The Silk Road was not a road and did not principally carry silk. It was a shifting network of overland routes linking northern India, Iran, Central Asia, and China, along which moved horses, glass, paper, spices, slaves, diseases, technologies, and — with unusual success — Buddhism.

For something like a thousand years, from roughly the first century CE to the tenth, the oasis cities of Central Asia were among the most vigorous Buddhist centres anywhere. Almost all of it is gone. What survives is art, ruins, and one of the greatest manuscript hoards ever found.

Gandhara and the Greek inheritance #

The transmission begins in Gandhara, the region around modern Peshawar and the Swat valley in Pakistan, extending into eastern Afghanistan.

Gandhara had been part of the Achaemenid Persian empire, then conquered by Alexander, then ruled by Greek-speaking dynasties for two centuries. The result was a genuinely hybrid culture, and it produced the most consequential single innovation in Buddhist visual history: the image of the Buddha in human form.

Before roughly the first century CE, Buddhist art was aniconic. The Buddha was represented by symbols — an empty throne, a parasol, a pair of footprints, a wheel, the bodhi tree — but not by a figure. Whether this reflects a doctrinal prohibition or simply a convention is debated, and the older confident claims about a “ban” have been questioned.

What is clear is that Buddha figures appear in Gandhara and at Mathura at around the same time, and that the Gandharan version is visibly Hellenistic: wavy hair, a heavy toga-like robe with deeply carved folds, the proportions and modelling of Greco-Roman sculpture. This is the ancestor of every Buddha image in East Asia. The ushnisha, the cranial protuberance signifying wisdom, begins here as a topknot of bound hair.

The Kushan empire, which controlled this region and much of northern India from the first to the third century CE, was the political frame that made the transmission possible. Its most famous ruler, Kanishka, is remembered in Buddhist tradition as a second Ashoka — patron of the Fourth Council in Kashmir, builder of a colossal stupa at Peshawar that later Chinese pilgrims described with astonishment, and issuer of coins bearing the Buddha’s image alongside Iranian, Greek, and Indian deities. Those coins are a fair summary of the religious environment.

The Gandharan manuscripts #

In the late 1990s, a group of birch-bark scrolls acquired by the British Library was identified as Buddhist texts in the Gandhari language and Kharoshthi script, dating to around the first century CE. Further finds followed. These are the oldest Buddhist manuscripts known and among the oldest South Asian manuscripts of any kind.

Their importance is hard to overstate. They give direct access to a Buddhist literature in an Indian vernacular other than Pali, from a school other than the Theravada, several centuries earlier than any previously available manuscript. They include versions of discourses known from the Pali canon, Abhidharma material, and early Mahayana texts, and their study has refined the picture of how the schools related to one another.

The Tarim basin #

East of the Pamirs, the routes skirted the Taklamakan desert, north and south, through a string of oasis kingdoms watered by snowmelt: Khotan, Kashgar, Kucha, Karashahr, Turfan, and, at the Chinese end, Dunhuang.

These were not way-stations but Buddhist cultures in their own right, with their own languages — Khotanese and Tumshuqese (Iranian), Tocharian A and B (a remarkable Indo-European branch with no close relatives), Uighur, Sogdian — their own monasteries, and their own literature.

Khotan, on the southern route, was a major Mahayana centre with a foundation legend connecting it to Ashoka. Khotanese monks and manuscripts were significant in the transmission to China, and the Book of Zambasta is a substantial original Buddhist work in Khotanese.

Kucha, on the northern route, was Sarvastivadin and produced the single most important figure in the history of Chinese Buddhist translation. Kumarajiva (344–413), son of an Indian father and a Kuchean princess, was taken captive when Chinese forces sacked Kucha, held for years in Liangzhou, and eventually installed at Chang’an in 401 as head of a state-sponsored translation bureau with reportedly hundreds of assistants.

Kumarajiva’s translations — the Lotus Sutra, the Diamond Sutra, the Heart Sutra in its most-used form, the Vimalakirti, Nagarjuna’s Mulamadhyamakakarika — became the standard Chinese versions and remain in liturgical use across East Asia sixteen centuries later. He translated for sense and for literary quality rather than word by word, and the fluency of his Chinese is a substantial part of why these particular texts became central.

Sogdians, Iranian-speaking merchants from around Samarkand, were the dominant trading community along the routes and important carriers of religion, transmitting Buddhism alongside Manichaeism, Zoroastrianism, and Nestorian Christianity.

