The Silk Road: A New History

by Valerie Hansen

  After crossing the Tianshan Mountains, Xuanzang and his companions arrived at Lake Issyk-kul in Kyrgyzstan. Issyk-kul means “hot lake” in Turkic, and the sea is fed by warm springs that prevent it from freezing over. The Chinese also called Issyk-kul the “warm sea.”5 Near the modern town of Tokmak, at a site called Ak-Beshim, near the lake’s western edge, Xuanzang met the leader of the Western Turks, the kaghan, who wore a fine green silk robe and wrapped a ten-foot-long silk band around his head, allowing his long hair to fall down his back.6

  At the time, in 630, the kaghan headed a confederation of Turks that controlled the territory all the way from Turfan to Persia. He did not rule directly but left local rulers, like those in Turfan, Kucha, and Samarkand, in place as long as they paid him tribute, provided troops when requested, and obeyed his commands. For several days the kaghan, like the Gaochang king, tried to persuade Xuanzang to stay at Tokmak and not go on to India. When Xuanzang did not consent, the kaghan finally gave in and provided the monk with an interpreter, fifty pieces of silk for travel expenses, and letters of introduction to the various rulers who accepted him as their overlord. From the town of Tokmak, Xuanzang and his party traveled westward, going through lovely mountain pastures, and then across the barren Kizil Kum desert before reaching Samarkand.

  In his Record of Travels to the West, a detailed account of the different countries of the Western Regions, Xuanzang sketched the basic traits of the Sogdians.7 They did not write characters; instead they used an alphabet of some twenty letters in various combinations to record a broad vocabulary. Their clothing was simple, made of fur and felt, and, like the Turkish kaghan, the men bound their heads with cloth and shaved their foreheads. This struck Chinese observers as unusual, because they viewed hair as an integral part of the body, a gift from one’s parents that should not be cut.

  Xuanzang gives voice to a widely held Chinese view of the Sogdians: “Their customs are slippery and tricky, and they frequently cheat and deceive, greatly desiring wealth, and fathers and sons alike seek profit.”8 The compilers of the official history of the Tang dynasty echoed this prejudice in their description of how the Sogdians raised their sons to be merchants: “When they give birth to a son, they put honey on his mouth and place glue in his palms so that when he grows up, he will speak sweet words and grasp coins in his hand as if they were glued there.… They are good at trading, love profit, and go abroad at the age of twenty. They are everywhere profit is to be found.”9

  Unfortunately, few Sogdian-language sources are available to correct these stereotypes. The climate in and around Samarkand is not as dry as that of the Taklamakan Desert, the soil is more acidic, and many materials were destroyed after the Islamic conquest of the early eighth century. Only two important groups of Sogdian-language documents survive: the first, the eight Sogdian “Ancient Letters” from the early fourth century, was found by Aurel Stein outside Dunhuang, while the second, nearly a hundred documents from a castle under siege, is from the early eighth century and was discovered in the 1930s outside Samarkand. Other Sogdian-language materials are limited to inscriptions on silver bowls or textiles, captions on paintings, and the many religious texts found at Turfan, which say little about the history of the Sogdians.10

  The first traces of the Sogdians in the archeological record come from the earliest levels of habitation at Samarkand, which date to the seventh century BCE. Writing several centuries later, the biographers of Alexander of Macedon commented on the ferocious resistance of the inhabitants of Marakanda, the Greek name for Samarkand, who ultimately surrendered to Alexander. After Alexander’s death, various dynasties took power, and, for much of the time, a confederation based in what is now Tashkent controlled the city.11

  Until recently, scholars have thought that the abandoned mailbag that Aurel Stein unearthed near Dunhuang in 1907 was the earliest surviving body of Sogdian-language materials, but between 1996 and 2006 archeologists working at the site of Kultobe (near Chimkent on the river Aris) in southern Kazakhstan found ten scraps of baked brick plaques with Sogdian letters on them. After careful examination of these materials, Nicholas Sims-Williams, the world’s leading expert on the Iranian languages of Central Asia, has determined that they predate the mailbag letters. At least four Sogdian city-states existed at the time this wall was built, but the texts are so fragmentary that it is difficult to learn much from them.12

