Tibet Archaeology

and all things Tibetan

Flight of the Kyung

February-May 2018

John Vincent Bellezza

Welcome to Flight of the Khyung and another exciting journey to ancient Tibet! In this special quarterly issue, we comb the caves, cliffs and boulders of Upper Tibet in search of history and the elusive other. Tibetan texts and folklore tell us that before the domination of Buddhism on the Plateau there were religious traditions known as bon. Historical materials describe early encounters between bon and Buddhism but many of these are partisan accounts. A more candid source for understanding the religious history of Tibet are words registered in stone at the time of those encounters. This is the actual record of individuals who participated in the remaking of the heart and soul of Tibet, recounting an epic tale of religious conviction and conflict that reverberates through to the present day.

Written in Stone: A comprehensive inventory of early rock inscriptions in Upper Tibet (ca. 700–1400 CE)

Table of Contents

This monograph is organized as follows:

Acknowledgements

General introduction to Upper Tibetan rock inscriptions

Part One: The archaeology and history of Upper Tibetan rock inscriptions

The archaeological context
The variable development of Tibetan epigraphy on the Western Tibetan Plateau
Rock inscriptions of the Vestigial period

Part Two: Paleographic analysis of Upper Tibetan rock inscriptions

Paleographic methods of dating
Paleographic typology
Physical qualities of the inscriptions

Part Three: A statistical and religious profile of rock inscriptions in Upper Tibet

Statistical analysis
Doctrinal significance of mantras
Buddhist and non-Buddhist contestation

Catalogue of Upper Tibetan rock inscriptions

Introduction
Bkra-shis-do, Gnam-mtsho
Other sites at Gnam-mtsho
Other sites on the Byang-thang
Far Western Tibet

Addendum

Bibliography

Acknowledgements

The writing this monograph was made possible through the generous support of the Tise Foundation (Chicago), an institution dedicated to advancing the artistic and scientific study of Tibet. My old friend, Bob Brundage (Petaluma), an expert in the material culture of traditional Tibet, also supplied funding for the draft preparation of the work in 2017.

I also gratefully acknowledge the institutions that provided financial support for the launching of my expeditions over the period of 18 years upon which this work is based. They include the Shelley & Donald Rubin Foundation (New York), Asian Cultural Council (New York), Tise Foundation (Chicago), Tibetan Medical Foundation (Weslaco), American Council of Learned Societies/Henry Luce Foundation (New York), Expeditions Council of the National Geographic Society (Washington, DC), Unicorn Foundation (Atlanta), Philadelphia Theravadin Meditation Center (Philadelphia), Spalding Trust (Stowmarket), and Shang Shung Institute (Merigar).

The expeditions to document rock inscriptions and other archaeological assets in Upper Tibet were also enabled by the generosity of individual donors whom I acknowledge with great pleasure: Mickey Stockwell (Boulder), Joseph Optiker (Burglen), Michael White (Nashville), and Lishu Shengyal Tenzin Gyaltsen (Gyalrong). I am also very grateful to Yungdrung Bon scholars who took the time to consider inscriptions with me. These individuals are Lopon Tenzin Namdak, Khenpo Yundrung Tenpa, Nagru Geshe Gelek Jinpa, Ashak Geshe Lungrig, Geshe Kelsang Shedup Gyatso, Tsultrim Namgyal, Yundrung Tenzin, and Geshe Monlam Tharchin. 

My heartfelt thanks go out to Mara Arizaga for her excellent organization and support. I would also like to heartily thank Angela Clyburn for processing more than 60 digital files of rock inscriptions with DStretch© image enhancing software. Her efforts permitted more complete readings to be made.

Finally, I hasten to add that any mistakes or other shortcomings in this work are entirely my own responsibility.

General introduction to Upper Tibetan rock inscriptions

This monograph catalogs the ancient rock inscriptions of Upper Tibet and analyzes their cultural and historical significance. These inscriptions were written in Tibetan or Tibetanized Sanskrit using a variety of Bodic scripts. The work furnishes comprehensive documentation of epigraphic resources as a tool for better understanding the religious complexion of the western reaches of Tibet between circa 700 and 1400 CE.* Rock inscriptions were carved and painted on boulders and cliffs and in caves and other parietal structures throughout the inhabitable southern half of the vast region. This work also examines ancient inscriptions written on the walls and interior structures of ruined temples and cave sanctuaries belonging to non-Buddhist religious groups.†

Certain aspects of the epigraphy of Upper Tibet are treated in Bellezza 2008, pp. 186–189; 1997a, pp. 186, 187, 195, 215, 216, 250; 1997b; 2000. Also see January and April 2011 and November 2012 Flight of the Khyung.

This work does not extend to inscriptions found in Buddhist monasteries and cave temples of Upper Tibet, nor to those on stelae (with one exception) or portable materials. For matters related to a classification of Tibetan epigraphy more broadly, see Scherrer-Schaub 2013. A bibliography of works on inscriptions of the Western Tibetan Plateau (mostly western fringe regions under Indian jurisdiction) has been published on the website of the University of Vienna. This bibliography is part of an announced project called “Inscriptions of Western Tibet”, by Christian Luczanits, Cristina Scherrer-Schaub, Jaroslav Poncar, Helmut Tauscher, and Kurt Tropper. See http://www.univie.ac.at/Tibetan-inscriptions/sub/bibliography.htm. Accessed in July of 2017.

This survey of rock inscriptions is the result of extensive fieldwork carried out between 1995 and 2013 on twenty different expeditions to Upper Tibet. Over the course of the surveys I documented more than 550 epigraphs, most of which can be assigned to the Early Historic period (ca. 650–1000 CE) and the Vestigial period (ca. 1000–1300 CE). The latter period is so named because it represents the final era in which rock art and inscriptions using older techniques of production were regularly made in Upper Tibet.

Upper Tibet refers to much of the western third of the Tibetan Plateau, minus the western and southern fringe regions comprised of Ladakh (La-dwags), Spiti (Spi-ti), Baltistan (Sbal-te), Dolpo (Dol-po), Mustang (Lo Smos-thang), etc. Upper Tibet consists of the interlinked geographic regions known as the Byang-thang and Stod. It is part of the territory now called the Tibet Autonomous Region of the Peoples Republic of China. Upper Tibet is characterized by expansive tablelands, broad valleys, lofty mountain ranges, and numerous lakes, which extend north and west from Lhasa all the way to the modern borders of northern India and western Nepal. This sprawling region is cut off from Central Tibet by the Transhimalaya ranges (called Gnyan-chen thang-lha and Gangs-dkar by modern geographers), but also extends into the upper Brahmaputra valley (Mar-tshang gtsang-po), the badlands of Gu-ge on the north Indian border, and Spu-rang at the trijuncture of Nepal, India and Tibet. Although Upper Tibet is extremely high in altitude (averaging around 4600 m) and sparsely populated, it has supported advanced cultural activity, including sedentary patterns of settlement, for approximately three millennia.

Much effort has been expended to make this survey of epigraphs as thorough as possible but in an area as huge as Upper Tibet it is inevitable that an unknown number of early inscriptions and epigraphic sites were overlooked. Yet, sustained search of the region and the wide distribution of sites covered renders this work representative of the general epigraphic status of Upper Tibet (in terms of contents, paleographic characteristics and methods of production). Nevertheless, the discovery of additional rock inscriptions should serve to augment the historical and cultural picture of the region.

The recording of inscriptions was conducted in conjunction with a survey of rock art. Of the 77 rock art sites I have studied in Upper Tibet, 41 (53% of the total) also contain rock inscriptions. There are also six sites in this work that just have rock inscriptions and no ancient rock art. The proclivity of epigraphers to rely on preexisting sites indicates that inscriptive activities were viewed as a natural carryover from rock art, the epigraphic tradition succeeding the tradition of making figures and symbols. Epigraphers were content to use the same sites and in many cases the same rock panels and boulders as earlier artists. The geographic and cultural prominence of most pre-established sites, including those on well-used routes, in conspicuous locations and with sacred connotations, was clearly an attraction to inscribers. The recording of words on stone surfaces was implicit recognition of more ancient rock art figuration and its social and religious functions. Although the mode of communications changed radically from preliterate to literate times, the medium for expression remained the same: stone.

Beginning in the Early Historic period, the rock canvas of artists was reimagined as the rock parchment of epigraphers, which is explicitly acknowledged in rock art and rock inscriptions made by the same individuals. It must be pointed out that no system of writing predating the Early Historic period has been discovered in any medium in Upper Tibet, be it alphabetic, syllabic, ideographic, pictographic, or in any other form of visually encoding speech.* Had a system of writing existed, the most durable medium for its dispersal would be stone, fixed and portable.†

In earlier works (Bellezza 2008, p. 187; January 2011 Flight of the Khyung), I suggested that the Tibetan letter A may possibly have been introduced into Upper Tibet as a magical symbol or cipher before the Early Historic period. Having had the opportunity to more closely examine the epigraphy of the region since then, I now believe this scenario is unlikely. Rather, the oldest examples of the letter A appear to represent an elementary style calligraphy dating to the Early Historic period.

Yungdrung Bon holds that there was a full-fledged literary tradition connected to the language, kingdom and religion of Zhang-zhung long before the 7th century CE. For textual sources upholding the customary view of the development of the Zhang-zhung written language and its inventor Ston-pa gshen-rab, see Norbu 2009, pp. 149–166. However, no substantive textual, archaeological or epigraphic evidence supporting these beliefs has come to light. The literary culture and scripts purported to be from Zhang-zhung (smar-chen, smar-chung, etc.) are only described and illustrated in post-11th century CE Yungdrung Bon sources. Most crucially, the so-called Zhang-zhung scripts (smar-yig) do not possess paleographic traits of pre-7th century CE systems of writing, as recorded in India or Central Asia, where epigraphic traditions were well established before that time. For a weighing of the evidence concerning the existence of a Zhang-zhung written language, see Blezer et al. 2013; van Schaik 2011, pp. 65–67; Bellezza 2008, p. 187. It is not unreasonable to suspect that the smar-yig scripts, as presented in Yungdrung Bon works, were derived from or inspired by ornamental Tibetan scripts such as lan-tsha and vartu, which appeared in the 11th and 12th centuries CE. According to the T’ang Annals, Tibetans had no writing for official purposes, relying instead on knotted cords and notched tally sticks for record keeping (observations applicable to the mid-7th century CE or earlier; see Snellgrove and Richardson 1968: 29; Bushell 1880: 440). These objects are related to ritual implements of considerable antiquity known as ju-thig/cu-thig (knotted cords used for divination in the Phya-gshen theg-pa of Yungdrung Bon) and khram-shing (wooden boards with carved markings used in wrathful ritual magic). On the use of notched boards for record keeping among ancient Tibetans, Li-su and Chinese, see Lacouperie 1885, pp. 420, 421, 429, 430; on the use of notched boards and knotted cords for keeping accounts in Zhang-zhung, see ibid., 427.

