1This thesis is the first attempt at a comprehensive study of a dozen Tibetan Buddhist monuments painted with murals in the Western Himalayan region of Ladakh (India) during the 14^th century. It comprises a text volume in English with a substantial resume in French (565 p.); an illustration volume providing photographs, maps, and drawings to accompany the text (431 p.); a corpus volume presenting each monument of the corpus systematically, with the addition of annotated plans of the murals and readings of the inscriptions for the most important ones (348 p.); and an appendix volume including short essays, diagrams, and tables related to the subject (115 p.).

Geographical and historical frames

2The northernmost territory of India, bordering Pakistan and China, Ladakh is crossed by the Indus River and surrounded by the Himalaya and Karakorum. Far from being cut off from neighbouring regions by these mountain ranges, it was until the 1950s a major trade hub at the crossroads of Tibet, the Tarim Basin, the Pamir Mountains, Kashmir, and North-western India.

3Ladakh stands out in the Tibetan cultural sphere for the density of its surviving Tibetan Buddhist heritage, notably painted monuments of the mid-13^th to late 15^th century related to the Drigung Kagyü school of Tibetan Buddhism. Indeed, in the Tibet Autonomous Region of China, only a few monuments of this period survived the Cultural Revolution, such as Shalu monastery and the religious compound of Gyantse, both in Tsang. Drigungthil, the mother-monastery of the Drigung school, located in Ü, was, in particular, razed to the ground a second time after its first destruction and restoration at the end of the 13^th century. More mid-13^th to late-15^th century painted monuments are still extant in the Western Tibetan regions but they are seriously vulnerable to theft. The situation may be less problematic in the border regions of Spiti, Dolpo, Mustang, Bhutan, and Minyak. Still, none of them appear to house such a diverse heritage as Ladakh.

4The political, religious, and art history of Ladakh remains, however, little-known. At the turn of the 13^th and 14^th centuries, in particular, Ladakh was surrounded by three major powers: the Delhi Sultanate to the south, the Yuan Mongols and their Sakya vassals to the east, and the Chagadaid Mongols, who supported the Drigung school, to the west and north. Which of these two Mongol powers ruled over the region, however, is debated.

The “Alchi group” and the “Wanla group”

5Historically and historiographically, the “Wanla group of monuments” under consideration in this thesis follows the “Alchi group”, named after the Sumtsek, or “three-storey [temple]”, inside the religious compound of Alchi, dated to the early 13^th century. As is well known, the latter temple attests to the peak of Kashmiri art in a Tibetan Buddhist milieu, as well as to a subtle shift from the practice of the Yoga Tantras to that of the Unsurpassed Yoga Tantras under the growing influence of the Drigung school. As for the Alchi group, it is an art-historiographical construction based in part on an iconographic and stylistic comparison of the sculptures and murals of the Sumtsek in Alchi with those of a dozen temples and gateway-stupas in Alchi and its neighbourhood, dating from the late 12^th to the first half of the 13^th century, and in part on the identification of a local network of patrons behind their foundation. Over the past decades, the foremost monuments of this group have attracted the attention of scholars and tourists alike. In contrast, the mid-13^th to late-15^th century Tibetan Buddhist monuments of Ladakh still lacked a comprehensive survey.

6The starting point of this thesis is another key monument in the political, religious, and art history of Ladakh: the Tashi Sumtsek, or “Auspicious three-storey [temple]” in Wanla (fig. 1). Probably founded in the first decades of the 14^th century by Bhag Darskya, a powerful local chief devoted to the Drigung school, it contains a very informative foundation inscription as well as an abundance of original ensembles of sculptures and murals (fig. 2). The construction of such a monument must have required gathering numerous skilled craftsmen, possibly from foreign regions. The foundation inscription of the temple relates that it was, moreover, followed by several more constructions. It is therefore expected that the same network of patrons and painters were responsible for sponsoring and painting a number of temples and gateway-stupas elsewhere in Ladakh, which might be called collectively the “Wanla group of monuments”.

