Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Asian History. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 19 June 2021

The Classical Silk Road: Trade and Connectivity across Central Asia, 100 bce–1200 cefree

The Classical Silk Road: Trade and Connectivity across Central Asia, 100 bce–1200 cefree

  • Valerie HansenValerie HansenYale University


The Silk Road refers to all the overland routes connecting the major oasis kingdoms of Central Asia including Dunhuang, Turfan, Khotan, and Samarkand to their neighbors: the Chinese landmass, the Mongolian grasslands, the Iranian plateau, and the Indian subcontinent. The best-known routes ran east-west, but the north-south routes to the nomadic states of the Asian grasslands were also important.

In the popular view of the Silk Road, extensive camel caravans carried goods over long distances, but this was rarely the case. Usually peddlers carried mostly local goods short distances. Government shipments to provision armies profoundly affected the region’s economy, because they involved much larger quantities than in the peddler trade.

Rulers regularly exchanged envoys who carried gifts, exchanges that continued even when private trade fell off. Whatever the reason for an individual’s trip, almost everyone—whether envoy, missionary, artist, craftsman, or refugee—bought and sold goods to pay for travel along the Silk Road.

Silk was not the primary commodity traded on these routes. Goods traveling east included ammonium chloride, paper, silver, gold, glassware, and aromatics such as spices, incense, and fragrant woods. Goods traveling west out of China included bronze mirrors, other metal goods, and paper, in addition to silk. Between 300 and 1000 ce, the most important function of silk was as a currency, not as a trade good, although it remained an important export throughout the period.

A vibrant series of cultural exchanges occurred alongside these commercial exchanges. Technologies, medicine, plants, music, and fashion all moved in both directions across Central Asia. Multiple religions also entered China during this time.

The term Silk Road may not be the most accurate term for these commercial and cultural exchanges, but, despite its flaws, the term has secured a firm place in both scholarly works and the popular mind.


  • Archaeology
  • Borderlands
  • Central Asia
  • China

Historical Overview

The core Silk Road region extended from Transoxiania, between the Amu Darya and the Syr Darya Rivers largely in modern-day Uzbekistan in the west, to the Tarim River basin in the Xinjiang Autonomous Region of China in the east.1 In some periods neighboring land empires, such as the Xiongnu steppe empire, the Kushans of India, or the Han and Tang dynasties of China, exercised indirect or direct political control over multiple oases. At other times the region’s oasis states retained their independence or managed to conquer some of their neighbors.

Three shorter periods comprise the classical period of the Silk Road: (1) 100 bce–300 ce, (2) 300–800, and (3) 800–1200. In the first period the Han dynasty (206 bce–220 ce) vied with the Xiongnu Empire in modern-day Mongolia for control of the Tarim Basin. After the collapse of the Xiongnu and the weakening of the Han dynasty in the late 1st century ce, the Kushan Empire in northern India influenced the Kroraina kingdom, which extended from east of Khotan to Loulan.2

Figure 1. The main routes of the Taklamakan. Travelers along the Silk Road chose between the route south of the Taklamakan Desert and that going north, which became more popular after 500 ce.

Source: Map courtesy of Oxford University Press. Valerie Hansen, The Silk Road: A New History with Documents (New York: Oxford University Press, 2017), map 2, inside back cover.

In the second period, between 300 and 800, the three most important powers on the Silk Road were Tang-dynasty China, the Western Türks, and the city states of Sogdiana, including Samarkand (in post-1991 Uzbekistan and Tajikistan). Sogdiana was home to the Sogdians, the most important Silk Road traders.

In the third period, between 800 and 1200, no external power controlled any part of the Silk Road. The two Tarim Basin oases of Khotan and Dunhuang were the most influential before around 1006, when the Qarakhanids, a tribal confederation based in Kashgar, conquered Khotan. Unlike the previous Silk Road powers who had been Buddhist, the Qarakhanids were Muslim, and their rule initiated the Islamicization of the Tarim Basin.3

The Opening of the Silk Road, 100 bce–300 ce

The Silk Road trade began with a noticeable increase in the movement of goods across Central Asia in the 1st century bce. This uptick coincided with an extended period during which the Xiongnu peoples of the Mongolian grasslands and the Han dynasty fought to control the Tarim Basin.4

Between 400 and 200 bce exchanges of different goods were already occurring in certain small regions, particularly Sogdiana, mostly in a north-south direction—but not across all of Central Asia.5 Different goods have been found in Xiongnu tombs, such as bronze mirrors, lacquer vessels, and silks from China; compound bows, arrows, dagger sheaths, and horse gear from the steppes; and faience, glass, amber beads, and Roman glass bowls from the Black Sea area. Through careful study of these goods, scholars have identified a preliminary period of exchanges before 100 bce, the peak of the trade between 100 and 1 bce, and a subsequent period of decline.

