The Silk Road: A New History

by Valerie Hansen

  Like markets all over China today, the Turfan market offered a wide variety of flours and grains, in addition to vegetables like onions and scallions. Other daily-use items, like cauldrons and pots, as well as livestock, including horses, camels, and cattle, were all for sale. One could even buy a cartload of human excrement used as fertilizer for 25/22/20 coins.

  The market also offered a variety of goods imported from the Iranian world. Many of these overlap with those in the scale-fee records: ammonium chloride, aromatics, sugar, and brass. The market register lists more than seventy different kinds of medicine. Many of the imported goods are small and light, because they had to be carried overland, but there are some heavier items, including brass-inlaid high-quality iron swords for 2,500/2,000/1,800 coins, which were for sale alongside much cheaper, locally made knives, which went for 90/80/70 coins. The largest goods from the west were animals: gelded Turkish steeds and Persian camels, which could have been walked to Turfan and would have found ready customers among the officers in the Tang armies stationed there. The horses sold for 20/18/16 rolls of silk; the camels for 33/30/27 rolls of silk.79 The different textile stalls offered specialized silks made in Sichuan, Henan, and other interior Chinese provinces, which were precisely the tax silks paid to soldiers.

  The market register portrays a market supplied by small-scale traders traveling in small caravans of ten to twenty animals, the same level of commerce documented by the scale-fee documents and the guosuo travel passes. The major player in the Central Asian economy—which diverges from the prevailing image of the Silk Road trade—was the Tang government. Starting in the 630s with the campaigns against the Western Turks, the Tang administration poured funds into the Western Regions to support its military efforts. To finance their campaigns, the Tang state collected bolts of cloth in central China and then shipped them to Wuwei and Qinzhou (modern Qin’an), also in Gansu Province, and from there to places farther west, closer to the frontier.80 Over twenty examples of this type of tax cloth originating from central China have been found in Xinjiang.81

  Immediately after the 640 conquest, the Tang forces in Turfan probably numbered several thousand. Although we speak of Tang armies, many of the soldiers were not Chinese but local.82 The losses of territory in the northwest, including Kucha, to the Tibetans from 670 to 692 resulted in ever-increasing military expenditures in the eighth century. Du You (735–812), the author of the first comprehensive institutional encyclopedia, put the costs of defending the frontier at two million strings of coins in 713, ten million strings in 741, and fourteen to fifteen million strings in 755. Tang officials combined strings of coin, piculs of grain, and bolts of cloth to create an aggregate accounting unit whose value eludes anyone who has tried to make sense of the internally contradictory figures that survive.83

  However one understands these figures, the outlays by the Tang state are staggering. Even individual payments dwarf all the transactions recorded in the Turfan documents. In the 730s or 740s, the central government sent 900,000 bolts of silk each year to four military headquarters in the frontier regions of the Western Regions: Hami, Turfan, Beiting, and Kucha. By 742 some five thousand Tang soldiers were stationed in Turfan, yet the tax receipts from local inhabitants covered only 9 percent of their expenses.84 The Tang state’s subsidy for the military injected vast sums of money, in the form of silk, into the local economies of the Silk Road oases.

  These massive expenditures by the Tang central government came to a sudden halt with the An Lushan rebellion. The rebellion forced the Tang dynasty to withdraw from Central Asia and nearly brought the dynasty down. Born to a Sogdian father and a Turkish mother, the leader of the rebellion, An Rokhshan, worked his way up from buying horses for the Chinese army to being the general in charge of three different originally separate military commands.85 (An’s given name, Lushan, was the Chinese transcription of the Sogdian “Rokhshan.”) Fearful that his own troops might join the rebels, the emperor Xuanzong (reigned 712–56) accepted their demand that he strangle his consort Yang Guifei, rumored to be romantically involved with An Rokhshan, and then abdicated his throne to his son Suzong, who reigned from 756 to 762. With the largest provinces in central China under the rebels’ control, the tax receipts of the Tang state plummeted after 755, forcing the Tang military to cease payments to the armies in the northwest.86 The Tang emperor had no choice but to hire Uighur mercenaries to fight the rebels. Only in 763 did the much-weakened dynasty succeed in putting down the rebellion.