Corpus and methodology

7To verify this postulate and to assess the chronology and the significance of the Wanla group for the political, religious, and art history of Ladakh and Tibet, this thesis considers a corpus of 34 monuments located over an area extending on nearly 250 km from south-east to north-west, essentially along the Indus River (fig. 3). It examines more specifically the connections between the three-storey temple in Wanla and a dozen of these monuments from a multidisciplinary perspective, based on an extensive photographic documentation carried out during annual field trips to Ladakh from 2010 to 2017.

8Information collected on each monument is analysed with a constant view of the whole corpus, so as to appreciate and to quantify the iconographic and stylistic similarities and variations manifested in the murals of separate monuments, and even within each monument itself. As a result, connections between distant murals can be identified and groups of monuments can be defined. In some cases, epigraphic evidence further makes it possible to substantiate the historical circumstances of their creation. Finally, radiocarbon dating is used to verify and to anchor the relative chronology obtained by the methods of stylistic, iconographic, and epigraphic analyses.

Scope of the chapters

9The thesis comprises an introduction setting the subject in its geographical, historical, and technical contexts, and five chapters.

10As a prelude to address the questions of the composition, chronology, and significance of the Wanla group, the first chapter reviews the literature, introduces my own fieldwork and corpus, and presents in detail the multidisciplinary approach taken for analysing this corpus.

11The second chapter initiates the study of the Wanla group by examining the murals of the three-storey temple in Wanla, arguably the central monument of this group. It aims at distinguishing the different painting areas and phases of its three storeys, which are requisites for the study of the other monuments of the Wanla group. To define the iconography and the style of the murals, it also draws comparisons with a number of earlier and contemporary monuments of Ladakh.

12The third chapter examines three monuments, whose murals I assume to relate in part to each other and to the three-storey temple in Wanla, with a view to dating the latter temple and to setting an initial boundary for the Wanla group of monuments. In the process, it contributes to our knowledge about the situation of the Drigung school in Western Tibet following the destruction of Drigungthil about 1290.

13The fourth chapter examines four more monuments whose murals I assume to relate either directly or indirectly to the three-storey temple in Wanla. It explores the network of patrons and painters who may have been behind their creation.

14Finally, the fifth chapter examines several later monuments with a view to setting a final boundary for the Wanla group. Among them, two are studied separately on account of their close relation to the Wanla group, while the others are considered altogether in an epilogue surveying the legacy of the Wanla group in the 15^th century.

Composition of the Wanla group

15In this way, the thesis demonstrates that the Wanla group of monuments is a historiographical construction as solid as the established Alchi group. In addition to the three-storey temple in Wanla, this group comprises three temples in Lingshed, Lamayuru, and Kanji, two others in Alchi, two gateway-stupas in Nyoma and Lamayuru, and a three-storey tower, also in Wanla.

16Most monuments of the Wanla group show a clear affiliation to the Drigung Kagyü Tibetan Buddhist school. There is often an emphasis on Avalokiteśvara and triads related to the Nyingma school that is absent from the iconographic programmes of earlier monuments in Ladakh. For the first time in the region, too, the most esoteric assemblies of the Sarma (Kagyü) and Nyingma schools are openly promoted alongside each other.

17Furthermore, the murals of the Wanla group show no consistent reference to the Kashmiri style of the earlier Alchi group. Rather, as in the case of most late-13^th to 14^th-century mural sites elsewhere in Ladakh, they follow the styles en vogue in Central Tibet. The repertoire of motifs associated with the Eastern Indian style is predominantly used, while that of the Nepalese style is reserved for a secondary stylistic mode, a principle hitherto undocumented outside the Wanla group.

18The reasons for these shared characteristics pertain to the creation of the Wanla group by a network of related patrons and painters. With two possible exceptions, all the monuments of the group were founded within an area of the Purik and Sham subregions that was controlled by the chief of Wanla, Bhag Darskya, from an early stage in his rise to regional power. The participation of his dynasty and his vassals in the creation of the other monuments of the Wanla group is suspected, if seldom confirmed by epigraphic evidence. The temple in Kanji was in particular founded as the family temple of a local nobleman named Chepabum, probably a vassal of the Wanla chiefs, while the gateway-stupa in Nyoma was almost certainly erected as a commemorative monument for the sake of Bhag Darskya himself.