  The eight Sogdian-language letters found by Stein are largely intact and thus far more informative. Stein discovered the abandoned mailbag 56 miles (90 km) northwest of Dunhuang. One of the letters was addressed to a resident of Samarkand, an indication that the letter carrier was on his way to Samarkand when the letters were lost. In 1907 Stein’s men unearthed the letters while exploring a series of watchtowers, spaced two miles (3.2 km) apart, built by different Chinese dynasties to defend the frontier. Twenty feet (6 m) or more high, these guard towers often were attached to small dwellings for the guards.13 At one tower, which Stein numbered T.XII.A (T for Tun-huang, a variant spelling of Dunhuang), Stein noticed nothing special and assigned a team to clear out the passageway while he went to explore another tower. When he returned in the evening, the workmen showed him what they had discovered: some colored silks, a wooden case, Chinese documents dating to the early first century CE, a piece of silk with Kharoshthi script on it from before 400 CE, and “one small roll after another of neatly folded paper containing what was manifestly some Western writing.”14 The script resembled Aramaic, and Stein remembered that he had found other similar examples at Loulan. Only later was the unfamiliar script identified as Sogdian.

  These eight sheets of paper proved to be extraordinarily revealing, even though the letters are difficult to decipher and many words are missing. The handful of the world’s scholars who can read Sogdian continue to debate the meaning of each sentence, even now still occasionally explicating a phrase that has puzzled everyone for the past century. Four of the five intact letters have been translated into English.15 Stein’s excavation methods were advanced for his day, but not perfect; his workmen did not record which materials they found at which level in the collapsed tower, a real problem, given that the letters were undated.

  Crucially, one letter provides a clue to the time of composition: “And, sirs, the last emperor, so they say, fled from Luoyang because of the famine and fire was set to his palace and to the city, and the palace was burnt and the city destroyed. Luoyang is no more. Ye [Zhangde fu, Henan] is no more!”16 Attacks on Luoyang occurred in 190, 311, and 535. Most scholars of Sogdian concur that the letter refers to the events of 311 and was written in either 313 or 314.17 The author of the letter refers to the invading army as “Huns,” and indeed their leader, Shi Le (274–333), belonged to one of the peoples in the Xiongnu confederation. This is one of the main pieces of evidence linking the Xiongnu with the Huns of Central Asia who invaded Europe in the late 300s.18

  The eight letters were not in envelopes but “folded up into neat little convolutes,” as Stein called them, measuring 3.5–5 inches (9–13 cm) long, 1–1.25 inches (2.5–3 cm) across. Although originating in different Chinese towns, the letter paper had similar dimensions, of roughly 15.5–16.5 inches (39–42 cm) by 9.5–9.75 inches (24–25 cm), suggesting that even then sheets of paper were made in standardized sheets—a rather rapid development given that paper came into widespread use in China only in the third century CE. Three letters were placed in individual silk bags; a fourth, letter 2 (shown at the beginning of this chapter), was in a silk bag with a linen covering that said “bound for Samarkand” but with no return address. The other letters were not addressed, suggesting that the person delivering them may have known the recipients for whom they were intended. Letters 1 and 3 were written by a woman living in Dunhuang to her mother and husband who probably lived in Loulan, while Letter 5 was sent from Wuwei.

  Already in the early fourth century, the letters reveal, Sogdian communities existed in Luoyang, Chang’an, Lanzhou, Wuwei, Jiuquan, and Dunhuang. T
he second letter mentions settlements of forty Sogdians in one place and one hundred “freemen” from Samarkand in another (both locations unfortunately illegible); the Luoyang settlement included both Sogdians and Indians. As soon as a community of Sogdians reached a certain size, perhaps forty, they erected a fire temple. The sabao performed ritual functions, namely tending the fire altar and presiding over Zoroastrian festivals, and as headman adjudicated disputes.

  In Iran Zoroastrianism evolved toward monotheism, with Ahura Mazda as the supreme deity, but in Sogdiana its adherents worshipped many deities, including Ahura Mazda.19 Zoroastrian teachings forbade both Chinese burial in the ground and Buddhist cremation since both were polluting: burial polluted the earth, and cremation polluted fire. Instead, the Zoroastrians exposed the corpses of the dead, allowing scavenging animals to clean the meat from the bones before burying them in clay urns, called ossuaries.

