Some Observations on silver objects and costumes from the Tibetan Empire (7^th-9^th century) ^[A]

July 18, 2002

(click on small images for large images with captions and descriptions)

Fig. 1

At present, a silver jug stands in a wooden frame in one of the chapels of the Lhasa Jokhang, traditionally regarded as the oldest temple in Tibet (see figs 1-7). This jug is approximately 80 cm in height. It was hammered from silver sheets, cut and assembled in four parts, two hemi- spherical sections joined at the diameter of the circle, a long thin neck, surmounted by an animal head with round mouth from which liquid can be poured. It weighs some 35 kg when full of liquid, and monks fill it daily with offerings of chang, Tibetan barley beer.[1] The gilded designs on the upper bowl of the jug are raised scrolling in heart shaped medallions, while on the lower bowl, there are three scenes representing Central Asian people, two lively solo dancers and three men in drunken revelry. The people represented on the jug reflect Tibetan familiarity with their neighbors' appearance and customs as we will discuss below.

The Tibetans believe this jug to be associated with Songtsen gampo (Srong btsan sgam po - died 649 A.D.), the first historic ruler of  Tibet. This sovereign is traditionally revered for his foundation of the Jokhang temple and for his numerous military campaigns to unify Tibet and expand Tibetan territory. His military conquests proved crucial to the establishment of the Tibetan empire in Central Asia which lasted for more than two centuries, during which time there was frequent Tibetan communication and trade with the outside world via the lucrative and multi-cultural Silk Routes threading between China and the Mediterranean basin. These military conquests are important to consider here because they help explain the formative foreign influences on Tibetan art and civilisation during the imperial period.

Fig. 4

Under the leadership of Songtsen his descendents, the Tibetans directed their forces to the north-east, completely conquering ca. 660 the plains around the Kokonor leading to the highly active trade routes between China and Central Asia. They then extended their thrust westward all along these trade routes, north and south of the Tarim basin. By 670, Tibet had successively occupied the southern oases of Khotan and the four northern fortified cities of Qocho, Kucha, Turfan, and Kashgar. In 704 Tibetan troops were west of the Pamir mountains, cooperating with western Turks in Tirmidh, a strategic city controlling the entry to Sogdiana. [1a] Western Tibet had already been taken over before Songtsen's death, and served as an additional haven for passage of troops and merchants from central Tibet towards the Pamirs and the Karakorum. Bolor /Gilgit was first subdued ca. 660, then again ca. 720, with a Tibetan princess sent as wife to the king of western Turks in ca.740 to cement the political alliance. [2] Tibetans also engaged in diplomatic and trade relations with Nepal and India as of the mid-seventh century. Tibetan salt was highly prized in this age before refrigeration, every season conveyed along “Salt Routes” from Kokonor to Lhasa then south to the Himalayas and eventually to India.

Fig. 5

Fig. 6

Fig. 7

Thus by the early 8^th century, although Tibetan political control would remain  intermittent in some regions of the Silk Routes until the middle of the 9^th century, trade was active on all fronts. (Uniterrupted rule over Dunhuang returned to the Tibetans from 787-846). In terms of trade, the road from Lhasa to Dunhuang passed northeast via the Kokonor, and there linked to the northern silk route to Dunhuang. To go west, this meant caravans leaving Lhasa travelled due west via Ladakh or first northeast towards Kokonor/ Qinghai where the caravan linked to the southern silk route towards Khotan and then  towards Ferghana, Samarkand in Sogdiana (modern Tajikistan and Uzbekistan), ultimately  towards Byzantium. To the east, from Lhasa the principal road went northeast via Kokonor towards central China, where the capital of the Tang dynasty at Chang'an  (modern Xian) was a bustling cosmopolitan city comprising a population of one million of many nationalities. Trade missions and military campaigns also left Lhasa traveling due east through Khams to reach present day Yunnan and Sichuan in southwest China.

Fig. 8

Through the constant trading missions to China - starting from 634 A.D. and continuing for over two hundred years, there had been on the average one mission either way every sixteen months - the Tibetans received bales of silks, and many spices, in exchange for Tibetan musk, wool, yaktails and horses. [3] Early records document Tibetan export of armor and weapons and salt as well. The Tibetans were so skilled in metal craftsmanship that the 3000 meter gorges of the Mekong river were crossed by Tibetan iron-chain suspension bridges by the early 8^th century. [4] Tibetan chain-mail and lamellar armour was renowned in the Tang Annals and judged to be invincible. [5] The Tang historians wrote, “The men and horses all wear chain mail armor. Its workmanship is extremely fine. It envelops them completely, leaving openings only for the two eyes. Thus, strong bows and sharp swords cannot injure them”. [6] Also famous was lamellar armour from Sogdiana which Sogdian envoys presented as tribute to the Chinese throne in 718. [7] Emulating Sasanian models of design for armour and metalwork, the Sogdians also developed high levels of skill, as documented by these examples of gilt silver bowls (see figs 8 and 9). [8] China excelled in silk production in several regions, Sichuan and central China near the capitals of Chang'an and Luoyang. Thus the Chinese exported their silks towards the western oases, where the weaving of cotton and wool was practiced as well as silk. The Chinese offered bales of silk with each tribute mission to central Tibet.

