yarlung drumze Hali, November 2010
von THOMAS WILD

Yarlung Drumze, monastic rug, Lhoka region, central Tibet, C-14 dated to 1460- 1650 AD.
Sheep's wool loop pile on a yak hair foundation, 1.00m (3'2") square. collection Private

IN THE FIELD OF CARPET STUDIES Tibetan rugs are usually perceived as a rather late product. For a number of reasons they have been considered to be barely independent of, and strongly influenced by, neighbouring pile-weaving traditions,1 or they have simply been mistaken for Chinese carpets. The first examples recorded in the West were not acquired until the 1880s,2 and written documentation of Tibetan carpets as trade goods is lacking, since they were primarily traded within the Himalayan world. This perception is strengthened by the very late appearance of Tibetan rugs on the Western collector market, probably only really starting with sales by Tibetan exiles in the 1950s. Given this context, the first 'tapetological' investigations concentrated on manufacture and usage, as well as motif descriptions.3
A few historical clues may be found in Tibetan literature – it is possible to follow traces of an old pile-weaving tradition back to the 10th/11th centuries.4 The existence of such a tradition is more than plausible, since sheep and yak herds represented the most important basis for life on the high, barren, Himalayan Plateau where, since time immemorial, inhospitable climatic conditions have necessitated a wool-based textile tradition: wool production is considered to be the major traditional source of wealth in Tibet.

Why then has it been so difficult to find any truly old Tibetan rugs, when foreign-made weavings have long been preserved on the Plateau, whether 18th-century rugs from Khotan, Ningxia carpets of the 17th century, or even spectacular key pieces such as the small group of early Anatolian animal carpets?5 The usual argument for this lack, that Tibetans wear out their rugs over a lifetime, seems very weak. Could it really be the case that there are no longer any genuinely old, original carpets in this country, so close to the sky, that is broken up into valleys and mountains like few other regions on earth, where monasteries – built on the most vertiginous cliffs – let centuries drift past like clouds?

Following their instincts, enthusiasts and experts have found connections to early Mongolian felts,6 or assigned speculative pre-1800 dates to pieces that gave the impression, visually and stylistically, of being old, without any way of providing genuine proof of their age.7 The question as to origin has also remained largely unanswered.

Armed with no more than hearsay and literary references, in 1995 I embarked on a search for possible places of origin. The key turned out to be the localisation of the specific type of monastic rugs generally known in Tibet as 'Wangden drumze' (Tibetan: dbang ldan grum ze) to the upper Wangden Valley, a tributary of the Nyang (myang) Valley between Gyantse (rgyal brtse) and Shigatse (gzhis ka rtse) in Tsang (gtsang) Province.8 Through contact with the people of Wangden, I soon realised that not all so-called 'Wangden' rugs could come from there. From their present-day products it soon became clear which 'Wangden' types are traditional to the region, and which are not. Since then, whenever I have been able, I have systematically visited monasteries in central Tibet. Every journey has brought new discoveries, confirmed some assumptions and thrown up new questions.

It is fortuitous that rugs for religious use opened the windows of knowledge, since many monasteries are still furnished with original weavings, and the monks are able to provide information about them. In the secular sphere it would be more difficult as, unlike monasteries which, like treasure-houses, seek to keep their assets, the weavings held by nomads, farmers and the aristocracy were subject to adaptation and fashion over the longer time period.

'Wangden drumze' is a catch-all term that roughly translates as 'carpet from Wangden', 'drumze' meaning a carpet from Tibet, and 'Wangden' being the name of the best-known production area for these distinctive weavings.9 In fact, rugs of this structural type can be found all over central Tibet and beyond, in temples and monasteries of all schools of Tibetan religion, whether of the Buddhist or Bon traditions. They were made as throne seats and for the rows of benches for monks to sit and meditate 3, 4. The technique is reminiscent of that used for light, multi-panelled tsuktruk (btsugs phrug) carpets, but woven on a vertical loom in contrast to the horizontal loom used for tsuktruk; such thick, heavy, pieces could only be managed on stable, upright looms. The exceptional thickness of the material insulates the sitter during lengthy periods of meditation or ceremonial, and is surely a modification specifically for this purpose. The motifs in the centre, represented in unmistakably clear-cut fashion, are intended to 'support' the monk, grounding him and giving him strength for the relevant rites. This is why it is so important to understand these monastery rugs as part of Tibetan religious culture.

