IV. Archeological Evidence for the Date of the Introduction of K'o-ssuWhen was this revolutionary new technique for making pictorial silk tapestry introduced to China? Since no examples of Chinese silk tapestry of pre-Sung date have been preserved in the land of their origin, we are forced to look for evidence in the outlying areas of Chinese civilization.
Claims have been made about the discovery of pictorial silk tapestry from the Han Dynasty, but these have been proven false, and no examples from that early period have ever been found. To be specific, two writers have published what were supposed to be figured silk tapestries from the Han. The first of these, found by Stein at Lou-lan in Chinese Turkestan, was of wool, and its motives were Western Asiatic in spite of a ridiculous attempt to prove that at least one of the elements was "Chinese" (see below). The second, found by Kozlov at Noin Ula in Outer Mongolia, was indeed Chinese, but it was silk damask, and not of tapestry weave. Unfortunately, both of these statements were accepted and repeated by other writers, making for much confusion.
The fragments in tapestry weave recovered by Stein at Lou-lan are believed to date from the first century B.C., or the latter part of the second. They are all of wool, and in Stein's own words, show in the style of their design unmistakable Hellenistic influence, suggesting production west of China. He assumed, moreover, that because of the importance of wool-raising in the Tarim Basin, plus the fact that one of the motives on an otherwise Hellenistic example was attributed to a "Chinese origin", all of these must have been made in that general region of Chinese Turkestan.
This seems very far-fetched. In the first place, all the motives were Western Asiatic or Hellenistic in feeling. Some were strikingly so, notably the chiaroscuro portrait of an individual of pronounced European type. Even the so-called "Chinese motive" is a Western one. This chimera, composed of the head and wings of a bird, horse's forelegs, and another creature for its tail, bears no remote resemblance to the bird-headed winged horse at Wu-liang-Tzu, with which it is said to have "the closest connection”. Beyond the fact that they are both composite monsters, they have almost nothing in common. Phyllis Ackerman says the chimeras are Scytho-Sarmatian in spirit, which seems very reasonable.
If the one "Chinese" element is not Chinese – and there is not the slightest reason to assume that it is – the fact that these Lou-lan tapestries were made of wool is no evidence of local manufacture. Wool tapestry was made throughout the Near East and into Western Asia. Moreover, as Lou-lan and the other oasis cities of the Tarim Basin owed their importance to being on the great trade routes between the empires of Rome and China, they must have been dependent on trade goods rather than on local manufactures. Middle men are seldom producers, and there is no reason to believe that these were. Highly technical arts like tapestry weaving do not flourish under such conditions. If all the silks found at Lou-lan were acknowledged to have been imports from China, why could not these technically high-quality, wool tapestries have been imported too, from the opposite direction? Pfister and Dr. Ackerman have called them "Parthian". but that term does not adequately account for the one with the European-style portrait. This must have come from somewhere even further west, unless it was made in one of the centers of Greek culture in Sogdiana, or Bactria, which we think is very probable.
Sogdiana, at the Eastern end of the Iranian plateau, was conquered by Alexander the Great in 327 B.C. and was governed by a succession of Greek rulers for nearly two centuries after that, until the Tochari, or Scythians, took it over. Under Greek rule this region seems to have enjoyed an almost purely Hellenistic culture, to judge by objects found as far away as Taxila in Northern India. Among other culture traits the Greeks must have brought with them the art of wool tapestry. If so, the portrait fragment found at Lou-lan would probably represent the period of actual Greek rule, while those with the more Scythian motives, notably the one with the chimeras, would date from the period after the Scythians had taken over, when the Sogdians presumably continued the art of wool-tapestry weaving, under their new masters. Aside from the single, tapestry shoe that Stein recovered at Astana in the Turfan oasis, and ascribed to the early T'ang, he found no examples in this technique between the Han period wool tapestries from Lou-lan, and the "late T'ang" silk tapestries at Tun-huang. This shoe is very interesting because it had a warp of vegetable fibres-long since rotted away-which were probably of hemp. Thus it must have been a forerunner of the "chu-ssu" used for shoes, belts, etc., in the Yuan and Ming dynasties. At the same time, aside from being made of silk, with a somewhat more sophisticated pattern, this is much like the earlier shoes of wool tapestry on hempen warps found at Lou-lan and elsewhere in Chinese Turkestan, indicating that it was the culmination of a long tradition.
This shoe from Astana was considered by Stein and Andrews as having Chinese characters on the toe, which led them to say that it was a Chinese shoe. These characters do not show in the photograph, and as neither Stein nor Andrews could read Chinese, we are inclined to be sceptical. However, since all the other motives are obviously non-Chinese Central Asian ones, it is likely that even if these are authentic characters, they are auspicious ones, like the shou for longevity, which were considered lucky by the "barbarians" as well, and frequently copied from Chinese things. Thus we see no reason to believe that this was not made in Central Asia rather than in China.