The archeological record does not reveal why these exchanges began, but it seems most likely that successful Xiongnu leaders bestowed prestige gifts on their followers, who then displayed a range of goods to broadcast their success.6 The top Xiongnu leader, the Shanyu, headed a complex hierarchy of subordinates ranging from lower-ranking kings of entire regions all the way down to commanders of ten soldiers.7 Chinese-language documents mention small parties of Xiongnu traders as well as raiders who traveled south to obtain bronze implements, animals, and people for their leaders,8 and Xiongnu-run workshops also produced goods their leaders could distribute among their followers.9

In 200 bce the Xiongnu defeated the Chinese, who signed a treaty requiring them to make annual payments in silk, foodstuffs, and wine.10 These payments continued off and on for the next three hundred years, and the two empires exchanged envoys at regular intervals with each other and also with neighboring smaller kingdoms. The exchange of envoys was the main channel for diplomacy among these ancient states; in addition to the gifts the envoys carried to other courts and received on behalf on their rulers, the envoys also gained valuable information about both friendly and enemy countries. Because many of these envoys engaged in trade themselves, a commercial trade arose alongside the official exchanges of envoys.11

Chinese-language historical records confirm that Silk Road exchanges did not begin when the Han dynasty first moved into Central Asia in the 2nd century bce. The first written description of the Silk Road trade (written at least 150 years later) concerns Zhang Qian (d. 113 bce), a Chinese envoy sent from the Han-dynasty capital of Chang’an (in modern Xi’an, Shaanxi province) to Central Asia during the reign of Emperor Wu of the Han dynasty (reigned 140–87 bce). Emperor Wu hoped that Zhang Qian could persuade the Yuezhi people, who lived in the Ferghana region of modern-day Uzbekistan, to ally with the Chinese against their common enemy, the Xiongnu.

On his way to the Yuezhi realm, Zhang passed through Xiongnu territory and was taken prisoner for ten years before he was able to escape and resume his journey. After his return to the Han-dynasty capital of Chang’an (modern Xi’an), sometime around 126 bce, he gave Emperor Han Wudi the first detailed eyewitness description of the peoples living in Central Asia.

Zhang Qian discovered that Chinese merchants and trade goods had arrived in Central Asia before he did. In the markets of Bactria, in modern-day northern Afghanistan, he observed bamboo and cloth woven in the Chinese province of Sichuan, several thousand miles away.12 His comments reveal that private commercial exchanges predated the start of Chinese diplomacy in the region, just as the archaeological evidence from Xiongnu tombs suggests.

The Chinese goods Zhang Qian saw probably traveled overland. Merchants transported them over long distance—most likely dividing the route into shorter segments—because they stood to make a profit. Archaeological finds show that the overland trade between China and Central Asia began much earlier, at least circa 1200 bce, indicating that trade was occurring in an eastern circuit and in a western one before the two circuits were connected in the 1st century bce.13

The main written records about the Silk Road region, the Chinese dynastic histories, report that more than fifty “kingdoms” existed in Central Asia during the Han dynasty. Surrounded by walls, these were more like city-states than kingdoms, with populations ranging from several hundred to several thousand people.14 As part of their ongoing diplomatic efforts to strengthen themselves against the Xiongnu, the Han dynasty welcomed envoys from these city-states and dispatched Chinese envoys, often in groups of one hundred or more, to visit their rulers.15 By the year 100 bce, the Han dynasty had conquered much of the Gansu corridor, the natural pathway connecting China with Central Asia, and stationed garrisons at regular intervals to strengthen their control.

When travelers going west from China arrived at Dunhuang, a major entrepot located in modern Gansu province, they had to decide whether to proceed north or south around the Taklamakan Desert. Documents found at the Xuanquan postal station located forty miles (64 km) east of Dunhuang document the arrivals and departures of an entire diplomatic corps of envoys from the far western edge of the Tarim Basin, modern Uzbekistan, Pakistan, and Afghanistan.16

The farthest west that Chinese goods traveled in this period was to Palmyra, Syria, where silks with Chinese characters woven into them have been found (the earliest is dated 40 ce).17 They were most likely woven in the imperial workshops of the Han dynasty. There is no evidence of direct contact between China and Rome. The famed account of a Roman envoy presenting gifts to the Chinese emperor in 166 ce probably concerns a merchant claiming to be Roman, rather than a genuine native of Rome.18

As Chinese power waned in the 2nd century ce, the Kushan Empire arose in modern-day north India, Pakistan, and Afghanistan. It never directly ruled any of the Tarim Basin kingdoms, but a group of refugees from the Gandhara region (in modern-day Pakistan and Afghanistan) crossed the Pamir Mountains in the late 200s and settled the small oasis town of Niya, on the southern route around the Taklamakan, where they introduced the Kharoṣṭhī script, different types of wooden documents, and elements of Indic law.19

Nearly one thousand documents from Niya include royal orders, private letters, and records of court disputes offering an unusually vivid glimpse of life in a small Silk Road community between around 240 and 330.20 They portray a subsistence society with a mixed population of locals, Indian migrants, Khotanese, Chinese, war refugees from Khotan, and raiders (called Supis and “hill people”).

