| articles

by Mary Shepherd Slusser

February 07, 2006

The author comments on an article also found in Please see:
The Lhasa gtsug lag khang: Observations on the Ancient wood Carvings by Amy Heller.
Here you can read a Preface responding to Mary Slusser's article(Link will open in new window)

(click on the small image for full screen image with captions. All photos by author unless otherwise noted)

Fig. 1


The role of Newar carpenters (silpakār, siṁkaḥmi) from the Nepal Valley in decorating the interior of the gTsug lag khang, Lhasa’s revered Jokhang, has been long recognized [1]. Traditionally, the Nepali carvings are dated about the middle of the seventh century - perhaps 639 -, corresponding to the temple’s foundation by the Tibetan ruler, Srong btsan sgam po (r. 629-649). While it is clear that several other parties had a hand in the decorative carvings of the temple, the relief carvings on the paired lintels above the portals of the oldest chapels, some of their jambs, and certain pillars can only have been created by Newars. In them they expressed the unmistakable aesthetics that characterized their homeland, politically the domain of the Licchavi dynasty - and sometimes Ābhīra Gupta - from about 300 to 850. Indeed, in the India and Nepal volume of Buddhist Sculptures in Tibet, von Schroeder discusses the Jokhang lintels purely as Nepalese works of the Licchavi Period (vol. I, pp. 407-31).

A recent paper by Dr. Amy Heller purports to offer specific Licchavi comparisons to the Jokhang chapel carvings. [2] But given the extensive range of choice from the sublime corpus of Licchavi art, it is to wonder at the peculiar Licchavi — and even non-Licchavi — examples selected for this purpose. Vishnu’s feet will hardly suffice (even were they correctly dated), [3] nor will a post-Licchavi sculpture do by labeling it "ca. 650" [4]


Fig. 2

The latter sculpture adorns a fountain at the temple of Vajrayogini seated on a hilltop above Sankhu village, a few miles east of Kathmandu (Figs. 1-4). It was selected as an example of Licchavi gaṇa, dwarfish figures popular as imaginary supports for waterspouts and architectural members. However, stylistically the sculpture is manifestly not a work of the Licchavis and lacking inscriptional evidence would certainly be dated to the thirteenth or fourteenth century. One observer, deeming it to be an “architectural fragment,” even assigned it to circa seventeenth century [5]. However, in this case the sculpture can be dated exactly thanks to the donor of the fountain; he recorded his gift on the side of the central makara spout which the figures support (Figs. 1, 3). For the welfare of his deceased parents and their descendants, the inscription reads, he had the fountain constructed in the winter of 1168 [6]. Although the inscription is not on the figures themselves, an examination of the fountain makes clear that they are an integral part of it, composed in such a way that they appear to support the spout above them, a typical gaṇa duty (Fig. 4). Since there are almost countless examples of bona fide Licchavi gaṇa it is a pity that one of them was not chosen instead (Fig. 5).

Fig. 3

Fig. 4

Fig. 5

If one is searching for examples from Licchavi Nepal to compare with the Jokhang carvings, an appropriate start would be with the diminutive stone caityas peppering the Kathmandu Valley (Figs. 6, 7) [7]. One of their ubiquitous motifs, a pilaster with bracket capitals employed to compartmentalize design elements, must have been popular with carpenters too. As an examination of the lintel carvings reveals, they often used it in the Jokhang and for the same purposes (Fig. 8) [8]. Many other motifs common to Licchavi art may also be discerned in the Jokhang carvings.

Licchavi reliefs also provide excellent comparisons; like the Jokhang carvings, they feature diverse figures in rocky settings (Figs. 9-11). As a bonus, one relief includes a female attendant adorned with tubular ear ornaments (Fig. 10) like the Jokhang example mentioned by Heller [9]. As for joyous companies of music-making gaṇa one need not go so far afield as Ajanta [10] since they can be found nearer to Lhasa in the art of Licchavi Nepal (Fig. 12) [11]. [12]

Fig. 6

Fig. 7

Fig. 8

Since one of the principal issues of Heller’s paper is to show that the Jokhang lintels were carved by Nepalese, it would have been pertinent to compare them to Nepalese wood carvings, in particular the remarkable toraṇa of Yetkhabāhā in Kathmandu (Fig. 13). Although it is a work of the Transitional Period it reflects certain of the Jokhang carvings, for example the upper lintel of the portal of the Protector Maitreya chapel (Fig. 8). In each case the carvings are divided into five panels separated by rocky formations (at Yetkha) or by the characteristic Licchavi bracketed pilasters (the Jokhang). In each of these compositions the central figure is a seated Buddha who, like all the accompanying figures, is depicted within a rocky cave.