The patterns of Tibetan trade with the west was not extensively recorded in the Tibetan annals, but recently archeologists have revealed many artefacts which clearly show the results of such commerce with western Asia as well as their impact on Tibetan artisans (see fig. 33, and fig 36 below). Textiles, silver objects and gold coins unquestionably demonstrate importation in antiquity. A Byzantine gold coin was found in Qinghai province north of the Dulan county tombs of the 8^th century, in Wulan county, summer of 2000. It may be dated confidently to 538-542 A.D, issued in Constantinople by Justinian I (Fig. 10, below). This was one of many coins from the west which reached Tibet, and some eventually reached China. The Annals of the Sui dynasty (581-618) tell us that west of the Yellow River, it was current to use the gold or silver coins of Western countries and this was not forbidden by the local administration. [9] The gold coin found in Dulan was treasured as an amulet and mounted into a ring by the Tibetan artistocracy, long after the coins had lost value as currency.  It has been slightly clipped down and then presumably mounted in a ring mount which has erased some of the details around the edge. However the style and letter forms are perfectly regular. It is official coin, a genuine specimen of the ruler Justinian I's three types of solidus, which has been dated to 538-542. [10] It was obviously not regarded as a coin to be used for commerce, but as an artifact in precious metal to be treasured by using it as jewelry. It was found with several other coins which were not from Constantinople but issued in the 6th century from Sasanian Iran (see below), and in fact these coins were so esteemed that there were contemporary Chinese copies made from genuine coins. Artefacts excavated from the 8th to 9th century Tibetan tombs at Dulan, Qinghai, further confirm trade and cultural relations with western Asia. [10a] Archeologists recovered a gilt silver finial (3.3 cm high) with a two-faced head like the legendary Roman hero Janus, capable of simultaneously seeing the present and the future, thus represented with two heads (Fig. 11, below). Although this silver head is now isolated, it is very similar to one on a gilt silver ewer (Fig. 12, below) decorated with people in costumes inspired by classical Greco-Roman styles which was excavated much further east in Ningxia from a mid-sixth century tomb (see Luo Feng 1998 ; Marshak 2001). Quite possibly the gilt head found at Dulan was a finial on a similar object. Youths in ancient Greek costume and grapevines decorate a Bactrian or Sogdian silver platter found in Tibet before 1950 (see fig. 13). [10b] Among artefacts from Dulan, hybrid animals and floral motifs also indicate design influences from western Asia. Archeologists at Dulan excavated several gilt silver panels forming a Buddhist reliquary, approximately 15 x 10 cm, 15 cm high. The silver is carved in honeysuckle vine tendrils with fleur-de-lys, flower buds, and hybrid birds blending the claws of an eagle, body of a pheasant and head of a female phoenix (see fig. 14, below) A fleur-de-lys motif in turquoise is visible on a gold belt buckle also excavated from the Dulan tombs (Fig. 15, below) Sogdian design influence is indicated by the honeysuckle motif, but the gilding technique on the silver panels is cladding, rather than mercury gilding (see Xu Xinguo 1996). Textiles and fragments of garments recovered in the Dulan tombs tell us that the silks and brocades from Sasanian Iran, Sogdiana as well as China were made into garments for the sophisticated Tibetan market (see fig. 36, below) [11]; the Tibetan taste for foreign textiles is also documented by Dunhuang mural paintings decorated under Tibetan patronage. For example, in Dunhuang cave 158, the Tibetan sovereign is portrayed wearing a robe patterned with large roundels and the Buddha's head rests on a pillow painted with ducks inside pearl medallions (see fig. 18, below); both of these textiles are probably inspired by Sogdian motifs. There were also Sogdian weavers installed in the Turfan oases under Tibetan sovereignty in 8^th to 9^th century, as well as weavers of Sogdian descent in Sichuan. [12]

Fig. 9

Fig.10

Fig. 11

Such multiple cultural influences constantly bombarded Tibet from the seventh century onwards and led to a very early eclectic mix of styles and esthetics. Indeed, it became the hallmark of great Tibetan art to mix regional styles – sometimes several simultaneously -  as can be recognized  first in their ancient metal work, and later in  painting and appliqué silks, as well as in sculptures which they made to honor the Buddhist religion.

Fig. 12

It would also appear that the Tibetans' skill in metal was very competent, whether for smelting iron to construct chain links for bridges or making metal armour, as well as fine metal skills, in both gold and silver, with hammered and repoussé decoration. Historically, we know about Tibetan ewers and metalwork of the period thanks to the Chinese Tang Annals. There is record of a spectacular gift received from Tibet in 641 A.D., a goose-shaped golden ewer seven feet high, capable of holding sixty litres of wine. In 648, a miniature golden city decorated with animals and men on horseback was presented as a gift. [12a] As the Tang historian Edward Schafer has written, “Though Iran may have been the ultimate source of the art of beating golden vessels and the ultimate inspiration of many of the designs worked on them by the artisans of Tang, it appears that Tibet must also be given an important place among the nations whose craftsmen contributed to the culture of Tang. To judge from the records of tribute and gifts from Tibet to  Tang  which over and over again list large objects of gold, remarkable for their beauty and rarity and excellent workmanship, the Tibetan goldsmiths were the wonder of the medieval world.” [13]

Fig. 13

It is interesting that the Tang annals record the goose-headed ewer because among the Tibetan historical accounts, there is the tradition that Srong btsan hid several silver chang pots with animal and bird heads as treasures for the Jokhang . Roberto Vitali [14] records this tradition observing that the later Tibetan literature describes horse-headed pots, while his personal observation inclined him to think of a camel's head. His description is corroborated  by the  rGyal po ka thang, written ca. 1345, which describes ten jugs  hidden by Srong btsan,  of which  three are camel head pots (rnga mong gi mgo can) and seven are duck head pots (ngang pa'i mgo can).  The similarity of head shape for ducks and geese is quite clear. Tang ceramic pots with phoenix heads have been excavated from tombs in western China [15] and other isolated examples are known,  such as this blue glazed pottery phoenix head ewer, of a far smaller dimension, only 33 cm high. [16] (Fig. 16, below) When the Tibetan ruler  the Fifth Dalai Lama composed a description of all the relics and images of the Jokhang in 1645, he  described the silver jar as the dngul dam rta mgo ma, the great silver vessel with a horse head, which the renowned Tibetan lama Tsong kha pa (1347-1419) had offered to the Jokhang once he had discovered it as a  terma,  a religious treasure (gter ma). [17] In the realm of Tibetan spirituality, the discovery of  terma  is a very important event. Terma  can be visions of saints who reveal their teachings to later spiritual descendents or terma may be objects or texts which were hidden long ago by great religious heroes like Padmasambhava, for revelation to the karmically appropriate person at a later auspicious date. Thus, while the Fifth Dalai Lama did not directly associate the jar  historically with Songtsen or the Pugyel (sPu.rgyal) dynasty, he did emphasize its antiquity and its great importance. Indeed, when the modern Tibetan historian T.W. Shakaba re-affirmed the Fifth Dalai Lama's description of a silver jar with horse-head finial, Shakabpa supplied the information that it was offered by Tsong kha pa to the statues of Songtsen and his two foreign wives in the Jokhang because it was attributed to the time of Song tsen. [18] The silver jar thus would have arrived in its current location as of 1409, when Tsong kha pa started the Monlam chen mo festival. [19] Another interpretation of the silver jar was given in early 20^th century by Si tu Panchen, a Buddhist master from Eastern Tibet who wrote a detailed description of his visits to sanctuaries in central Tibet from 1918 to 1920. He recorded that the silver  jug has a deer head finial, it is a holy vessel  from the Tibetan sovereigns of the Pugyel dynasty used to contain chang beer offerings at celebrations of marriage or birth, and had been found as terma treasure  at  Brag Yer pa near Lhasa and subsequently offered to the Jokhang. [20]  Brag Yer pa is famous in Tibetan history as a meditation center of  Padmasambhava in the 8^th century, thus a treasure recovered from this site was regarded as potentially a relic of great antiquity for Situ Panchen. We will discuss the meaning of the deer head finial below. Su Bai, professor of archeology in Beijing university, recently studied this jar also, and he identified the animal head finial as a sheep head. [21] In 1980, the former caretaker of the Lhasa temple published his map of the sanctuary, in which, most simply, the silver jug is recorded as “ the Chang pot of Songtsen” [22], thus the traditional origin is maintained.  Incidentally, all of the animals identified as the head of the jug are native to Tibet, including the Bactrian camel which roams wild in herds in the Kokonor region.

