Foreign Devils on the Silk Road

by Peter Hopkirk

  Next day he began to excavate the ruined structures that had helped to save his life the previous night. The first building to be cleared of sand, a small Buddhist temple, yielded a number of interesting frescoes and painted panels, as well as fragments of further manuscripts. More important, however, were the finds which came to light in the ground-floor rooms of the next building they tackled. Here they extracted from the dry sand a small cache of neatly rolled documents in Chinese. Some of these were badly decayed due to the damp which many years earlier, before the town’s water resources finally dried up, must have seeped up through the mud floor. Others, fortunately, were in a good state of preservation. These were later translated by two of the leading sinologists of the day, Edouard Chavannes and Sir Robert Douglas. Two transpired to be bonds for small personal loans – one money, the other grain – made by one Chien-ying, described as a priest of the Hu-kuo monastery. The names of the borrowers, together with those of their guarantors, are also appended as pledging the whole of their household goods and cattle as surety. Both documents bear the date of the same year: 782.

  As Stein points out, the Chinese designation of the monastery (Hu-kuo, literally ‘country protecting’) and the Chinese names of the superintending priests recorded in a third document ‘leave little doubt as to the nationality of the monkish establishment’, adding: ‘That the population which supported it was not Chinese is plainly indicated by the … names of the borrowers and their sureties.’ To Stein the value of these finds lay in their very triviality. He explains: ‘Unimportant in character and insignificant in size and material, it is highly improbable that these documents should date back to a period preceding by any great length of time the final abandonment of the building.’ All bore dates ranging from 782 to 787, thus suggesting that Dandan-uilik was deserted around the end of the eighth century.

  In the same building that yielded the cache of rolled-up documents, Stein also found three finely painted wooden panels, one of which depicted a man riding a horse and another astride a two-humped Bactrian camel. Its art-historical significance was at once recognised by Stein as he brushed the sand off it. Here indeed was the clearest proof he had yet found that his theories were correct. Not only did the drawing and composition demonstrate the high standard which Silk Road artists of the seventh century had already achieved, but its unmistakable ‘mix’ of Indian, Persian and Chinese influences provided a text-book example of how Serindian art had evolved during its slow passage eastwards.

  Stein describes this exquisite little painting thus: ‘The rider of the horse, whose handsome, youthful face shows an interesting combination of Indian and Chinese features, wears his long black hair tied in a loose knot at the crown.… The feet are cased in high black boots with felt soles, very much like those still worn by men of means in Chinese Turkestan, and are placed in stirrups.… From a girdle hangs a long sword, nearly straight, and of a pattern that appears early in Persia and other Muhammadan countries of the East.’ Referring to the harness and trappings, he adds: ‘We could not have wished for a more accurate picture of that horse millinery which in the eighth century flourished throughout Turkestan as much as it does nowadays.’

  In all Stein excavated a total of fourteen buildings during nearly three weeks at Dandan-uilik. He also made a detailed survey of the site, noting its ghostly orchards and avenues of poplars, their gaunt, splintered trunks half buried in the sand. Here and there he found traces of old irrigation channels ‘evidently constructed after the fashion that still prevails in the country’. He concluded that Dandan-uilik was not abandoned because of any sudden catastrophe, and suggested two possible explanations. Either political troubles had led to neglect of the community’s irrigation system – without which life could not be sustained – or the streams supplying that system had dried up over the years, giving the inhabitants no choice but to leave. All the archaeological evidence, he argued, pointed to a gradual abandonment and nothing had been found to support suggestions, contained in local legends, that this once-flourishing caravan city had met with a Sodom and Gomorrah style fate.

  On January 6, 1901, after paying off his labourers, Stein set off eastwards across the desert with his treasure-laden caravan towards the Keriya river, intending to follow it upstream to the oasis of Keriya. Of Dandan-uilik, which had so triumphantly confirmed what he had long believed about the Taklamakan, he wrote: ‘It was with mixed feelings that I said farewell to the silent sand-dunes amidst which I had worked for the last three weeks. They had yielded up enough to answer most of the questions which arise about the strange ruins they have helped to preserve, and on my many walks across these swelling waves of sand I had grown almost fond of their simple scenery. Dandan-uilik was to lapse once more into that solitude which for a thousand years had probably never been disturbed so long as during my visit.’ To reach the river they had to cross a succession of dawans, or sandhills, some rising as high as a hundred and fifty feet. Finally they reached the river with its ‘glittering ice’, and turned southwards towards the relatively modern oasis of Keriya, where Stein hoped to learn of other ruins in the neighbourhood.

  To the modern archaeologist, who is prepared to devote years of his life to one site, Stein’s rapid progression from one excavation to another would be unthinkable. But in his day, it has to be remembered, scientific archaeology was still in its infancy. Moreover, in this inhospitable region excavations could never be more than lightning raids, their duration limited by the supplies that could be carried. Added to this, Stein was no Schliemann, backed by a huge personal fortune. He had had to wring the small budget required for his expedition from reluctant bureaucrats, and would have to justify himself to them when he got back. He was, furthermore, a civil servant himself, with a job to return to within a stated period, so time was all important. Only by covering as much ground as possible and showing that spectacular finds were to be made in the Taklamakan, could he hope to obtain support for future expeditions.