Although the discovery process for the entirety of 'Wangden' rugs is ongoing, and a concise overview remains to be done, I would like to give a brief outline of my findings to date, as this will contribute to understanding of the C-14 dated piece 1.

'Wangden drumze' were all made in the same technique and the patterns are based on a shared design pool. Despite many inconsistencies and contradictions, one can identify three groups that differ in terms of knot density, the use of various wool types (goat and yak, in addition to sheep), and the emphasis on specific motifs. The largest group includes carpets from the Nyang region in southern Tsang Province which, with the rugs from the Wangden region in the middle Nyang Valley, probably forms the centre of this pile-weaving art.10 A second group has so far only been observed in the Buddhist Drigung Kagyü (bri gung bka' brgyud) line,11 which offers interesting hints of links between religious schools in the choice of their monastic rugs. The third group, into which I place the tested rug, can be localised in the region of Lhoka (lho kha) in southern Tibet,12 and can be provisionally divided into a southerly variant from Luntse (lhun brtse rdzong) district and a northern variant from the Yarlung Valley region.13 Correctly, one should also add a fourth group of rugs whose source has yet to be localised, either because they combine elements from several groups, or are simply unique and have no further connections apart from the shared pile-weaving technique.

Tsakli, votive picture, Tibet, 13th/14th century. Courtesy Nicholas Wright

As a result of my observations in the field, I am able to distinguish truly old pieces from more recent ones. I therefore decided to have the rug described here 1 radiocarbon tested in Zurich, yielding a clear result of 1460-1650 AD (95.4% probability).14

This fragmented 'Yarlung drumze' (yar klung grum ze) – as I call it – was originally around a metre square in size. The substantial format can be explained by its use as a monastic throne seat. The dyed sheep's wool pile is looped into the sturdy, gleaming blackishbrown yak hair foundation. The thick loosely-twisted wefts are covered by the harder, thinner warps, so that the back of the rug is warp-faced. In the 'Tibetan loop' technique, the pile threads are quite loosely inserted into the warp weave.15 The pile of the inner field, shaved to a length of about two centimetres, is surrounded by a five centimetre wide shaggy fringe to protect the sitter from draughts.

Monk sitting on a Wangden drumze in a Bon Temple in Tangna, Tibet. Photograph by Leslie Weir, ca.1930, after Tibet, Heiliger Raum, Fotografien 1880-1950, Frankfurt 1990, p.75

Seated Buddhist monks in Nanying Monastery, near Gyantse, Tsang Province, central Tibet. Photograph by Charles Bell, 1920-21, after John Clarke, Tibet, Caught in Time, Beirut 1997, p.105

A triple outline makes the central motif, a grid network, stand out against the shining raspberry ground. The colour sequence of the framing lines – blue, aubergine and green – is repeated from bottom to top in the horizontal rows of the grid. Set off by one row of knots in ground colour, a yellow grid row completes the square of the motif, giving this colour a particular importance. Each row consists of four small squares with a central point. The blue row has yellow centres and the aubergine row green ones, while the green and yellow rows both contain blue centres.16 Wangdeners to whom the symbol of squares with a central point is still familiar call it chumichen sum (chu mig chen gsum), literally 'three great springs', symbolically standing for a 'holy' or 'pure' source.17 It is thought to be an ancient pre-Buddhist symbol andis seen almost exclusively on monastic rugs and carpets – a further indication of the religious aspect of this rug. The dimensions of the chumichen sum grid are of particular interest, as 35 centimetres square roughly corresponds to the Tibetan ell (skang-gang). Legend tells of a high lama who meditated on a rug of just this size: he plays a special role in my story of the Tibetan monastery rug.

Even without radiocarbon testing, anyone who has seen this piece must have an inkling that they are faced with something very, very old and quite possibly unique: the lustrous sheen of the yak hair, the saturated, oily sheep's wool and the extraordinary patina of the colours! I have never before seen the deep, bright raspberry red (perhaps Indian lac?) in this intensity. The composition too, with its potent graphic austerity, is unsurpassed.