Fig. 14

Fig. 15

The jar bears a partially effaced  inscription in Tibetan letters carved on low relief on the neck of the animal, stating  “In the year 1946 on this very blessed religious treasure  … was put on” : rab ‘byung bcu drug pa me khyi lor'….di nyid gter ma byin can la brten bkal….(see Fig. 17, below)I interpret this inscription to mean that the jar was recovered, possibly re-gilded, in 1946; Richardson thought the original jar was recovered with an exact replica at this time, while von Shroeder considered the jar was “re-discovered”  in 1946. [23] H. E. Richardson lived many years in Lhasa himself and had studied the Jokhang. He wrote “one object which may date from the seventh century has survived. A great silver wine jar in Srong btsan sgampo's meditation chapel is decorated with repoussé figures of drinking scenes…which show Central Asian Sasanian influence.” [24]

Fig. 16

Indeed, the lower spherical portion of the jug has three scenes in gilt repoussé, separated by thick scrolling foliation which forms a heart in the center of the scrollwork (see figs. 1 - 3). In one scene, there are three men, a big, rotund man with thick eyebrows, long beard, mustache and curly hair, who is carried by two younger men, one of whom is crouching to carry the big man's leg (see fig. 5). Their robes all have a small scrolling leaf pattern scattered randomly all over the fabric. The man being carried appears so drunk as to be asleep. He wears a short cape over his shoulders, extending to the elbow level of the sleeve which has been lengthened to hang far beyond the wrist. This type of short cape and overlong sleeves are already represented as Tibetan costumes in the mural paintings of the Dunhuang caves painted during the Tibetan occupation of the late 8th to mid-9th century (Fig.28, below). [25] The man on the Lhasa jug, in typical Tibetan nomad fashion, has one arm hidden inside the long sleeve , and he has freed the other arm, which is clasped in the hand of the man carrying him. The other sleeve hangs loosely at his side, just as Tibetans drape their coats today (see fig. 19, below). The man's inert position allows us to examine how the fabric is gathered to mark the area of the wrist above the narrow cuff. (see fig. 20, below)

Fig. 17

The other two scenes show men with similar long sleeved robes of thick fabric, who are in the midst of performing the dance known in China as the “Sogdian whirl ” (see figs. 6-7) [26] As they raise one leg during their dance, it is possible to see buffoon pants stuffed into boots which tightly fit around their calves and ankles. [27] Dancers in similar costumes and in this distinctive position were represented in Chinese ceramics and incised carvings in stone on doors (see fig. 21). [28] According to the Tang Annals, in 718 when the Sogdians presented their coat of mail, a caravan set out from Samarkand bringing carpets, brass, precious rings, mats, lions and dancers – the female performers of the “Sogdian whirl”. [29] Instead of women, on the silver jug in Lhasa, the dancers are men whose robes are tied at the waist by a thin ribbon belt with a purse attached.  These  men are all smiling, exuding good humour in their revelry. The facial traits do not seem Tibetan at all, for the eyes are round, lacking the extra fold of the eyelid found in Asian ethnic groups. Their thick hair, long curls and full beards seem to indicate Central Asian or Mediterranean ethnic groups. Round pendant earrings are worn by the biggest man, and by the two dancers.

Fig. 18

Fig. 19

Fig. 20

Fig. 21

Fig. 22

Fig. 23

In terms of theme, Von Schroeder stated "the ritual function shows an obvious relationship with the Dionysus cult originating in ancient Greece. Dionysus or Bacchus …was worshipped as the god of the vine and its cultivation" On the Lhasa silver jug, in the opinion of this writer, we see scenes of human celebration, rather than worship of deities; a connection with a Bacchus cult seems unlikely. We have examined earlier a Bactrian or Sogdian platter found in Tibet with episodes based on Greek mythology (see fig. 13). If the silver jug of the Jokhang was originally intended for Dionysian rituals, one would expect that the costumes would reflect this, showing a semblance of Greek toga and draperies as found in metalwork of the Hellenized East or Central Asian metalwork influenced by the Greco-Bactrian civilization. This is not the case here. Von Schroeder also observed that “ the musicians  have a sun and crescent moon motif on a ribbon in their unruly hair ”, associated with the Sasanian and Sogdian rulers, and suggested that they might be royal dancers. [30] Indeed, comparison with late 5^th  to 6^th century Sasanian coins found in the Wulan county of northern Tibet shows how the sun and crescent moon are integrated in the crown, as the upper portion above what appear to be two festoons attached to a helmet section (figure 22, above, left). Each of the Sasanian rulers had a very distinct crown which was represented on the coins, but many had the sun and moon as a finial. This particular coin and crown belongs to the reign of Peroz, 459-485. [31] In the opinion of this writer, it is unlikely that the dancers were royalty, or that they were really depicted wearing crowns. Rather, it seems probable the artisan who designed the dancers and their costumes for the silver jar was attempting to  emphasize the Central Asian origin of these dancers spinning on one leg. The sun and moon in the hair would appear to be a hair ornament adapted from the Sasanian emblem of royalty - but it is not a faithful copy of the Sasanian crown, it is a distortion of the crown with the sun and moon. Tibetans long knew the sun-crescent moon as symbolic of the celestial elements, for these are found carved on the capitals of stone stele erected at the construction of the Tibetan royal tombs and the first monasteries in Tibet during the 8th-9th century. (figure 23, above, right) 

Fig. 24

Further, it is certain that the Tibetans appreciated this dance, as evidenced by the representation of a dancer performing the "Sogdian Whirl" painted during the Tibetan occupation of Dunhuang in cave 112 of the Dunhuang grottoes (see Fig. 24, left). Here we see the same dance, but the dancer wears draperies and long flowing scarves similar to those of the Bodhisattva ! And this dancer carries above the shoulder the same instrument as the dancer on the Lhasa silver jug. This instrument is a pipa, a Chinese guitar, rather similar in shape to the oud, the lute of  Islamic lands, but the pipa is more elongated than the oud.

