Written in the language called Gandhari, the texts are adding a new dimension to the Buddhist canons of ancient Sanskrit and the living traditions of Pali, Chinese, and Tibetan. A team of linguists and historians at the University of Washington are reconstructing these texts in an NEH-supported project with the British Library.

What the translations reveal is akin to discovering that the world is round, at least in the world of Buddhist studies. Tradition purports that Buddhist texts form a single line of development from the Buddha: as the religion spread, the Buddha's teachings were heard by disciples, then retold word for word, generation by generation.

"The Gandhari texts indicate that this is not quite accurate," explains Collette Cox, an expert on early Buddhism for the project. "We now have a multiplicity of different versions of the teachings from a very early period, which indicates that there may have been much more acceptance among Buddhists of presenting the teachings in their own way.

"It's like Greek philosophy if we only knew Plato through Socrates, and we didn't know any other Greek philosophers," continues Cox. "This is filling in a lot of gaps and adding complexity, making our tasks as scholars much more difficult, but also making our scholarly work much closer to what may have actually happened historically."

Buddhism originated in India around the fifth century B.C.E., when Siddhartha, the Buddha, developed his belief system. Although Buddhism in India had essentially disappeared by 1200 C.E., varying strains of the faith survive today in Sri Lanka, South Korea, Nepal, and other Asian countries. The big question, says Cox, is "how did it get from point A to point B?"

Scholars have long speculated Gandhara, located in northern Pakistan and eastern Afghanistan, to be the area from which Buddhism flowed into east and central Asia beginning around 100 C.E. Temples and shrines date back to the period and the region was near the Silk Road, providing the opportunity to travel and proselytize. But this theory was unsubstantiated because there was scant written evidence of Buddhists in the area.

"If you look at the map of Asia and study the trade routes and the history and so forth, you see the likelihood that Gandhara was the funneling point from which Buddhism came out of India, went up into central Asia," says Richard Salomon, the project's director and head linguist. "And this time of the earliest manuscripts--first, second century C.E.--is when you first start seeing the presence of Buddhism in China. We think Buddhist monks went to China as missionaries, and probably went through Gandhara and carried manuscripts very similar to ones that we're looking at now."

This long-standing theory is closer to being proven because of these new manuscripts, says Cox. "Even though this hypothesis has been around for a while, we had absolutely no information about what Buddhism was like in that area. Until recently, we had archaeological remains and inscriptions but only one text. It was sort of like a black hole. . . . These documents actually connect Indian Buddhism to East Asian Buddhism."

Similarities are found in titles and content, and some Gandhari words appear in a transformed fashion in Chinese. "Sometimes names are translated literally in the Chinese texts," says Cox, "but at other times they are translated using Chinese characters to mimic the pronunciation of the original language. In the cases of these translations, these transliterations suggest the original texts were written in the Indic dialect of Gandhari."

Unlike religions such as Islam or Judaism, in which language is critical to the scripture, the Buddha said that monks should speak to the people in their own language.

"We think that as Buddhism spread across India and beyond, the missionaries would pick up the local language. Presumably, Buddhist communities would write down the texts in their own language. So we can imagine there were versions in other local dialects that haven't survived," Salomon says.

The only previous Gandhari text was discovered in central Asia in 1892 and published by John Brough in the 1960s. No other manuscripts surfaced until 1994 when the British Library bought a collection of twenty-nine Gandhari scrolls with a mysterious history.

"The British Library purchased the scrolls at auction in the U.K. with no firm information about their provenance," says Salomon, who doubts we will ever know exactly where they were found. He places them in Gaandhara because of several reports from the past two hundred years of the existence of similar documents, and of hundreds of reliquaries inscribed in Gandhari that have been unearthed near Buddhist monasteries in the region.

The manuscripts were buried in spherical clay pots. "Most likely the Buddhist monks dropped the manuscripts in these," says Salomon, "then they'd put a plate or dish on the pot's neck and maybe seal it with wax."

The pots were buried in stupas, large mounds or stone structures that, among other things, hold sacred objects. But there was one major flaw: water seeped into the pots. Documents stored along the edges rotted, while those that were inside, on top, or in the middle were better preserved.