  During the Tang campaign to regain control from the mutineers, Uighur mercenaries occupied Luoyang in 762. There, in a fateful encounter with far-reaching implications for Turfan, which came under Uighur rule fifty years later, the leader of the Uighurs encountered a Sogdian teacher who introduced him to the basic teachings of Manichaeism.87 Manichaeism, a religion founded in Iran by the prophet Mani (ca. 210–76), held that the forces of light and darkness were engaged in a perpetual battle for control of the universe. The Uighur kaghan adopted Manichaeism as the official religion of his people and recorded his decision in a trilingual inscription (in Sogdian, Uighur, and Chinese) on a stone tablet.88 This was the first—and the only—time in world history that any state named Manichaeism its official religion.

  The Tibetan Empire seized on this moment of Tang vulnerability during the rebellion to expand its power. During the 780s Tibetan armies moved into Gansu, conquered the Beiting (Beshbalyq) Protectorate immediately to the north of Turfan in 786, and in 792 took Turfan as well. In 803 the Uighurs wrested control of Turfan from the Tibetans. The Uighurs in Mongolia were then defeated by the Kirghiz in 840, and some of these Uighurs withdrew to Turfan. There, between 866 and 872, they established a new state called the Uighur Kaghanate, with its capital at Gaochang City.89 A second Uighur kaghanate was based to the east in Ganzhou (Zhangye, Gansu).

  Under the Uighurs, the local people of Turfan continued to record their purchases and sales of land, slaves, and animals in contracts, but they used Uighur, not Chinese, as their written language.90 The thirteenth- and fourteenth-century contracts from Turfan in Uighur show that the economy returned largely to barter, with people exchanging animals and land for fixed measures of grain or cloth, often cotton, which replaced silk as a currency.

  The documents in the Uighur language reveal much about the religious life of the community. Under the Tang dynasty, the residents of Turfan had worshipped Buddhist, Daoist, and Zoroastrian deities as well as local ones. Under the Uighurs, devotees of two new religions worshipped in Turfan as well: Christianity and Manichaeism.

  Evidence of Christianity was discovered in the early twentieth century by the second German expedition to the region. Outside the eastern walls of Gaochang City, archeologists found a small Christian church, from which they salvaged one mural showing Palm Sunday worship. At Bulayik, a site to the north of Turfan, they excavated Christian manuscripts in Syriac, Sogdian, Middle Persian, modern Persian, and Uighur. One manuscript even gave a line in Greek before translating a psalm into Sogdian. Syriac was the primary language of worship, but some psalters and hymn collections have Sogdian headings in them. These Sogdian headings indicate that the Christians of Bulayik were mainly Sogdian speakers, though the presence of Turkish names and linguistic features in the Sogdian texts suggests that they were gradually giving up Sogdian in favor of Uighur. The dating of these manuscripts is uncertain; most likely they date to the ninth and tenth centuries, when Turfan was the capital of the Uighur Kaghanate.91

  MANICHAEAN WALL PAINTING FROM BEZEKLIK

  A tree of life heavily laden with fruit, with three intertwined trunks, dominates this large painting from cave 38 at Bezeklik, which stands 5 feet (1.5 m) high, 8 feet (2.4 m) wide, making it one of the world’s largest surviving Manichaean artworks. The Uighur-language prayers at the base of the tree give the name of the donor, who requests the protection of guardian deities. The female donor wears an unusual bird headdress and kneels at the right of the tree; two guardian deities stand behind her, a
nd three others kneel next to her. The opposite side of the painting shows her husband, partially effaced, wearing a similar headdress. This copy was made in 1931, when the mural was already severely damaged.

