The Silk Road: A New History

by Valerie Hansen

  CHINESE SILK FROM NIYA

  Artfully woven into the fabric were Chinese characters saying “kings and lords shall be married for thousands of autumns and tens of thousands of years; it is right that they bear sons and grandsons.” This is one of thirty-seven textiles found in a single burial at Niya, making it one of the most important finds along the Silk Road. Courtesy of Wang Binghua.

  The couple lay underneath a single-layered coverlet of blue silk brocade with a pattern of stylized dancers in red, white, and brown. The deceased wore complete sets of clothing.

  A slightly later tomb, numbered M8, also contained a deceased couple, as well as some textiles with Chinese characters on them and a simple clay vessel with the Chinese character for king on it.44 The use of the words “king” and “lord” in the textiles from tombs M3 and M8 suggests that these are gifts from the Chinese central authorities to a local king. At some point after 48 CE the Shanshan Kingdom “incorporated” the Jingjue Kingdom, we learn from the History of the Later Han.45 And so the Niya site, which had served as the capital of the Jingjue state, became part of the larger Shanshan Kingdom.

  A contemporaneous burial from the site of Yingpan (southwest of Loulan) contrasts sharply with the Niya burials, because the corpse was buried in wool, not cotton or silk.46 The deceased man wears a red woolen robe with an elaborate design composed of pairs of facing pomegranate trees, animals, and human figures. Naked cherubic figures brandish swords and lassos as they confront one another in a combat stance. With two interwoven layers, this textile is too complex to have been made by local weavers. It was probably made in Bactria, far to the west, where local artisans modified the Greco-Roman motifs first introduced to the region by the armies of Alexander of Macedon in the fourth century BCE.47 Archeologists have speculated about the identity of this beautifully dressed corpse. The former director of the Xinjiang Archaeological Institute, Wang Binghua, suggests the he may have been a ruler of yet another small oasis kingdom mentioned in the official histories, the Kingdom of Shan (literally “The Hill Kingdom”), whose southeastern border abutted the Shanshan Kingdom.48

  YINGPAN BURIAL

  Interred in a painted wooden coffin, the deceased wore a white mask, made of layers of hemp glued together, with a rectangle of gold foil on the forehead above the eyes. He was buried with two miniature sets of clothing (for use in the afterworld?): one at his left wrist, the other on his stomach. Grave M15, excavated 1995. Courtesy of Wang Binghua.

  Whether or not the deceased in Niya tomb M3 or the Yingpan tomb were actually local kings (still the most likely possibility), they certainly were among the wealthiest people living at these settlements, and their tombs offer a vivid picture of the local economy. The region’s inhabitants buried the dead with grains such as millet, barley, and wheat, and fruits grown in orchards like grapes, pears, peaches, pomegranates, and dates. They viewed a whole leg of lamb as the ultimate treat, the centerpiece of a feast in the next life, and garments made from imported textiles as the most suitable clothing for that life.

  Most analysts concur that the material evidence from the Niya, Yingpan, and Loulan sites dates to somewhere between the second and fourth centuries but are not sure exactly when. In contrast, the written evidence from Loulan is clear: documents in both Chinese and Kharoshthi reveal that Chinese armies were stationed at Loulan in the late third and early fourth centuries CE.

  Most of the documents in Chinese from Loulan date to 263 to 272, with a few remaining examples from 330.49 This was the time when the Shanshan Kingdom ruled the region from Niya to Loulan and when the different Chinese dynasties based in north China who succeeded the Han dynasty, primarily the Wei (220–65) and the Western Jin (265–316), stationed garrisons in Loulan. Loulan produced about fifty Kharoshthi-script documents, but over seven hundred Chinese documents, either short texts (usually no more than ten characters) on wood slips or on tiny scraps of paper.50 The Chinese often recorded private transactions on paper, while garrison officials tended to use wooden slips for their records, an indication that private individuals used paper before government officials did.51