It was in all likelihood made in or near the Yarlung Valley in the central Tibetan province of Lhoka. I will brief ly list the features that allow an attribution to this region. First, the surviving rugs I saw there are all remarkably old; purely local demand and production had probably more or less died out by the beginning of the 20th century.18 Second, the yak hair foundation, only found here, is typical.19 Third, the rug shares its highly original composition with later weavings from this region. The omission of all decorative embellishment directs the attention to the central motif. Overall, this group has a massive, austere and pure appearance, that may represent the traces of an imperial style that had survived in the hinterland of Lhoka, an area of long-standing cultivation that is regarded as the cradle of Tibetan civilisation. This was where the Yarlung dynasty was founded in the 7th century by King Songtsen Gampo (srong btsan sgam po, 617-649), who led the country into a glorious imperial era that ended in the 9th century.20 During this time Buddhism became the state religion and the first Buddhist temples and monasteries were built. Then, at the height of the Buddhist renaissance (950-1200), the region came into the sphere of influence of the Phagmo Drupa dynasty (phag mo gru pa, 1150-1435), which established its capital at Nedong (sne gdong) in the Yarlung Valley, south of the Tsanpo (Brahmaputra) River, and played a key role in the shaping of Tibetan Buddhism. Although challenged by the Rinpungpas (rin spings pa, 1435-1565) and the Tsangpa kings (gtsang pa rgyal po, 1565-1642), the dynasty retained cultural supremacy until Lhasa again became the capital under the fifth Dalai Lama (1617-1682) and Tibet was united and centralised. This brief historical panorama roughly covers the date period of our 'Yarlung drumze'.

Since only Wangden has survived as a monastic pile-weaving centre, any information obtained there must be applied as a model for other regions. At the heart of this pile-weaving tradition stood the Bahulhakang (ba 'ug lha khang), the temple reputedly founded by King Songtsen Gampo in the 7th century.21 Until its destruction during the Cultural Revolution it belonged to the Shalupa (zha lu pa) school.22 The people living around the Bahulhakang are able to relate some legends centered on the temple. Thus they tell of the lama who meditated in their temple on his 35 centimetre square rug covered with sanctified symbols. After his death a text giving details of its pattern and working instructions was found. Now missing, this still forms the guideline for the weavers.23 Producing rugs according to a sacred instruction is seen as a religious duty for the weavers, connecting them spiritually to their patron.

In general, one may say that making pile-woven monastic rugs of this type was more a vocation than a profession in an agricultural or semi-nomadic society. Through it one could not only pay one's dues in the material world but also perform a sacred duty in the spiritual one. As for the conditions under which monastery rugs were produced, the general assumption is that they were made as a sideline by certain villages or groups and not in professional pileweaving workshops.

Tibet is surrounded by carpet and rug cultures, including those of Mughal India, Turkestan, Xinjiang and China, traces of which go back a long way. The discovery and attribution of a Tibetan rug at least 350 years old, the earliest example published to date, opens up one of the few remaining terra incognita of the carpet world. Having certainty about an old piece sharpens the vision and makes it rewarding to investigate pictorial sources such as wall-paintings, manuscript illustrations, and the votive pictures called tsaklis 2. Like its neighbours, Tibetan carpets can now be grounded in history.

Pile rug fragment, Bactria, Central Asia, C-14 tested to 380- 550 AD. Wool, loop- pile, 'warp-faced back' technique, 37 x 25cm (1'2" x 8"). Private collection

In the search for points of contact one can find related warpfaced weaving techniques in adjacent areas of Central Asia, the socalled 'warp-faced-back' group,24 to which our dated fragment also belongs. An astonishing early comparison piece also exists – a 1,500 year-old carpet fragment from the so-called 'Bactrian world' 5, that exhibits almost identical workmanship, both in the technique used for the pile loops and in their size and density.25

Technique as a bridge leads us into pre-Islamic, Bactrianpermeated Central Asia, with a large number of religious influences and cults, where popular Mahayana Buddhism developed and was disseminated via the Silk Road. But having vanished elsewhere as a result of the emergence of Islam, this Bactrian/Buddhist culture was able to survive in a form of its own on the neighbouring Tibetan Plateau.26 The Wangden story thus permits us to draw inferences about this Bactrian world, in which there were pile-weaving deities, now only preserved as archaeological artefacts.27

We may assume that over the centuries 'drumze' for the ecclesiastical sphere followed the path from folk belief to state religion and were adapted for the respective rites. The weaver-lama relationship, which we now see only in Wangden, paints us a vivid picture of those bygone days when the art of pile weaving enjoyed divine protection.28

NOTES