Most of the more than 550 ancient epigraphs documented in Upper Tibet concern religious matters. These inscriptions are devotional, exhibitive and instructional in nature and with sectarian overtones. The body of religious inscriptions is comprised of non-Buddhist and Buddhist mantras, prayers, dedications, and memorials. Many Buddhist inscriptions are readily identifiable through an examination of content but also by the style of script and grammatical conventions used, as well as by accompanying rock art. Non-Buddhist epigraphy is identifiable relying on the same set of multifaceted criteria. However, the precise religious affiliations of non-Buddhist art and inscriptions are not often obvious, thus the use of a blanket term in this work to denote examples distinct from Buddhist types.*

The term ‘non-Buddhist’ was chosen to contend with ambiguities concerning the religious identity of many rock inscriptions presented in this work. This label denotes several forms of ancient religions practiced in Upper Tibet. Firstly, ‘non-Buddhist’ refers to archaic religious cults of the Early Historic period, which are sometimes associated with bon in Tibetan literature composed in that time. Secondly, ‘non-Buddhist’ denotes Yungdrung Bon (G.yung-drung Bon), a lamaist religion that emerged in the late 10th and 11th centuries CE. Due to uncertainties in dating Upper Tibetan rock inscriptions, it is not always clear which examples predate or postdate the late 10th century CE, demanding the application of a general term when denoting religious traditions that are not Buddhist. Moreover, there is some historical and epigraphic evidence to suggest that from the late 10th to 13th century CE there were localized religious cults in Upper Tibet that did not see themselves as belonging to either Buddhism or Yungdrung Bon, the pan-regional Tibetan religious movements of that time. This category of heterodoxic practitioners is likely to have included those still performing animal sacrifices. An edict of Lha bla-ma ye-shes ’od (early 11th century CE) explicitly abolished the sacrifice of livestock in Far Western Tibet (Karmay 2014; May 2015 Flight of the Khyung), but how quickly it had the desired result is difficult to judge. The biography of the Buddhist saint Mi-la ras-pa (ca. 1052–1135 CE) records a bon ritual in which 100 yaks, 100 sheep and 100 goats were to be sacrificed (Chang 1999: 241–258). For other accounts of animal sacrifice in bon rituals in the testament of Mi-la-ras-pa, see Stein 1972, pp. 238–240. In this work, ‘non-Buddhist’ is also applied to any Upper Tibetan heterodoxic groups of the Vestigial period. On the use of the term bon/Bon in an archaeological nomenclature, see Bellezza in press-b.

There are rock inscriptions in Upper Tibet that are not manifestly religious in nature, but ‘secular’ content is uncommon in the region’s epigraphic tradition. Non-religious inscriptions include proper names of people, clans and places, partial alphabets, and labels identifying rock art. There are also a few inscriptions that refer to economic and political matters but only in an abbreviated fashion. Absent from the region are epigraphs containing administrative measures, legal codes, decrees, treaties, and trade records, the kinds of cultural, social, economic and political information preserved in ancient inscriptions in other areas of the world.

Upper Tibet in the wider region. Map by Brian Sebastian and J. V. Bellezza.

Map of sites in Upper Tibet with rock inscriptions. Note that modern international boundaries are not designated. Map by Brian Sebastian and J. V. Bellezza.

Site No. Site Name County Type
1 Bkra-shis do-chen ’Dam gzhung Painted
2 Bkra-shis do-chung ’Dam-gzhung Painted
3 Thang-lha’i rten-khang ’Dam gzhung Carved
4 Rta-mchog ngang-pa.do Dpal-mgon Painted
5 Khyi-rgan gag-pa do Dpal-mgon Painted
6 Lug do Dpal-mgon Painted
7 Ra-ma do Dpal-mgon Painted
8 Stong-shong phug Dpal-mgon Painted
9 Se mo do South / Srin mo do South / Nang do South Dpal-mgon Painted
10 Rigs lnga do Dpal-mgon Painted
11 Lce do Dpal-mgon Painted
12 Lha-ris sgrub-phug Dpal-mgon Painted
13 Slob-dpon phug Dpal-mgon Painted
14 Rta-ra dmar-lding Shan-rtsa Painted
15 Chos-lung O-rgyan bsam-gtan-gling Shan-rtsa Painted
16 Dar-lung phug-pa Shan-rtsa Painted
17 Skyid-sgrom dgon-pa Shan-rtsa Painted
18 Chu-ro Shan-rtsa Painted
19 Stong-shong brag-khung Shan-rtsa Painted
20 Bshag-bsangs Shan-rtsa Carved
21 Gnas kun-bzang Shan-rtsa Painted
22 Am-nag Nyi-ma Painted
23 Dar-chung Nyi-ma Carved
24 G.yung-drung lha-rtse Nyi-ma Carved
25 Gu-ru rin-po-che phug Mtsho-chen Painted
26 Do dril bu Mtsho-chen Painted
27 Do dril-bu, Da-rog mtsho ’Brong-pa Painted
28 Lha-khang dmar-chags ’Brong-pa Painted
29 Rdzong pi-phi ‘Brong-pa Painted
30 Ri-rgyal Sger-rtse Carved
31 Gong-ra Ru-thog Carved
32 Chu-mkhar gyam sgrub-phug Ru-thog Carved
33 Skabs-ren spungs-ri Ru-thog Carved
34 Tham-ka-can Ru-thog Painted
35 Mchod-rten sbug sna-kha Ru-thog Carved
36 Glog-phug mkhar Ru-thog Carved
37 Brag-gdong Ru-thog Carved
38 Rwa-‘brog ‘phrang Ru-thog Carved
39 Sna-kha sogs Ru-thog Carved
40 Nag-skyom Ru-thog Carved
41 Rdu-ru-can Ru-thog Carved
42 Ri-mo gdong Ru-thog Carved
43 Rno-ba g.yang-rdo Ru-thog Carved
44 She-rang mkhar-lung Ru-thog Carved
45 Ru-thog rdzong Ru-thog Carved
46 Gser-sgam Ru-thog Carved
47 Brag-gyam Ru-thog Carved

Part One: The archaeology and history of Upper Tibetan rock inscriptions

The archaeological context

After the demise of the Zhang-zhung kingdom in the 640s CE, Buddhism and literacy were gradually introduced in Upper Tibet, as well as other critical cultural and technological developments, fundamentally altering the character of the region.* The rise of a pan-Tibetan civilizational order in the Imperial period (ca. 625–850 CE) led to the decline of older cultures in Upper Tibet and other parts of the Plateau. Associated with polities, cultures and languages traditionally known as Zhang-Zhung, Sum-pa, and Mon, archaeological evidence indicates that prehistoric (pre-7th century CE) Upper Tibet remained locked in a preliterate stage of development (this also appears to be true of Central Tibet and most other areas of the Plateau). Moreover, no evidence for early epigraphic traditions in foreign languages and scripts has been discovered in Upper Tibet.†

There are numerous scholarly works pertaining to the collapse of Zhang-zhung and/or the spread of Buddhism in Tibet. For example, see Bacot et al. 1940–1946; Beckwith 1987; Bellezza 2014c; Dotson 2009; Kapstein 2006; Richardson 1998; Scherrer-Schaub 2002; 2014; Uray 1972a.

See Bellezza 2008, pp. 95, 574; 2014c, pp. 192, 298.

The new civilizational order that emerged in Tibet in the Early Historic period was marked by momentous changes in religion, art, architecture, artisanal production, social customs, folk traditions, economic systems, and political structures. Cultural and technological innovations sweeping across the Tibetan Plateau were spurred on by and increasingly modulated through Buddhism and the written word. Among the greatest of cultural feats was the development of a Tibetan script and the capability of recording in written form the collective Tibetan experience.* Manuscription and textual dispersal gave rise to a literary tradition in which the public and private fund of information and imagination could be set down and disseminated.

There are many Tibetological works that touch upon the introduction of writing in Tibet and the labors of the minister Thon-mi bsam-bho-ṭa. As per tradition, he was sent to India by the 33rd king of Tibet, Srong-btsan sgam-po, to develop a Tibetan script based on those already in use there in the early 7th century CE. On the invention of Tibetan writing in major Buddhist historical sources (Chos-’byung) such as Dba’ bzhed and Rgyal rabs gsal ba’i me long see, for example, Wangdu and Diemberger 2000, pp. 26–29; Sørensen 1994, p. 167; van Schaik 2011. Also see Narkyid 1982 and his restatement of Dge-’dun chos-’phel’s ideas on the origin of the Tibetan script. Based on his paleographic study, Schuh (2013: 172–174) is persuaded to see Thon-mi bsam-bho-ṭa as relating to the adaption of an Indic script for the needs of the Tibetan administrative apparatus. On the facility of writing as a state-ordained activity in the Old Tibetan Chronicle and Old Tibetan Annals, see van Schaik, ibid., pp. 52, 53. For a view based on a reading of early historical sources, which holds that Thon-mi bsam-bho-ṭa was either a minor personality or a later fabrication, see Miller 1976.

The appearance of rock inscriptions in Upper Tibet in the Early Historic period was one offshoot of the epic project of bringing writing to Tibet. Although of limited magnitude when seen in the full scope of literary activities taking place in Tibet at that time, rock inscriptions are one of the most enduring and ubiquitous manifestations of the acquired power of literacy. Moreover, these epigraphs appear to have had considerable local significance, charting the religious aspirations and belief patterns of key locales in Upper Tibet.

The approximately 550 ancient inscriptions I have documented in Upper Tibet are disbursed over 47 sites in an area covering almost 350,000 km². Given the size of the region, a relatively small number of inscriptions have been documented, and of these, just a tiny percentage contain explicit historical information. Although this is disappointing to the historian, it is also culturally revealing. Sea-swell religious changes played an important role in the reduced repertoire of subjects available for written expression. Literacy was bundled up with new forms of piety and formalism inspired by Indian tradition, in which indigenous rock art themes had a diminished role. The disinclination of Upper Tibetans to chronicle a wide range of activities in rock inscriptions contrasts with earlier pictorial rock art, which is replete with hunting, battles, martial contests, mytho-ritual displays, and many other kinds of compositions.*

I have documented at least 10,000 petroglyphs and pictographs in Upper Tibet. Refer to Flight of the Khyung and my printed works for hundreds of images and analyses.

The harsh climatic conditions of Upper Tibet may have played a part in dissuading individuals from writing in open environments. Nevertheless, the paucity of rock inscriptions in Upper Tibet also shows that literacy (yig-shes) was slow to develop in the region. This is supported by the general lack of calligraphic and grammatical refinement exhibited by the corpus of inscriptions, even in the post-1000 CE era. It does not seem that writing came readily to most Upper Tibetan residents, many of whom we can infer remained illiterate (yig-rmongs).* Even among those who composed rock inscriptions, standards of literacy were not very high.†

In fact, most of the Upper Tibetan population could not read or write until the institution of a modern system of education under the Chinese Communists. Yet, only a fraction of this new literacy project is directed towards native Tibetan languages.