The cave temple complexes #

The routes are marked by rock-cut monasteries and shrines with painted and sculpted interiors, several of them on an extraordinary scale.

Bamiyan in Afghanistan had two colossal standing Buddhas, 38 and 55 metres, carved into a cliff in the sixth century, surrounded by painted caves. They were destroyed by the Taliban in March 2001. Subsequent analysis of the surviving cave paintings identified them as among the earliest known uses of oil-based binding media in painting, predating European oil painting by centuries.

Kizil, near Kucha, has hundreds of caves with murals in a distinctive Central Asian style.

Dunhuang, at the desert’s eastern edge where the routes rejoined before entering China, has the Mogao caves: nearly 500 decorated caves cut over a millennium from the fourth century, with some 45,000 square metres of wall painting. It is the largest surviving body of Buddhist art anywhere.

The Dunhuang library cave #

In 1900 a Daoist caretaker at Mogao, Wang Yuanlu, found a sealed chamber behind a wall. It contained perhaps 50,000 manuscripts and paintings, deposited over centuries and walled up around 1000 CE for reasons that remain unclear — a response to an approaching threat, or the ritual burial of worn-out sacred texts.

The contents are in Chinese, Tibetan, Sanskrit, Khotanese, Uighur, Sogdian, Hebrew, and more. They include Buddhist scripture in quantity, but also contracts, letters, medical texts, divination manuals, Nestorian Christian and Manichaean documents, and a printed copy of the Diamond Sutra dated to 868 — the earliest dated printed book in existence, complete with a frontispiece and a colophon recording that it was made for free distribution.

The Tibetan material from Dunhuang is the primary source for the earliest period of Tibetan Buddhism, predating the later chronicles and frequently contradicting them.

Between 1907 and the 1920s, expeditions led by Aurel Stein, Paul Pelliot, and others removed the great majority of the collection to London, Paris, Beijing, Tokyo, and St Petersburg. The removals were legal under the arrangements of the time and are regarded in China as plunder; the resulting dispersal is now partly addressed by the International Dunhuang Project, which has digitised and reunited the collection online.

Why it ended #

Central Asian Buddhism declined between roughly the ninth and fourteenth centuries. The reasons are regional and cumulative: the collapse of the Tang dynasty’s western reach and the Tibetan empire’s control; the Uighur migration into the Tarim basin; the gradual spread of Islam eastward, with Khotan falling to the Karakhanids around 1006; and the long-term shift of trade to maritime routes, which drained the oasis economies.

By the time European explorers arrived in the nineteenth century, the cities were sand-covered ruins and the languages were extinct. Tocharian and Khotanese are known entirely from manuscripts recovered in this period.

Common questions #

Did Greek art really produce the Buddha image? Gandharan sculpture is unmistakably Hellenistic in technique and idiom, and the chronology places the first Buddha figures in a Hellenised milieu. The older claim that Greeks alone invented the type has been tempered: figures appear at Mathura, in a wholly Indian idiom, at about the same time, and the relationship between the two is debated. Influence is certain; sole authorship is not.

Was there really a “Greco-Buddhism”? There was sustained contact and real syncretism in art and coinage. The Milindapanha, a Pali text, records a dialogue between the monk Nagasena and the Indo-Greek king Menander, who is presented as becoming a Buddhist. How far this reflects genuine philosophical exchange rather than a literary device is argued about, but Menander was a real ruler and his coins are common.

What happened to Bamiyan? The Buddhas were dynamited in 2001. Fragments have been conserved, the cave paintings partially stabilised, and reconstruction has been repeatedly proposed and repeatedly deferred; the site remains a UNESCO World Heritage site in danger.

Can I see the Dunhuang manuscripts? Yes. The International Dunhuang Project provides free high-resolution images of the dispersed collections online, and the Mogao caves themselves can be visited under strict conservation controls.

Sources & further reading #

Richard Foltz, Religions of the Silk Road (Palgrave Macmillan, 2nd ed. 2010) — the best short synthesis. Susan Whitfield, Life Along the Silk Road (University of California Press, 2nd ed. 2015) — reconstructed lives, based closely on the Dunhuang documents. Richard Salomon, Ancient Buddhist Scrolls from Gandhāra (University of Washington Press, 1999) — the introduction to the Gandharan manuscript finds. Erik Zürcher, The Buddhist Conquest of China (Brill, 3rd ed. 2007) — on Kumarajiva and the translation bureaus. Roderick Whitfield, Susan Whitfield & Neville Agnew, Cave Temples of Mogao (Getty Conservation Institute, 2nd ed. 2015).