Few of the Niya documents mention trade. At one point, when a dispute about the price of silk occurred, local officials urged those involved to wait for Chinese merchants to resolve the matter, an indication that Chinese traders did not live in the valley full time but visited the village on an occasional basis. Two reports of goods stolen from “runaways” (the documents provide no other information about the victims)—various woolen and silk textiles, jewelry (including seven strings of pearls), and a mirror—are the only indications of long-distance trade.21

Between 100 bce and 300 ce, envoys bearing gifts and payments to supply armies constituted most of the Silk Road trade; private trade existed but not on a large scale.

The Sogdians and China, 300–800 ce

Between 300 and 500 Central Asia was divided in multiple political units with no single power in place for long; few historical records in Chinese or other languages survive from this period. But eight letters written in the Sogdian language—found in a discarded mailbag in the early 1900s by Aurel Stein fifty-six miles (90 km) northwest of Dunhuang—reveal that the migration of Sogdians to China and their participation in the long-distance trade across Central Asia had already begun. The Ancient Sogdian letters show much about the organization of the Sogdian trade, the types of business in which the Sogdian traders engaged, and the commodities they traded.22

The Ancient Sogdian Letters

One of the letters provides a crucial clue to the time of their 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 place was burnt and the city destroyed. Luoyang is no more.”23 In 311 the Xiongnu commander Shi Le conquered the city of Luoyang, the capital of the Han dynasty between 25 and 220 ce, and founded the Later Zhao dynasty (319–351), one of the many regional dynasties that ruled the Chinese region between the fall of the Han dynasty in 220 ce and the conquest of north China (the Yellow River region) and south China (the Yangzi and Pearl River valleys) by the Sui dynasty in 589.

The letters document the existence of Sogdian communities in Luoyang, Dunhuang, and Chang’an (modern Xi’an), and three other Chinese cities. The sabao headmen of these communities handled community matters and judged disputes; they also tended fire altars and performed other rituals of Zoroastrianism, which the Sogdians had brought from their homeland.

The author of the second letter is an agent representing his employers, a father and son based in Samarkand, in two towns: the agent commissioned multiple households to weave woolen and linen textiles. In Dunhuang, he also sold musk, a perfume fixative processed from a small ball-shaped gland in front of the penis of the Tibetan musk deer that sprayed an odorous secretion.24

Another letter mentions additional commodities that the Sogdians traded: camphor, pepper, silver, and “white,” most likely ceruse, a cosmetic with a white lead base. Camphor and pepper were Southeast Asian products, which the Sogdians probably purchased at Chinese markets. The letters do not refer to silk (a word possibly meaning something “derived from a silk worm” occurs only once), but it is likely that Chinese silk was a common trade good at the time.

The letters are difficult to read—scholars often publish an article explaining a new interpretation of a single word—and the value of the stater coins and units of weight mentioned in multiple letters is uncertain. The quantities of the various commodities are not large—ranging from 3.3 lb (1.5 kg) to 88 lb (40 kg)—and could have been carried by a single pack animal. A woman stranded in Dunhuang complains that five caravans passed through the oasis on their way to Samarkand in the course of three years, an indicator of how infrequent caravan travel was.

For all the difficulties the letters pose to those who hope to decipher them, they demonstrate conclusively that a private, for-profit trade run by Sogdians existed in northwest China by the early 300s.

Sogdian Migration into China

In the centuries after the Ancient Sogdian letters, the Sogdians continued to move into China. Their rate of migration increased after 500, when the region of Sogdiana experienced a sustained agricultural boom.25 At this time, traffic around the Taklamakan Desert largely shifted to the northern route. One of the largest Sogdian communities where they settled was Turfan, on the northern edge of the Taklamakan Desert. The Sogdians called Turfan “Chinatown” because the king and his subjects followed so many Chinese customs, including writing all documents in Chinese characters, but, in fact, the Gaochang kingdom was an independent oasis state between 500 and 640 that had formally accepted the Western Türks as its overlord.

Although the Sogdians are famous for being traders, the Sogdians living in Turfan pursued other occupations as well: farming the land, serving in the army, running inns, painting, working leather, and selling iron goods. The discovery of several tombs of deceased Sogdian headmen in modern Xi’an, Taiyuan, and other northern cities has provided considerable information about the daily life of Sogdians. Twelve painted reliefs in the tomb of An Jia (also spelled An Qie) in Xi’an show scenes of parties where dancers perform the Sogdian whirl, a dance that became extremely popular all over China, and multiple consultations—most likely visits by envoys—between Sogdians and Türks.26

Many Sogdians practiced the Zoroastrian religion, then dominant in Iran, but some subscribed to Manichean teachings, which were banned in Iran, and others belonged to the Christian Church of the East. Some also embraced Buddhism, as did large numbers of Chinese. Buddhism was the only religion originating outside China that attracted a large Chinese following: Zoroastrianism, Manicheism, and Christianity had mostly non-Chinese adherents.