  Like their Han-dynasty predecessors, the Loulan garrison belonged to the Chinese system of military colonies whose inhabitants were expected to grow their own food while they stood ready to serve in the army. The soldiers in Chinese employ used draft animals, like cows and horses, to plow the land on which they raised wheat, barley, and millet. They were not necessarily Chinese; the Chinese military recruited from among the local people as well. The farmer-soldiers also introduced agricultural techniques, most notably irrigation for watering crops. They experimented with cattle-drawn ploughs, and they used new types of iron spade and sickle, the first metal tools employed in the region.52

  Chinese government regulations stipulated that each soldier was entitled to 1 peck (dou) 2 pints (sheng) (approximately 2.6 quarts [2.4 L]) of grain each day—but local officials could not always provide the stipulated amount, and the rations sometimes dipped as low as half that.53 When the grain grown by the farmer-soldiers ran short, Chinese officials, surviving documents reveal, bought extra grain from the local people using coins and colored silk. The Loulan garrison received funds, in both coin and silk, from military units based to the east in the city of Dunhuang or possibly Wuwei, both in Gansu. The silk came in various colors and in two lengths, long and short. One bolt of plain tabby silk found by Stein at Loulan in 1901, shown in color plate 5A, is the only surviving example of silk used as money from such an early period.54

  Many documents give the exchange rates for conversions among three different types of currency: coins, colored silk, and grain. Officials used silk to buy grain and horses for their men, and the soldiers themselves also exchanged silk and grain for shoes and clothing. They regularly converted prices from one currency to another.55

  The Loulan documents mention a few much larger transactions as well. One wood slip, dated 330, reports that the Sogdians, traders originally from the Samarkand region, presented ten thousand piculs (each picul was approximately 1/2 bushel, or 20 L) of something (the word is missing), most likely of grain, and two hundred coins (qian) to the authorities.56 Although the back of the slip has the seals of two Chinese officials, the document does not explain why the Sogdians made these payments. This was either a tax payment or one of an ongoing series of transactions to provide the Chinese troops with food. Another fragment records a large payment of 319 animals in exchange for 4,326 bolts of colored silk.57 This, too, appears to be a payment by Sogdian merchants to the Chinese authorities, and we know from fragments of two Sogdian-language documents found by Stein that Sogdians were active at Loulan at this time.58 In later centuries, Sogdians played a key role in supplying Chinese armies, and it is quite likely they had already begun to do so in the early fourth century CE at Loulan.

  The Chinese documents found by Stein and Hedin at Loulan come from only a few spots.59 Still, they give the overwhelming impression that the transactions at Loulan exclusively involved the garrison collectively or soldiers individually, using grain, silk, and coins to obtain grain, horses, clothing, and shoes from the local people. In sum, the indigenous subsistence economy occasionally supplied the Chinese garrison with locally produced commodities. As Ito Toshio, who teaches at Osaka Kyoiku University, has concluded after a thorough survey, the documents do not mention any profit-seeking activities.60 The only evidence of merchants—and it is highly fragmentary—indicates Sogdian merchants who worked for the military authorities.

  The Kharoshthi-script documents from Niya and Loulan are much richer than the Chinese-language documents. They portray a broader range of society, from the lowliest cultivators to the ruler himself, involved in a host of activities, some utterly mundane. Accordingly, they make it possible to glimpse life along the Silk Road in a way that the Chinese documents do not.

  Some of the Kharoshthi documents give the name of the current king, the year of his reign, and occasionally the name of his predecessor or successor. I
n 1920, using such clues, Rapson and his collaborators drew up a list of five kings who ruled for a total of some ninety years. Yet no one knew exactly when those local kings had ruled. In 1940 Thomas Burrow produced a comprehensive translation of all those Kharoshthi documents whose meaning could be understood, but they still remained undated.

