TABLE 6.1 THE RULERS OF INDEPENDENT DUNHUANG, 851–1002
Like the Tibetan rulers before him, Zhang Huaishen and his successors sponsored the construction of many new caves. Building a cave was an intensely religious act: when one ruler decided to dig a cave, he and his wife ate vegetarian food for an entire month and then lit lamps, burned incense, and paid monks to pray and copy sutras, all in the hope of generating Buddhist merit. Only then did they begin construction of the cave.76
Some of the caves at Dunhuang contain portraits of Zhang Yichao and the rulers who succeeded him: around 925 Cao Yijin, who had taken over from the Zhang family in 914, commissioned a series of portraits of his predecessors in cave 98. On viewing these portraits, any visitor might imagine, as the Cao-family donors surely desired, that the order of succession had been smooth. In fact it was anything but. Zhang Yichao’s nephew Zhang Huaishen succeeded him on his death in 867 and ruled until 890, when his cousin, a son of Zhang Yichao, killed him, his wife, and six children. The new ruler, Zhang Huaiding, ruled for a year before dying a natural death; the next ruler came to the throne as a minor and was overthrown by his guardian Suo Xun. In 894 the former ruler regained power and stayed on the throne until 910. The last years of Zhang-family rule coincided with the closing years of the Tang dynasty, a time of great political uncertainty, when the emperors were first held prisoner and then overthrown in 907.77
In 915 Cao Yijin, the son-in-law of the last Zhang-family ruler, took the throne, and his family ruled until 1002. After that year records do not refer to any Cao by name, suggesting that the Uighur Kaghanate based in Ganzhou (now Zhangye, Gansu Province), had gained control of Dunhuang. In the eighth century the Uighurs had originally been united, but broke apart in 840 following the Kirghiz destruction of the unified Uighur Kaghanate, which prompted Uighurs to flee to both Turfan (the Western Uighur Kaghanate included Beiting, Gaochang, Yanqi, and Kucha) and Ganzhou, the site of a second, smaller kaghanate.78 In 1028 the Uighur Kaghanate at Ganzhou fell to the Tanguts, and in the 1030s Dunhuang followed, both becoming part of the Xi Xia realm, which included part of northwest China. We know little about the power struggles after 1000, simply because no materials from cave 17 or any other excavated documents describe these events in detail.
Between 848 and 1002, just as during the preceding period of Tibetan rule, the travelers who appear most often in the documentary record are envoys and monks. The Zhang and Cao families maintained diplomatic relations with all of their neighbors; they sent and received gift-bearing missions from the Tang capital at Chang’an and from other closer rulers as well, most notably the rulers of Khotan and the Uighur Kaghanates.79 Although many documents record the coming and going of emissaries, few detail what gifts they presented and what they received in return. For that reason one inventory of gifts given and received by a delegation that traveled to Chang’an in 877 is particularly important.
In 877 Zhang Huaishen, Zhang Yichao’s nephew, had been ruling Dunhuang for ten years, but the Tang emperor had not yet recognized him as a legitimate successor. He sent the delegation with a request for a formal banner that would acknowledge him as the military governor of Dunhuang, the title that his uncle had held before him. The 877 delegation presented one ball of jade (weight not specified), one yak tail, one antelope horn (presumably for medicinal use), and one letter to the Tang emperor.80
Hosting the delegation for nearly four months (they arrived on the twenty-seventh day of the twelfth month and left on the eleventh day of the fourth month), the Chinese divided the group into three levels (the top three officials, thirteen lower-level officials, and thirteen carriers), and gave different gifts to each group. For example, the three men in the highest rank received fifteen bolts of cloth (the documents do not specify what kind), one silver bowl, and one suit of brocade. Those in the second and third ranks received correspondingly less: the thirteen men in the second tier were given ten bolts of cloth (not fifteen), a silver cup (not a bowl), and a suit, while the thirteen men in the bottom tier received eight bolts of cloth, a suit, and no silver. Combining these with the gifts from other branches of the government, the group collectively was given a total of 561 bolts of cloth, five silver bowls, fourteen silver cups, and fifty suits. In addition, each member of the group received forty-three bolts of cloth to cover travel expenses—literally, “the cost of camels and horses,” a total of 1,247 bolts, more than twice the amount of cloth that the delegation received. Once they had gathered all the gifts together, they drew up an inventory, placed everything in leather bags with wooden tags, and then stamped the tags, sealing them shut. The gifts were opened only on arrival in Dunhuang. In 878 the delegation returned without the banner; only in 888 did the Tang court grant the desired honor.81
Although the Tang emperor did not grant the desired banner, he bore all the expense of hosting the delegation and the cost of the many gifts it bestowed on the members of the delegation. For the entire history of the Silk Road—going back to Sogdian delegations recorded at Xuanquan–members of tribute delegations had, in addition to the presentation of the formal gifts that was their duty, engaged in private trade on the side. We do not know how much the individual members of the trade delegation benefited from any dealings they were able to conduct—they did not record such transactions—but the amount of silk clothing given to each constituted a major gift.