Von Schroeder considered Tajikistan as the place of manufacture for this jug, dating it to ca. 8^th  century. Certainly the Central Asian physiognomy points to that direction, but the misunderstanding of the crown would be unlikely if the artisan was from Sogdiana, which corresponds to the area of modern Tajikistan and Uzbekistan. The costumes of Central Asians and Tibetans during the 7^th to mid-9^th century are somewhat similar. Tibetan costumes are documented by painted scrolls and wall paintings from the Dunhuang caves  and Tang tombs, also by one painting which is a Sung copy of a Tang official court painting made by Yan Li Ben, ca. 641 A.D. The Central Asians' costumes are known principally from Dunhuang paintings and from clay figurines made in Tang China, while Sogdian aristocratic garments are known from the mural paintings of Afrasiab near Samarkand (Uzbekistan) and Penjikent (Tajikistan). On the whole, the Sogdian garments are close fitting, while the Central Asian and Tibetan garments are ample.  The Sogdian garments close in the center, with large lapel at the round collar; the coat has a border of separate fabric emphasizing the central closure and the hemline (see figs. 26a, b, below). This central closure is distinctive in Sogdian garments, and appears frequently in the Afrasiab mural paintings, but it is interesting that the upper jacket of the Tibetan gya-lu-che costume, now conserved in the Tibetan Collection of The Newark Museum (Fig. 27, below), has a central closure. This garment was worn by government non-monastic officials  for New Year ceremonies, and it is said to represent the ancient kings' costume. [32] Tibetans garments usually closed to the left but this jacket was apparently at least one  exception to the rule. Perhaps an ancient version of the Gya lu che costume is visible in the group of the Tibetan sovereign and his attendants represented in Dunhuang cave 159 (see fig. 28, below). At the far right of the group, this person wears a dark hip-length jacket with v-neck collar, over a white skirt with dark fabric from knee to ankle. While it is not a central closure, this garment appears to be less draped to the left than the other garments represented. Perhaps there were other Tibetan garments which closed at the center but our knowledge is too fragmentary. The Chinese vestimentary protocol indicated that Westerners (including both Tibetans, Central Asians and in principle Sogdians) closed their garments right over left, while the Chinese were not to do so but rather, closed their garments left over right. In 612, the edict by Gao Pe, ruler of Gaochang/Turfan was made as follows: “formerly, when our kingdom was in the wild country of the border, we wore our hair floating on the back and we buttoned our clothes on the left. Now that the Suei dynasty is governing, the universe is pacified and united. Thus…the common people and all aristocrats should all take away their braids and adopt Chinese coiffure and re-align the lapel of their robes.”  [33] Even today Tibetans mostly button their shirts on the upper right shoulder just under the collarbone (fig. 29b, below). The 8th -10th century mural paintings of Tibetan royalty and attendants from Dunhuang and nearby Yulin show robes draped to the side instead of central closure. Their principal garment is a coat, called chu-pa in Tibetan, of ample cut, draped to the left, with very long sleeves. It is only distantly related to the ancient Greek coat, the candys, with overlong sleeves and narrow fit, central closure, which is closer to the dancers' close-fitting garments on the Lhasa jug. The robes of the two dancers  are open at the neck and draped to reveal fabric cut to make high collars bordered with a lapel. One dancer has a collar on which a ribbon is sewn above the lapel, each side  closes with a round button to be knotted to  close the robe. Their robes are draped from right to left, forming a v-neck collar. The painting of Yan li Ben (see Fig. 25, below) is said to depict a Tibetan envoy according to the colophons of the Sung dynasty, and his garment has round collar, central closure, side slit from waist to hem, relatively close to the description of the candys robe, but is the envoy in reality a Tibetan or is he from Sogdiana or western Central Asia?

Fig. 25

Fig. 26a

Fig. 26b

Fig. 27

Fig. 28

In fact, I am indebted to Dr. Elfriede Knauer, costume historian, and Professor Boris Marshak, specialist in Sogdian archeology and civilisation, for independently informing me of their opinion that the portrait in the Yan li Ben painting must represent a Central Asian person, probably a Sogdian envoy, due to the facial features, the pose and the garment. Indeed, this garment is very similar in cut and fabric to those represented in Afrasiab or Penjikent. The garments of the Tibetans in Dunhuang  usually resemble those on the Lhasa jug with loose fit and overlong sleeves. But there are a few exceptions. In addition to the Gya lu che costume, a Tibetan figure wearing v-necked coat, with double lapel, over long sleeves with cuff in contrasting fabric, central closure, belt and side slits is represented in a Tibetan Buddhist text from Dunhuang, now in the Bibliothèque Nationale. [34] The Tibetan person beside this figure also has a v-neck double lapel with a round button at the edge of the collar just like the dancer of the Lhasa silver jug, and this robe too appears to have a central closure (see Fig. 30, below). An ancient rock carving in the Leb Khog valley (near Jeykundo, in Khams, eastern Tibet) presents a group of Tibetan nobles worshipping a Buddha. Here also one figure wears a garment with the side slit from the waist to the knee-length hem, very much like the garment worn by the envoy to the Tang (see Fig. 31, below). These examples from different regions are not conclusive but tend to confirm that the central closure is not completely absent from Tibetan clothing of the Pugyel dynasty.  It is also important to recall that an envoy to the Tang court received official garments, for according to one recent Chinese study, the robe of the envoy in Yan Li Ben's painting reflects precisely such a robe-of-state made in Sichuan of fabrics emulating Sasanian and Sogdian designs. [35] Only when all of the numerous colophons of the Yan Li Ben scroll have been fully analysed will we be able to determine if indeed this man is the Tibetan envoy or a Sogdian envoy. But already we can be sure that the people on the Lhasa jug are wearing garments popular in Tibet during the Tibetan empire.

Fig. 29a

Fig. 29b

Tibetan silver objects recently excavated by archeologists or found in Tibet seem to bear significant points of comparison with the Jokhang jug. In terms of design and techniques of manufacture, these objects clearly bear the influence of Sogdian workmanship of the 7^th-8^th century, yet there are significant differences as well which we will examine below. For example, although the scale is vastly smaller for this cast silver vase 17 cm in height now in the Ashmolean museum, [35a] it provides a very close model of certain  designs on the Lhasa jug. Divided into lower and upper sections by a raised petal motif in proximity with a raised line of small beading, the upper section has hearts formed inside of  the raised petals design. On the Lhasa jug, there are also hearts as intermediary motifs between the scenes with humans, as well as on the upper section of the jug, which may have been made in two portions, with the head made separately as a third element. The scrolling edges of the heart design are florid, as befits the scale of the jug. The separation of the upper and lower sections alternates three different designs of scolls and geometric forms, which are separated by thin plain bands in relief. The principal designs on the Ashmolean vase represent a pair of phoenix chasing each other : each scene is separated by floral and leaf designs. The lower edge is decorated in beading, above which tendrils of small vines are carved. The beading of the lower edge appears frequently in the silver of the Pugyel period now known. In the Cleveland Museum of Art, Dr. Martha Carter has studied a silver repoussé vase 22 cm in height very similar in shape and proportions with fluted neck and beading on the lower edge of the vase (see Fig. 32, below).[35b] The base of this vase was soldered on and the beading was attached by welding. [36] This vase is decorated with a roaring lion, a turbaned bird-man, a phoenix, and a dragon. [37] The vase was found with a cup and a rhyton. All three vessels all have very similar designs and gilding techniques. However, the vase and rhyton are made in repoussé with the beaded band soldered on the rim while the decoration of the cup is carved into the metal, leaving the interior smooth. The cup bears an inscription in Tibetan letters with characteristic ancient punctuation and vowel marks which has been attributed to the 7th century bride of Songtsen gampo (see fig. 34, below). [38] While the interpretation of the inscription regarding the identity of the owner may be subject to question, as the inscription seems to refer to the three vessels and rhytons are generally used by men and not by women, the chronological attribution of mid to late 7th to 8th century made by Dr. Martha Carter is well founded.