Spelling errors (yig-skyon) are quite common in the rock inscriptions of Upper Tibet, as is sloppy writing (yig-long), irregular spacing between letters (yig-bar mi-mthun) and other errors in composition (yig-lhod).

As noted, Upper Tibet did not host inscriptions in other languages and scripts predating Old Tibetan epigraphs of the Early Historic period. A lack of familiarity with writing prior to the invention of the Tibetan alphabet must have contributed to slow progress in establishing a literary culture in the region. On the other hand, the epigraphic resources of Lower Ladakh and Northern Pakistan are much richer and varied. Inscriptions in a dozen languages and scripts have been chronicled in those regions, dating perhaps from as early as the 1st century CE through the entire first millennium. In addition to Tibetan, the corpus of inscriptions in Ladakh and Northern Pakistan includes examples in Brāhmī, Śarādā, Kharoṣṭthī, Sogdian, Arabic, Hebrew, Chinese, Bactrian, and other scripts and languages.*

For an introduction to the cosmopolitan array of inscriptions in Northern Pakistan, see Thewalt 1985; Jettmar 1989; 1982; for the multilingual inscriptions of Ladakh, see Bruneau 2011.

The absence of multilingual inscriptions in Upper Tibet during the Protohistoric period (ca. 100 BCE to 625 CE) suggests that this region was not heavily affected by foreign invasion, colonization, or alien intellectual advances before the advent of Buddhism.* It appears that the native cultural fabric and civilizational trajectory of the region remained largely undisturbed with little or no formative impact from beyond the Tibetan Plateau. The endogenous development of Upper Tibet in the Protohistoric period is reflected in its unique assemblage of archaeological monuments. These residential, ceremonial and burial monuments are characterized by peculiar forms, many of which do not seem to have changed much over time. Although the chronology of archaic monuments in Upper Tibet is still being formulated, the assemblage is dominated by all-stone corbelled residential structures and funerary pillars and enclosures with morphological traits and spatial formats found nowhere else on the Tibetan Plateau.†

The Protohistoric period comes after an era of seminal Eurasian cultural and technological interactions in the Late Bronze Age and Iron Age, which strongly impacted Upper Tibet and other regions of the Plateau. For the archaeological evidence supporting these exchanges, see articles in the October and November 2014, February 2015, March 2016, and January 2017, etc. Flight of the Khyung. Also see Bellezza 2008; Bruneau and Bellezza 2013.

On these monuments, see Bellezza 2001; 2002a; 2008; 2011; 2014a; 2014b; 2014c.

The variable development of epigraphy on the Western Tibetan Plateau

Even after the invention of Tibetan writing in the 7th century CE, the epigraphic evidence suggests that Upper Tibet remained largely shunted off from cosmopolitan currents swirling around the Indian Subcontinent and north Inner Asia. With two later exceptions, no rock inscriptions in non-Tibetan scripts have been documented anywhere in Upper Tibet. This indicates that Upper Tibet continued to be a kind of cultural backwater in which few travelers from abroad passed or took the time to leave written signs. This is not so surprising given the extreme altitude, geographic isolation, harsh climatic conditions, and lack of many natural resources in Upper Tibet. The warlike nature of the native inhabitants may also have played a part in keeping palpable foreign influences at bay.

The only non-Tibetan inscriptions discovered in Upper Tibet occur on opposite ends of the region at Gnam-mtsho and G.ya’-ma mchod-rten. In the Brag phying-dkar phug of the Bkra-shis do-chen headland, there is a group of red ochre inscriptions superficially resembling old Chinese ideograms, however, their cultural affiliation has not been identified. The inscriptions (displaying varying degrees of chirographic proficiency) appear to be related to scripts used by Altaic peoples known as Khitans (Liao dynasty) and Jurchens (Qin dynasty). There is decidedly non-Tibetan rock art associated with these enigmatic inscriptions consisting of anthropomorphic figures, horse, sun, and house (one of the only domiciles depicted in Upper Tibetan rock art), etc.* G.ya’-ma mchod-rten is situated at the foot of the Himalaya Range in southwestern Tibet. There are many rock mchod-rten and shelters at this high-altitude site, which according to the local oral tradition were constructed over a long period of time by Bhotia traders coming from the Indian district of Darchula.† There are several Devanāgarī inscriptions carved into stones at G.ya’-ma mchod-rten.

On these inscriptions and rock art, see Bellezza 2000, p. 49; 1997a, pp. 208–210; Suolang Wangdui 1994, fig. 200. A closer examination of this non-Tibetic epigraphy and art will appear in an upcoming issue of Flight of the Khyung.

On this site and its inscriptions, see Bellezza 2014a, p. 46.

More surprising than the absence of foreign language inscriptions in Upper Tibet is the scarcity of examples in Tibetan dating to the Early Historic period. Only around 6% of all those included in this survey are confidently attributed to that time (although an additional 17% possibly predates 1000 CE). What inscriptions do exist from the Early Historic period are restricted in content, lacking with few exceptions the political, military and administrative references of Tibetan language rock epigraphs in Ladakh and Northern Pakistan.*

On the Tibetan epigraphy and mchod-rten dedications of Ladakh, see Orofino 1990; Denwood 2007; 1980; Denwood and Howard 1990; Takeuchi 2012b; Francke 1902; 1903; 1905a; 1905b; 1906a; 1906b; 1907; 1914. On Buddhist rock inscriptions in Eastern Tibetan attributed to the Early Historic period, see, for example Heller 1997; Karmay 1997. On the rock inscriptions of Spiti, see Thakur 2001; October (figs. 18.30, 18.31) and November 2015 (figs. 24.2; figs. 1–3) Flight of the Khyung. For Tibetan epigraphy and accompanying mchod-rten in Northern Pakistan, see Denwood 2007; Jettmar and Sagaster 1993; Jettmar 1990, pp. 808–810; Jettmar 1989; Mock 2013; Francke 1928; Sha-bo mkha’-byams 2016. On Tibetan inscriptions and mchod-rten in Wakhan, see Mock 2016; forthcoming; Sha-bo mkha’-byams ibid. It is held that in Northern Pakistan the production of Buddhist rock art was halted after the 12th century CE (Jettmar 2008: 85). Tibetan variants in the Gilgit Basin appear to have been made during or after Little Palūr was part of the Tibetan empire, extending from the second half of the 8th century to the middle of the 9th century (Jettmar 1993: 98). As regards Baltistan, Jettmar (1990: 810) observes that by the second diffusion period of Tibetan Buddhism (ca. 1000 CE), Buddhist rock carvings were supplanted by more elaborate reliefs. Jettmar’s dating of the technical transition from simple line engravings to relief carvings also appears to hold true for Ladakh.

Many of the Tibetan inscriptions in Lower Ladakh and Northern Pakistan constitute personal mchod-rten dedications. Some of these can be dated to the Early Historic period but others belong to the Vestigial period.* Mchod-rten dedications typically furnish a clan or vocational title, the personal name of the maker, and record the act of establishing (bzheng) or inscribing (bris) the religious monument. This establishment or inscription refers to the actual carving of the mchod-rten. Some inscriptions in Lower Ladakh and Northern Pakistan provide the astrological year (duodecimal system based on twelve different animals). References to administrative or military personnel (e.g., blon, stong-dpon, dmag-dpon) of the Tibetan imperium and later times are quite common in the Ladakh inscriptions (despite many being destroyed in recent years). These kinds of inscriptions are also known in a few areas of Northern Pakistan and in Wakhan.

As pertains to Old Tibetan inscriptions in Lower Ladakh, Takeuchi (2012b: 51) suggests that they either date to the Imperial period or to the time of the founding of the Ladakh kingdom (post-Imperial period to 11th century CE). Takeuchi is more inclined to date them to the later period; however, he does acknowledge that the occurrence of Tibetan inscriptions in Gilgit and Baltistan may favor production of some Lower Ladakh epigraphs in the Imperial period (ibid. 55). On Tibetan rock inscriptions dated to the Imperial period in Baltistan, see Schuh 2013. To this list of outlying territories with a Tibetan epigraphic record must be added the even more remote region of Wakhan. It is not likely that Tibetan inscriptions with administrative and military nomenclature were created so far afield after the withdrawal of imperial occupying forces in the mid-9th century CE. Based on Old Tibetan linguistic characteristics and styles of mchod-rten, Orofino (1990: 177) maintains that at least some of the inscriptions from Lower Ladakh recorded in the Tucci collection probably date to the time of the Tibetan imperial occupation, ca. 750–850 CE. Denwood (1980: 163; 2007: 45) also suggests that such inscriptions and mchod-rten belong to the Imperial period. A comparable date for the inscriptions of Lower Ladakh was accepted by Richardson (Orofino 1990: 179 [n. 11]). Tibetan inscriptions recently discovered in Lower Ladakh by Quentin Devers and Viraf Mehta warrant a periodization variously in the Imperial period, post-Imperial period and Vestigial period, as part of an evolving Tibetan epigraphic tradition. For a dedication and mchod-rten in Lower Ladakh dated to the early 11th century CE, see Devers et al. 2016; Devers forthcoming. The Classical Tibetan grammar and paleography of this inscription contrasts with those in Lower Ladakh composed in Old Tibetan, helping to establish a relative chronology of the region’s Tibetan epigraphy. Takeuchi dates Classical Tibetan inscriptions documented by Francke at Ba-lu mkhar and Khalatse to the 11th and 12th centuries CE (2012b: 51), and Denwood (ibid. 48) dates an inscription at Wanla to the 12th or early 13th century CE.

Only in Ru-thog and Sgar, districts bordering Ladakh, are there inscriptions comparable in content to the dedicatory types in Ladakh and Northern Pakistan described above. Counterparts in Upper Tibet are very uncommon with just a few examples being documented. At Brag-gyam, a site on the border with Ladakh, two mchod-rten dedications have been discovered that provide clan (rus) and personal (ming) names, as is found in more northern regions (see infra, figs. 329, 330). There is also what appears to be a dedication of one or two carved ritual thunderbolts in Ru-thog, which records a personal name as well as a Tibetan imperial district known as Spyǐ-ti sde (fig. 355). Additionally, there is a non-Buddhist swastika dedication engraved in Ru-thog (fig. 349). To my knowledge, swastika dedications are not found in the Tibetan epigraphy of Ladakh or Northern Pakistan.

The paucity of mchod-rten dedications in Upper Tibet in the format seen in Ladakh and Northern Pakistan is a curious phenomenon. The historical factors explaining the epigraphic gulf between Upper Tibet and its northwestern neighbors remain uncertain, but I will expound upon those I think best account for the facts on the ground. Inscriptions with clan, administrative and military nomenclature in Ladakh are concentrated in Lower Ladakh (below the confluence of the Indus and Zanskar rivers). They hardly occur in Upper Ladakh. This encourages one to look for political or cultural factors to account for major differences in the epigraphic record of Upper Tibet/Upper Ladakh and Lower Ladakh/Northern Pakistan.* In fact, there is mounting archaeological evidence for an ancient cultural divide between Upper Ladakh and Lower Ladakh.†

Tibetan inscriptions in Lower Ladakh occur at A-lci, Ba-lu mkhar, Khalatse, and many other locations. Those in Northern Pakistan are found in the regions of Chitral, Punyal, Indus Kohistan and Baltistan, etc.