The documents were rolled and folded before they were stored-Salomon likens the ancient scrolls to squashed cigars. Before unrolling a document for translation, British conservators had to moisturize it by placing it in a humidor. The first book published by the project, Ancient Buddhist Scrolls from Gandhara, explains the process: "Due to the inherent fragility of the birch bark with its horizontal striations, each layer of the scroll tended to form a separate fragment of varying length as it was unrolled, but it was possible to preserve in very large measure the original sequence of layers and their texts. These fragments were then encapsulated between layers of glass. . . ."

The resulting manuscripts vary wildly in size and condition, ranging from a legible two-foot by eight-inch section to manuscripts that have broken into hundreds of pieces. The library shoots high-detail photographs of the fragments, then scans them into a computer. "About 80 to 90 percent of what we do is on computer," Salomon estimates. The University of Washington group reconstructs the manuscripts on-screen, moving the pieces around like jigsaw puzzles with decomposed and missing pieces. "Some of the documents are in terrible shape," he says. "I was just looking at one of the fragments published a couple years ago, and there are 105 pieces of it."

The British Library enlisted Salomon, one of only a handful of people familiarwith Gandhari, to translate the scrolls. Apart from the 1892 text, the language was known in fragmented fashion through coins, legal and administrative documents, and inscriptions on vessels that held sacred objects, such the bones of the Buddha.

The reliquaries "often have dedicatory or ritual inscriptions on the outside, and they were usually in a basic formula: So-and-so, the son of so-and-so and wife of so-and-so, dedicates these relics in such-and-such a place," Salomon says. "There's not much variation, so you get a small body of vocabulary." But with the new documents the language has begun to take shape. "It's now much richer and more diverse. There's a huge amount of new vocabulary and different styles of writing."

The writing sometimes bears striking resemblances to other known canons, such as in the first verse of the Rhinoceros Sutra, an early text found in several languages advocating solitary asceticism as a path towards enlightenment. Although the manuscript is fragmented, Salomon and his team have used the Pali text in order to fill in missing words and to provide the final line found in the other canons of Buddhism:

"laying aside violence toward all beings,
not harming even one among them,
benevolent and sympathetic/
with a loving mind,
one should wander alone/
like the rhinoceros."
The materials are classified in three basic categories: sutras, which are teachings of the Buddha; avadanas, or pious legends, which concern the Buddha or his disciples; and scholastic materials which contain debates over doctrine. The Gandhari manuscripts are distinguishable from the other canons in their ordering of material, the multiple hands transcribing the documents, and what seems to be the monks' willingness to alter the texts.

Several avadanas are particular to northwest India and mention local political figures. "From that we might speculate that every region had its own collection of these stories," says Cox. "These characters do not appear in the past life stories of the Pali or Chinese texts, so it suggests we have a local composition. It seems to be an attempt to make them locally relevant."

"The material as a whole contains large amounts of both general Buddhist literature and local materials," says Salomon. "In the long run, it's going to show us the relationship between common, pan-Buddhist tradition and local, individual tradition."

In addition there appears in the Gandhari texts a previously unknown school of the faith. Cox has studied one doctrine in the collection written from a perspective different from any other that exists in doctrinal treatises.

The opponent in this text is familiar. "This opponent is called Sarvastivada, or 'those who claim that everything exists.' If we look at Chinese translations, we have a number of Sarvastivada doctrinal texts that represent the position of the Saravastivadans in which they're arguing against other people. But we don't have any texts in which they're the objects of criticism. Here we have a text that's clearly written from the opponent of the Saravastivadans."

Cox believes that many schools composed doctrinal treatises with their own variations, but only a few remain. "We assumed those were the dominant schools and therefore the dominant position, and that they formed the basis for later Buddhist doctrinal interpretation in India and in East Asia and so forth. This may not be true.

"This provides more proof that there was greater variety in the doctrinal, scholastic world than we knew about previously," says Cox. "In the Christian tradition, the Gnostic texts impact our knowledge of early Christian documents," she continues. "The Gandhari texts provide similar evidence on the Buddhist side, forcing us to examine assumptions based on only the other sets of early Buddhist documents. We're no longer looking at Buddhists as transmitting a single body of material that came to be different over time."

------------------------------

Matthew Summers-Sparks is a writer in Washington, D.C.

The Early Buddhist Manuscript Project, a joint program of the University of Washington and the British Library, has received $317,500 from NEH to transcribe, translate, and publish its collection of Buddhist scrolls. Four volumes have been published and a dictionary of Gandhari is underway.