Fig. 30

Fig. 31

Fig. 32

Another Tibetan cast silver vase, 19 cm in height, 437 grams, in a private collection, has parcel gilding in the recess of the raised designs (see fig. 35, below). This vase presents virtually the same shape and proportion as the vases in the Ashmolean and Cleveland museums of art, but the principal designs represent three hybrid creatures, for example, a  part bird-part horse prancing inside elaborate branches and foliage. In the upper register, this vase follows the model of entwined creatures rather than confronted pairs of birds or lions which are frequent designs adapted from Sogdian textiles. [39] This vase has an inscription in Tibetan language presenting the distinctive characteristics of the punctuation and spelling used during the Pugyel dynasty, known from Dunhuang documents (see fig. 39, below), as well as carved inscriptions on stone stele and on silver objects such as the Tibetan cup (see fig. 34, below). The inscription details the weight of the vase using the ancient Tibetan measure (srang) of silver. Although there is an archaic style of writing and spelling which persists in certain  Buddhist texts, this is different in vocabulary and punctuation from the inscription carved on this vase. In the opinion of H.E. Richardson, due to this inscription, a date of later than 9^th century can be ruled out (personal communication March 2000). The hybrid animals on the bowl of this vase and the entwined necks of the pairs of hybrid birds on the upper bowl of the vase also appear to be typical of Tibetan designs of the dynastic period, influenced by Sogdian textile designs which favor confronted hybrid animals or birds inside pearl medallion, as well as floral or hearts as intersitial designs (see fig. 18, above and fig. 36, below), as well as floral or hearts as intersitial designs. Entwined or confronted animals are also frequently seen in the Tibetan amulets called togcha (thogs-lcags) cast in copper alloys (see fig. 37 and fig. 38, below). The exquisite craftsmanship, the finesse of the numerous designs on this silver vase and its historic inscription make it particularly significant and it will be studied in detail in a forthcoming publication.

Fig. 33

Fig. 34

Fig. 35

In the ancient mythology related to the early kings, among the ancestors were hybrid humans with bird horns, webbed fingers and toes, who traveled on bulls with wings, and these silver vessels provide concrete examples of hybrid creatures. [39a] Another object decorated this way is a seal used on a Dunhuang document dated mid-8th to early 9th century(see fig. 39, below). When first studied, this animal was erroneously understood to be a lion emblem seal (a winged lion) but in fact, it is a hybrid : this animal has a dog or jackal head, wings, and a lion's body, legs and tail. [40] These hybrid animals appear  to be Tibetan variations on the fantastic creatures  of Central Asian and Chinese repertories. In fact, Dr. Carter has identified the head of the rhyton of the Cleveland Museum of Art (see fig.33 and fig. 46, below) as a one-horned deer, known in China as the qilin (“heavenly deer”) (see fig. 40, below), but this animal was also found on Sogdian silver of the seventh and eighth centuries, and earlier still on Sasanian metalwork. [41]  It was in China that the deer's single antler underwent mutation and developed into a  mushroom-like protuberance . In Dr. Carter's  opinion, on the head of the Lhasa silver jug, it is precisely this mushroom crown which is seen. [42] Thus when Si tu Panchen, the Tibetan historian writing ca. 1918, described  the head on the Lhasa silver jug as that of a deer, he was in fact referring  to the distinctive head of a “heavenly deer”. Due to his great interest in art history,  perhaps he was familiar with this ancient motif. It is understandable that the other Tibetan historians were ignorant of the existence of this particular fantastic creature !

Fig. 36

In addition to the cylindrical shape of the cup in the Cleveland Museum of Art, the corpus of Tibetan silver objects attributed to the Pugyel dynasty also includes one-handle cups of a bowl shape. We will examine here four different examples, from Nizhne Novgorod Museum, Gorky, Russia, from the Uldry Collection in the Rietberg Museum, Zurich, Switzerland, from the Metropolitan Museum of Art, New York, and from a private collection. These present a distinctive shape, proportions, and handle, although the diameter varies considerably. The Russian cup measures 11 cm in diameter (see fig. 40). On the base, in relief  the head of a man with thick hair and beard on the base of the cup and on the upper side of the handle, while the bowl of the cup has raised scrollwork interspersed  with recumbant hybrid, fantastic animals. There is a line of beading defining the design of the cup near the upper rim while the base is not defined by beaded edge. [43] Professor Marshak has informed me that on the thumb piece of that cup there is a head of a bald bearded man between two elephant heads. The similar pair of the elephant heads one can see on the Sogdian cup supposedly found in Luoyang. [44] The Tibetan cup in the Uldry collection (fig. 41) is similar in proportion and in beading along the upper rim to the cup found in Russia, but here the base is defined by the beaded edge. Contrary to the Cleveland vessels with soldered beading, the decoration is incised from the same sheet of silver.[45] Rather than animals, the decoration scheme presents stylized floral and leaf motifs carved out. The shape of the ring handle is also as if a coin had been inserted to make the measurement for the handle's diameter. [46] The Tibetan cup in the Metropolitan Museum of Art (see fig. 42) is the largest of the four, 15.2 cm in diameter. [46a] The ring handle is quite different : instead of a coin or a face, a rampant lion appears. The bowl of the cup has series of leaf tendrils whose stems entwine alternating with floral designs. It is inscribed with a Tibetan letter. The lion motif is perfectly in context for the period . But it is striking to see this on the handle, with virtually no wear in the lion, as the finger held the cup by the edge of the handle rather than the upper portion. Our last example of this type cup combines several features of the other cups in a most distinctive manner. This cup (fig. 43, below) has dancing rampant lions as the side decoration, in between beading on the base and a beaded side decoration beneath a plain rim. The lions are separated by two series of floral interstitial motifs. On the base of the cup, a deer stands, with its head facing backwards, in a position which is frequent in Sogdian silver work.

Fig. 37

Fig. 38

Fig. 39

An unusual bowl (see fig. 44, below) was presented in the antiquities markets in 2001. [47] This model of bowl is larger still, diameter 16.4 cm. It is distinguished by the support of the caryatids, which is typical of Nepalese sculptures of this period and recalls the contemporary wood carvings in the Jokhang temple traditionally attributed to Nepalese as well as Tibetan artists. Quite probably Nepalese silversmiths were also working in Tibet at this time. There is a gazelle incised on the base, scratching its head with its hoof, a motif also seen in Sogdian silver designs. There is a gazelle incised on the base, scratching its head with its hoof. The sides of the bowl ower in particular have small geese flying in the sky above a leaf design, inbetween which arcs of a circle surround a diamond in relief. The lower edge again has the beading. These bowls and cups all appear to be related in terms of technique of manufacture. Although they share design motifs with the repoussé work, the system of carving out was apparently preferred for bowls and cups in this period.