Many envoys traveled to Dunhuang in the period of Cao-family rule, we learn from accounts detailing the beer and provisions provided to them.82 One register, probably from 964, records the beer consumption of fifty different visiting diplomats in just under seven lunar months: one from the Song dynasty, fourteen from Tibet, eleven from Khotan, one from the Turfan Uighur Kaghanate, seven from the Yizhou Uighur Kaghanate, and seventeen from the Ganzhou Uighur Kaghanate.83 Most of the visitors stayed a few days, but one group stayed 203 days, which must have been a burden for their hosts, who provided flour for the morning and evening meals and a round, flat wheat cake at lunch.
As these registers show, Dunhuang officials hosted people of all social ranks during this period: the crown prince of Khotan, envoys, monks, workmen, scribes, artists, and even a single “walking non-Chinese,” a term that most likely denoted an itinerant trader of some type. A similar register mentions a “Persian monk” and a “Brahman,” both again apparently traveling on their own.84 Because of the detailed registers, we know about these particular individuals, but not the many other travelers to and from Dunhuang who have left no traces in the historical record.
Other groups, including refugees and robbers, were traveling, too. Thieves are perhaps the least documented group of all. The pilgrim Xuanzang was once robbed of everything, including his clothing. Travelers frequently allude to the risk of crime and often traveled in groups to avoid being robbed.
Members of these official delegations were so sure that they would profit from participating in a tribute mission that they took loans out to rent camels for their journeys. Cave 17 preserved five such loan contracts.85 The contracts anticipate various reasons why the borrower might not return the camel: the animal might fall sick or die en route, might get lost or be stolen, or might be stolen by the envoy himself.86 All follow the same format: they explain that the person taking the camel was participating in a tribute mission, give the rental price of the camel—always in bolts of cloth—to be paid on the borrower’s return, and then specify a penalty should the borrower not return. Gone are the standard bolts of silk used by the Tang dynasty; these contracts specify the length and width of each bolt of silk that is borrowed, further confirmation that the Dunhuang economy of the ninth and tenth centuries operated differently from when the Tang was at the peak of its power before 755. It is not simply that no coins circulated; even the standard-sized bolt of silk fell from use as the Silk Road economy reverted to a subsistence economy.
A ROBBERY ON THE SILK ROAD
Here a brigand holding an enormous blade menaces a group of merchants, who cower before him with
their unloaded goods laid out on the ground as their donkeys look on. A rare depiction of robbers on the Silk Road, this depicts a miracle performed by the Buddhist deity Guanyin, who drove away the thieves in response to prayers by her devotees. Drawing by Amelia Sargent.