19The patrons of the Wanla group appear, moreover, to have called in part on the same painters for executing the murals of their monuments. Due to its large dimensions, the creation of the three-storey temple in Wanla required the gathering of ten or more painters – from masters to apprentices –, at least some of them composing small workshops. The study of the personal styles and iconographic repertoires of these anonymous painters suggests that several of them participated in the creation of other monuments of the Wanla group. This is most notably the case of two masters, one trained locally and the other an expert in Drigung iconography, who also worked respectively in Lingshed and Kanji, and in Nyoma and Lamayuru.

Chronology of the Wanla group

20Thanks to the identification of the aforementioned network of patrons and painters, as well as the reexamination of other art-historical, architectural, and historical evidence, the period over which the monuments of the Wanla group were painted with murals is better defined.

21The murals of the three-storey temple in Wanla were painted in two to four successive phases, from the ground floor to the gallery and the lantern, which explains the absence of an overall artistic programme taking into account the elevation of the temple. The murals of most other monuments of the Wanla group can be dated after the first of these painting phases, and occasionally also between themselves.

22If the foundation of the three-storey temple in Wanla appears to have been crucial for the creation of the Wanla group, there is little internal evidence to date it precisely, even from a detailed analysis of its lineages of Drigung abbots. Nonetheless, two pieces of evidence make it possible to establish an initial boundary for the Wanla group. The first one consists of the savant Drigung iconography of the niche-shrine in Lingshed, which can be dated to the years 1303-1314, during the tenure of the abbot Dorje Rinchen over Drigungthil. Since this iconography was copied in the nearby temple, a relatively early monument of the Wanla group, the beginning of Dorje Rinchen’s tenure may also be considered as a valid terminus post quem for it. The second piece of evidence consists of the representation of a Tibetan master with a Nyingma hat inside the temples in Wanla and Lamayuru. The probable identification of this master as the abbot Dorje Gyälpo, the younger brother of Dorje Rinchen, suggests that the Wanla group developed during his tenure over Drigungthil in the years 1314-1350.

23The end of the Wanla group is more arbitrarily determined than its beginning. Still, the temple in Mulbek and the third temple in Alchi may be considered to mark a break with it since their patrons had no obvious connection to the dynasty of Bhag Darskya and their painters departed from the artistic principles at the core of the Wanla group, either voluntarily or involuntarily. Crucially, epigraphic evidence suggests that the temple in Mulbek was founded in commemoration of an otherwise unknown king of Purik who passed away in 1332 or 1392. Iconographic and stylistic comparisons with 15^th-century mural sites make it possible to narrow down the dating of this temple to the few years after 1392 as well as to situate the painting of the third temple in Alchi around the late 14^th century. Thus, the final boundary of the Wanla group can be set at that time.

The place of the Wanla group in the political, religious, and art history of Ladakh and Tibet

24Now that the chronology of the Wanla group has been clarified, its place in the artistic, political, and religious history of Ladakh and Tibet can be reassessed.

25The comparison between the Alchi group and the Wanla group attests to some major religious and aesthetic changes that occurred in Ladakh between the mid-13^th century and the early 14^th century. Highly esoteric teachings were now openly promulgated by some patrons in Ladakh. The murals of the Wanla group especially provide precious evidence of a synthesis of the Kagyü and Nyingma teachings among the followers of the Drigung school during the 14^th century. The gradual religious alignment of Ladakhi patrons with Central Tibetan Buddhist schools over the 13^th century may be reflected in the aesthetic change from Kashmiri- to Central Tibetan-derived art that occurred in the second half of that century. It may also have become inevitable owing to a lack of available craftsmanship from Kashmir following the brutal conquest of that region by the Mongols in the mid-13^th century, which was said to have led to a reduction of its population from 3,6 million to only 1,3 million houses.

26Overall, the three-storey temple in Wanla and the gateway-stupa in Nyoma can be considered respectively the most elaborate Tibetan Buddhist temple and gateway-stupa dated to the 14^th century still extant in Ladakh and Western Tibet. The creation of these and other monuments of the Wanla group had a tangible influence on later mural painting in Ladakh, with some painters active in the first decades of the 15^th century carrying on the artistic principles associated with these monuments.