  Then, in 1965, John Brough announced that he had found the key to dating the Kharoshthi documents: the Chinese title shizhong, literally meaning “palace attendant,” corresponded to the term jitumgha. In 263, the Kroraina king Amgoka used the new title for the first time. He may have received it from the ruler of the Western Jin dynasty (265–316), a regional Chinese dynasty based in Luoyang city, Henan Province, but this was two years before the Western Jin decisively defeated the dynasties that preceded it. The forms of address for the rulers also changed: before the seventeenth year of Amgoka’s reign, letter writers used long strings of titles for kings; after that year, the titles became noticeably shorter and include the word jitumgha.61

  The year 263 was the seventeenth of King Amgoka’s reign. Once this single year was fixed, it was simply a matter of assigning calendar years to all the kings’ reigns. Brough’s original chronology has since been extended slightly because of the subsequent discovery of Kharoshthi documents naming other kings.62 Not everyone accepts Brough’s dating, but there is general agreement that the Kharoshthi documents date from the mid-third to the mid-fourth century, give or take twenty years. This period overlaps with the Chinese documents from Loulan, which date to 263–330. And because the Kharoshthi documents refer to no specific external event, there is little chance of dating them more precisely.

  Since the local inhabitants lacked their own written language, the Kharoshthi script served to record people’s names, very strange-sounding names indeed. The approximately one thousand proper names and 150 loanwords that appear in the Kharoshthi documents indicate that the local language of Niya was not Chinese and totally unlike the language spoken by the refugees from Gandhara. Writing in 1935, Thomas Burrow suggested that the indigenous Niya language was related to Tocharian, an Indo-European language spoken on the northern route, but his suggestion has not been widely accepted, nor has it prompted further study.63 It appears that, before the arrival of the immigrants, the indigenous peoples had their own language but no means of writing it, which is why they adopted the Kharoshthi script.

  While rulers tend to have names of local origin like Ly’ipeya, many scribes have names of Sanskritic origin like Buddhasena, meaning “The one whose lord is the Buddha.” As we often see today, names are not always a reliable indicator of someone’s ethnic background; immigrant parents do sometimes draw the names of their children from the culture of their new home. Yet often the only identifying trace that survives of a given individual living along the Silk Routes is his or her name.

  In reconstructing the migration of the technologically more sophisticated Gandharans to Niya, one might expect the migrants to overthrow the local rulers and found their own states. Interestingly, the names of the rulers and the scribes indicate that, while many scribes were Gandharan, the rulers continued to be local men. The scenario of refugees from north India, migrating in groups of no more than a hundred at a time, seems likely.

  The Kharoshthi documents do not record what happened when the first migrants from India arrived. A later ruler instructed local officials to receive refugees, “who are to be looked after as if they were your own.” He also stipulated that the refugees be given land, houses, and seeds “so that they can make copious and plentiful cultivation.”64 Not all refugees fared so well; some were assigned to work as slaves for the residents. The later treatment of migrants is important because it suggests how the migrants from Gandhara may have been treated on their arrival.

  The refugees taught the indigenous peoples how to write their script and to store their documents in archives, the first of which Stein and a man named Rustam, whom Stein called “the most experienced and most reliable of my old diggers of 1901,” discovered in 1906. The two men returned to room 8 of house 24, because, as Stein explained:

  Already during the first clearing I had noticed a large lump of clay or plaster near the wall where the packets of tablets lay closest. I had ordered it to be left undisturbed, though I thought little of its having come to that place by more than accident. Rustam had just extracted between it and the wall a well-preserved double-wedge tablet, when I saw him eagerly burrow with his hands into the floor just as when my fox terrier “Dash” was at work opening rat holes. Before I could put any questions I saw Rustam triumphantly draw forth from about six inches [15 cm] below the floor a complete rectangular document with its double clay seal intact and its envelope still unopened. When the hole was enlarged we saw that the space toward the wall and below its foundation beam was full of closely packed layers of similar documents. It was clear that we had struck a small hidden archive.65

  The marking of the site with a lump of clay or plaster, Stein felt, showed that the original owner had been obliged to leave the village in a hurry, but with the intent to return.

  This single find produced nearly eighty documents, of which twenty-six were “double rectangular tablets” with their seals intact.66 Stein used this term for a specific type of document: like a shallow drawer, the upper piece of wood slotted into the bottom rectangular piece of wood, and the two pieces of wood were tied together with string and sealed.

