The Silk Road: A New History

by Valerie Hansen

  In thinking about the removal of the Dunhuang documents to other countries, we should resist the urge to judge Stein by modern standards rather than by those of his own time. Today many observers support the return of the Elgin Marbles to Greece. Consider, though, that Stein and the other explorers were operating at the height of imperialism before World War I. The European powers and Japan all sent teams to Xinjiang to excavate, and few contemporary observers voiced scruples. Among the few who did were Albert Grünwedel of Germany and the Russian scholar S. F. Ol’denburg, who both criticized Le Coq and others for removing wall paintings from their original sites.16

  Foreign visitors had legitimate grounds for concluding that library cave materials were safer if taken away from Dunhuang. The caves at Dunhuang had suffered damage during the Muslim uprisings of 1862–77, and Stein was acutely aware of how restive the local population was.17 Only one month after Stein left, in June 1907, the town exploded in a riot over grain prices.

  Chinese views of Stein’s conduct have softened over the years. During the Cultural Revolution, he was a thief, pure and simple. Even in the mid-1980s, when I was in graduate school, a Chinese classmate bristled with rage when our professor said that, if he were a Dunhuang document, he would prefer to have been taken to either Paris or London, because the conditions of preservation were so much better than those in Beijing. In 1998 the full Chinese translation of Serindia, including Stein’s detailed account of the negotiations at Dunhuang, appeared with a thoughtful preface by a prominent Chinese archeologist, Meng Fanren. Serindia, with its team of authors who translated the different materials Stein found, represented “the very highest level of research in this field before the 1920s,” yet Stein’s individual actions, Meng concluded, constituted “plundering behavior which deserves severe and justified criticism.”18

  Advances in publishing have made the Dunhuang documents held in foreign countries increasingly available to Chinese scholars: first came the distribution of microfilm in the late 1970s, then the publication of multivolume sets with clearly legible photographs in the 1990s, and now the ongoing loading of photographs onto the website of the International Dunhuang Project based in London.19 In 2005 Professor Rong Xinjiang of Peking University, one of China’s leading historians of the Tang dynasty, published an article in China’s leading academic history journal, Lishi Yanjiu (Historical research), in which he contrasted the actions of Stein, who did not tell Chinese scholars of his finds, with those of Pelliot, who gave Chinese colleagues photographs of the materials he had purchased and shipped to Paris. Professor Rong reminded his readers of one indisputable fact: for all their calls to protect the Dunhuang documents, no early twentieth-century Chinese scholar ever left the comfort of his own home. Not one followed Stein and Pelliot’s example and personally visited the site of Dunhuang. The wholesale removal of the Dunhuang documents was the result.20

  Still, even by the standards of his own day, Stein’s actions still seem deceitful. He claimed to be a devotee of Xuanzang. He knowingly paid a fraction of the market price for the scrolls and paintings. He took extreme measures to maintain secrecy, doing everything at night, and telling a minimal number of people what he was doing. One cannot help wondering why Stein writes so openly about being so clandestine.

  Although Stein does not specifically mention William Matthew Flinders Petrie in his discussion of Dunhuang, his other writings frequently acknowledge his influence.21 Petrie, the leading British archeological excavator in Egypt, came to visit Stein in 1902 after Stein had returned from the First Expedition. In the preface to Ancient Khotan, Stein called Petrie an “archeological explorer of unequalled experience.”22 In 1904 Petrie published Methods & Aims in Archaeology, a step-by-step guide through all stages of an excavation: equipping an expedition, digging in the field, and the publication of results. Having excavated in Egypt, Petrie instructs archeologists how to work in poorer countries, explaining how to use small amounts of money to ensure that workmen submit their small portable finds to the excavator rather than sell them on their own: “Nothing can ensure better care than paying for it,” he concludes. Petrie also advised his readers to publish their results in two volumes, a cheaper version with fewer plates for “students and the general public” and “a magnificent edition for libraries, book-collectors, and rich amateurs.” Stein closely followed his advice; even the layout and typeface of his books replicated those of Petrie.23