27From the political perspective, it is significant that the Wanla group of monuments comprises most 14^th-century mural sites of Ladakh. The prominence and the distribution of this group from the Purik subregion to that of Changthang substantiates the claim made in the foundation inscription of the three-storey temple that a strong political power centred on Wanla existed during that period, which is otherwise little known in the history of Ladakh. The historic recognition of a subregion named after Wanla may also reflect the political authority once exerted by the chiefs of Wanla over the surrounding villages and subregions. The rule of the Wanla chiefs, however, cannot have exceeded a few generations.

28From the perspective of religion, it is significant that all monuments of the Wanla group with a clear religious affiliation relate to the Drigung school, attesting to the resilience of that school following the destruction of Drigungthil by Sakya and Yuan troops about 1290. In light of this evidence, the previously proposed hypothesis that the Sakyapas appropriated most of the region by 1280 and controlled it intermittently for a century does not seem sustainable. The chances are that the foreigners ruling over Ladakh at the turn of the 14^th century were not Yuan but Chagadaids sponsoring the Drigung school. It is yet to be determined whether Bhag Darskya ruled over part of Ladakh under the suzerainty of the latter Mongols or the Kashmir sultans.

Methodological significance, limitations, and future work

29While addressing the three research axes above, this thesis has explored several methodological issues relative to Tibetan art history. Its methodological advances and limitations are reassessed here along with avenues of future research.

30To start with style, this thesis has reconsidered the notion that the murals of the Wanla group and related monuments represent a “regional” or “transitional” style in the sense that they would depart from the Eastern Indian repertoire and incorporate Nepalese motifs without distinction in a local synthesis. In fact, some of the painters of the Wanla group clearly appear to have discriminated between these two stylistic repertoires, which they combined in part intentionally to achieve a distinction between principal and secondary deities as well as variation.

31Beyond the distinction of stylistic repertoires, the tentative attribution of artworks to anonymous painters in a formal approach has contributed to a refined classification of the murals within and between the monuments of the Wanla group. This classification suggests the existence of various types of work organization among painters. As expected, however, there remain several questions regarding the potential variation and transmission of the personal styles of the painters, the coordination of the master and senior painters among them, and the participation of assistants and apprentices.

32Turning to iconography, the evidence gathered emphasizes the importance of repetitions and copying in Tibetan Buddhist mural painting and their potential interest for Tibetan art history. The high consistency of style and iconography in the corpus of some masters indicates that the painstaking development of iconographic repertoires by the painters throughout their training and career led to repetitions, either because the patrons tended to hire experienced painters for painting specific complex iconographic themes or because the painters tended to propose iconographic themes they mastered whenever they took part in the composition of the murals.

33What is more, the identification of probable pairs of models and copies suggests that, when asked to represent complex iconographic themes, some painters resorted to copying existing artworks, including murals of the vicinity, as a means to ensure their accuracy. Thus, together with stylistic analysis, the study of iconography provides crucial information for attributing murals and dating them relatively. Accounting for these and other issues, this thesis has reconsidered the analysis of lineages for dating the Wanla group of monuments. This method should be applied with closer attention to the specifics of the murals, especially with some knowledge of the iconographic repertoires of the painters who executed them.

34The multidisciplinary approach of this thesis, combining the methods of stylistic and iconographic analyses with both epigraphy and the first extensive radiocarbon dating of early murals of Ladakh, has made it possible for a substantial refinement of the art history of the region during the 14^th century. In addition to the stylistic and iconographic issues discussed above, future avenues of research in the pursuit of this refinement include conducting further technical analyses of the Wanla group of monuments as well as gathering more external art-historical and epigraphic evidence related to this group. Finally, it should be stressed that the multidisciplinary approach to Tibetan mural painting taken in this thesis can contribute to a deeper understanding of other groups of monuments as well. In fact, it has already enabled a more nuanced art history for more than twenty other 13^th-15^th-century mural sites in Ladakh without direct connection to the Wanla group of monuments, which I wish to describe more thoroughly in the future.