  The caravan halted for five days at Keriya where Dash, Stein’s terrier, managed to make enemies of the local dog population. ‘We had no little trouble’, wrote Stein, ‘in protecting him from the large village dogs which he persisted in provoking by his self-assertive behaviour.’ The day after their arrival, as the result of local enquiries, Stein heard stories of another ruined city in the desert north of Niya, the next major oasis eastwards along the old Silk Road. This was confirmed by an old man who said that he himself had seen these ancient ruins, some ten years before, half buried in the sand.

  On January 18, Stein set out for Niya. ‘Scarcely two miles beyond the town we were again in barren sands, the outskirts of the great desert northwards,’ he wrote. In Niya, itself an ancient oasis, he received unexpected confirmation of the great antiquity of the site they were proposing to investigate some seventy miles to the north. One of his men fell into conversation with a villager who owned two inscribed wooden tablets which he said had come from this kone-shahr, or ‘old town’. On examination these proved to be written in Kharoshthi, an ancient script used in India’s extreme north-west a few centuries before and after the beginning of the Christian era. Further enquiries produced the man who had actually found them in a ruined house there the previous year while digging for treasure. He had come across several, thrown some away, including the two brought to Stein for examination, and given the rest to his children to play with. Stein generously rewarded the man who had rescued the two from the roadside – much to the chagrin of the original finder, a man called Ibrahim. However, he signed on Ibrahim as their guide, with promises of rich rewards if he could lead them to the ruined houses where he had found the panels. But he issued instructions to his men to watch Ibrahim closely lest he change his mind and slip away, preferring perhaps to keep this potential goldmine to himself. ‘It had been impossible’, explained Stein, ‘to hide from him the value which I attached to these tablets, and … he subsequently seemed to regret not having himself made a haul of them.’

  Aft
er travelling down the frozen Niya river for five days, with the snow-capped Kun Lun sparkling in the distance behind them, the party sighted the first two ruined houses, looking not unlike the ghostly structures of Dandan-uilik. But Stein soon realised that this site (which has come to be known as Niya, although far to the north of the present town of that name) was a good deal older, judging from the style of some finely carved pieces of wood he came upon in one house in which an early Gandhara influence was clearly visible.

  By the end of the first day, it was evident to Stein that Ibrahim had brought them to the right spot. Eighty-five inscribed tablets were found in one room alone, with many more emerging from rooms subsequently cleared of sand, most of them excellently preserved. All, significantly, were wooden, for at that time paper had yet to reach Turkestan from China, where it had been invented in AD 105. The tablets were mostly in wedge-shaped pairs, from seven to fifteen inches long, and held together with string. Where the pairs had remained together, with their Kharoshthi texts facing inwards, the black ink looked as fresh as if it had just been applied. Rather like a modern envelope, the outside surface carried only a brief text, suggesting an addressee. A number of the tablets bore clay seals, the secrets of which were to emerge later. After examining a number of the mysterious documents in his tent that night, Stein decided that they probably represented official orders or letters written in an early Indian Prakrit language, but using the Kharoshthi script.

  Stein now turned his attention to the first two houses they had passed on their arrival. During the next few days his men, reinforced by labour from the nearest village to the south, unearthed from these a number of interesting relics, including a beautifully carved stool, which today can be seen in the British Museum. Other finds in these and other houses included an ancient mousetrap, a boot last, a stout walking-stick, part of a guitar, a bow still in working order, another carved stool, a piece of a rug designed in elaborate geometric patterns and dyed in harmoniously blended colours, as well as many other everyday household objects. However, as at Dandan-uilik, Stein came upon little of intrinsic – as against historical – value. This site too, it appeared, had been evacuated gradually, rather than as the result of some catastrophe, giving the inhabitants time to remove their valuables. This possibility apparently never occurred to the native treasure-seekers, who remained ever hopeful of finding hastily abandoned gold or other valuables at these ruined sites.

  While excavating the two isolated houses, Stein suddenly realised that he was standing in the middle of an ancient bostan, or garden. The trunks of poplars, dead for many centuries, could still be seen rising eight to ten feet above the sand and forming avenues and little squares. ‘It was with a strange feeling, obliterating almost all sense of time,’ he wrote, ‘that I walked between two parallel rush fences that still form a little country lane just as they did nearly seventeen centuries ago.’ Beneath the sand, beside the boles of the white poplars, his diggers found the remains of fruit trees, including apple, plum, peach, apricot and mulberry, the wood of which they recognised from their own village.

  Stein next moved his operations to an area two miles north of his camp. For there he had spotted half a dozen more groups of ruined buildings scattered over an area several miles square. Spread on the sand around one crumbling ruin he came upon a number of wooden tablets, their inscriptions bleached off by the sun. He decided to dig here, and soon turned up a narrow slip of wood bearing Chinese characters. Before long he realised what he had unearthed – an ancient rubbish heap. ‘For three long working days,’ Stein wrote afterwards, ‘I had to inhale its odours, still pungent after so many centuries, and to swallow in liberal doses antique microbes luckily now dead.’ For embedded in an ancient morass of broken pottery, rags, straw, bits of leather ‘and other less savoury refuse’, he came upon layer after layer of inscribed wooden tablets.