The "late Tang" fragments of silk tapestry found by Stein and Pelliot at the "Cave of the Thousand Buddhas" (Ch'ien-Fo-tung) in Tun-huang are very well known. Tun-huang was a station on the old Silk Road, in Northwestern Kansu on the frontier of Chinese Turkestan, but it was also noted for its Buddhist shrines in great, rock-cut caves. After the westward expansion of the Han Empire, in the second century B.C., and the opening of the route across Turkestan, Tun-huang was ruled by the Chinese, off and on for almost a thousand years. However, about 847 A.D. in the later T'ang period, it became part of the Uighur Empire, and the Uighurs held it until about 1031, during the early Sung, when it was taken from them by the Tanguts, a Tibetan people who were founding the Hsi-hsia Empire. The few pieces of silk tapestry found there by Stein and Pelliot were discovered among a rich store of manuscripts and paintings that had been walled up in a side chapel off the main cave, sometime shortly before the Tangut invasion. Since, according to Stein, the large majority of the dated documents found in the chapel belonged to the tenth century of our era, it would seem likely that the tapestries were of approximately the same date, which would make them relics of the Uighur occupation at a Period equivalent to the Five Dynasties or Early Sung in China.
Stein reported that all the silk tapestry fragments found at Ch'ien-Fo-tung were of exceptional fineness in technique, and all were hand-made with the needle.69 He pointed out that the great value attached to such work was illustrated by the fact that twice he found small pieces of the identical fabric forming part of the mountings for different head-pieces to paintings and manuscript-roll covers.
Silk tapestry would indeed have been rare in Tun-huang, as this was primarily a religious center at that time, and not a place where textiles would be produced. Such luxuries had to be brought from long distances. We do not, however, accept Stein's statement that the tapestries must have been brought from China because the motives on them were "purely Chinese". The motives are certainly not typical of Chinese art – though some of them show traces of Far Eastern feeling – and their closest relatives appear to be in the fragments of Uighur tapestry found by Stein and von le Coq at various sites in the region of Turfan. None of the latter were precisely dated, but they are believed to have come from the late T'ang or Five Dynasties period. In fact, since Tun-huang belonged to the Uighurs, and was nearer to their oasis-centers in Eastern Turkestan than it was to any of the larger cities in China proper, it probably drew most of its gifts and supplies from Uighur patrons, rather than from Chinese. Thus, the tapestry fragments recovered from Tun-huang were probably Uighur, and like the earlier ones discovered at Lou-lan, were undoubtedly imported from regions further to the west of the place where they were found.
Where did the Uighurs get this art? Clues to the probable answer are given by Stein in Serindia. Reporting his discovery of Sogdian Buddhist texts at Ch'ien-Fo-tung, he says that some were written on the back of Chinese scrolls which manifestly belonged to T'ang times and were put to new use locally. This, he says, is proof that men of Sogdian origin were in Tun-huang, during the period of Uighur domination; while Pelliot's researches have proved that there were colonies of Sogdian emigrants at Turfan and other Uighur centers at that time. Then, in another place, Stein remarks, "In the case of those few silk fabrics from the West which found their way into the walled-up chapel at Tun-huang, local production in that old Sogdian region, which by that time must have grown its own silk, just as it does at present, appears to me on general grounds by far the most likely solution". He was not speaking about the silk tapestries as he believed them to be Chinese, but we feel quite sure that this statement applies to them above all. In our opinion, these fragments found at Tun-huang were scraps from tapestry fabrics either woven by Sogdians for the Uighurs, or by the Uighurs, who had obtained the raw silk and learned the technique from the Sogdians. In any case, they were probably the lineal descendants of the wool tapestry fragments from Lou-lan. For, assuming that the Sogdians were the weavers of the Lou-lan wool tapestries, it seems likely that after they obtained silk of their own, they adapted the old technique to the new medium.
Before we leave the Tun-huang fragments, we should consider the suggestion made by Miss Simmons that a study of these shows that, like the contemporary Coptic and Peruvian weaves, this form of tapestry had not yet reached the pictorial stage. This would seem to be a dangerous generalization. In the first place, since Tun-huang was a rather remote outpost of the Uighur Empire, the finest pieces would not have been apt to reach there, even in the form of scraps. Secondly, von le Coq's finds in Turfan seem to indicate that by the 10th century the Uighurs – or their Sogdian artisans – already had the concept of large compositions, and the technical ability to carry them out.