How old is a Tibetan rug?

The age of Tibetan rugs remains a riddle. On one hand, the craft is unquestionably ancient. Legend dates the introduction of rug weaving to Tibet to the 7th century; Tibetan sources use the modern word for pile rug (drumze) from the 13th century; and Western travellers have reported on pile rugs in Tibet since the 18th century. On the other hand, Arthur Leeper’s forty year-old dictum still holds: ‘Were we in fact presented with a rug many hundred years older than those tentatively dated so far, it is uncertain that we would recognize it as such.’

The difficulty of dating Tibetan rugs has many causes. In contrast to rugs from other weaving regions, Tibetan rugs exhibit no datable design discontinuities before the turn of the 20th century; literary references and visual representations in other arts are formulaic and rare; surviving local inventories do not record rugs; and documented Western acquisitions do not commence until the late 19th century. Consequently, most observers simply divide Tibetan rugs into those woven before and those woven after 1900; they summarily assign apparently older Tibetan rugs—having traditional nonfigurative designs and lacking synthetic dyes and cotton—to the 19th century. Within the small surviving corpus of arguably pre-20th century rugs, it is exceedingly difficult to differentiate between those woven during the 19th century and those woven in earlier centuries.

In the absence of historical documentation and art-historical evidence, radiocarbon analysis is the only available tool for dating antique Tibetan rugs before the dye and design revolutions of the early 20th century.

Berlin-based researcher and dealer Thomas Wild pioneered the approach when he dated the fragment of a jewel-toned triden monastic seating square, now in the Molacek collection (HALI 161, pp.66-69), to between 1460 and 1650 (95.4% confidence). Surprisingly, only one other Tibetan rug of similar age has come to light since then—another triden square of the same design, in better condition, and with a brighter palette.

Believing that more truly ancient Tibetan rugs must have survived, fellow collector Rudi Molacek and I submitted sixteen Tibetan rugs from our respective collections for radiocarbon dating at the IRPA-KIK laboratory, Brussels, the largest such effort to date. Seven of these rugs were dated from the 15th to 17th centuries, including examples belonging to different design groups, formats and religious affiliations from the only two previously described Tibetan rugs of similar vintage. Technical details are available at hali.com and are described at greater length in the author’s chapter in Part 2 of the recently published two-volume set Tibetan Rugs: The Molacek Collection.

Radiocarbon dating

Radiocarbon analysis is a scientific dating technique that exploits the presence of the radioactive carbon isotope 14C in the atmosphere. Organisms ingest carbon and maintain the ambient 14C concentration inside their bodies for as long as they live. When wool is harvested, the carbon cycle stops, and the radiocarbon concentration dwindles through radioactive decay. Scientists can then measure the remaining 14C concentration using mass spectroscopy, and estimate after how many years wool shorn in 1950 (the conventional reference year) would contain as little 14C as the sample. This is the sample’s so-called radiocarbon age, reported as ‘radiocarbon years before present’ (BP).

Although the logic of radiocarbon dating is straightforward, the technique has limitations. Most importantly, radiocarbon years BP do not equal calendar years of age, because the concentration of atmospheric 14C fluctuated throughout recorded history. Consequently, wool shorn in a year when the atmospheric 14C concentration was low may contain the same amount of radiocarbon today as older wool shorn when the atmospheric 14C concentration was higher. For example, wool samples from 1493, 1600 and 1610 would all have the same 14C concentration today and hence share the same carbon age BP. Especially with younger objects (produced in the past 500 years or so), radiocarbon dating often returns multiple possible calendar years, all of which are equally plausible and hence indistinguishable without additional evidence.

Radiocarbon dating also suffers statistical uncertainty, so that each of the possible calendar dates associated with a given radiocarbon age BP is reported as a range of possible dates. By convention, these calendar ranges are constructed so that analysts can be 95.4% confident that the true age of the sample is contained in the combined calendar ranges. The total percentage is apportioned across calendar ranges. The width of each calendar interval (and hence its reported percentage), however, does not correspond to the likelihood that the interval contains the true date of production.

Because of these limitations, radiocarbon dating—with some exceptions— typically divides Tibetan rugs into two broad groups: those woven ca. 1420–1650 and those woven ca. 1650–1950. Supplementary art-historical comparison sometimes allows analysts to rule out part of the estimated age range.

Radiocarbon dating thus advances the dating of antique Tibetan rugs. Whereas conventional methods (e.g., inspecting for modern designs and synthetic dyes) can only determine which rugs were plausibly woven before the turn of the 20th century, radiocarbon testing can further identify rugs woven earlier than ca. 1650.

Tibetan rugs from the 15th– 17th centuries

Seven of the sixteen rugs newly tested for this study were dated to the 15th–17th century. Together with the two similarly dated Tibetan rugs previously described by Thomas Wild, this brings the total number of Tibetan rugs of confirmed antiquity to nine.

Several commonalities stand out among them. First, all Tibetan rugs determined to pre-date the 18th century are woven in warp-face back wangdrum technique, wherein the design of the face of the rug is not visible verso. Wangdrum rugs are extremely thick and coarsely woven and often have a distinctive bushy pile fringe to ensconce the sitter on all four sides. Second, all are monastic rugs, in a variety of formats: square triden throne seats, rectangular khadens, and long tsokden bench runners that furnished the assembly halls of Tibetan monasteries. Third, all, to some extent, have so-called ‘hard-brown backs’, composed of tightly spun, dark brown yak wool warps.6 Fourth, all probably originate from the northern Lhokha province of southern Tibet, as identified by established criteria of materials and design.