Fig. 9

Actually, the Yetkha toraṇa is closer in time to the Jokhang carvings than previously thought. Hampered by the gospel that no extant wood carving in Nepal predates the thirteenth century — or at the earliest the twelfth — scholars usually have adopted that dating for the toraṇa. [13] However, as part of my research on the antiquity of extant Nepalese wood carving, those dates can now be revised. Radiocarbon testing places the toraṇa sometime between 900 and 1050. [14] This date proves doubly interesting because it corroborates the tradition that the monastery was founded in the reign of King Bhaskara Deva who ruled the Nepal Valley from 1045 to 1048 [15].


Fig. 10

As the title of Heller’s paper indicates, it is primarily concerned with the Jokhang wood carvings and their Licchavi inspiration, but the author also briefly addresses the architecture of the building which houses them. Although she recognizes that the “most ancient section of the gTsug lag khang seems remarkably faithful to the square form of ancient Nepalese models” she proposes that its nucleus was actually a secular building, a residential Tibetan fortress-tower like those traditionally associated with Srong btsan sgam po. [16] This may be so — and others, as Heller writes (note 26), have thought so too — but such a daring hypothesis deserves far more support than is offered here. Although at the outset the author promises a discussion of her idea, the only discussion appears in the these few words at the end: “the four lateral chapels and the central sanctuary [form] a sort of core tower around which additional chapels seem to have been constructed later.” Meager as this “discussion” is, it seems a lot to swallow. Where are the sophisticated drawings to support the proposal? Where are the architectural investigations revealing physical modifications of the core tower into a vihara configuration? What is to preclude the existence of two buildings, one the secular palace mentioned in the Tang Annals, and one, the temple, they did not mention? What evidence to conflate the one with the other? These questions arise and many more. If the author has answers we need them to better understand and assess this startling hypothesis.

Fig. 11

The core tower hypothesis has a major impact on the dating of the Jokhang wood carvings. If we accept the proposed date of ca. 710 to 750 for the remodeling of a fortress-tower into a temple, then the Newar-inspired carvings which decorate it must be similarly dated. If they were indeed carved in the eighth century, then the Jokhang’s fame as the repository of Nepalese Ur-carvings would be challenged by those in Nepal that are now radiocarbon dated to the sixth-seventh century. Then, too, what about radiocarbon testing in the Jokhang itself that establishes a seventh-century date for the columns in front of the Jobo shrine? [17] If so, can such evidence be ignored? Further, an eighth-century temple would postdate the apogee of Nepalese direct influence at Lhasa when the exiled court of the Licchavi Narendradeva refuged there in the 640s. [18] Whether the company included a Nepalese wife for the Tibetan king, as tradition affirms, is of little consequence. With or without her, the presence of a large elite Nepalese entourage at the Tibetan court would have been the most likely time to have introduced architectural and artistic concepts of their homeland. [19] Perhaps more sophisticated studies may prove the hypothesis sound and the original gTsug lag khang is only an old fortress-tower refurbished as a Buddhist vihara. But until otherwise shown, I think it would be prudent to cling to a mid-seventh century date backed as it is by a considerable body of evidence. [20] Finally, from a cultural viewpoint, it seems almost unthinkable that those undertaking such an enterprise as the gTsug lag khang would refurbish an old building rather than offering the gods the sparkling altogether new one they would expect. Not much merit there.


Fig. 13

Fig. 12

With this paper Heller has made an important contribution in revealing the hitherto hidden friezes of amorous couples above the doorway to the central chapel, the Jokhang proper, occupied by the Jobo Rinpoche image. [21] From what can be observed in the published sample, the carvings do not seem to reflect Newar art and thus open a new channel of inquiry. Heller’s discovery that some of the lintel panels can be moved and inserted elsewhere by virtue of interlocking edges (p. 6) is also intriguing and could have important implications respecting their dating.