This cultural divide is corroborated by the post-doctoral field research of Quentin Devers. Based on the contrasting monumental, rock art and epigraphic assemblages in Upper Ladakh and Lower Ladakh, Devers has identified a major cultural and political divide around the confluence of the Indus and Zanskar rivers. According to his findings, Upper Ladakh came under powerful Tibetan cultural influence while Lower Ladakh was receptive to Kashmiri, Dardic and Turkic cultural inputs before the period of Western Tibetan political ascendancy (10th century CE). The plan and construction of fortresses types in the two parts of Ladakh reflect this cultural watershed (as do types of rock art figures). For example, it is only in Upper Ladakh where all-stone corbelled buildings were raised, an architectural tradition closely associated with ancient Upper Tibet. See Devers forthcoming; Devers et al. 2016; June 2013 Flight of the Khyung.

In the Imperial period, an indeterminate portion of Upper Tibet (centered in Western Tibet) and possibly Upper Ladakh belonged to the Zhang-zhung territorial division of the Tibetan empire.* Like other outlying but integral parts of the Tibetan imperial polity, Zhang-zhung was partitioned into civil cum military units known as stong-sde. Zhang-zhung adjoined the core imperial entity of Central Tibet (Spu-rgyal), and the Brahmaputra (Gtsang-po) river valley provided an easily traversable conduit between them. Although the prehistoric archaeological record indicates that there were significant cultural differences between Upper Tibet and Central Tibet,† Old Tibetan and Classical Tibetan literature voice a longstanding cultural and ethnical kinship between the two regions. Tibetan historical and quasi-historical documents portray Zhang Zhung and Central Tibet as politically separate, but often as culturally and religiously complementary countries (Bellezza 2008). It is written in sundry Classical Tibetan sources that Upper Tibet and Central Tibet shared early religious ties in terms of the doctrines, personnel and institutional organization that came to be called bon. These two core regions had twin primary lineages in common, the Dmu and Phywa, suggesting an interlinked cultural and ethnical arrangement.‡ Moreover, an Old Tibetan ritual tradition maintains that Upper Tibet and Central Tibet, from the headwaters of the great rivers in southwestern Tibet to Rkong-po in the east, were constituent parts of a non-Buddhist ritual entity in distant times (cf. Dotson 2008: 56–60).

See Zeisler 2009-2010; Bellezza in press-b; June 2013 and May 2015 Flight of the Khyung.

The unique pre-7th century CE archaeological monumental assemblage of Upper Tibet indicates that the region was culturally distinct in many ways from Central Tibet prior to the Imperial period. See Bellezza 2008, pp. 24, 70, 569–578. The pilgrim Huichao on his return home in 727 CE noted that Yang-t’ung (Zhang-zhung) had completely different clothing, language and customs from those of Central Tibet (Scherrer-Schaub 2002: 271; Zeisler 2009-2010: 379). However, Huichao describes Yang-t’ung and two other countries (P’o-lü and So-po-tz’u) as ‘narrow’ and ‘rugged’ where grapes were cultivated and with Buddhist monasteries (Scherrer-Schaub ibid.; Zeisler 2009-2010: 387, 406, 407), hardly fitting the geographical or cultural character of Upper Tibet in the early 8th century CE. This description and the location of P’o-lü (Bolor: Gilgit and Baltistan) and probably that of So-po-tz’u (appears to be one of the northwestern lands commonly called Sog by Upper Tibetans) suggests the author had Lower Ladakh in mind when referring to Yang-t’ung (cf. Zeisler ibid.). If so, this region appears to have been incorporated into the Tibetan imperial administrative province of Zhang-zhung.

Dagkar (1999) furnishes examples of the regional interplay between the Dmu and Phywa/Phya lineages.

The cultural interrelationship between Upper Tibet and Central Tibet is likely to have become more pronounced as the latter strengthened politically in the late 6th century CE. Thus, it is possible that the annexation of Zhang-zhung by the Tibetan empire in the 640s CE may have been viewed by the conquerors more in terms of settling an internecine score rather than the vanquishment of an alien nation. Be that as it may, there is hardly any epigraphic evidence for troops and administrators garrisoned in Upper Tibet. Such personnel did not see fit to chronicle their presence there in rock inscriptions. The same can be said of Upper Ladakh and Central Tibet. On the other hand, the Tibetan epigraphic corpus of Lower Ladakh and Northern Pakistan (and Wakhan) appears to stem from military campaigns and the annexation of lands not intimately bound to the Tibetan cultural realm. Rather, lands to the north and west were probably considered foreign (non-Tibetic) territories, prompting those literate in Tibetan to leave their mark there in inscriptions in pursuance of their military and political mission. In doing so, they furnished information about their ancestry, rank and profession often in conjunction with religious activities.

There are not many examples of proper names of clans, people and places in the epigraphy of Upper Tibet. Several occurrences of the Khyung-pho (Khyung-po) clan in Early Historic-period inscriptions of Ru-thog strongly suggest that it was native to the region (figs. 332, 333, 337, 349, 358). This is supported by clan treasury literature (rus-mdzod) and the Upper Tibetan oral tradition, which localize the origins of the Khyung-po clan in Zhang-zhung (Bellezza 2008: 258–269, 288, 289).* As per epigraphs already published, this clan does not appear to be documented in the rock inscriptions of Lower Ladakh and Northern Pakistan.† However, it does occur in a Tibetan mchod-rten dedication from Wakhan that probably dates to the Imperial period (Mock 2016: 132).‡ There is one other prominent clan name recorded in an inscription in Far Western Tibet (Sgar) dating to the Early Historic period: Co-gru (fig. 329).§ This clan does not occur in the epigraphy of Lower Ladakh or Northern Pakistan of which I am familiar. Two other important clans, Sgro and Snyags, are referred to in an inscription in the Eastern Byang-thang that appears to predate 1000 CE (fig. 232). Moreover, very few personal names are seen in the inscriptions of Upper Tibet and those mentioned are mostly religious personalities recorded by third parties. The proper names of people are also found in one of two inscriptions that may be related to the ritual disposal of mortal remains (crematory ashes; figs. 123, 254). In addition to Spyǐ-sde noted above, place names mentioned in the inscriptions of Upper Tibet include Lug do (Sheep Headland) and the name for Central Tibet, Bod (figs. 134, 207, 210, 253, 355).

Khyung-po is still found in the valleys of Ru-thog, where it is classified as one of the eighteen major clans of the district (Korpon Lobsang Khenrap 2000: 53). In the text Bshad mdzod chen mo, Khyung is listed as one of the six clans that ruled Mnga’-ris. The other five are Cog-ro, Cog-rtse, Thang-dkar, Bre, and Wa-na. See Norbu 2009, pp. 67. There is an ancient inscription mentioning the Khyung-po clan in the cis-Himalayan region of Bharmour. See Thakur 2001, p. 22; June 2015 Flight of the Khyung. The Khyung-po clan was also dispersed in Central Tibet in the Imperial period. For instance, the famous minister Khyung-po spung-sad zu-tse had his stronghold at Khri-bom, which appears to be in or near Mang-dkar/Mang-mkhar, now in Lha-rtse county (Dotson 2012: 182).

Clans in the epigraphs of Lower Ladakh published thus far include Smer, ’Bro, Dbas, etc. See Takeuchi 2012b; Orofino 1990; Denwood 1980.

The Khyung-po clan is mentioned on a damaged wooden slip recovered in Mazar Tagh, Eastern Turkestan. This slip concerns a corporal (’og-pon) from the ’O-zho-pag regiment (sde) named {Snang}-kngo (= kong). For this slip, see Thomas 1933, p. 559. The ’O-zho-pag regiment can be identified with ‘O-co-bag, one of the five divisions (stong-sde) of Upper Zhang-zhung listed in Lde’u chos ’byung (ca. 12th century CE). This territory was probably situated in northwestern Tibet or farther west on the fringes of the Tibetan Plateau. On the Lde’u chos ’byung list of stong-sde, see Bellezza 2011, pp. 58, 59.

On this clan, see Vitali 1996, passim; Dotson 2012, pp. 185, 186.

Terms for rank and vocation are also rare in the epigraphy of Upper Tibet. Generic terms for ‘leader’ include dpon and mi’go (figs. 213, 253). The meagre reference to people, places, activities, and professions in Upper Tibet is explainable by the preponderance of strictly religious materials in the rock inscriptions, most of which are mantras (sngags). Hence, the epigraphy of the region served primarily as an instrument for the propagation of religious doctrines and sentiments and for staking out sectarian territory. Around half of the corpus of Upper Tibetan rock inscriptions belongs to non-Buddhist religious traditions, reflecting the significance of such groups in the region until the 13th century CE. For the Imperial period, there are few inscriptions that can be confidently assigned a Buddhist orientation. Many inscriptions of the Vestigial period herald encounters between Buddhist and non-Buddhist religions. As we shall see, the deliberate erasure and superimposition of inscriptions is graphic evidence for tensions between these two categories of religious practitioners.*

On this sectarian rivalry as seen in the epigraphic record, also see Bellezza 2008, pp. 188, 189; November 2012 Flight of the Khyung.

The tiny number of rock inscriptions in Upper Tibet pertaining to political activities, their brevity and the absence of official communications means that they have little relevance to the famous pillar edits of Central Tibet. The pillar edicts were erected in prominent locations such as Zhol and the Jo-khang in Lhasa, the confluence of the ’Phan-po chu and Skyid chu, and in villages along the banks of the Yar-lung gtsang-po. Pillar inscriptions often deal with matters of state; viz., military campaigns, treaties with foreign powers, and the divinity and majesty of Tibetan kings.* There are Buddhist pillar inscriptions in Central Tibet assigned to the 11th century (Richardson 1985: 149), but those upholding the prerogative of the btsan-po are assigned to the Imperial period, while imitative styles are generally dated to the post-Imperial period (ca. 850–1000 CE). There is also a Buddhist pillar inscription with a relief carving of the deity Spyan-ras-gzigs in Zhi-bde, Spu-rang, that may date to the early 10th CE.†

On the pillar edicts of Central Tibet see, for example, Richardson 1985; 1998; Iwao 2009; Walter and Beckwith 2010; Scherrer-Schaub 2014.

For a preliminary study of this pillar, see Tshe-ring chos-rgyal and Zla-ba tshe-ring 1994; Zla-ba tshe-ring et al.-a. 2000, p. 173; Vitali 1996, pp. 168, 169; Denwood 2007, p. 47. Vitali (ibid.) dates the pillar inscription to the mid-9th century CE, but given its paleographic characteristics and the stylistic traits of the carved deity, as well as the foundation period of the Buddhist Guge kingdom, a later date seems indicated.