  In a prescient chapter entitled “Ethics of Archaeology,” Petrie explains that once archeologists have finished digging at a site, nothing remains for future generations to find. The archeologist can place objects in museums, but these are bound to fold, ultimately leaving published books as the only record. “The only test of right is the procuring of the greatest amount of knowledge now and in future,” he concludes. Petrie disparages the governments that frequently pass regulations that keep foreign archeologists from digging but allow the “ignorant peasant” to “dig and destroy” as he pleases. Stein’s preface to Ancient Khotan quotes Petrie’s injunction that investigators had to “work with the fullest care and detail in recording, to publish everything fully.”24 Stein’s frankness about evading Chinese government regulations and negotiating with Daoist Wang perfectly embodies the pragmatic spirit of Petrie’s manual: like his mentor, Stein sought to procure “the greatest amount of knowledge now and in future” and had no scruples about shipping documents and artifacts out of China.

  In line with Petrie’s instructions, Stein tried to reconstruct the original context of cave 17 as best as he could. The layer-by-layer arrangement of materials in the library cave indicated that it was not a randomly preserved collection of documents and paintings. Clearly an individual or a group had deliberately placed this particular collection of materials in the cave. But why? The presence of scraps of paper prompted Stein to assume that the cave was a repository for wastepaper.

  Professor Rong has carefully cross-checked Stein’s account against both Chinese sources and Paul Pelliot’s account. Even though Stein did not have the opportunity to examine the library cave carefully, his writings remain the most detailed description of the cave, whose original arrangement was irretrievably destroyed when Daoist Wang opened the cave first for Stein and the following year for Pelliot. Challenging the wastepaper hypothesis, Professor Rong proposes a different explanation for the placement of the documents in the library cave.25

  In many ways, the term “library cave” that Stein used is misleading. The library cave is not a separate cave; it is a small storeroom, measuring just under 10 feet (2.90 m) square and under 9 feet (2.66 m) high. It was originally hidden from view, but Daoist Wang discovered it when he tapped on the wall of cave 16 and realized that an empty space lay behind it. Daoist Wang broke through the wall and found a storeroom.

  The library cave originated as a memorial to the powerful monk Hongbian, who was appointed controller-in-chief of Buddhist monks by the Tang emperor in 851. After he died in 862 his disciples placed items associated with him in the cave and came to visit the cave to pay their respects.26 Then sometime early in the tenth century, the monks began to use the cave as a storeroom for manuscripts.27 When he cleared out the storeroom sometime around 1900, Daoist Wang removed the statue. Subsequently, the Dunhuang Research Institute returned it to its original position, where it is today.

  THE NORTHERN WALL OF THE LIBRARY CAVE

  The walls of the hidden library cave were all blank except for the northern wall, which shows two acolytes and two trees framing the monk Hongbian’s statue on the center platform. To the right stands a nun holding a round fan decorated with phoenixes; to the left, a lay woman in a man’s robe clasps a walking stick. This painting dates to the period when the small cave was used as a funerary cave for the monk Hongbian.

  Many of the texts in the library cave have labels giving the names of the monasteries that owned them. During the tenth century Dunhuang was a Buddhist center with some fifteen monasteries, and the Three Realms (Sanjie) Mona
stery was one of the smaller ones there.28 Because its name appears more frequently on the texts found in the library, the Three Realms Monastery likely used the cave to store its manuscripts.

  An important clue to the purpose of the library cave comes in a preface to a Buddhist text written by a monk named Daozhen (active 934–87). He explained why he collected materials for the monastery library: “Having seen that among the contents of the storehouse of his temple the sets of scriptures and commentaries were incomplete, thereupon bowed his forehead to the ground and, with devout sincerity, took an oath and made prayers: I [Daozhen] will go carefully through the cartons and storehouses of all the families, seeking after old and decayed scriptural texts. I will gather them in the monastery, repair and patch them from beginning to end, and pass them down to other ages.”29 Sometime after 987, when Daozhen died, other monk-librarians continued to collect manuscripts for the Three Realms Monastery collection.