While some people, like envoys and monks, traveled to towns beyond Dunhuang, many more remained enmeshed in the local economy. Many of Dunhuang’s residents joined mutual aid societies. The charters they signed shed much light on their concerns. Small groups of local people, usually between fifteen and twenty individuals, joined these societies so that they could pool their resources. Some of the groups were largely social, meeting once a month, and had charters that required each member to bring a small contribution—a measure of grain or of beer—to the meeting. Other groups helped each other survive unexpected expenses. If one member of the group had to pay for the wedding or the funeral of a relative, he or she had use of that month’s revenues. People with roughly equal incomes formed groups because they could afford similar contributions.87 The wealthiest people in Dunhuang joined lay associations that sponsored the construction of new caves.88
The monasteries were the richest institutions in local society. They had sufficient surplus grain that they could make loans to the poor, and many of these contracts survive. Local people borrowed grain so that they could get enough seed to plant in the spring. Their very survival depended on these loans. The poor lived on the edge: parents were often forced to sell their children or put them up for adoption.89
The monasteries kept track of these loans and also maintained detailed inventories of all their property.90 These inventories document the holdings of the richest local institutions. Since the wealthy frequently made contributions to the monasteries to gain merit, the Buddhist monasteries—like their counterparts in Europe—had large collections of valuable goods. Yet because archeologists have not yet uncovered any monastic treasuries, we must rely on written inventories documenting monastic holdings. Many items are preceded by the word fan, meaning “foreign,” and some scholars have concluded that these items must indeed be of foreign manufacture. But this is not necessarily the case. French fries are not made in France; they are simply French in inspiration.91 Similarly, in the case of the items listed on the monastic inventories, without a given item in hand it is impossible to know if it was truly foreign made or just foreign in style.
The goods listed in the monastic inventories fall into four categories: textiles, metalware, incense and fragrance, and precious stones. Where some of the textiles are clearly of local manufacture (Khotanese felt, for example), others, such as “Iranian brocade” or “Merv silk,” appear to be imported. It is possible, though, that these textiles were not made outside of China but simply copied foreign silks. The same is true of the thirty-seven listed metal goods. A “silver censer with silver lions” is plausibly from the Iranian world, but an “Iranian-style lock” was too heavy and too utilitarian to transport overland over long distances. These locks were probably made by local metalsmiths. “Hufen,” often translated as “Iranian powder” occurs frequently on the list of fragrances; this is the name for ceruse, a white-colored, lead-based makeup powder, which also appears in the Sogdian Ancient Letters. The adjective “hu” often means “Iranian” or “Iranian-style” in Dunhuang documents, but in this instance it means “paste,” because one had to mix ceruse with water before applying it to one’s skin.92
Only one category in the monastic inventories is definitely of foreign origin: precious stones, including lapis lazuli (from Badakhshan in the northeast corner of Afghanistan), agate (from India), amber (from northeastern Europe), coral (from the ocean, most likely via Tibet), and pearls (usually from Sri Lanka). The foreign merchants in Tang tales almost always deal in precious stones, and precious stones were the one commodity that was sufficiently light that an individual merchant could carry a small bag on his person over long distances. Other materials from Dunhuang confirm our impression of an economy in which largely locally made goods circulated. These goods included silks of all different varieties, cotton, fur, tea, ceramics, medicine, fragrances, Khotanese jade, and draft animals.
Who brought these goods to Dunhuang? Many of the emissaries going back and forth conducted trade on the side, and they are the most likely agents of transmission. Dunhuang served as a center for envoys from surrounding towns, and delegations often presented goods, such as cotton textiles woven in Turfan or jade from Khotan, that they had purchased in the course of their travels.93 Envoys were one group whose movements are well-documented in the materials from cave 17, which rarely mention merchants. Interestingly, those that do are in non-Chinese languages, specifically Sogdian, Uighur Turkish, or “Turco-Sogdian,” a mixed language combining elements of both. These shed considerable light on the comings and goings of caravaneers.
Around the year 1000 Sogdian began to gradually die out. Sogdian ceased to be used as a written language, and many (but not all) former Sogdian-speakers turned instead to Turkish. A small group of documents from cave 17 offer a glimpse of this linguistic shift just as it was occurring. They are in Turco-Sogdian, which is basically Sogdian but with a strong Uighur influence, in the form of Turkic loanwords and, more importantly, Turkic constructions unknown in earlier Sogdian.94 This small group includes an account written by a low-level merchant who reports to his employer which different commodities he obtained from producers. The author, possibly a member of the Church of the East, moves from village to village, picking up pieces of woven cloth from different weaving households, most likely individual families, who produce the cloth. He records just how far he travels on this trip: 60 miles (100 km) to the village of Changle, which is 60 miles (100 km) northeast of Dunhuang and 30 miles (50 km) west of Guazhou, also in Gansu Province. This account accords with the dearth of coins reported in Chinese and Tibetan documents from cave 17.