The oldest rock inscriptions in Upper Tibet were written in Old Tibetan, a literary form of the Tibetan language that first appeared in the early Imperial period (7th century CE). In the post-Imperial period, Old Tibetan evolved towards Classical forms of the language which made their debut in the 10th and 11th century CE and still prevail today. The scripts (yig-rigs) used to write Tibetan evolved over time, reaching a mature stage of development with the advent of Classical Tibetan. Classical Tibetan scripts persist to the present day.

The chronology of the development of Old Tibetan and the transition to Classical Tibetan are still not well understood, as dated textual benchmarks and other historical indications are limited in scope. The pillar edicts of Central Tibet constitute the most comprehensively dated corpus of Old Tibetan materials, but they are largely restricted in content to special administrative and royal functions, reducing their utility as dating tools for other types of Old Tibetan materials. Based on grammatical, paleographic, semantical, and stylistic traits, discrete stages in the evolution of Old Tibetan have been put forward.* While these various approaches to understanding the evolution of the Tibetan language have considerable merit, studies carried out to date are preliminary in nature. Corroboration of the chronology and the utility of criteria set forth by researchers is still pending.

Takeuchi (2013) proposes a three-step evolutionary process for Old Tibetan: 1) formation of literary Old Tibetan (7th–8th centuries CE), 2) dissemination from the Tibetan Plateau to Central Asia (8th–9th centuries CE), 3) lingua franca status (9th–11th centuries CE). Although Takeuchi’s work appears to be based on a sound evolutionary pattern, it is mainly concerned with the geographic diffusion of Old Tibetan and only briefly focuses on text-internal indications (grammar, vocabulary, syntax, semantics, script, format, etc.). On codicological criteria usable in the dating of Old Tibet texts, see Scherrer-Schaub 1999. For a broader typological approach to dating Old Tibetan texts focusing on Ta-po monastery holdings, see Scherrer-Schaub and Bonani 2002. Radiocarbon analysis of text samples in the Scherrer-Schaub and Bonani study raises questions about the application of typological criteria to dating Tibetan manuscripts. On methodological parameters applied to dating Tibetan texts relying on radiocarbon analysis, see Helman-Ważny 2014. Bellezza 2013 reviews changing orthographic patterns and verb morphology in non-Buddhist funerary texts as an indicator of the chronological development of Old Tibetan, a preliminary step in such an analysis. Two studies by van Schaik (2011; 2014) are concerned with paleographic development as a method to date Old Tibetan texts, slips and inscriptions.

The practical development of a culture of literacy appears to have taken different paths on the Western Tibetan Plateau (WTP). The corpus of Tibetan language inscriptions from the Early Historic period in Lower Ladakh and Northern Pakistan demonstrate that a cross-section of the elite population (indigenous and/or occupying) had attained some knowledge of writing in the century or two after the invention of the Tibetan script. While the rock inscriptions of Lower Ladakh and Northern Pakistan are usually not long (most are less than fifteen syllables), they reveal a basic mastery of Old Tibetan grammatical, orthographic and calligraphic conventions. This mastery is less evident in the epigraphy of Upper Tibet in the Early Historic period.* As explained, there are few inscriptions there matching the content and style of those in more northern and western territories. Furthermore, Upper Tibet inscriptions in the Early Historic period tend to be shorter, fewer in number, and exhibit less calligraphic and technical refinement. The disparate epigraphic picture of the WTP supports an observation made by Scherrer-Schaub (2002: 276), in her study on the integration of Buddhism into the administrative apparatus of the Tibetan imperial state, that Buddhism came to the Tibetan Plateau from different regions and in successive stages.

The generally higher technical competence of engravers in Lower Ladakh and Northern Pakistan is also seen in facsimiles of mchod-rten accompanying Tibetan inscriptions. Many tend to be more elaborate and finely executed than mchod-rten in Upper Tibet. For a comprehensive inventory of rock art mchod-rten in Upper Tibet and comparison with Ladakhi and Northern Pakistan variants, see Bellezza 2017-2018; in press-a.

A variety of historical and cultural conditions seem to account for the better developed Tibetan epigraphic tradition in outlying regions of the WTP and beyond. As elaborated above, political and military elements underpinning territorial consolidation and organization must have been instrumental in the creation of rock inscriptions in Lower Ladakh and Northern Pakistan. The proximity of these regions to literary centers in northern India would also have been pivotal in stimulating interest in reading and writing among Tibetan speakers on the western margins of the empire. Another vehicle for the spread of literacy in Tibetan controlled regions was Buddhism,* as epigraphs belonging to this religion in Lower Ladakh and Northern Pakistan demonstrate.† Furthermore, epigraphers in Lower Ladakh and Northern Pakistan may have been inspired by older inscriptions in a variety of languages and scripts, many of which embody Buddhist themes.

For a review of the role of literacy and state-sponsorship in the transregional propagation of Tibetan Buddhism during and in the aftermath of the Imperial period, see van Schaik 2016.

For example, personal names such as Rdo-rje, Arya-dpal, Rgyal-ba ye-shes, Bdud-’joms, and Shes-rab (see Takeuchi 2012b, inscription nos. 10, 11, 17, 33, 47, 49, 50, 73) are indicative of Buddhist adherents, as is the use of the sexagenary cycle of dating and the style of some mchod-rten accompanying inscriptions in Lower Ladakh. Reference to the sexagenary cycle does not occur in the rock inscriptions of Upper Tibet. However, Takeuchi’s assertion (ibid. 55) that the carvings appearing alongside inscriptions in Lower Ladakh are “totally Buddhistic” is an over-generalization. In fact, there are different types of rock art mchod-rten and other kinds of tiered shrines associated with Tibetan inscriptions in Lower Ladakh (for a few of these, see Devers et al. 2016). Except where inscriptions feature Buddhist personal names, terminology or mantras, a non-Buddhist or syncretistic religious identity must also be considered. Rudimentary Mchod-rten with trident-like finials in Upper Tibetan rock art are almost exclusively connected to non-Buddhist religious traditions. Analogous rock carvings are found in Lower Ladakh and Northern Pakistan, and they too may have been created by non-Buddhists. See Bellezza 2017-2018; 2002a pp. 128, 129, 144, 145. That such mchod-rten in Northern Pakistan are likely to be the work of a ‘folk religion’ coexisting with Buddhism or of ‘Bon’ practitioners, see Jettmar 1982, p. 29; 1985. Francke attributes ‘primitive’ mchod-rten rock art in Ladakh to a people who were probably illiterate (1903: 362).

The occurrence of dozens of Tibetan non-mantric inscriptions some dating to the Imperial period in Lower Ladakh and Northern Pakistan but few in Far Western Tibet raises crucial questions about how literacy might have spread onto the Tibetan Plateau in the 7th century CE. The invention of the Tibetan alphabet occurred ca. 630–650 CE and began to be used officially sometime thereafter.* The Tibetan script was modelled on the northern Brahmic family of Indo-scripts. The genesis of the Tibetan alphabet coincides with the use of cognate Brahmic scripts in the early 7th century CE, and probably predates the emergence of proto-Śarādā, a western variant of a late Gupta appearing in Kashmir, Gilgit and the southern Tarim Basin in the 7th or 8th century CE. Marshaling textual, epigraphic and paleographic evidence, van Schaik (2011) settles upon the late Gupta script of eastern Indian lands such as Magadhā as the most likely prototype for the development of the Tibetan script. Another potential source for Tibetan scripts is a northern Brahmic variant from the plains of the Indus, adjoining hilly tracts and beyond. In a painstaking comparative study of graphemes of Brahmic and early Tibetan inscriptions, Schuh (2013) shows that the scripts of Kashmir, Gilgit, eastern Afghanistan and the southern Tarim Basin are likely to have contributed much to the creation of Tibetan writing.† Schuh (ibid., 176) holds that a combination of innovation and the adoption of templates from the Indian Buddhist cultural world, particularly in the north, account for the development of Tibetan writing. Quite convincingly, Schuh (ibid., 172) suggests interregional trade and affairs of state as major mechanisms for this adoption by Tibetans.

On this subject, see further Hill 2010, p. 110; Takeuchi 2013, pp. 3, 4.

According to the Old Tibetan Annals, the defeat of the Zhang-zhung kingdom occurred ca. 644 CE (Dotson 2009: 82; Uray 1972b: 35, 41). It is widely agreed that the Tibetan conquest of Ladakh took place in the early 8th century CE and that of Northern Pakistan by the middle of the 8th century CE (for details of these conquests see, Beckwith 1987; Orofino 1990; Jettmar 1993; Petech 1977; Tucci 1977). On Tibetan imperial military campaigns and the occupation of Wakhan, see Beckwith ibid.

The history of Tibetan occupation of western lands indicates that Upper Tibet had considerably more chronological leeway than Lower Ladakh, Northern Pakistan and Wakhan for mastering Tibetan literacy, an opportunity that appears to have been largely missed. While Tibetan literacy infused northwestern regions, Upper Tibet resisted the trend to create rock inscriptions, a good indication that Stod and the Byang-thang enjoyed significantly less literary output in the Imperial period.

Despite lying nearer to the Imperial-period scribal centers of Buddhist institutions in Central Tibet, fewer rock inscriptions on the Byang-thang than in Far Western Tibet can be confidently dated to the Imperial period. This suggests that the high tablelands were more insulated from literary induced cultural changes than was Ru-thog, which was closer to the well-established Indo-Buddhist cultural realm of northern and western lands.

The types of scripts and the presentation of rock inscriptions in the Early Historic period vary in the east and west of Upper Tibet (for discussion, see “Paleographic typology”, infra). Regional religious and paleographic contrasts suggest that different geographic sources were involved in the formation of the Tibetan alphabet. A northern variant of Brahmic was the likely prototype for the range of scripts seen in the epigraphy of Far Western Tibet (Sgar and Ru-thog), Ladakh, Northern Pakistan, and Wakhan. This northern tradition of Tibetan writing was closely bound to the diffusion of Buddhism in the wider region, as the presence of a few early Buddhist epigraphs in Far Western Tibet indicates. On the other hand, the abecedarian scripts of the Eastern Byang-thang epigraphic record may owe their inspiration to an eastern Indian source. This group of elementary scripts is not directly linked to Buddhist thought or practice. These observations on possible bifocal origins of scripts, derived from the areal distribution of rock inscriptions in Upper Tibet, are lent support by the paleographic findings of Schuh (2013), who concludes that multiple geographic sources probably influenced the development of Tibetan scripts for use in various dialects.