  The Dunhuang monasteries all maintained lists of the texts they hoped to acquire, an indication that they were still collecting texts and paintings in the years before the closing of the cave. The earliest text in the cave, a Buddhist text, is dated 405, the latest to 1002.30 The scrolls in the library cave included far more than simply Buddhist texts.31

  Because paper was expensive at Dunhuang, students at monastery schools practiced writing characters in the unused margins or on the reverse side of Buddhist texts that had been discarded. The monastic schools trained students to read and write; some became monks, some did not.32 These students began just as students of Chinese today do, by copying out individual characters over and over, gradually progressing to study more difficult texts. The Dunhuang manuscripts contain many errors, since the students were not all advanced; often teachers crossed out a student’s mistaken character and inserted the correct one next to it. Students copied all kinds of materials in their quest for literacy: Buddhist texts, of course, but also contracts, model letters, literary exercises (such as a dialogue between Water and Tea), and long narratives called “transformation texts.”33

  The most famous text from the library cave is the Diamond Sutra, which was not copied by hand but printed with woodblocks. The Chinese first developed this method in the early eighth century when they realized that they could take a sheet of paper with characters on it, glue it face-down on a block of soft wood, cut out the wood around the characters to form a reverse image, and then print the positive image using that block. The Diamond Sutra from Dunhuang consists of seven woodblock-printed sheets that have been glued together (shown on page 236).

  The dedication explains that a Buddhist devotee, on behalf of his parents, commissioned the text for the benefit of all. Such an action generated merit for his parents and for himself. The Diamond Sutra bears the date of the fifteenth day of the fourth month in 868. Cave 17 contained fragments of books printed on woodblocks before 868, including an almanac from 834, but the Diamond Sutra is the earliest complete printed book in the world.34 Scholars recognize that Dunhuang, unlike Sichuan, was not a printing center; the vast majority of the texts in the cave were manuscripts copied by hand.

  CALL NUMBERS FOR A BUDDHIST LIBRARY

  The Chinese scrolls in cave 17 were divided into bundles of a dozen or so and then each covered with a separate wrapper. Unusually, one label on the upper right is still visible; it gives the name of the Buddhist text along with its number from the Thousand-Character Classic, which served as the equivalent of a modern library call number. Courtesy of the Board of the British Library.

  The monk-librarians at Dunhuang used a sophisticated method to catalog Buddhist texts. They consulted catalogs of large Buddhist monastic libraries in the Tang capital of Chang’an that divided all Buddhist texts into categories such as sermons, regulations, or histories.35 A primer called The Thousand-Character Classic consists of one thousand characters with no repetitions, making it a kind of Chinese alphabet. The Buddhist librarians assigned a single character to each Buddhist text and then grouped many of the Chinese-language scrolls into what Stein called “regular library bundles.”

  One thousand and fifty of these bundles contained twelve or so scrolls in Chinese; in addition, there were eighty packets and eleven large texts written on leaf-shaped pages called pothi in Tibetan, the language introduced in 786.36 In that year, the Tibetans helped the Tang dynasty to suppress a rebellion, but, when the Tang failed to make the promised payment, the Tibetans conquered Dunhuang. The bundles all originally had outside wrappers, but because none of the people who first saw them—Daoist Wang, Jiang Xiaowan, Stein, or Pelliot—grasped their significance, only a few survive.

  In addition to the regular bundles of Chinese- and Tibetan-language materials, the custodians of the library cave also included another type of bundle that Stein called either “miscellaneous” or “mixed” bundles.37 These comprised Buddhist texts written on palm-leaf-shaped pothi slips or scrolls in Sanskrit, Khotanese, Tibetan, Uighur, and Sogdian. Some of these were complete copies of Buddhist texts; others were fragmentary. They also placed paintings (almost always of Buddhist deities), scraps of paintings, damaged ends of sutra rolls, and individual sheets of paper in the cave. In addition, they saved scraps of material that might be of use in future repairs to Buddhist manuscripts. While those in charge of larger monastic libraries might have discarded these scraps, the small size of the Three Realms Monastery collection made the custodians more cautious. Since any unknown materials might come in handy in the future, everything with writing on it was worth saving. For these reasons, cave 17 contains a huge variety of material. Unlike the recycled shoe soles from Turfan, this is not a random collection of documents; everything in cave 17 had some kind of connection to Buddhism, either because it was copied on the back of a Buddhist text or because students in monastic schools wrote it.