  Like most Christians in Central Asia, the Christians at Turfan belonged to the Church of the East, which was based in Mesopotamia, and the liturgical language was Syriac, a dialect of Aramaic. The teachings of the Church of the East held that Christ had two natures—divine and human—and, furthermore, that Mary was the mother of the human Jesus but not the divine Christ. Their opponents sometimes called them Nestorians in an effort to associate them with Nestorius (ca. 381–ca. 451), a Syrian patriarch in Constantinople from 428 to 431, who had been expelled from the church, but members of the church do not refer to themselves as such.92

  After the kaghan’s conversion, Manichaeism was the official religion of the Uighur Kaghanate. One charter, 125 lines long, specifies how a Manichaean monastery should be run, and most likely dates to the ninth century. It is not clear whether the Uighur government of Turfan or the monastery’s own leaders issued the document, which charges different monastic officials with supervising fields, vineyards, and the monastery storehouse. Some of the titles, like “elect,” are unique to Manichaeism, but the monastery’s structure closely resembled that of Buddhist monasteries. Dependent workers tilled the fields and supplied the monastery’s residents with grain and clothing. The clergy conducted feasts and were responsible for the spiritual lives of the congregation, whose main obligation was to supply them with vegetarian food that they would eat and so increase the amount of light in their bodies.93

  Albert von Le Coq, the German excavator who was so active at Kucha, found some of the most interesting documents about Manichaeism in two buried monastic libraries dating to the period of Uighur rule. Texts of many Manichaean hymns survive: some in the liturgical language of Parthian, which Mani spoke, some in Uighur, the local language of Turfan by the year 1000. These hymns often celebrate the victory of the forces of light over the forces of darkness:

  All beings of Light, the righteous [elect] and the auditors, who have endured much suffering, will rejoice with the Father.…

  For they have fought together with Him, and they have overcome and vanquished that Dark One who had boasted in vain.94

  Hymns like this have permitted scholars to reconstruct the major tenets of Manichaeism, which would otherwise be unknown, since so few Manichaean texts exist anywhere in the world.

  Some of the texts Le Coq found were beautifully illustrated but so severely damaged by water that all the pages stuck together and could not be separated. One such fragment survives in the Museum of Indian Art in Berlin, which holds all the material brought back by the four German expeditions that survived the bombing during World War II. This miniature depicts the Bema festival, the most important holiday of the Manichaean year, in which the clergy, or elect, sang hymns, read aloud Mani’s teachings, and ate a meal, shown in color plate 11A.95

  Although Manichaeism was the official state religion of the Uighur Kaghanate, little Manichaean art survives onsite at Turfan. Only one cave painting at Bezeklik, all scholars concur, is definitely Manichaean.96 The mural has suffered great damage since 1931 when the copy shown on page 109 was made, and those managing the site rarely show it to visitors.

  Why does so little Manichaean art survive at Turfan and the surrounding cave sites? Sometime around the year 1000 the rulers of the Uighur Kaghanate chose to patronize Buddhism and not Manichaeism.97 Several surviving caves in Turfan, including cave 38 at Bezeklik, bear witness to this shift: close examination of the cave walls shows that the caves had two layers, often a Manichaean layer (not always visible) lies beneath a Buddhist layer. The Uighur court’s decision to support Buddhism apparently ushered in a new era in which only one religion was tolerated.

  In 1209 the Mongols defeated the Uighur Kaghanate of Turfan but left the Uighur kings in place. In 1275 the Uighurs sided with Khubilai Khan. When defeated by one of his rivals, the Uighur royal family fled and settled in Gansu in 1283. Although peasant rebels overthrew the Mongol rulers of China and established the Ming dynasty in the fourteenth century, Turfan remained outside of the borders of China and under the rule of the united Mongols and, later, the Chaghatai branch of the Mongols. In 1383 Xidir (Xizir) Khoja (reigned 1389–99), himself a Muslim, conquered Turfan and forced the inhabitants to convert to Islam, the prevailing religion of the region today.98 The region remained independent of China until 1756, when the Qing dynasty armies invaded.99

  The history of Turfan falls into three distinct periods: before the Tang conquest of 640, Tang rule (640–755), and after 803, when the Uighur Kaghanate was based in the oasis. In the periods both before and after Chinese rule, the economy was largely self-sufficient. Most of the documented movement along the overland routes was either by envoys or refugees. The high point of the Silk Road trade coincided with (because it was caused by) the presence of the Chinese troops. The Tang government injected vast amounts of both cloth and coin into the local economy, which resulted in high interest rates on loans even for poor farmers. But when the Chinese forces withdrew after 755 the local economy reverted to a subsistence basis. As coming chapters will show, much information about the spending patterns of the Tang government survives in other oases (particularly Dunhuang), but the overall pattern is clear. The Silk Road trade was largely the byproduct of Chinese government spending—not long-distance commerce conducted by private merchants, as is so often thought.