  Despite his cold-benumbed fingers and the stench raised by the fresh breeze, Stein kept a careful record of the stratification of each piece – some two hundred in all. Although disagreeable, this was essential to enable those whose task it would be to translate their texts to establish their chronology. In addition to the wooden tablets, he extracted two dozen Kharoshthi documents on leather, each neatly folded. These, he could see, were official documents of some kind. Most were dated, but frustratingly, only by the month and day.

  A number of the wooden tablets, like some of those he had found the first day, bore clay seals. On cleaning the first of these Stein was astonished to recognise the figure of Pallas Athene, with aegis and thunderbolt. Other seals also depicted Greek deities, including a standing and a seated Eros, Heracles and another Athene. A number bore portrait heads of men and women with barbarian features but executed in classical style. Here was powerful evidence of how western iconography, travelling eastwards along the Silk Road, had penetrated far into this remote corner of Central Asia. As if to symbolise this fusion of East and West, one wooden ‘envelope’ bore two seals. One of these, according to its Chinese inscription, belonged to the Chinese political officer in charge of the Lop district far to the east. The other, showing a portrait head, was clearly cut in classical western style.

  This rich hoard of documents, which is still being studied by scholars today, proved to consist of reports and orders to local officials and police: complaints, summonses, orders for safe conduct or arrest, lists of labourers, accounts, and other everyday matters in this long-dead community. The language used throughout was, as Stein had surmised, an early Indian Prakrit written in Kharoshthi script. No documents of such early date concerned with day-to-day life have yet come to light in India. Their discovery adds some credence perhaps to a local tradition, recorded by Hsuan-tsang and also found in ancient Tibetan texts, that the Khotan region was conquered and colonised by Indians from Taxila (today in Pakistan) about two centuries before the birth of Christ. Numerous coins they found dating from the later Han dynasty – which collapsed in AD 220 – and a document bearing a precise date corresponding to the year 269, when the Emperor Wu-ti II ruled the ‘western regions’, led Stein to conclude that the site was abandoned not long after this. ‘Great political and economic disturbances must have accompanied the withdrawal of Chinese authority from these parts,’ he wrote, ‘and with them one feels tempted to connect directly or indirectly the final abandonment of the site.’

  Although there were many ruined buildings still to yield their secrets, Stein and his men were utterly exhausted after sixteen days of continuous digging in the bitter Taklamakan winter. It was no surprise to him therefore when men sent out to spot undiscovered buildings lurking behind sand dunes failed – ‘for obvious reasons of their own’ – to find any. But, with the season of sandstorms about to begin, he knew it was time to move on again, for he had heard in Niya of yet another site, further still along the Silk Road, which he wished to explore before finally starting home with his many treasures.

  Leaving the Niya site on February 13, he struck eastwards across the desert towards the Endere, yet another of the snow-fed rivers which debouched into the thirsty Taklamakan. For it was on the far side of this river that the new site was said to lie. A week later, after reaching and crossing the frozen river, Stein came upon what was now becoming a familiar sight – rows of ancient wooden posts thrusting out of the sand, indicating that a once-thriving community had existed in this desolate spot. In addition, surrounding some of the ruined structures, was a huge clay rampart some seventeen feet high and thirty feet wide at the bottom. This was topped by a parapet of brick, adding a further five feet to its height. Obviously it had been built for protection, but against whom? While Stein was pondering this his labourers arrived from Niya, some one hundred and twenty miles away. As usual, his meticulous planning had paid off and he was able to commence work at once. During the next seven days, intensive excavations were carried out from early morning until, by the light of bonfires, late evening.

  It took nearly two days to clear the remains of what was once a
Buddhist temple standing within the ancient ramparts. Here, among the crumbling remains of life-sized stucco figures, Stein found what scholars later established to be the oldest known specimens of Tibetan writing. Written on tough, yellow paper they proved, on subsequent examination in London, to be sacred Buddhist texts. Scratched on nearby walls he also found graffiti in Tibetan which he carefully photographed.

  Another inscription, this time in Chinese, presented something of a puzzle. It recorded the visit, in 719, of a Chinese administrator. And yet when the Buddhist traveller Hsuan-tsang passed that way some seventy years earlier he reported seeing no inhabited place during his ten-day march. However, at the precise spot where Stein was now excavating (today known as Endere), Hsuan-tsang mentions that there was an abandoned settlement. Stein was to clear up the mystery on a subsequent visit to the site. Evidence he discovered in an ancient rubbish dump showed that, after being abandoned to the desert for several centuries, Endere was then reoccupied by the Chinese, but some time after Hsuan-tsang’s visit. The circular rampart had clearly been built to try to keep the warlike Tibetans at bay. The Tibetan graffiti found within the rampart bear witness to what is already known from the Chinese annals – that at the end of the eighth century the fierce Tibetans finally drove the Chinese from the area.