The scarcity of mchod-rten dedications and inscriptions of the Early Historic Period suggests that the acquisition of literacy was generally retarded in Upper Tibet. This must be related to the strength of non-Buddhist religious traditions in Upper Tibet, the region being fairly impervious to Buddhist inroads until the second half of the 10th century CE and even later.* The imposition of Buddhism as the state religion by King Khri-srong lde’u-btsan in the last quarter of the 8th century CE does not appear to have radically affected the Upper Tibetan religious scene. In any event, there is no indication in the epigraphic and archaeological records of the Early Historic period of all-out Buddhist activities in Upper Tibet. Tibetan historical texts do mention a few Buddhist monasteries founded in Far Western Tibet during the Imperial period, but these seem to have been outposts or peripheral to the main thrust of religious observance in Upper Tibet at that time.† As noted, there are few rock inscriptions in the region clearly identifiable as Buddhist which can be assigned to the Imperial period.‡

Even at the beginning of the 11th century CE, non-Buddhist religion or bon in western Tibet was well entrenched, requiring royal authority and concerted measures to contain it and permit Buddhism to gain traction. This is documented in the edicts of Lha bla-ma ye-shes-’od, the first biography of Lo-tsa-ba Rin-chen bzang-po and in historical texts such as Mnga’ ris rgyal rabs. For these references, see May 2015 Flight of the Khyung (“Archaic cultural traditions in the biography of Lotsawa Rinchen Sangpo” and “The decree of Lha Lama Yesheö”); Karmay 2014; Vitali 1996, pp. 112, 220–231. The Stag lung chos ’byung places the elimination of bon-po and other non-Buddhist groups at Gnam-mtsho in the first quarter of the 13th century CE (Bellezza 1997a: 169–173). Vivid accounts of the defeat of the bon-po in the 11th to 12th centuries CE are recounted in the Mi la ras pa’i mgur ’bum. On Mi-la ras-pa’s conversion activities, see Chang 1999, pp. 241–258. In one tale in this hagiography, the Buddhist saint Mi-la ras-pa (ca. 1052–1135 CE) sways bon-po by tailoring his songs to the structure and recitational style of the preexisting non-Buddhist liturgical tradition, beginning with a mock exposition of ritual origins (smrang). In his nine songs, Mi-la ras-pa replaces the augury, astrology, subjugation of demons, propitiation, sacrificial offerings, feast, and coda of the non-Buddhist tradition with Buddhist ethical, philosophical and spiritual teachings. While the exploits of Mi-la ras-pa are heavily mythicized and paradigmatic in nature (so holy was he that sometimes when walking his feet are said not to have touched the ground), his concerted efforts to bring bon-po to Buddhism bespeak major religious changes occurring in much of Tibet in the early Vestigial period.

The two best known Buddhist temples in Upper Tibet during the so-called bstan-pa snga-dar were Dpal-rgyas and Skra-bdun/Pra-dum. On these temples, see Vitali 1996, pp. 267, 276. On other early Buddhist temples in Upper Tibet, see Bellezza 2014a, pp. 311–322.

I have documented a few Tibetan rock inscriptions dating to the Early Historic period in Spiti, as have members of the Spiti Rock Art and Historical Society. None of these have explicit Buddhist content. Buddhist rock inscriptions in Spiti seem to arise with the bstan-pa phyi-dar and the founding of Ta-po monastery in the late 10th century CE. Two of the inscriptions recorded note Yul tapo/Yul ta-po.

In the Upper Tibetan inscriptions of the Early Historic period there is just one faint indication of a proper name that may be partially derived from Zhang-zhung, an early Tibeto-Burman language(s) (fig. 358). This lack of non-Bodic onomastic content can be related to the highly limited occurrence of personal names in the epigraphy of the region more generally. In Lower Ladakh the non-Bodic name Rye-shin (Takeuchi 2012b, no. 17) may possibly be of Zhang-zhung origins. Other personal names in the inscriptions of Lower Ladakh also appear be of non-Bodic origins, including Lca ang-ri-nu, Co-ku and Ka-shi (ibid. nos. 23, 51, 70). In Tibetan inscriptions from Wakhan documented by John Mock, we find Mu-ge ligs-cung and Ru-shin pya-ligs (Bellezza 2008: 182 [n. 191]). These are personal names, the syllables of which are attested in the Zhang-zhung language(s) However, the word elements (mu-ge, ligs-cung, ru-shin, pya-ligs) do not appear in extant Zhang-zhung lexicons. Thus, these may be Tibetan transliterations of names in other languages such as Turkic variants.* After 1000 CE, when the epigraphic output of Upper Tibet expanded, it appears that Zhang-zhung as a spoken language was already extinct.†

Mug-lig has been put forward as the Tibetan cognate of the Turkic Bükli. See Venturi 2008, pp. 22, 23 (after Clauson 1957; Ligeti 1971; Moriyasu 1980).

Nagano (2014a: 1, 2014b: 131) estimates that the Zhang-zhung language became extinct in the 9th or 10th century CE. It is generally supposed that it was gradually replaced by Tibetan after the Central Tibetan conquest of the Zhang-zhung kingdom in the mid-7th century CE (cf. Martin 2010: 5).

I find it puzzling that the epigraphic record of the Early Historic period does not contain definitive linguistic evidence for Zhang-zhung, Sum-pa or other archaic languages of the Tibetan Plateau. Related languages are still spoken on the western edge of the Plateau in Lahoul and Kinnaur (ancient inscriptions in the earliest native languages of these regions have not been discovered). Although earlier rock art in Upper Tibet was presumably created by peoples with such archaic languages, no attempt was made to assert their linguistic identity in rock inscriptions. These speakers resisted wholesale the adoption of the Tibetan system of writing for what must have been a set of highly persuasive cultural and political reasons.

Rock inscriptions of the Vestigial period

The ancient rock inscriptions of Upper Tibet constitute a chapter in the cultural history of the region that ended six or seven hundred years ago. A historical indication of the terminus ante quem of painted inscriptions is found at Gnam-mtsho, where more than 90% of these types in Upper Tibet are located. Biographies of Stag-lung bka’-brgyud masters (’Bring-ston shākya rin-chen, Kong-po dar-shes, etc.) in the Stag lung chos ’byung describe the victory of this subsect over non-Buddhists (called bon-po and ma-sangs) during the first half of the 13th century CE (Bellezza 1997a: 169–173). This work (written by Zhabs-drung Ngag-dbang rnam-rgyal in 1609 CE) is unambiguous in equating the triumph of the Stag-lung-pa with the elimination or displacement of non-Buddhist groups at Gnam-mtsho.* Approximately half of the inscriptions at Gnam-mtsho were made by non-Buddhists and many Buddhist examples added in response to them through erasure, juxtaposition and superimposition. Thus, inscriptions at Gnam-mtsho reflect the confrontation between Buddhists and non-Buddhists referred to in the Stag lung chos ’byung.

The seat of this sect was Dpal stag-lung thang, a major political, economic and religious center in northern regions, which enjoyed the support of prominent clans on the Byang-thang. See Akester 2016, pp. 37, 38.

Paleographic evidence suggests that the rivalry between the two major religious factions at Gnam-mtsho began by the late 10th or 11th centuries CE. That a religious struggle or transition was occurring in the early years of the Vestigial period is suggested by the recorded presence in the region of Buddhist luminaries such Mi-la-ras-pa, Rgyal-ba lo-ras-pa and Rgwa lo-tsa-ba in the 11th and 12th centuries CE (Bellezza 1997a). Similarly, farther east on the Byang-thang, the establishment of Gzims-phug monastery in 1085 CE appears to have led to the conversion or displacement of non-Buddhists (Bellezza 2014a: 446, 447). The encounter between the Buddhists and non-Buddhists at Gzims-phug is documented in its rock art and epigraphy. As discussed above, there are also Tibetan documents chronicling the tense relationship between Buddhists and non-Buddhists in Far Western Tibet in the 11th century CE.

The paleographic and grammatical features of the corpus of Upper Tibetan rock inscriptions indicate that the use of red ochre and other mineral pigments to write on stone surfaces fell out of favor by the end of the Vestigial period. Presumably seen as old-fashioned or too closely aligned with more ancient rock art, ca. 13th or 14th century CE, the regular production of painted rock inscriptions ceased.

The same chronology for the discontinuation of painted inscriptions is applicable to the termination of the old tradition of engraved rock inscriptions (yig-rigs brkos-ma) in Upper Tibet (and other areas of the WTP). Paleographic and grammatical evidence indicates that the carving of letters directly into stone surfaces was disfavored after the 13th or 14th century CE. Beginning in the late 10th or 11th centuries CE, relief carving of Buddhist mantras on portable and fixed stones surfaces became commonplace in Upper Tibet.* Unlike the cutting and grinding techniques used to make rock inscriptions, the new type of engraving produced raised letters, the surrounding areas of stone being removed from the composition. The more sophisticated relief carvings eventually came to replace the direct cutting of letters into stone surfaces as the primary epigraphic tradition on the Western Tibetan Plateau. As with inscriptions written with mineral pigments, it appears that the older carving techniques were perceived as obsolete, as likely as not for being too closely associated with the earlier non-Buddhist culture of rock art figuration.†

Buddhist and non-Buddhist mantras on portable stones produced using the more primitive technique of directly cutting letters are also found in Upper Tibet but they are rare. These artifacts can be dated to the Early Historic period. See April 2011 Flight of the Khyung; Bellezza 2014a, p. 312–314, 468. For an epigraphic survey of Ma-ṇi mantras carved on portable stones (and the construction of walls to house them) and their function for gaining merit in Spiti, see Laurent 2017.

It must be pointed out, however, that the simple line carvings of mantras on stone surfaces has continued in Tibet and other countries where Tibetan Buddhism spread to the present day. This represents a minor artisanal tradition displaying modern paleographic and physical traits. Of marginal importance to Tibetan epigraphy since ca. 1400 CE, the cutting letters directly into stone is usually associated with less skilled forms of inscription and craftsmanship than the ubiquitous relief engravings.

It is curious that only around 550 epigraphs created in the old technique have been recorded, even making an allowance for an indeterminate number still to be documented in Upper Tibet. This contrasts with the tens of thousands of relief carvings on rock surfaces, fixed and portable, made in the region since the 11th century CE (they are particularly common in Far Western Tibet). For an era that lasted 600 or 700 years that works out on the average to less than one inscription per year painted or carved in the old style in Upper Tibet. Clearly, creating these rock inscriptions was no everyday undertaking. Much of the epigraphic activity of the Vestigial period is linked to confrontation between Buddhist and non-Buddhist religions, suggesting a higher incidence than the overall average for epigraphic activity in places like Gnam-mtsho. Yet, even when focusing on this historical event, the fabrication of old style (snga-ma) rock inscriptions was a relatively rare activity in Upper Tibet.