Sogdian communities existed in almost every major Chinese city, and the migrants introduced many elements of their culture to their Chinese neighbors: crafts, music and dances, and food all came to China from the west.27 Interestingly, when local officials listed the names of all who lived in their districts on household registers, they used the same format for everyone and did not indicate whether they were Chinese or Sogdian.28 The large Sogdian population made Chinese society unusually open to outside influence between 300 and 1000.

A Snapshot of the Silk Road Trade in the Scale-Fee Tax Receipts c. 600

The Gaochang Kingdom officials collected an import tax of less than 1 percent on goods entering the Turfan oasis, which they assessed on the basis of weight.29 One rare set of scale-fee tax receipts, which records thirty-seven transactions over a single year sometime around 600, underlines the importance of Sogdian traders in the Silk Road trade: fully forty-one of forty-eight merchants’ names are distinctively Sogdian (the others came from different oases in the Tarim Basin). The register survives because after it was discarded, it was cut up to make shoe soles for the dead in Turfan. This reuse of paper is one reason that so much information from Turfan survives.

Mentioning many more goods than the Ancient Sogdian letters, this register is the single most informative document about the goods exchanged along the Silk Road. Brass, medicine, copper, turmeric, and raw sugar each appear once. Gold, silver, silk thread, and aromatics (such as spices, incense, or fragrant woods) are listed more frequently, as is ammonium chloride, which was used as an ingredient in dyes, to work leather, and a flux to lower the melting point of metals. The largest quantity of any goods carried by any single trader—aromatics—in a caravan was either 350 or 1050 pounds (160 or 480 kg)—an amount that a few pack animals could easily transport.30

This register does not mention woven silk, which was probably taxed by size or design and not by weight. Nor does it mention paper, which was one of China’s most influential exports. Paper was sufficiently costly that people purchased discarded official documents so that they could write on the blank side; such paper was exported as far as Mt. Mugh, in modern-day Tajikistan, where the locals also used leather and wooden sticks as writing material.31

Tang Rule in the Tarim Basin, 640–755

In 589 the Sui dynasty gained control of the Yellow and Yangzi River valleys, the core areas of the Han dynasty (206 bce220 ce), and in 618 a Sui general overthrew the Sui emperor and founded the Tang dynasty (618–907 ce). These developments directly affected the Silk Road region because the second Tang emperor, Tang Taizong (r. 626–649) dispatched troops to Central Asia to fight the Türks, who had formed alliances with some twenty-four oasis states in the Tarim Basin (see the article “The Kök Türk Empires”). In 640 Turfan, still under the rule of the Gaochang Kingdom, fell to the Tang forces, and in the following years Kucha, Khotan, Yanqi, and Kashgar did so as well.

The Tang created a new political unit, the Anxi Protectorate, over which it maintained intermittent control until 755. The Anxi Protectorate, which reached as far west as the modern town of Tokmak on the western edge of Lake Issyk-kul, Kyrgyzstan, marked the farthest extent of Chinese control in the west until the 1750s, when the Qing empire gained control of even more territory in modern-day Xinjiang (see the article “The Expansion of the Qing Dynasty of China and the Zunghar Mongol State”).

When under Tang-dynasty rule, the different oasis kingdoms of the Tarim Basin were subject to the same laws and regulations as prefectures in central China. Because of a chronic shortage of metal, only about 10 percent of the empire’s money supply took the form of coins: other forms of money included grain in fixed amounts and bolts of silk measuring 21 inches (54 cm) wide and 13 yards (12 m) long (just enough to make two suits with trousers and jackets).32 Starting in 200 ce, when the shortage of coins first occurred, various Chinese dynasties had used silk as currency, and silk played this role in various Central Asian economies as well. Silk’s main importance to the Silk Road economy was not as a commodity but as a currency.

A price list compiled by local officials in 743 of over 350 goods for sale in the Turfan market indicates which prices were given in silk and which in coins. The officials recorded three prices (high, medium, and low) for each item (it is uncertain whether the officials set the prices or were simply recording existing prices). In accordance with Tang law, the most expensive items were priced in silk; a Türk horse cost 20/18/16 bolts of silk, and a Persian camel, 33/30/27 bolts. Such items were few; only five of 350 prices are given in bolts of silk. All the others are given in coins, but it’s very possible that coin shortages prompted purchasers to substitute silk or grain for coins in actual transactions.