Fig. 40

A gilt-silver platter (see fig. 45, below) attributed to Tibetan workmanship of this period is now in the collection of the Miho Museum (Marshak 1996, Carter 1998). It is striking due to the composition of the four hybrid pairs of animals - wings, hoofs, horse bodies but each time different head - rams, or deers with big antlers, or the mythical deer. Each pair of animals has a double half-flower between the two, similar to the interstitial flowers on the Sogdian textile (see fig. 36, above). From the base to the rim, vines grow and progressively entwine to separate the pairs of each species of fantastic creatures. At the center, a centaur is holding grapes. In technique of metalwork it is congruent with the contemporary examples.

Let us return to the rhyton and cup in the collections of the Cleveland Museum of  Art to briefly consider the textile patterns on the Lhasa silver jug. In general, within scrollwork of vines and leaves, there are hybrid animals and birds, as well as two bird-men, wearing turbans and central closure jacket, whose lower bodies have bird forms (see Fig. 33, above). According to Dr. Martha Carter, the bird-man's costume reflects contemporary Tibetan costume and turban, and the small circles on the birdman's jacket represent the roundel designs so typical of Sasanian and Sogdian motifs and their derivatives (see fig.18 and fig. 36, above). [48] Professor Boris Marshak, an archeologist specializing in Sogdian art and civilisation, has informed me he considers the costumes on the Cleveland rhyton and vase to represent  the Tibetans' conception of a contemporary Sogdian robe. This line drawing of a Sogdian aristocrat from Penjikent mural paintings of late 7^th to early 8^th century (see fig. 26 a,b, above) allow us to understand the cut of the Sogdian robe with round collar and contrasting panels of fabric at the central closure and at hem. This is important to consider here, for although the robes on the Lhasa silver jug adopted the ample Central Asian cut and draping, the minute incised florets and small scrolls of cloud shapes are quite different from the Sogdian and Central Asian roundel designs, which the Tibetans knew and appreciated. Typically the Sasanian roundels represented an outer bead ring or rings with one or a confronting pair of birds or animals, while the Sogdian roundels generally add chains of hearts as part of the ring of pearls inside of which are confronted lions or creatures of fantasy, such as hybrid bird-horse or dragons-rams. The Sogdian duck is particularly frequent. The Tibetan sovereigns and their courtiers known from paintings among the Dunhuang manuscripts and the painted murals in the caves wear ample robes which do not fit close to the body, instead  they define the shape of the body by using lapels, closing draped from right to left, a separate textile border at the hem and fur trim (see figs. 18, 28, above and 47, below). Our understanding of Tibetan costume has been conditioned by these representations and by one other painting, to which we have already  referred above. This is a painting by the official court painter of the Tang dynasty,Yan li Ben (see Fig. 25, above). His original painting is no longer extant, now replaced  by a copy made during the Sung dynasty (12^th ce), whose authenticity is vouchsafed by no less than 19 colophons. One of these colophons identified a Tibetan emissary to the Tang emperor. The emissary is represented as a  small man, rather frail, with a short, sparse beard and moustache, chignon, a robe fitted very close to the body, with a round collar closing at the center of the coat, and from the waist side panels with slits to facilitate riding.  The robe is not ankle length, but only falls to mid-calf. The robe is made from two fabrics, the collar, cuffs, side fabric and front closure are confectioned from a  textile  having small roundels enclosing ducks. His physical appearance and costume differ from most of the Tibetans represented in Dunhuang. Although the present writer, following the previous identifications by several scholars has earlier identified this emissary as the Tibetan minister mGar who petitioned the throne to award a Tang princess as bride to Srong btsan sgam po, the contrast of the robe with the contemporary 8^th-9^th representations of Tibetans lead me to question the accuracy of this identification.  In consideration of the similarity of the costume of the bird-man and the emissary, thanks to discussion with costume historian Dr. Elfriede Knauer as well as Professor Boris Marshak,  it is possible to consider that the colophon identifying this emissary as a Tibetan reflects the 12^th century Sung reality of the importance of their Tibetan neighbours, while the memory of the emissaries to the Tang was somewhat less distinct. The Tibetans were then their principal neighbours to the west, while the Sogdian empire was but a distant memory, of less prestigious importance by that time. Thus one may consider that possibly this emissary represents a Sogdian petitioning the Tang throne.  It is also important to recall that an envoy to the Tang court received official garments, for according to one recent Chinese study, the robe of the envoy in Yan Li Ben's painting reflects precisely such a robe-of-state made in Sichuan of fabrics emulating Sasanian and Sogdian designs. [49] Only when all of the 19 colophons of the Yan Li Ben scroll have been fully analysed will we be able to determine if indeed this man is the Tibetan envoy or a Sogdian envoy.

In comparison with  the costumes of the bird-men on the silver vase and rhyton of the Cleveland Museum which represent Sogdian costume, or Tibetan variations thereof,  the men portrayed on the silver jug of Lhasa show the Tibetan familiarity with central Asian physiognomy, but they are portrayed wearing Tibetan robes.

Fig. 41

Professor Marshak has summarized several differentiations between Sogdian and Sasanian workmanship which are pertinent to recall : Sogdian vessels are usually less massive than Sasanian ones (which is visible in the weight of the Sasanian and Sogdian bowls in Fig. 8 and Fig. 9, above). Accordingly Iranian silversmiths cut away the background in order to emboss the figures on their plates and bowls, while the Sogdians preferred hammering and chasing. In theme, the Sasanian vessels more typically represent royal scenes as well as single hybrid animals while the Sogdian designs may confront pairs or groups of hybrid creatures. [50] Thus in the examples we have examined here, we have found both cutting away techniques particularly in the Tibetan cups as well as examples of repoussé, hammering and chasing. In terms of both technique and design, the examples of gilt silver panels (see fig. 14, above) from Dulan demonstrate Sogdian influences in their honeysuckle vines and composition, but Chinese influence in the typology of the hybrid phoenix. Rather than mercury foil gilding, like the Cleveland vessels, they are gilt by cladding; the design is cut out. These silver panels appear to be Tibetan workmanship, strongly reflecting Sogdian design and metallurgical techniques but somewhat combined with Chinese design influences. The Lhasa jug is a transmutation of Sogdian and Chinese designs, in a context more related to Sogdian metallurgical techniques. The design of the animal's necklace is derived from Chinese design, and this motif is repeated at the join of the neck with the bowl of the jug. The necklace is formed of small metal circles in repoussé, each with a square cut out at the center. This distintive shape is adapted from the shape of Chinese coins, such as the Tang coin in fig. 48, below (see the animal's necklace in fig. 4, detail). The Lhasa jar appears to be purely decorated by repoussé for the scenes with people and the intermediary heart motifs, while the ornamental bands of vines and leaves  at the center of the jar and the base of the neck appear to be added beaten and incised pieces which were soldered in place. Added pieces were common in Sasanian plates and some bowls to make design elements in high relief by fastening additional pieces of silver to the object. In the Lhasa jug, however, the added pieces are not the relief designs which appear to be purely repoussé but rather the closure of the “seams” of the two hemi-spheres or the neck sections. At  the top of the neck, the geometric design is incised and repoussé carving. The gilding on the Lhasa jug appears to be foil mercury gilding, rather than cladding, because the surface is now smooth and very shiny, as opposed to thick coat of cladding. However, this may have been the result of the restoration work performed in 1946, referred to in the inscription on the nape of the camel's neck.