  Sanskrit, Sogdian, Tibetan, Uighur, Khotanese: the various languages of the materials in the cave fully justify Stein’s label, “The Polyglot Library.”38 In some cases, a single sheet of paper reveals the existence of a religious community, or perhaps a lone traveler, about which nothing else is known. One sheet of paper from cave 17 is an eighteen-line prayer written in Hebrew; each line starts with a letter in the Hebrew alphabet followed by a selection from Psalms, shown in color plate 12. Folded over several times, it served as a protective talisman, which was sewn into a small pouch and worn by the bearer, most likely around the neck.39 Perhaps a Jewish traveler came to Dunhuang; equally likely, someone purchased the talisman (the shape of the letters suggests it was made in Babylon) and brought it to Dunhuang. Similarly, only two sheets of paper from the cave document the existence of a Sogdian-speaking Zoroastrian community. One sheet bears lines from the Avesta, the earliest Zoroastrian text; another shows two different Zoroastrian female deities facing each other.40

  The Iranian religion of Zoroastrianism is one of the “three different teachings,” a term Chinese scholars use to denote the two Iranian religions of Zoroastrianism and Manichaeism as well as the teachings of the Church of the East, based in Syria. All originated outside of China, entered the empire during the Tang dynasty, and did not survive the religious bans of 845. The librarians’ eclecticism makes cave 17 the single most informative repository of primary sources about the different religions of the Silk Road.

  The religious texts in the library cave indicate that people living at Dunhuang tolerated each other’s beliefs to an extraordinary extent. The monks putting aside these texts did not necessarily know the languages in which they were written and probably could not read them. But their willingness to preserve texts written in other languages underlines the cosmopolitanism characteristic of the Silk Road. Even though they lived in a small community of some 30,000 people, they respected other people’s languages, writings, and probably even their right to worship as they liked.41

  The sources from the library cave, like those from Turfan and the Christian monument in Xi’an, are particularly important because they offer the view of the devotees, no
t that of high-ranking clergy or the Chinese authorities, whose views so often shape the historical record about religion. As instructive as they are, the Dunhuang texts do not describe the congregations of these different religions, with the result that we know little of their size. If all the surviving texts from a given religion are in a foreign language, we can surmise that the church did not attract many Chinese converts; on the other hand, Chinese-language translations point to the presence of local converts.

  The survival at Turfan of Manichaean texts in multiple Iranian languages—Parthian, Middle Persian, and Sogdian—and old Turkic, and at Dunhuang in Chinese, have made it possible for scholars to study the teachings of a world religion otherwise known largely through the writings of Saint Augustine, who wrote in his Confessions about being a Manichaean before converting to Christianity.42 The library cave holds a total of three Manichaean texts written in Chinese characters.

  Even though some of the texts are written in Chinese characters, they suggest that most Manichaean believers spoke an Iranian language. The longest of the three, a hymn scroll, uses Chinese characters to phonetically record twenty different Sogdian-language hymns and prayers. Because the text does not translate these hymns, they remain incomprehensible to a native speaker of Chinese. Someone who knew how to speak Sogdian but couldn’t read it—the child of Sogdian migrants to Dunhuang, say—could use these pronunciation guides to sing along with the congregation.43 One hymn in the hymn scroll, “Praise of the World of Light,” appears to be a direct translation of a Parthian text found at Turfan. But the Chinese version equates the World of Light with the Amitabha Buddha’s Western Paradise. The Light World is a “world of perfect bliss,” where “everything is light, and no place is dark; where all buddhas and envoys of Light live” and “everything is clean and pure, eternally happy, calm and quiet, undisturbed and unhindered; one receives happiness and has no worry or affliction.”44 Mani urged his followers to use the terminology of existing religions in seeking converts. This text beautifully illustrates this chameleon strategy by naming Mani as one of the three most important teachers in China along with Buddha and Laozi; in this telling Mani has assumed Confucius’s position.