  Miwnay, the Dunhuang woman who wrote letters 1 and 3, was abandoned by her husband and saddled with his debts. The list of the men she approached for assistance provides a capsule sketch of Sogdian society in exile. Miwnay turned to a councilor (apparently an official who collected taxes), a relative of her husband’s, and then to a third man, apparently a business associate. Each refused to help on the grounds that it was her husband’s obligation, not theirs. Finally, she approached the “temple-priest,” who promised a camel and a male escort.

  Miwnay vents her frustration forcefully in her letter to her husband: “I obeyed your command and came to Dunhuang and did not observe my mother’s bidding or that of my brothers. Surely the gods were angry with me on the day when I did your bidding! I would rather be a dog’s or a pig’s wife than yours!”20 A postscript written by her daughter reports that the two poverty-stricken women were reduced to tending sheep. Stuck in Dunhuang for three years, Miwnay had five opportunities to leave with caravans but not enough money to pay the twenty staters she needed for her passage.

  Scholars are not certain of the value of a stater. Did it weigh .42 ounces (12 grams), as some of the stater coins circulating at the time did? Or was it a much lighter coin of .02 ounces (.6 grams) of silver circulating in Samarkand? (This is just one of the many conundrums still facing Silk Road researchers.)

  Far richer than Miwnay was the business agent who reported the fall of Luoyang. He had sufficient funds in Samarkand for him to authorize merchants handling his affairs to “take 1,000 or 2,000 staters out of the money” to assist an orphan in his care. In his letter, the business agent wrote to his boss in Samarkand about the different people he had hired in the Gansu towns of Jiuquan and Wuwei. His letter documents three levels in the company: the boss (a father and son in Samarkand); the agent (author of the letter), who supervises a network of weavers who work for them; and the weavers themselves.

  Letter 2 also mentions some of the commodities being traded at the time, namely woolen cloths and linen. The agent reports that he has sent 32 “vesicles”—a unit of uncertain value—of musk to Dunhuang. Musk, processed from the glands of the musk deer, was used as a scent and as a fixative. According to Étienne de la Vaissière, a prominent historian of the Sogdians, the musk probably weighed 1.75 pounds (.8 kg), an enormous amount of pure musk.21 The second letter also mentions woolen cloths and linen but gives no quantities.

  The fifth letter, from a more local level of commerce between Guzang and Dunhuang, is addressed to the caravan leader. It mentions much smaller amounts of money: the author claims that he received only four and a half staters from the twenty he was owed. He describes several goods being sent by caravan from Guzang to other destinations, most likely Loulan, some 900 miles (1,400 km) away: “white,” most likely ceruse, a cosmetic with a white lead base; pepper; silver; and “rysk,” a term whose meaning is not clear. Certain goods traveled great distances: pepper (letter 5) and camphor (letter 6) could be purchased only in Southeast Asia or India, while musk (letter 2) came from the Tibetan border with Gansu. In the sixth letter, only sections of which survive, the author asks the recipient to buy something, possibly “derived from the silkworm,” meaning silk cloth or thread. If this is not available, then the recipient is to purchase camphor instead. This is apparently the only mention of silk in the Sogdian letters.22 None of the quantities given in the Sogdian letters is certain, but most scholars concur that the amounts are small, most likely between 3.3 pounds (1.5 kg) and 88 pounds (40 kg).23 An animal, or a few animals, could easily have carried these amounts, suggesting that the Silk Road trade was a trade limited in scope, in what some scholars have called a “petty” trade.24

  Source: Étienne de la Vaissière, Histoire des Marchands Sogdiens

  (Paris: Collège de France, Institut des Hautes Études Chinoises, 2002), Map 3.

  The Sogdian Ancient Letters are significant because they are among the only Silk Road documents written by merchants, not by the authorities overseeing or taxing trade. They depict a diasporic group of Sogdians living peacefully as merchants, farmers, and even servants in lands still engaged in commerce and long-distance trade at a time when China was in chaos, with one dynasty having just fallen and another struggling to take over.