Many of the items for sale in the market, including vegetables and grains, were local. The highest concentration of imported items occurs in the list of fragrant woods and rare minerals; the prices of items such as malachite, azurite, and cinnabar were given in coins per ounce. This market register confirms that most of the Silk Road trade was local, and only a low number of goods, small in size and light in weight, were traded across long distances.33

The Tang government collected taxes in bolts of silk in central China and shipped the silk to Central Asia to finance its military presence. The payments were enormous. Each year in the 730s and 740s the central government sent a total of nine hundred thousand bolts of silk to four separate military commands in Central Asia.34 One record from Dunhuang gives the exact breakdown of two shipments sent in 745 to a garrison 435 miles (700 km) east of Dunhuang that totaled fifteen thousand bolts.35 Army officials used the silk to purchase supplies for the troops and to pay the soldiers, who then spent their pay, in the form of silk, in the markets. The Silk Road trade boomed in these years because of the high level of military expenditures.

But it came to a crashing halt in 755 when An Lushan, a general born to a Sogdian father and a Türk mother, led a massive uprising against the dynasty, which forced the Tang emperor to recall all of the troops from Central Asia to the capital. It took seven years for the Tang suppress the rebellion. The dynasty was forced to hire mercenaries from a new power in Central Asia—the Uyghurs (also spelled Uighurs)—to do so (see the article “The Uyghur Empire (744–840)”). Starting in 763, armies from another ascendant power—the Tibetans—looted the capital of Chang’an for two weeks every year, and the dynasty was powerless to stop them, an indication that the Uyghurs and the Tibetans had replaced the Tang as the major powers in the Tarim Basin (see the article “The Medieval Uyghurs of the 8th through 14th Centuries”).

Transoxiana also experienced a major change of leadership with the expansion of the Umayyad caliphate (661–750) to the west. Caliphate armies defeated the Sasanian empire of Iran in 651 and moved into Central Asia, taking Samarkand in 712. The newly conquered territory in Transoxiana shifted to face east; the spread of Arabic and the annual hajj pilgrimage further reinforced the region’s ties to Baghdad.

After the Tang Withdrawal from Central Asia, 800–1200

The withdrawal of the Chinese from Central Asia had both immediate and long-term effects. In the short run, some units of the Tang army were stranded and continued to live in the garrisons where they were stationed, but eventually they assimilated into the local populations or returned to central China. Central Asia broke into smaller political units, many centered on a single oasis, very few of which ever conquered their neighbors.

The Widespread Shortage of Coins, 800–1000

The Tang withdrawal directly affected the Central Asian economy. An insufficient supply of coins forced the Tang government since its founding to recognize grain and silk as currencies, and the shortage of coins became even more pronounced after 755.36 After 800 no coins circulated in eastern Central Asia for two full centuries, as contracts in multiple languages including Chinese and Tibetan indicate. (In western Central Asia, in contrast, the Samanid rulers minted large quantities of silver coins known for their purity, which have been recovered at many sites in Northern and Eastern Europe.)

The main repository of materials from this period is cave no. 17 at the Mogao cave complex outside Dunhuang, which had over five hundred caves. Cave no. 17 contained around forty thousand documents, half in Chinese, half in Tibetan, with smaller numbers of materials in other languages, such as Khotanese, Sanskrit, Sogdian, and Uyghur. These documents largely date to the period when Dunhuang was under Tibetan control (786–848) or under the rule of one of two Chinese families, first the Zhangs (848–913) and then the Caos (914–1002).37 Many of the documents concern Buddhism, the primary religion in the area, home to fourteen monasteries, but individual texts cover other topics, such as medicine, literature, and other religions including Manicheism, Zoroastrianism, and the Christian Church of the East.38

The Tibetan dynasty that conquered Dunhuang in 786, the Yarlung dynasty, had first unified Tibet in 617. After the supply of coins dried up after 800, the residents of Dunhuang used fixed measures of grain and bolts of cloth as currency, as contemporary contracts show. The visits of envoys to various rulers’ courts in India, China, Tibet, and Central Asia occurred despite the coin dearth.39 Buddhist monks also continued to make pilgrimages, financing their trips with grain and cloth. In 842 the confederation of peoples supporting the Yarlung dynasty imploded, and Tibetan control over Dunhuang weakened, allowing Zhang Yichao, the leader of the Chinese Returning-to-Righteousness army, to take control of the oasis in 848. Zhang declared his allegiance to the Tang emperor, but he was in fact fully independent.

Most of the people living in Dunhuang at this time worked in the fields, but one document records the movements of a merchant who traveled along a 150 mile (250 km) circuit, dropping off up to twenty pieces of white cloth to local households, and then returning to pick them up after they had been dyed red. This is a classic peddler’s itinerary that covered a limited route, dealt in locally produced goods, and—in the absence of coins—exchanged one good for another.40

Despite the shortage of coins, envoys continued to travel among various courts. A register of provisions and beer granted to the envoys who visited Dunhuang, most likely in 964, lists the names of fifty-one different emissaries who came to Dunhuang in the space of just seven months: one from the Song dynasty, which had taken power in China in 960, fourteen from Tibet, twenty-four from two Uyghur kaghanates at Ganzhou and Turfan, and eleven from Khotan, which was located in the southwest corner of the Tarim Basin. The rulers of Dunhuang hosted individuals as well: scribes, workmen, artists, monks, and the crown prince of Khotan, a list that gives a sense of who was traveling along the Silk Road in the 900s.