The scarcity of old rock inscriptions raises questions as to their function and the conditions under which they were produced. Here again we find parallels with the older rock art tradition. I estimate that more than 10,000 individual petroglyphs or pictographs were made in Upper Tibet over a time span exceeding two millennia (Bruneau and Bellezza 2013: 11). On the average, that is approximately five pieces of rock art per year (around 85% of all rock art was produced before the Early Historic period). Whatever the period of production, the themes and subjects depicted in rock art indicate that this mode of representation was largely reserved for chronicling more specialized functions and activities like ceremonial hunting, sacred displays, the communication of myths and legends, and other significant aspects of cultural knowledge. Similarly, special circumstances may be implicated in the production of old style rock epigraphs. This kind of epigraphy was also probably too closely wrapped up with the moribund tradition of rock art figuration. We might conclude that carving and painting inscriptions in the manner of rock art had come to be viewed as passé, dissuading epigraphers from working in the tradition.

To recapitulate: historical factors and the lack of a more recent corpus of rock inscriptions produced in the fashion of rock art support a tail off date of 1300 or 1400 CE for the old tradition of epigraphy in Upper Tibet.* This is supported by the disappearance of non-Buddhist adherents in most of the region by the mid-13th century CE. Moreover, parietal surfaces enrobed in the smooth, hard rock veneer favored by rock artists and epigraphers at Gnam-mtsho were full of mineral pigment applications by the time confrontation between Buddhist and non-Buddhists concluded in the mid-13th century CE. There were hardly any prime surfaces left for inscription after that time. Nevertheless, exceptions to the date proposed for the demise of the old tradition of rock inscription have been documented in this survey. On these later rock inscriptions, see infra, figs. xlix–li.

Part Two: Paleographic analysis of Upper Tibetan rock inscriptions

Paleographic methods of dating

This part of the monograph focuses on the scriptive and physical qualities of rock inscriptions and how they may be organized into a system of paleographic classification. In addition to historical methods of dating rock inscriptions in Upper Tibet, paleographic examination enhances significantly what can be gleaned about their age. Paleographic study permits inscriptions to be categorized in ways that reflect their chronological development. This exercise in seriation is primarily based on the styles of scripts used. Typological study of scripts also aids in the identification of ancient religious affiliations in Upper Tibet.

In recent decades, efforts have been made in the field of paleography to become more scientifically rigorous by precisely recording variations in graphic patterns. Ductus, or the configuration of the strokes of letters, has been assessed for Tibetan in preliminary studies of typological classification by van Schaik (2014) and Takeuchi (2012a). The strokes or lines making up letters are based on directional (straight and curved), pivoting and joining movements. A study of ductus may also be extended to quantifying the flow of ink left on a writing surface.

A stylistic approach is predicated on qualifying the forms of painted and engraved scripts (yig-rigs). It is dependent on an appraisal of the calligraphy (’bri-lugs), which is defined through the formation of letters and grammatical marks (yig-ris).* Stylistic analysis is best combined with an assessment of the physical condition, contents and historical context of inscriptions.

For an exposition of the traditional art of Tibetan calligraphy, see Jamgӧn Kongtrul Lodrӧ 2012.

The physical condition of inscriptions furnishes some indication of their age. Gauging the ablation and browning of pigments as well as the erosion and re-patination of carvings serves as a marker of relative age, particularly for inscriptions on the same rock face. This approach can be extended on a site by site basis, supplying some idea of how rock inscriptions degrade over time. A study of the superimposition of rock inscriptions is another important tool in relative dating. An appraisal of the orthographic, grammatical and lexical traits of inscriptions also aid in better placing inscriptions chronologically and serve to differentiate Old Tibetan and Classical Tibetan traits. Text-based indications are an obvious source for chronological information; however, due to the brevity of most rock inscriptions in Upper Tibet, an analysis of prose and semantics is of limited utility. Finally, as we have seen, Tibetan historical literature can assist in broadly dating inscriptions, supporting paleographic and grammatical study.

Despite the application of a broad set criteria to dating rock inscriptions in Upper Tibet, the development of a firm chronology remains elusive. As none of the Upper Tibetan inscriptions have been dated through independent means such as (radiocarbon analysis), the chronology presented in this work must be viewed as provisional and open to modification should new sources of information require it.

Old Tibetan of the Early Historic period was written in non-cursive block letters (gzab-ma) known as dbu-can (letters with heads). There were several types of dbu-can scripts (yig-gzugs) in use in the Imperial period and post-Imperial period, which were inscribed on various types of writing surfaces (’bri-gzhi), including stone, wood, metal, paper, and plaster. The choice of scripts in the Early Historic period varied according to the purpose for which they were intended (van Schaik 2011; 2014). Many rock inscriptions assignable to the Early Historic period in Upper Tibet display florid scripts, typically set down in lines of syllables (and words) that are unruly or irregular. These kinds of scripts are also encountered in non-Buddhist mytho-ritual texts of the Dunhuang collection, the earliest group of which on paleographic, grammatical, lexical, ideological, and text format grounds can be assigned to the Imperial period with a fair degree of confidence.

I refer to the full range of more primitive scripts of the Early Historic period as “rudimentary dbu-can”, in contradistinction to standard types of dbu-can scripts as seen in canonical works, official documents, etc. Rock inscriptions in rudimentary forms of dbu-can represent a fledging stage in the evolution of writing and literacy in Upper Tibet.* With notable exception, these inscriptions tend to be short (under 10 syllables) and without punctuation or head marks. Rudimentary dbu-can is related to a class of scripts called by van Schaik (2013, 2014) the ‘square style’, which was used to pen a heterogeneous corpus of mytho-ritual manuscripts and other works in the Imperial period.

For a traditional view of the historical development of dbu-can scripts in the Imperial period see, Jamgӧn Kongtrul Lodrӧ. 2012, pp. 251, 252.

Rock inscriptions in rudimentary dbu-can typically display Old Tibetan orthographic and grammatical conventions. Although there are relatively few inscriptions attributable to the pre-1000 CE period, they are varied in contents and include clan and personal names and some ritual matter. However, unlike inscriptions of the post-1000 CE period, most of those assigned to the Early Historic period are not pietistic or devotional in character and hardly include complex mantras or professions of faith.*

In my study of non-Buddhist funerary ritual manuscripts from Dunhuang assigned to the Imperial period, the use of mantra (sngags, snying-po) is not observed. It is with the mytho-ritual texts of Gathang Bumpa of the post-Imperial period that mantras first appear in non-Buddhist ritual dispensation, albeit in very limited numbers. See Bellezza 2013, p. 252; 2014d, p. 192.

Paleographic typology

Early Historic period

In this work there are few rock inscriptions attributed specifically to the Imperial period; most early examples are more broadly dated to the Early Historic period (particularly to the post-Imperial phase). Reliance on a broader chronological category is necessitated by historical and paleographic uncertainties militating against finer grain periodization of Old Tibetan inscriptions in Upper Tibet.

Fig. i. Inscription engraved in the ‘epigraphic style’, Ru-thog

This inscription is related to what has been called the ‘epigraphic style’ of dbu-can, a script assigned to the Imperial period (see fig. 355).* This style of script consists of upright square letters most famously associated with the pillar edicts of Central Tibet of the Imperial period. Located in Ru-thog, this rock inscription also exhibits more irregular features associated with van Schaik’s ‘square style’, a script that may have evolved from the ‘epigraphic style’. The Ru-thog inscription was neatly engraved on an open rock face and conforms quite closely to the more refined state sanctioned calligraphic conventions used for official communications in the Imperial period. It was written in a more steady and schooled hand than rudimentary dbu-can rock inscriptions in Upper Tibet, suggesting, as does the contents of the inscription, that the epigrapher may have been acting in an official capacity.

In addition to the ‘epigraphic style’ and ‘square style’, in his typology of Tibetan scripts of the Imperial period, van Schaik (2014; 2013) includes the ‘sutra style’, ‘official’ style’ and ‘monastic style’. These are not found in the rock epigraphy of Upper Tibet. Through a more thorough study of graphemes including ductus, it should be possible to subdivide van Schaik’s epigraphic and square styles, yielding a more refined system of paleographic classification.

The Ru-thog inscription was indited on a rock face already hosting rock art, one of several such sites in the Ra-bang-Khul-pa valley. It is found in a conspicuous location in which most traffic through the valley must pass. The inscription may therefore have been of regional significance, composed perhaps by an individual passing through the east-west running valley on official business. On a grander scale, the longer and more adeptly inscribed pillar edicts of Central Tibet were tied to the organs of royal power and enjoyed wide visibility. The text of the Ru-thog inscription seems to confirm an Imperial-period date. The personal name belongs to the omasticon of that period and mention of the District of Spiti (Spyǐ-ti sde) accords with what we know of the administrative organization of the WTP during the Tibetan empire. The inclusion of petroglyphs of two thunderbolts (rdo-rje) that appear to have been engraved along with the Ru-thog inscription indicate that these carvings were made by a Buddhist adherent. As such, this is one of the oldest Buddhist rock inscriptions documented in Upper Tibet.

Fig. ii. Another inscription engraved in the epigraphic style, Ru-thog

Like fig. i, this inscription is found in Ru-thog and features paleographic traits of the epigraphic and square style scripts (see fig. 333). Thus, on a paleographic basis, it can probably be attributed to the Imperial period. This dating is supported by the Old Tibetan orthography of the inscription (see clan name: Khyung-pho).

Fig. iii. An inscription in a modified epigraphic script, Ru-thog

This inscription is a rougher rendering of the epigraphic style script (see fig. 358). Like figs. i and ii, it may date to the Imperial period but this is less clear. In any case, the Old Tibetan paleography and orthography of the inscription indicate that it belongs to the Early Historic period.

Fig. iv. Red ochre inscription with primitive letters, Gnam-mtsho

This is another rock inscription that is probably attributable to the Imperial period (see fig. 136). It is found on the opposite end of Upper Tibet along the shores of Gnam-mtsho, on the headland of Bkra-shis do-chen. In this highly eroded inscription, the formation of the letters exhibits archaic characteristics, some of which are so unusually constituted that their reading is questionable. Hardest to read are the final three characters of the second line.

There are other rock inscriptions in Upper Tibet consisting of just the mystic syllable A or Oṁ, which were also written in a rudimentary dbu-can script. These inscriptions are tentatively assigned to the Early Historic period, but caution in their dating is warranted. An inscription of just one syllable (yig-’bru) furnishes little graphic or grammatical evidence on which to construct a paleographic typology. Furthermore, primitive looking does not necessarily mean archaic; crudeness can just be a sign of poor penmanship or carelessness. Still, despite these admonitions, some mystic syllables in the rock inscriptions of Upper Tibet appear to be precursors of the tradition of writing mantras that fully developed in the Vestigial period. These specimens concentrated in the Eastern Byang-thang exhibit distinctive styles of calligraphy not seen in later inscriptions.

Fig. v. A row of letters all reading A written in a rudimentary fashion, Gnam-mtsho

These letters, while superficially resembling the A written in the dbu-med script, seem to represent a bona fide archaic script (see fig. 86). This inscription is sandwiched between large pictographic mchod-rten that were almost certainly painted by non-Buddhists. An Early Historic period date for this inscription is suggested by the style of writing and wider rock art context.

