât Tekin, Orhon Türkçesi Grameri (2003)

Cite this article

Yavuz, Emre-E. “Uyghur Khaganate.” Turkic Atlas, revised 8 November 2026. https://turkicatlas.org/polities/uyghur-khaganate/

Last revised: 8 November 2026