One letter begins by giving the total amount of cloth the agent is carrying: one hundred pieces of “white” and nineteen pieces of “red” raghzi cloth, which is used to make warm winter clothing.95 (Raghzi is a Sogdian word, meaning either wool or some other type of cloth made from fur.) The pieces dyed red are worth more than undyed pieces; the agent usually trades three pieces of undyed cloth for two of dyed, and four dyed pieces of cloth are worth a single sheep. On his next transaction, he picks up four dyed pieces and twenty-one undyed. He records each transaction carefully, and they are all similarly small. This is a classic peddler’s itinerary: covering a limited terrain, dealing in locally produced goods, and consisting largely of trading one item for another.
The author of this letter, which dates to the late ninth century, was equally comfortable writing in both Sogdian and Uighur. Writing in the mid-eleventh century, the lexicographer Mahmud of Kashgar described the Sogdian residents of Semirech’e, in what is now Kazakhstan, as speaking both Sogdian and Uighur, but within two hundred years the Sogdian language had died out.96
Another group of documents, in Uighur, complements this picture of the peddler trade in the Turco-Sogdian documents. Uighur was the language of the Uighur kaghanates. Few documents in cave 17 are in Uighur—some forty or so.97 These include religious texts, lists of merchandise, letters, and legal decisions, which mention various locally made goods: cloth (including silk, wool, and cotton), slaves, sheep, dye, camels, lacquer cups, combs, casseroles, steel small knives, pickaxes, handkerchiefs, embroidery, whey, and dried fruits. A few goods, such as silver bowls or silver quivers, may have been of foreign manufacture, yet only musk and pearls are definitely foreign (one letter mentions 117 pearls, the most valuable single item).98 The authors of these materials describe a world bounded to the east by Suzhou, modern Jiuquan in Gansu Province; to the north by Hami, Xinjiang, and Ötükän on the upper reaches of the Orkhon River; to the west by Miran, near the Tibetan border; and to the southwest by Khotan. The Uighur materials portray a commercial world exactly like that of the Turco-Sogdian materials: local peddlers traveling within a circumscribed area
and trading one locally manufactured good for another.
Other scholars have seen these Turco-Sogdian and Uighur documents as evidence of a thriving Silk Road trade.99 The mere mention of trade confirms their expectations. Although the documents refer only to small-scale transactions involving largely local goods, those predisposed to see a large-scale Silk Road trade see this as sufficient evidence. But all the documents this book has examined—with the exception of the government documents listing massive payments to troops stationed in the northwest—point to a small-scale, local trade, rather than a thriving long-distance trade.
When Aurel Stein first arrived in Dunhuang on March 23, 1907, he encountered a merchant named Sher Ali Khan from Kabul, Afghanistan. His caravan of forty camels traveled from Afghanistan to Khotan and then on to central Gansu; on his way home, he took the Southern Silk Road route as well. His business model was simple. He sold British textiles purchased in Kashmir and Yarkand to the Chinese and bought Chinese silk and tea to sell on his return to Kabul. Sher Ali Khan offered to carry Stein’s mail to Kashgar, and Stein, always glad of a chance to correspond with his friends, immediately began drafting letters and finished writing only at 3 in the morning. Stein then set off to explore the watch-towers of Dunhuang, where he found the Sogdian Ancient Letters. To Stein’s dismay, one evening, as he returned to his camp, he glimpsed Sher Ali Khan’s caravan, which had covered “in eleven days less than eighty miles.” It turned out that the caravan had hired an inexperienced guide who had lost his way in the desert. The disappearance of two valuable ponies delayed the caravan further still. Stein said goodbye to Sher Ali Khan a second time, and, to his surprise, the letters actually arrived in England. His friends received them in the end of September, about six months after he had written them.100
The Silk Road: A New History Page 24