  A LETTER TO SAMARKAND

  One of eight folded pieces of paper from an abandoned mailbag, this letter was written on a sheet of paper and then folded into a small silk bag and labeled “Bound for Samarkand.” These letters, dating to 313 or 314, are among the most important surviving documents about the Silk Road trade, because they are written by private individuals, including businessmen, and not by government officials. Courtesy of the Board of the British Library.

  CHAPTER 4

  Homeland of the Sogdians,

  the Silk Road Traders

  Samarkand and Sogdiana

  In 630, when the Chinese monk Xuanzang left Turfan, he took the most traveled route west. After stopping in Kucha, he crossed the Tianshan Mountains, visited the kaghan of the Western Turks on the northwestern edge of Lake Issykkul in what is now Kyrgyzstan, and then proceeded to Samarkand in modern Uzbekistan. From Samarkand, travelers could go west to Syria, return to the oasis states of the Taklamakan, or proceed south to India, as Xuanzang did. Samarkand was then the major city of the Sogdians, the Iranian people who played such an important role in the Silk Road trade and who formed China’s largest and most influential immigrant community during the Tang dynasty.1 The Sogdians spoke a Middle Iranian language called Sogdian, a descendant of which is still spoken in the remote Yaghnob Valley in nearby Tajikistan (see the document on the facing page in Sogdian).

  In Samarkand, Xuanzang entered the cultural sphere of the Iranians, whose languages, religious practices, and customs, although equally old and equally sophisticated, differed profoundly from those of the Chinese. The modern traveler following in Xuanzang’s steps crosses a different kind of border, just as distinct, between China and the former Soviet Union. Jokingly referred to as the “Steel Road” by Chinese, this treacherous highway is strewn with overturned trucks and metal debris from dismantled Soviet-era factories being carted off to China.

  The seventh-century route posed real dangers. After waiting two months for the snow to melt, Xuanzang left Kucha and headed toward the Tianshan Mountains. Provided with camels, horses, and guards by the Kucha king, Xuanzang traveled only two days before encountering more than two thousand Turk (Tujue) robbers on horseback. They did not rob Xuanzang, his disciple and biographer Huili explains, since they were too busy dividing previously looted goods among themselves.

  The travelers then reached the towering Tianshan Mountains, where Mount Ling made a profound impression on Xuanzang:

  The mountain is dan
gerous and precipitous, so steep that it rises up to the sky. Since the road was first opened, the icy snow accumulates in places where it remains frozen, not melting in either the spring or summer. Huge expanses of ice meet and join the clouds. The snow is so blindingly white that one cannot see where it ends and the clouds begin. The icy pinnacles fall and lie across the road. Some are a hundred feet high, some several cubits [zhang, roughly 10 feet, or 3 m] across.

  The trip was extremely arduous, Huili continues:

  It is difficult to proceed on the rough, narrow paths. Add the snowy wind flying in different directions, and even fur-lined garments and boots cannot prevent a battle with the chill. Whenever one wants to sleep or eat, there is never a dry place to stop, so the only thing to do is to hang up a cauldron and cook, and to lay one’s mat out on the ice and sleep.

  After seven days the survivors in Xuanzang’s group finally left the mountains. Three or four out of every ten people in the group had died from hunger or cold, and the losses among the horses and cattle were even greater.2

  These fatalities were unusually high, prompting some to wonder whether Xuanzang and his companions were caught in an avalanche unmentioned by his biographer.3 Due to the extraordinarily dry climate, ice formed only at the mountaintops of the Tianshan Mountains, far above the timberline, with the result that a band of dirt and sand lay immediately below the ice. When chunks of ice broke off, they hurtled down dirt, not ice, creating truly terrifying avalanches. Avalanche or not, this was certainly the most perilous crossing of Xuanzang’s entire trip to India.4