Fig. 42

Fig. 43

Fig. 44

Although this study is in no way exhaustive, in consideration of the workmanship and the marked affinities of the Lhasa jug with other Tibetan vessels, it is proposed to concur with the opinion that this jug is indeed a Tibetan creation, manufactured during the Pugyel dynasty. An attribution to the pre-650 A.D. reign of Songtsen gampo does, however, seem perhaps too early for the Tibetans to have assimilated both the diverse metallurgical and esthetic elements which adorn this jug. It must be recalled that the first Tibetan forays and occupations of the western Silk Road oases occurred in the late seventh century. The apogee of Tibetan participation in the complex network of trade linking China and the Mediterranean occurred during the late seventh to mid-ninth century, linking foreign artists and their wares with the Tibetan markets. The impact of this commerce is just starting to be understood due to concrete artefacts such as those studied here. To establish a more precise chronology for the silver jug of the Lhasa Jokhang, comparison with several firmly dated Tibetan silver objects is necessary. As the systematic archeological investigations of Tibetan tombs progressively yield more artefacts and data, a typology and chronological analysis of the production of metalwork in Tibet will eventually be possible.

Fig. 45

To conclude, in the opinion of this writer, the silver jug now in the Jokhang presents significant differences from Sogdian workmanship and does not faithfully copy the Sasanian or Sogdian costumes and interstitial designs. The scale is remarkable and the work is consistent throughout, demonstrating the high skill of the craftsmen. Having examined ancient and modern Tibetan costumes, Sogdian and Sasanian costumes, if we now look again at the silver jug of the Lhasa temple, it is apparent that the people represent Tibetan depictions of Central Asians but wearing robes very similar to Tibetan robes of the Pugyel empire period - and even to modern Tibetan robes! No Sogdian carver would put a crown on a dancer, and their dancers didn't carry pipa Chinese guitar over their shoulders - it is the Tibetans who created a transposition between the Chinese musicians and the Sogdian dancers, and represented it in the mural paintings made under Tibetan patronage during their occupation of Dunhuang, as well as in the silver jug of the Jokhang. This transmutation brings striking results, in the Lhasa jug, and leads us to consider that  there is a degree of amalgamation which becomes characteristic  of  all of the Tibetan silver vessels examined here. This capacity to harmoniously import several different items of esthetic vocabulary (hearts, geometrics, people) as well as emulating foreign techniques of manufacture, but adapting these esthetics and techniques to Tibetan taste -  this would seem to be the hallmark in the development of Tibetan esthetic sensitivities.

Fig. 46

Fig. 47

Fig. 48

This silver jar of the Lhasa Jokhang thus is a product of the Tibetan genius to bring together several different influences, blended so harmoniously that a new esthetic idiom is created. It must have been used at wonderful banquets!

Amy Heller,  May 3, 2002

Selected Bibliography

Anninos, Tony, "Tokches, Images of Change in early Buddhist Tibet", Orientations 29/9: 93-98 (1998).

Azarpay, Guitty, Sogdian Painting, Berkeley, 1981.

Backus, Charles, The Nan chao Kingdom and Tang China's Southwestern Frontier, Cambridge, 1981.

Beckwith, Christopher, “ Tibet and the Early Medieval Florissance in Eurasia : a preliminary note on the economic history of the Tibetan Empire ” in Central Asiatic Journal, 21/2, 1977 : 89-104.

Beckwith, Christopher, The Tibetan Empire in Central Asia, Princeton, 1987.

Béguin, Gilles et  Marie Laureillard, Chine, La Gloire des Empereurs,  Réunion des Musées Nationaux, Paris, 2000.

Belenitsky, Alexandre, L'Asie Centrale Archeologia Mundi, Nagel, Genève, 1968.

Bunker, E. "The enigmatic Role of Silver in China" Orientations 1994, vol. 25/11: 73-78.

Carter, Martha L., “ Three Silver Vessels from Tibet's Earliest Historical Era : a Preliminary Study ”, Cleveland Studies in the History of Art, vol. 3, 1998 : 22-47.

Chinese Numismatics, 2001, vol 4 (issue no. 75). Yan Lin, “ An excavation of Byzantine gold coins in Wulan county, Qinghai province ”. p. 40 and plate.

Chegini, N. N.  and A.V. Nikitin, “ Sasanian Iran – economy, society, arts and crafts ” in Litvinsky, B. A., Zhang Guang-da, R. Shabani Sambhabadi (eds) vol. 3, History of Civilizations of Central Asia, Delhi, Motilal Banarsidass 1999/ UNESCO 1996

Demiéville, Paul,  Le Concile de Lhasa, Paris, 1987.

Fifth Dalai Lama, Catalogue of the Lhasa gTsug lag khang (in Tibetan), written in 1645 and republished in  Three dkar chag, pp.1-45, 1970, Delhi.

rGyal po bka' thang, recovered by Orgyan gling pa in 1349, reprinted in Lhasa, 1990.

Haarh, Erik, The Yar-lung Dynasty, Copenhagen, 1969.

Harper, Prudence and Pieter Meyers, Silver Vessels of the Sasanian Period, The Metropolitan Museum of Art, 1981.

Heller, Amy, “An Eighth century Child's Garment of Sogdian and Chinese Silks ” in E. Knight (ed.), Chinese and Central Asian Textiles, Orientations, Hong Kong, 1998 : 220-222.

Heller, Amy, Tibetan Art, Tracing the development of Spiritual Ideals and Art in Tibet, 600-2000 A.D., Jaca Book, Milano, 1999.

Juliano, Annette and Lerner, Judith, Monks and Merchants, Silkroad Treasures from Northwest China, Asia Society, New York, 2001.

Karmay, Heather (née Stoddard), Early Sino-Tibetan Art, Warminster, 1975

Karmay, Heather (née Stoddard), “ Tibetan costume, Seventh to Eleventh Centuries ” in A. Macdonald and Y. Imadea (eds.), Art du Tibet, Paris, 1977 : 65-81.

Kelley, Clarence W., “Tang Dynasty Gold and Silver”,  Orientations, 16/3, 1985 : pages 10 - 17

Knauer, Elfriede R. ”Toward a history of the Sleeved Coat ”, Expedition (The University Magazine of Archaeology, Anthropology, University of Pennsylavania) vol. 21 (1) :18-36. 1978.

Knauer, E. R. The Camel's Load in Life and Death, Acanthus, Zurich, 1998.