The crown princes of Khotan visited Dunhuang often because the ruler of Khotan had married the daughter of the ruler of Dunhuang sometime before 936, and following that the Khotanese maintained a residence for the crown prince and his entourage in Dunhuang. A group of seven Khotanese princes set off on a trip to China, sometime in the 10th century, carrying saddles, harnesses, and other horse tack in addition to 800 pounds (360 kg) of jade. Khotan was a large wealthy oasis particularly known for its fine green jade, which it had been exporting to central China since at least 1200 bce.

It took the group eighteen days to travel the 950 miles to Dunhuang on horseback. An unusually detailed group of documents in Khotanese from the Dunhuang documents cave reveals that the Dunhuang ruler refused to permit the princes to continue because he was concerned about their safety; he did allow three monks to proceed, because they were not carrying any gifts. Everyone in the delegation—the princes, higher- and lower-ranking envoys, monks, lay people—engaged in impromptu exchanges of locally produced goods for other locally produced goods. Sometimes they paid with a bolt of silk, but they also traded an antelope skin or a live sheep. This type of exchange typified Silk Road exchanges in the 9th and 10th centuries.

In 970 a Qarakhanid army attempted to invade Khotan. Originating in the steppe grasslands to the north, the Qarakhanids had established a base in Kashgar, some 300 miles (500 km) northwest of Khotan. Sometime before 955 the Qarakhanid leader Satuq Bughra Khan converted to Islam, which marked the farthest point east that Islam had spread by that point. Initially, the Khotanese repelled the attack and captured Kashgar, but the Qarakhanids regrouped and conquered Khotan sometime before 1006, when they launched a major campaign toward the west.

Eventually, the Qarakhanids gained control of all the territory stretching from the Aral Sea in the west to Khotan in the east. They engaged in diplomacy by exchanging envoys just as earlier Silk Road rulers had, and they also sent envoys to the Islamic powers to the west such as the Ghaznavids in Afghanistan as well as to the Kitan peoples who controlled portions of northern China and the eastern section of the Eurasian grasslands.

The Qarakhanids were the primary trading partner of the Kitan peoples, who founded the Liao dynasty (907–1125 ce).41 The main goods traveling from the Islamic world to the Liao realm—many carried by envoys as gifts to local rulers—were horn, frankincense, wrought iron weapons, treated hides, three types of silk, glass, ammonium chloride, and gems such as jade, pearls, amber, and agate.42 Many of these items had been traded in earlier periods as well, and almost all, with the exception of perishable hides and textiles, have been found in Liao-dynasty tombs. The Liao paid for them with silver and Chinese silks (both of which they received in large quantities each year after signing the Treaty of Chanyuan with the Song dynasty in 1005).

Under the Qarakhanids, the rulers of other Tarim Basin oases did not convert to Islam. Initially, the Uyghur rulers of Kucha and Turfan patronized both Manichaeism and Buddhism and opted for Buddhism sometime after 1000. Dunhuang remained Buddhist. When the Liao dynasty fell in 1125, a group of Liao leaders migrated to Xinjiang; known as the Western Liao, they patronized both Buddhism and the Christian Church of the East.

In 1211 a leader named Kuchluk assumed leadership of the Western Liao (see also the article “The Qara Khitai”). A convert to Buddhism, Kuchluk was completely opposed to Islam and forced the residents of Kashgar and Khotan to give up Islam and to choose between Buddhism and Christianity. He was the last Tarim Basin ruler to ban Islam. In 1218 Chinggis Khan defeated Kuchluk’s forces and rescinded the ban on Islam. Although Chinggis did not support Islam, many later Mongol rulers did, and the faith eventually became the dominant religion of the Tarim Basin.

The Mongol conquest brought the classical period of the Silk Road to an end, even as it initiated a new series of exchanges. The exchange of gifts carried by envoys from one court to another was a key component of the Silk Road trade from the 1st century bce to the 12th century ce. A more commercial trade arose alongside the envoy trade because so many of the emissaries traded goods privately. Whenever the major neighboring powers stationed armies in Central Asia, they had to provide them with grain and other provisions, and these shipments contributed to the overall volume of trade. The trade peaked during the 7th and 8th centuries when the Tang central government ruled the Tarim Basin directly. When the Tang armies withdrew from the region after 755, the commercial trade subsided, leaving envoys as the main travelers along the Silk Road.