It is distressing to write this critique, but the apparent weakness of the core tower hypothesis reflects not only my opinion but that of several scholars, Nepali and Western, who discussed Heller’s paper informally in Kathmandu this past November. Because the revised dating of the Jokhang’s initial construction phase impacts my research on ancient Nepalese wood carving, it seems to have fallen to me to open the core tower hypothesis to further discussion by a greater circle of our peers. It is too important a subject, I think, to not get it right.


Addendum: Since writing the above, further corroboration of a seventh century date for the Jokhang is provided by André Alexander's just published study, The Temples of Lhasa (Serindia, 2005). As he writes (p. 54) "two separate carbon datings place [a carved wood cornice lion] with very high probability into Songtsen Gampo's time." So did the previous test mentioned above. Thus, even if Heller's core tower hypothesis should be substantiated, its enlargement as a vihara took place in the seventh century, not the eighth as she proposes, and confirm the traditional date of the Nepalese carvings which embellish it.

© Mary Shepherd Slusser


1. Tucci, Giuseppe, 1952. A Lhasa e oltre. Rome. p. 77; and 1987, To Lhasa and Beyond. Ithaca, New York: Snow Lion Publications) (first published as A Lhasa e oltre, Rome, 1952), p. 119; von Schroeder, Ulrich, 1981. Indo-Tibetan Bronzes. Hong Kong:Visual Dharma Publications, p. 400; Vitali, Roberto, 1990. Early Temples of Central Tibet. London: Serindia Publications. p. 71, p. 75, p. 77. p. 81; Chan, Victor, 1994. Tibetan Handbook, A Pilgrimage Guide. Chico, California: Moon Publications, p. 62, p. 75, pp. 77-78, p.81 passim; von Schroeder, Ulrich, 2001. Buddhist Sculptures in Tibet. 2 vols. Vol. 1, India and Nepal. Hong Kong: Visual Dharma Publications, pp. 407-09, pp. 412-431; Slusser, Mary S., 2004. “Richesse et diversité de la sculpture Népalaise,” in Dossiers d’Archéologie, no. 293, May 2004, pp. 46-57.

2. Heller, Amy, 2005. “The Lhasa gtsug lag khang: Observations on the Ancient Wood Carvings,” The Tibet Journal, vol. xxix, no. 3 (Autumn 2005), pp. 3-24.

3. The Jalaśayana Vishnu at Budhanilkantha is not “dated by inscription 643 A.D.” but was carved about 641, a guess date based on two Licchavi śīlapatra, dated respectively 640 and 641, which grant favors to the citizens who dragged the huge uncut stone across the Valley (Slusser, Mary S. and Gautama Vajracharya 1973: "Some Nepalese Stone Sculptures: A Reappraisal Within Their Cultural and Historical Context," Artibus Asiae, vol. 35, nos. 1/2. pp. 84-87).

4. Heller, Amy, 2005. “The Lhasa gtsug lag khang: Observations on the Ancient Wood Carvings,” The Tibet Journal, vol. xxix, no. 3 (Autumn 2005), pp. 3-24. Figs. 2a, 13.

5. Bangdel, Lain S, 1995. Inventory of Stone Sculptures in the Kathmandu Valley. Kathmandu: Royal Nepal Academy, p. 471.

6. Regmi, D.R.1966. Medieval Nepal. 4 parts. Part III. Calcutta: Firma K. L. Mukhopadhyay. p. 7 inscription xiii; Abhilekha-saṁgraha Part II, p. 12.

7. Gutschow, Niels, 1997. The Nepalese Caitya: 1500 years of Buddhist Votive Architecture in the Kathmandu Valley. Stuttgart/London: Axel Menges. Figs. 257-60, 281-82, 288-89, 327.

8. von Schroeder, Ulrich, 2001. Buddhist Sculptures in Tibet. 2 vols. Vol. 1, India and Nepal. Hong Kong: Visual Dharma Publications, Pls. 132 A-C, 133 A-D, 134 A-F.

9. p. 5, Fig. 7b; von Schroeder Pl. 132C.

10. Heller, Amy, 2005. “The Lhasa gtsug lag khang: Observations on the Ancient Wood Carvings,” The Tibet Journal, vol. xxix, no. 3 (Autumn 2005), pp. 3-24. Fig. 3a, b.

11. Pal, Pratapaditya, 1978. The Arts of Nepal. vol. 1, Sculpture. Leiden: Brill. Figs. 129, 147, 148; Slusser, Mary S. 2002. “The Exhibition Galleries,” Patan Museum Guide. Patan Museum. p. 68.