The largest design group of confirmed 15th- to 17th-century rugs displays the distinctively monastic chumik chensum design, which is thought to pre-date the advent of Buddhism in Tibet and translates to ‘three great springs’ of a holy source.8 Both previously identified ancient Tibetan rugs share this design. Within the group of newly identified ancient Tibetan rugs, the runner (tsokden, 8) renders the chumik chensum design most conventionally, as a regular grid of uniformly coloured square frames surrounding a central dot. (This rug is a composite, assembled from two runners that once formed a pair; the pile fringe has been shorn.) The other rugs in this group present variants of the design. An imposingly large triden square mingles source motifs of different colours (2), and the chumik chensum motifs of the smaller triden square (3) lack a central dot. The latter is the only completely preserved rug in this group and stands out for its pleasing light palette.

The second group of 15th-to 17th-century Tibetan rugs belongs to a different design family, displaying prominent framed swastikas (Tib. yungdrung). No other Tibetan swastika rugs have previously been attributed to this early period. All three rugs in this group are in excellent condition for their age, with only minor native repairs (4, 7) or shorn fringe (6).

The sole khaden format rug in this group, (4), impresses with a rare and austere design of three muted colours (red, green and blue) and a central field composed of three large swastikas within a shared line frame. Its design has become less legible with age, adding visual intrigue. The two blue outer swastikas are barely visible against the green ground, owing to surface decay of its vegetal yellow component that has rendered the green ground bluer. The eye is further confounded by two short horizontal bars above and below the main field that share the colour of the red fill between the arms of the outer swastikas. Most importantly, the swastikas turn counterclockwise, which identifies this rug as belonging to the Bön religion. Identifiable Bön weavings are rare in general, and this example is the first Bön rug dated to the 15th–17th centuries.

The two red-ground swastika rugs resemble each other strongly in colours, structure and design, although the swastika on (7) is blue, and the swastikas on (6) are yellow. In contrast to the Bön khaden, the swastikas on (6, 7) turn clockwise, indicating use in a Buddhist monastery. Also in contrast to the Bön rug, each swastika on the tsogden bench runner (6) is enclosed in its own line borders, forming individual seats.

Unusually, the last seat at the top of the runner differs from the others, omitting the swastika and comprising two adjacent green and blue chumik chensum motifs surrounded by a rainbow border. Distinct end seats are rare in Tibetan temple runners, indicating reserved seats of honour for higher-ranking lamas.

The evident similarities of (6) and (7) suggest a shared provenance, perhaps from the same monastery, and possibly a shared age. If so, their ages would fall at the intersection of their individually estimated age, i.e., somewhere between 1520 and 1640.

Later and inconclusively dated Tibetan rugs

Radiocarbon analysis dates eight rugs in this study to the long 300-year span between about 1650 and 1950. Since radiocarbon testing cannot differentiate within this range, any further attempts at dating these relatively younger rugs must resort to heuristics. Most plausibly, rugs clearly resembling any one of the securely dated rugs in all respects, including structural details, design, colours, handle, and regional origin, are strong candidates for very old age. The clearest candidate is the large wangdrum triden throne seat (9) from northern Lhokha province with multicoloured chumik chensum design and glossy hardbrown back, which may date to the later 17th or 18th centuries.

Disappointingly, radiocarbon analysis sometimes returns inconclusive results. One example is the unusual and by all appearances very old wangdrum triden throne seat (5), with its double-dorje medallion and oversized running swastika border, probably from the Wangden area of central Tsang province. Three separate radiocarbon tests of two different samples from this rug returned non-overlapping estimates ranging from 1220 to 1950 perhaps owed to sample contamination or unrecognised repairs. While further tests may yet converge on a more precise estimate, at present, the age of this rug is undetermined.

These conflicting estimates across successive tests not only indicate the pitfalls of radiocarbon dating for this particular rug, but also impute uncertainty to the results of the singly dated rugs. That said, however uncertain any individual radiocarbon test may be, the large number of positive radiocarbon tests in this study aggregates to near certainty that some surviving Tibetan rugs are indeed more than 350, and possibly over 500, years old.

Conclusion: The irony in Tibetan rug scholarship is that the oldest rugs that have been found in Tibet are not Tibetan but Anatolian. A few well-known examples include the famous 13th-century ‘Potala Compartment Carpet’ fragment as well as five figurative Seljuk-period rugs from the 14th and 15th centuries.10 Against the conventional practice of assigning all evidently antique Tibetan rugs to the 19th century,11 this study contributes to filling the gap in the historical record. Among the sixteen Tibetan rugs that have been submitted for this radiocarbon analysis, seven were determined to originate from the 15th–17th centuries. These seven rugs, include additional examples of the only previously known type of pre-18th-century Tibetan rugs, as well as rugs from new design groups, formats and even religious affiliations.

While age may not matter much for appreciating rugs as art, it matters for art history and for understanding the evolution of Tibetan weaving and its connections to other artistic traditions. For connoisseurs, confirmed age delivers the delights of a palpable past.