Discussion of the Literature

In 1877 Ferdinand von Richthofen (1833–1905) articulated his conception of the Silk Road(s) (Seidenstrasse[n]), which he saw as an east-west artery along which Roman coins traveled east and Chinese silk traveled west.43 His reading of different classical sources, in particular Pliny the Elder (23–79) railing against Roman women wearing diaphanous silks, underpinned this supposition, which 150 years of subsequent research have cast into question.

No one has been able to identify with certainty the source of the silks Pliny complained about. They may have come from China, but silk was also produced in India and on the Greek island of Cos.44 Another objection: no coins minted in Rome have ever been found in China. The earliest coins in China from the Roman empire are Byzantine coins dating between 409 and 450 and buried after 500. Many more silver coins minted by the Sasanian Empire of Iran, and copies of them, have been found in Xinjiang, the earliest dating to the 300s, and they came into wider use after 500.45 Finally, some art historical evidence of the exchange of motifs between Rome and China survives from after 200 ce, but those motifs moved from the eastern edge of the Roman empire in modern Syria to the western edge of the Chinese empire in Xinjiang, not from Rome to the Han-dynasty capital at Chang’an.46

The term Silk Road poses still other problems. No one in Central Asia used it in the past. Instead they simply said the road to Turfan or whatever the next stopping point was. Silk was not the most important commodity traded along it, and when silk was important, it was largely as a medium of exchange used by the Tang authorities.

The focus on east-west routes across Central Asia obscures the significance of north-south routes linking Central Asia with the Eurasian grasslands to the north, where the trade most likely originated. It may seem that everyone agrees on the basic definition of the Silk Road—the route through Eurasia that linked east and west—but close scrutiny of Silk Road scholarship shows wide disagreement about what individual authors mean by east and west. Some focus on China’s links to the Tarim Basin or the ties between Transoxiana and Iran; no one has found evidence linking China directly with Rome as von Richthofen proposed (see also the articles “The Early Silk Road(s)” and “The Concept of the Silk Road in the 19th and 20th Centuries”).47

Von Richthofen assumed that the volume of the trade was significant, and many subsequent analysts have reasoned that any mention in the sources of trade through Central Asia constitutes evidence of a long-distance, for-profit, high-volume trade.48 But the evidence that survives—which is admittedly scarce—shows the continuing importance of official exchanges of envoys who conducted private exchanges on the side. Shipments to supply the Tang armies were far larger than recorded individual transactions. The quantities in almost all documented commercial transactions are not large; a few pack animals could transport them easily.

The main topic for future research is further probing the extent and nature of the overland trade through Central Asia. Comparative studies with other trade routes may offer helpful comparisons.49 Future excavations may unearth documents with new information; the main source of documents in the 21st century has been Khotan. Although many of the documents found there so far concern the provisioning and staffing of military guardposts, a letter from a stranded Jewish trader writing in Judeo-Persian (the New Persian language written in Hebrew script) points to a diaspora of Jews in Khotan, as does an earlier letter from the same merchant found in 1901.50

Primary Sources

Transmitted Sources

Historically Central Asia was surrounded by literate neighbors: the Parthian and Sasanian empires of Iran, the Byzantine Empire, the Abbasid Caliphate, and various China-based states, each with its own history-writing traditions. Because most of the histories that survive from these societies portray the past from the ruler’s vantage point in the capital, they say less about Central Asia, but they are important because comparable sources do not survive from within the region.51

Starting with Sima Qian’s Record of the Grand Historian (Shiji), written after 100 bce, each of China’s dynastic histories contains a section about neighboring countries that gives a capsule history of each place, its geographic location, its population, a list of the gift-bearing missions sent to the Chinese emperor, and a description of the inhabitants.52 With few exceptions they have not been translated into English, but they provide an excellent starting point for the study of any Silk Road oasis.53

In contrast, the Arabic-language chronicle of al-Tabari (839–923 ce) has been translated in its entirety twice, once in the 19th century, and again between 1987 and 1997.54 Its detailed description of the conquest of Samarkand provides crucial information that allowed historians to delineate the events leading to the burial of the Mount Mugh letters in the early 700s.55 The Byzantine source, The History of Menander the Guardsman, illuminates the exchange of envoys and trade among the Byzantines, the Türks, the Iranians, and the Sogdians.56

Archaeological Sources

Foreign explorers were the first to excavate in Central Asia, starting in the 1890s. Aurel Stein is most famous for removing thousands of documents from cave 17 at Dunhuang, but he dug at many sites throughout Central Asia, including Niya, and many of the most important early discoveries were his.57

Following the Russian revolution of 1917 and the Chinese revolution of 1949, different archaeological traditions arose in Central Asia: in Transoxiana Soviet-style archaeology prevailed even after the break-up of the Soviet Union, and Chinese-style archaeology was the norm throughout Xinjiang.