12. Typical of the guesswork that characterizes the dating of Nepalese art, Pal assigns the threshold of Figure 12 to the eleventh century and I think it is classic Licchavi.

13. Despite this convention, certain that the toraṇa had to be older, I proposed to place it about the twelfth century (Slusser, Mary S., 1982: Nepal Mandala. 2 vols. Princeton: Princeton University Press. II, Fig. 165). On the basis of carved struts which I had dated similarly early (Slusser, Mary S., 1979: "Indreśvara Mahādeva: A Thirteenth-century Nepalese Shrine," Artibus Asiae vol. 41, nos. 2/3, pp. 185-225. Figs. 35-40), one scholar objected that such dates were “doubtful” and proposed end of the fourteenth century to the beginning of the fifteenth as more realistic (Rau, Heimo, 1985: “Workshops of Traditional Newar Woodcarving: A Stylistic Analysis,” Journal of the Nepal Research Center, vol. 7, pp. 144-182. Kirtipur: Tribhuvan University. p. 15). The revised dating was based on the assumption that everything prior to 1350 had been destroyed by the Muslims who briefly raided the Kathmandu Valley in that year.

14. I have been able to radiocarbon test several other “thirteenth century” wood carvings with equally surprising results to be discussed in a forthcoming work. The validity of radiocarbon dating of wood has been questioned on the basis that “an old beam” might have been used, thus skewing the date, and that it would also be dependent on which part of the tree it was taken from (Luczanits, Christian, 2005: "The Early Buddhist Heritage of Ladakh Reconsidered," in Ladakhi Histories, Local and Regional Perspectives. ed. John Bray. Leiden/Boston: Brill. p.84). With respect to Nepalese wood carvings these misgivings may be laid to rest. There are no old beams of sufficient size to be salvaged from the typical Newar style building and if there were they could not, and would not, be used. śilpakar affirm that they are unable to do fine carving in old wood, using only the most recently felled tree for that purpose. Further, it would be inconceivable that they would offer the gods old, recycled material. If it cracks later as it dries, no matter. Fresh flower offerings fade too. For the moment the gods have been given the best and what happens later is of slight concern. As for the C14 date being dependent on which part of the tree the sample was taken, the sal tree (Shorea robusta), the tropical hardwood typically used by Nepalese carvers, matures at about one hundred years. The wood chosen for carving lies between the heartwood and the exterior rings so would have an age of approximately fifty years, a time span of modest effect on a radiocarbon date.

15. Śākya, Hem Raj, 1978. Śrī Bhāṣkarakṛti Mahāvihāra, Yetakhābāhā. Kathmandu: V.S. 2035. p. 21, p. 23; Wright, Daniel, ed. 1966. History of Nepal. Translated from Parbatiya by Munshi Shew Shunker Singh and Pundit Sri Gunanand, with an introductory sketch of the country and people of Nepal by the editor. First published 1877, 3rd edition, reprint Calcutta: Ranjan Gupta. pp. 13-14.

16. Heller, Amy, 2005. “The Lhasa gtsug lag khang: Observations on the Ancient Wood Carvings,” The Tibet Journal, vol. xxix, no. 3 (Autumn 2005), pp. 3, 8.

17. Chan, Victor, 1994. Tibetan Handbook, A Pilgrimage Guide. Chico, California: Moon Publications. p.63.

18. Vitali, Roberto, 1990. Early Temples of Central Tibet. London: Serindia Publications. pp. 70-74; Slusser, Mary S. 1982. Nepal Mandala. 2 vols. Princeton: Princeton University Press. Vol. 1, pp. 29-30, pp. 32-36.

19. Nepalese temples, it may be noted, were not “rendered in wood,” as the author writes (p. 3), but in brick with wooden details. Further, I am not aware that per se they ever fulfilled “political functions.”

20. Vitali, Roberto, 1990. Early Temples of Central Tibet. London: Serindia Publications. p. 70; Chan, Victor, 1994. Tibetan Handbook, A Pilgrimage Guide. Chico, California: Moon Publications. p.63.

21. Heller, Amy, 2005. “The Lhasa gtsug lag khang: Observations on the Ancient Wood Carvings,” The Tibet Journal, vol. xxix, no. 3 (Autumn 2005), p. 7, Fig. 15a.

Bibliography | articles