  In the following centuries the Sogdians continued to speak their own language but modified their clothing and hairstyles to conform with the demands of their new, nomadic conquerors—the Huns, Kidarites, Hephthalites, and Turks, who gained control of Samarkand, sometimes with the assistance of the Sasanian empire (224–651), far to the west with their capital at Ctesiphon, near modern-day Baghdad. In 509 Samarkand fell to the Hephthalites, a confederation of Iranian and Turkic peoples who are occasionally called the White Huns and who lived in northern Afghanistan.25 Later, around 560, the Sasanians allied with the newly formed Turkish confederation to defeat the Hephthalites.26 After 565 Samarkand came under the rule of the Western Turks, which is why Xuanzang continued on to Samarkand after meeting the kaghan at Tokmak. Although the Turks developed a written language of their own in the eighth century, they often wrote in Sogdian, and cultural ties between the Turks and Sogdians were close.

  During these centuries of frequent political shifts, the Sogdians gradually expanded out from Samarkand and Bukhara. In the fifth century and later, Sogdians put up Sogdian-style buildings and irrigation works as they designed new settlements around the Zerafshan River. Economic growth accelerated in the fifth century, and by the sixth and seventh century, Sogdiana had become the richest country in Central Asia, as demonstrated by the ever larger houses and more elaborate paintings found by archeologists at Panjikent.27

  Panjikent, 37 miles (60 km) east of Samarkand in Tajikistan, is one of the most important archeological sites on the Silk Road, as first Soviet and now Russian archeologists from the Hermitage Museum in St. Petersburg have been excavating the site every summer since 1947.28 Unlike many of the excavations in China, archeologists are not digging up individual tombs but painstakingly excavating a small city, house by house, quadrant by quadrant.

  The excavations at Panjikent have so far uncovered about 15–17 acres (6–7 hectares), or one-half of the small city. Built sometime in the fifth century, the city reached its largest size in the seventh century. Panjikent fell to Arab armies in 722, revived briefly in the 740s, and then was completely abandoned between 770 and 780.29 Between five and seven thousand inhabitants lived there, surrounded by a city wall dating to the fifth century. Their city contained several streets, alleyways, two bazaars, and two temples, one with a side fire altar, the other with images of at least ten different deities.30 This temple had a room with a separate entrance that housed a statue of the Indian god Shiva sitting on a bull holding a trident. His trident and erect penis conformed to Indian prototypes, but his boots were Sogdian.

  Commercial granaries and bazaars show that a retail trade existed in Panjikent. Although archeologists have not found evidence of permanent buildings to house caravans, called caravanserai in Persian, anywhere in Sogdiana, some modern historians believe that the caravanserai originated in the reg
ion. The geographer Ibn H. awqal described the ruin of a giant building that could house up to two hundred travelers and their animals, with food for all and room to sleep as well.31 Several Panjikent houses had courtyards large enough to house a caravan, and the word “hotel” in Sogdian (tym) was borrowed from the Chinese word “inn” (dian).32

  Caravans passed through Panjikent, since it was on the road between Samarkand and China, which crossed the Ammoniac Mountains, a major source of ammonium chloride in the Tianshan Mountains between modern-day Tajikistan and China.33 Yet few artifacts found in Panjikent can be identified as coming via the caravan trade; small glass vessels dating to the seventh century are one important exception. Local production of glass began only in the mid-eighth century.34

  More evidence of trade lies in the thousands of bronze coins found in the city, many apparently loose change discarded or mislaid at the market place. Silver coins, from the Sasanian Empire, also circulated during the sixth century in smaller quantities. The earliest locally minted coins date to the second half of the seventh century. Apparently the central authorities granted local workshops the right to make coins. In the seventh century, the period of greatest contact between Sogdiana and China, the residents of Panjikent used bronze coins with the same shape as Chinese coins—round coins with a square hole—some with Chinese characters on them, some without.

  As at Turfan, residents sometimes used gold coins. Between 1947 and 1995, archeologists found two genuine gold coins from the Byzantine Empire and six extremely thin imitations. Five, found in houses, indicate that the coins and the imitations were used as currency.35

  Similarly, imitation gold coins also served as burial goods. Two of the gold coins (and possibly a third) were found inside naus structures the Sogdians built to house the dead, usually members of the same family. These buildings were small, square, and made of mud brick; they held the cleaned bones in ossuaries.36 Zoroastrian texts do not mention naus buildings, which first appear in the Samarkand region—but not in central Iran—in the late fourth or fifth centuries.