Soviet researchers excavated archaeological preserves, some established quite early, where they worked to identify the major structures of important cities like Samarkand. The longest-running and most influential Soviet-style excavation was of Panjikent, a small city west of Samarkand in modern-day Tajikistan. There, archaeologists have uncovered more than half of the entire city, revealing a Zoroastrian fire altar, houses with elaborately painted walls, and shops.58 Other sites surrounding Samarkand, such as Kafir Kala 7.5 miles (12 km) south of the city, have produced evidence about Sogdian communities in the 500s.59

Although the Chinese have excavated sections of some cities, including the markets of Chang’an, the vast majority of excavations are of tombs. The burial goods from tombs make it possible to trace trade routes from where certain minerals were mined to wherever the tomb with that good is located, and the carved reliefs on stone of Sogdian daily life have been unusually illuminating about the urban existence of Sogdians residing in China.

Excavated Documents

The most unusual source for Silk Road history is the documents found underground. Some of these, like the Niya judgments and police reports of stolen goods, were deliberately buried underground for archival storage. Some, like the Ancient Sogdian letters, were lost and recovered centuries later by archaeologists. And still others survive because they were recycled to make items of funeral clothing such as hats and shirts for the dead; these include contracts, official orders, and the scale-tax-fee receipts that reveal so much about the Silk Road trade.

Figure 2. Recycled paper goods from the Astana Graveyard.

Source: Aurel Stein, Innermost Asia (Oxford: Clarendon Press, 1928), plate XCIII detail.

Figure 2 shows some of the goods made with recycled paper that Aurel Stein excavated from tombs at Turfan: a hat decorated with flowers (top), a rolled up flag (right), a string of coins (bottom center), and shoes (bottom left and right). The arrow marks the Chinese characters still visible on the inside of the shoe. Documents preserved in this way have provided many insights into how commerce was conducted on the Silk Road.

Links to Digital Materials

Further Reading

  • Brosseder, Ursula. “Dynamics of Communication and Exchange along the ‘Steppe Highway’ in the Centuries around the Turn of the Era.” In The Complexity of Interaction along the Eurasian Steppe Zone in the First Millennium AD. Edited by Jan Bemmann, et al., 199–332. Bonn: Vor- und Frühgeschichtliche Archäologie, Rheinische Friedrich-Wilhelms-Universität Bonn, 2015.
  • Chin, Tamara. “The Invention of the Silk Road, 1877.” Critical Inquiry 40, no. 1 (2013): 194–219.
  • Giele, Enno. “Evidence for the Xiongnu in Chinese-Language Documents from the Han Period.” In Xiongnu Archaeology: Multidisciplinary Perspectives of the First Steppe Empire in Inner Asia. Edited by Ursula Brosseder and Bryan K. Miller, 1–27. Bonn: Vor- und Frühgeschichtliche Archäologie, Rheinische Friedrich-Wilhelms-Universität Bonn, 2011.
  • Hansen, Valerie. The Silk Road: A New History with Documents. New York: Oxford University Press, 2017.
  • Høisæter, Tomas Larsen. “Polities and Nomads: The Emergence of the Silk Road Exchange in the Tarim Basin Region during Late Prehistory (2000–400 bce).” Bulletin of the School of Oriental and African Studies 80, no. 2 (2017): 339–363.
  • Hyun Jin Kim. The Huns. New York: Routledge, 2015.
  • La Vaissière, Etienne de. Sogdian Traders: A History. Translated by James Ward. Boston: Brill, 2005.
  • Miller, Bryan K., and Ursula Brosseder. “Global Dynamics in Local Processes of Iron Age Inner Asia.” In The Routledge Handbook of Archaeology and Globalization. Edited by Tamar Hodos, 470–487. New York: Routledge, 2017.
  • Wang, Helen. Money on the Silk Road: The Evidence from Eastern Central Asia to c. AD 800. London: British Museum Press, 2004.
  • Wen, Xin. “The Kingly Exchange: The Silk Road and the East Eurasian World in the Age of Fragmentation.” PhD diss., Harvard University, 2017.
  • Whitfield, Susan. Silk Roads: Peoples, Cultures, Landscapes. Berkeley: University of California Press, 2019.
  • Yoeli-Tlalim, Ronit. ReOrienting Histories of Medicine: Encounters along the Silk Roads. London: Bloomsbury, 2020.
  • Yu, Taishan. LiangHan WeiJin Nanbeichao yu Xiyu guanxi shi yanjiu. Beijing: Zhonghua Shuju, 1995.
  • Zhang, Guangda, and Rong Xinjiang. “A Concise History of the Turfan Oasis and Its Exploration.” Asia Major 3rd series, 11, no. 2 (1998): 13–36.
  • Zhang, Zhan. “Secular Khotanese Documents and the Administrative System in Khotan.” Bulletin of the Asia Institute 28 (2018): 57–98.