  Local officials archived these documents and retrieved them when needed. In one instance, a monk sold a plot of land to a man named Ramshotsa for three horses; twenty years later, when someone encroached on Ramshotsa’s land, officials consulted the earlier rectangular tablet before deciding that the land indeed belonged to Ramshotsa.67 In all, over two hundred double rectangular tablets were excavated at Niya, most ending with a statement of the penalty to be charged should either party challenge the terms of the exchange, and a variation of the assertion that the document’s “authority is a thousand years, as long as life.”68

  EAST MEETS WEST ON A KHAROSHTHI DOCUMENT

  This wooden document from Niya survives intact: the upper drawer has been slotted into the lower holder, and cords have been wrapped around both pieces of wood through notches and then sealed with clay. The seal on the left is in Chinese; that on the right shows a Western-looking face, most likely a Greek or Roman deity, which are often shown on Gandharan seals. These double rectangular tablets record the exchange between two parties of various types of property—slaves, livestock, and land—and give the names of the officials who recorded the transaction.

  Suspecting that the different shapes of documents had distinct purposes, Stein proposed that a second type of document, “wedge-shaped tablets,” was for royal orders or policy decisions. He found nearly three hundred documents of this type. The wedge-shaped documents consisted of two pieces of wood of the same size, 7–15 inches (18–38 cm) long and 1.13–2.5 inches (3–6 cm) wide, that were placed face to face, tied together with string, and then sealed. The seals portrayed Greek gods like Athena, Eros, and Heracles, who were familiar to the migrants from the Gandhara region, who had worshipped them for centuries.69 The outside piece of wood gave the name of the recipient, and on the inside were the king’s orders, most of which followed the same opening formula:

  To be given to the cozbo Tamjaka.

  His majesty the king writes, he instructs the cozbo Tamjaka as follows:70

  These official orders came from the king of Kroraina to the highest-ranking local official, or cozbo, the equivalent of governor.71 Assisted by a group of lower officials, the cozbo heard and adjudicated local disputes.

  This wedge-shaped tablet was addressed to Tamjaka, the cozbo of Cadh’ota, the name used in the Kharoshthi documents for the settlement at Niya. The king asked the cozbo to investigate a complaint from a resident of the town that soldiers from a neighboring district had stolen two of his cows. They had eaten one, he claimed, and returned the other. The royal orders often addressed i
ntensely local concerns like this.

  If the king had a more urgent order, he wrote it on leather. Only a few of this type of document survive. Other shapes of documents from Niya were used for private correspondence or for lists. The Japanese scholar Akamatsu Akihiko, who teaches Indic languages at Kyoto University, has suggested that the different types of Kharoshthi documents had their origins in the bureaucratic system of the Mauryan dynasty of northern India (ca. 320–185 BCE) as recorded in the Arthashastra.72 This text, while it may be based on earlier texts, dates to the second to fourth centuries CE.73 Attributed to Kautilya, the Arthashastra is a prescriptive text packed full of instructions about how to govern. Presuming that the ruler will issue written orders to his subordinates, it lists “the characteristics of a good edict” and “the defects” of bad edicts. It also gives the sources of law as dharma (a Sanskrit term usually understood as meaning correct conduct according to law or custom, but sometimes specifically indicating the teachings of the Buddha), evidence, custom, and royal edicts. Since the royal edicts are assumed to coincide with dharma, they take precedence over the other sources of law.

  The Arthashastra lists nine types of royal edicts (some with subtypes) that do not correspond one-to-one with the Niya documents, but the overlap is noticeable. Many of the Kharoshthi documents from Niya, for example, seem to fit the category of “a conditional order,” which instruct the recipient, “If there is any truth to this report, then the following shall be done.”74 The resemblance is not surprising: people familiar with South Asian bureaucratic norms in the third and fourth centuries CE wrote both the Arthashastra and the Kharoshthi documents.