Knauer, Elfriede R. “ Le vêtement des Nomades Eurasiatiques et sa Postérité ”, Académie des Inscriptions et Belles-Lettres, Comptes rendu des Séances de l'Année 1999, novembre-décembre, Paris, 1141-1187.

Knauer, E. R. "A Man's Caftan and Leggings from the North Caucasus of the Eighth to Tenth Century: A Genealogical Study", Metropolitan Museum Journal, vol. 36/ 2001: 124-154.

Lin Meicun “ Some Persian and Central Asian Silver wares with Inscriptions from China ” Wenwu, 1997 (9) : 55-65.

Luo, Feng, “A Central Asian style Gilt-silver Ewer ” Orientations 29/7: 28-33 (1998)

Litvinsky, B. A., Zhang Guang-da, R. Shabani Sambhabadi (eds) vol. 3, History of Civilizations of Central Asia, Delhi, Motilal Banarsidass 1999/ UNESCO 1996

Mahler, Jane Gaston,  The Westerners among the Figurines of the Tang Dynasty of China, Romae, Serie Orientale Roma (20), Roma, 1959.

Marshak, Boris I. “ Sogdian civilisation ” in Litvinsky, B. A., Zhang Guang-da, R. Shabani Sambhabadi (eds) vol. 3, History of Civilizations of Central Asia, Delhi, Motilal Banarsidass 1999/ UNESCO 1996

Marshak, Boris I. “ A Sogdian Silver Bowl in the Freer Gallery of Art ”, Ars Orientalis 19 : 103-110 (1999).

Marshak, Boris I. “ The Decoration of some Late Sasanian Silver Vessels and its Subject-Matter ” in V. S. Curtis, R. Hillenbrand, and J. M. Rogers (eds), The Art and Archeology of Ancient Persia, London, 1998 : 84-93.

Marshak, Boris I. "Ewer showing Greek Mythological scenes" in Juliano, Annette and Lerner, Judith, Monks and Merchants, Asia Society, New York,
2001: 99-100.

Meyers, Pieter, "Technical study" in Harper, Prudence O. and Meyers, Pieter, Silver Vessels of the Sasanian Period, volume one: royal imagery, The Metropolitan Museum of Art, New York, 1981: 143-184.

Meyers, Pieter, "Technical Examinations of Sasanian silver objects in Riggisberg" in Otavsky, Karel (ed.) Entlang der Seidenstrasse, Riggisberg,
1998: 239-246.

Otavsky, Karel (ed.) Entlang der Seidenstrasse, Riggisberger Berichte, Abegg Stiftung, Riggisberg (Bern) 1998.

Overlaet, Bruno (ed.) Splendeurs des Sassanides. L'empire Perse entre Rome et la Chine, Musées Royaux d'art et d'histoire, Brussels, 1993.

Qiang, Ning “ Public Showcase Private Residence : The Early Tang Tombs revisited ” in Hali, Stone and Slate, 60-74, 1998.

Qiang, Ning. “This Life and the After-life. A reading of the Early Tang Tomb Paintings ” in Gisèle Croës,  exhibition catalogue Maastricht/ Bruxelles, 1996.

Rawson, Jessica. The Ornament on Chinese Silver of the Tang Dynasty, British Museum Occasional Paper 40, London, 1982.

Rawson, J. "Central Asian Silver and Its Influence on Chinese Ceramics" in Bulletin of the Asia Institute, 1991 vol.5: 139-151.

Reynolds, Valrae. From the Tibetan Realm, Treasures of the Newark Museum Tibetan Collection, Prestel, Munich, 1999.

Richardson, Hugh E.  “ The Jo-khang “ Cathedral ” of Lhasa ”, originally published in A. Macdonald and Y.Imaeda (Eds) Art du Tibet, Paris,  1977 : 157-188, reprinted in Michael Aris , ed. High Peaks, Pure Earth, London, 1998.

Richardson, Hugh  E. “Some Monuments of the Yarlung dynasty”  in P. Pal, ed. The Path to Void, Mumbai, 1996 : 26-45, reprinted in Michael  Aris, ed. High Peaks Pure Earth, 292-302.

Rowan, Diana. Oriental Art, vol 47, no. 1 : 31-36 “ Identifying a Bodhisattva Attribute : Tracing the long history of a small object ”, 2001.

Schafer, Edward. The Golden Peaches of Samarkand. Berkeley, 1963.

Schloss, Ezekial. Foreigners in Ancient Chinese Art, China House Gallery, New York, 1969

Shakabpa, T. W. D. Guide to the Central Temple of Lhasa ( in Tibetan), Delhi, 1982.

Sheng, Angela “ Woven Motifs in Turfan silks : Chinese or Iranian ”, Orientations 30/4 April 1999 45-52.

Si tu Pan chen, Guide to Central Tibet (in Tibetan), written in 1920 and reprinted in Lhasa, 1999

Su Bai, “  Two important Cultural relics discovered in Tibet related to the cultural exchange between Ancient China and other countries ” in UNESCO : Land Routes of the Silk Roads and the Cultural exchanges between East and West before the 10^th century, New World Press, Beijing, 1996 : 409-411.

Taring, Zasak J. The index and plan of Lhasa Cathedral in Tibet, (Lhasa Tsug lag khang gi sata and Kar chhag), Delhi, 1980 (no publisher indicated).

Tucci, Giuseppe, To Lhasa and Beyond, Roma, 1956.

Uldry, Pierre. Chinesisches Gold und Silber, Die Sammlung Pierre Uldry, Museum Rietberg, Zürich, 1994.

Vitali , Roberto. Early Temples of Central Tibet, London, 1990.

Von Schroeder, U. Buddhist Sculptures in Tibet, Visual Dharma Publications, Hong Kong, 2001.

Watt, James C. Y. and Anne E. Wardwell, When Silk was Gold, The Metropolitan Museum of Art, New York, 1998.

Wenwu (8), 1992 : 1-11, “Excavation of the Northern Wei Tombs at the Southern Suburbs of Datong City, Shanxi Province”

Wenwu 1996 /5 : 33-49, “ Some Remarks on Paintings engraved on the Stone coffin of Li Shou , early Tang noble” by Sun Ji.

Yokohari, Kazuko, An essay on the Debut of the Chinese Samit based on the study of the Astana Textiles, in: Bulletin of the Ancient Orient Museum 1991/ vol 12: 41-103.

Xu Xinguo, " A Conclusion of Gold and silver-plating vessels found in Tubo tombs in Dulan belong to Sogdi system" in in Zhongguo Zangxue, 1994/ 4: 31-45.

Xu Xinguo, " A Silver Casket from a Dulan Sacrificial Horse Trench" translated by Bruce Doar in China Art and Archeology Digest, 1996: 37-49.

Zhao Feng, Treasures in Silk, an Illustrated history of Chinese textiles, Hong Kong, 1999.

Zhao Feng, "Silk Roundels from the Sui to the Tang" in Hali 1997/vol 92: 80-85, 130.