Fig. vi. The letter A or possibly the syllable Oṁ inscribed in a rudimentary fashion, Gnam-mtsho

This style of the letter A (possible vowel sign notwithstanding) is commonly seen in other rock inscriptions in Upper Tibet (see fig. 16). Always consisting of just a single letter or syllable, the elementary script form anticipates the evolution of dbu-med. It appears to represent a precursory type of writing datable to the Early Historic period which developed primarily in the Eastern Byang-thang. The class of rudimentary scripts represented in this inscription and the following two specimens (figs. vii, viii) have not been documented in Far Western Tibet, delineating the paleographic divide described in Part One of this work.

Fig. vii. Two examples of the letter A written in a rudimentary script, Gnam-mtsho

This rock inscription is highly eroded and the red ochre has browned considerably with age (see fig. 184). This is another example of the letter A in an archaic script of the Eastern Byang-thang, which can be dated to the Early Historic period and perhaps to the earliest phase of Tibetan writing.

Fig. viii. The letter A inscribed two times in a rudimentary script and accompanying rock art, ’Brong-pa

Here there is not only the primitive letter A but rock art forming a single composition (as seen in the quality, hue and application of the red ochre used and in the lineation of the letters and figures; see fig. 316). These are the most westerly examples of the letter A written in rudimentary scripts closely associated with the Eastern Byang-thang. Like almost all eastern counterparts, this epigraph was painted in red ochre. The separated sun and moon on the left side of the photograph represents an older style of figuration, as does the rendition of the tree, both of which are known in other rock art of Upper Tibet. The archaic form of the letters and artistic qualities of the figures support an Early Historic period date for the inscription.

Fig. ix. An inscription that appears to include an incomplete Buddhist Ma-ṇi mantra, Sger-rtse

This inscription also exhibits paleographic qualities attributable to the Imperial period or post-Imperial period (see fig. 318). If it belongs to the first phase of the Early Historic period, like Fig. i, it is one of the oldest Buddhist rock inscriptions in Upper Tibet. The word kar appears to be the Old Tibetan variant of the word dkar (white). Note how the letter A assumes a more complex form than the two proceeding examples in figs. vii and viii. The more refined calligraphy and Buddhist orientation of this inscription reflects its placement further west in Upper Tibet. A yak was carved along with the inscription (only the rear legs and belly fringe are visible in the photograph), another indication of its archaic status. Although symbols were sometimes added to inscriptions of the Vestigial period, most zoomorphic rock art produced in conjunction with inscriptions is synonymous with an older tradition.

Fig. x. Mantric inscriptions indited in the square style of script and superimposed on earlier zoomorphic rock art of tigers attacking wild yaks, Gu-ge

This Ma-ṇi inscription (partially visible in the photograph) and Sanskritic affirmatory word were rendered in a script closely related to the official square style of the Imperial period (see fig. 323). However, the variable thickness of the strokes of letters is a calligraphic refinement that may suggest a later date. Thus, creation in the 11th or 12th century CE cannot be ruled out.* This is an excellent example of the more sophisticated script conventions associated with Buddhism that sprung up in Far Western Tibet.

On the underlying rock art, see October 2017 Flight of the Khyung, fig. 33.

The first cursive scripts emerge prior to the post-Imperial period in the Tibetan textual and probably epigraphic tradition as well. I refer to these prototypes of the dbu-med script as ‘rudimentary dbu-med’. These kinds of scripts are fairly well represented in the ochre rock inscriptions of Upper Tibet. Van Schaik (2013: 423) traces the origins of ‘early dbu-med’ to a group of texts attributed to the last phase of the Tibetan empire in the 830s and 840s CE. However, Schuh (2013: 168–172, 174) charts the evolution of dbu-med as already divergent before that time. A form of rudimentary dbu-med commonly seen in documents of the post-Imperial period, such as non-Buddhist mytho-ritual documents of the Gathang Bumpa collection,* seems most relevant to a comparative study of Upper Tibetan rock inscriptions.

For these documents, see Pa-tshab pa-sangs dbang-’dus and Glang-ru nor-bu tshe-ring 2007; Bellezza 2010; van Schaik 2014, p. 322. Takeuchi (2012a; 2013) calls this kind of script semi-cursive or the post-Imperial style, one of several types of scripts that arose after ca. 850 CE.

Fig. xi. An unusually long inscription made in the rudimentary dbu-med script, Gnam-mtsho

The attenuated heads, long vertical legs, flattened vowel signs, and large looped strokes of the letters identify this inscription as an early example of a script precursory to dbu-med (see fig. 182). I am inclined to date this inscription to the post-Imperial period, but an earlier date cannot be ruled out.

Fig. xii. The longest rock inscription in Upper Tibet belonging to the Early Historic period, Gnam-mtsho

This extensive inscription has largely worn away and is mostly illegible (see fig. 135). It is another example of a script presenting the first signs of cursivization. It can be dated to either to the Imperial period or post-Imperial period, the latter phase seemingly more likely.

Fig. xiii. Early inscription with cursive characteristics and head marks, Gnam-mtsho

This inscription is another early example of the early cursivization of Tibetan (see fig. 207). It is best dated to the post-Imperial period or possibly a little later.

Fig. xiv. Early example of a cursive influenced script, Gnam-mtsho

This inscription with its large superscribed la (la-mgo) and wedge-shaded ca may to date to the post-Imperial period (see fig. 4).

There are inscriptions in other types of non-cursive scripts that are assigned to the Early Historic period, which may date either to the Imperial period or post-Imperial period. These scripts are common in Far Western Tibet and more westerly areas of the Tibetan Plateau.

Fig. xv. Early example of the dbu-can script from Ru-thog

This inscription with its Old Tibetan orthography (reads: Khyung-pho) can be attributed to either the Imperial period or post-Imperial period (see fig. 337).

Fig. xvi. Another dbu-can inscription of the Early Historic period from Ru-thog

Again, it is not certain whether this inscription should be attributed to the Imperial period or post-Imperial period (see fig. 329). Perhaps an even later date needs to be considered which would make this inscription a product of the bstan-pa phyi-dar.

There are mantric inscriptions in dbu-can scripts that potentially predate 1000 CE. These include both Buddhist and non-Buddhist variants that can be tentatively assigned to the 10th century CE. The Ma-ṇi mantras in this category are all located in Ru-thog. If dating prior to 1000 CE is warranted, it appears that Buddhist influences emanating ultimately from Kashmir were responsible for their creation.* This class of more cultivated Tibetan scripts is a product of Far Western Tibet and adjoining fringe regions of the Plateau.

There are also inscriptions of Ma-ṇi mantras and the syllable Oṁ in Lower Ladakh that may predate the 11th century CE. For example, see Takeuchi 2012b, pp. 43 (text 73), 44 (text 79), 67 (figs. 50, 53). Takeuchi (ibid., 51) also considers the possibility that such inscriptions might predate the 11th century CE.

Fig. xvii. Ma-ṇi mantra with accompanying mchod-rten

This is an example of a Ma-ṇi mantra that may possibly date to the late 10th century CE, at the beginning of the bstan-pa phyi-dar or perhaps a little earlier (see fig. 350).* However, keeping in mind uncertainties regarding the evolution of paleographic traditions in Upper Tibet, a post-1000 CE date can also be tendered for this epigraph.

As pertains to the Dunhuang manuscripts, Takeuchi (2013: 7 [n. 21]) dates the earliest occurrences of the Ma-ṇi mantra to the post-Imperial period. Although Takeuchi chronological assignment may well be accurate, he prescribes dates for Old Tibetan texts in this article with a minimal amount of evidence presented.

Fig. xviii. Ma-ṇi mantras and autograph that may possibly predate 1000 CE

This is another example of a Buddhist composition in Ru-thog that potentially predates the 11th century CE (see fig. 369). Like figs. xvi and xvii, this inscription is part of a paleographic and artistic tradition centered in Far Western Tibet, Ladakh and Northern Pakistan.

Fig. xix. An early type of Ōm syllable from Gnam-mtsho

This may possibly be yet another example of a mantric inscription that predates the 11th century CE (see fig. 18). The V-shaped superscribed vowel sign and ma written as a vertical line, as well as the truncated lower extension of the A, indicate that this inscription is of considerable antiquity, possibly warranting attribution to the 10th century CE or even earlier. Inscriptions in related scripts are found in Far Western Tibet and more westerly regions of the Plateau.

Fig. xx. Oṁ syllables and the letter A of indeterminate age, Gu-ge

With so little text to go on, the dating of such sacred syllables in earlier style scripts is a moot point (see fig. 324). I include them in this typology as potential examples of mantric inscriptions predating the 11th century CE.

Fig. xxi. Another Oṁ syllable from Gu-ge

The dating of this elementary inscription is unclear, save that it was made no later than the close of the Vestigial period (14th century CE) and the termination of the tradition of carving letters directly into stone substrates. Nevetheless, with its underdeveloped subjoined A (a-chung), assignment of this inscription to the Early Historic period must also be considered.

Fig. xxii. Two non-Buddhist mantras, Gnam-mtsho

These two inscriptions exhibit the A and Oṁ syllable without a subjoined a-chung, a less common occurrence in the Vestigial period than the use of the so-called a-chung, especially in non-Buddhist epigraphs (see fig. 119). This and the formation of the letters may suggest assignment to the Early Historic period. The flaming jewels symbol that appears to have been composed with the mantra on the right (not pictured) is of an archaic type and supports an early date.* As with inscriptions in figs. xvii–xxi, categorization in a chronological category intermediate between the Early Historic period and Vestigial period may be indicated. The most peculiar paleographic feature of the inscription on the left is the curved lower ligature representing the vowel sign zhabs-kyu, another indication of considerable antiquity.

On this symbol, see Bellezza 2002a, pp. 130, 207 (fig. XI-13a).

Fig. xxiii. Abbreviated Sale-‘od mantra, Gnam-mtsho

To conclude this analysis of the paleography of Upper Tibetan rock inscriptions potentially belonging to the Early Historic period, are three intriguing inscriptions, beginning here with the mantra Yungdrung Bon calls Sale-’od (see fig. 54). Its age is questionable but the forms of the sa, le and ’od are associated with scripts of the Early Historic period. If this is an inchoate rather than an incomplete example of the mantra, an Early Historic period attribution is strengthened. The letter A written three times in the upper line is in an unconventional handwriting found only at Gnam-mtsho (also see fig. xxiv). Nevertheless, the well-formed lower A is of a more modern appearance, possibly suggesting instead that the inscription belongs to the Vestigial period or to a time intermediate between the two major periods delineated in this work.*

The continued usage of Old Tibetan orthographic and paleographic conventions in the 11th century CE and through much of the Vestigial period is well known. For a list of these compiled from temple wall inscriptions in the Western Himalaya dated to the 11th century, see Thakur 1997, p. 969. Also see Scherrer-Schaub and Bonani 2002.

Fig. xxiv. A non-standard letter A with the addition of other red ochre marks, Gnam-mtsho

This unusual letter A is found in the same small cave as fig. xxiii (see fig. 56). Both these inscriptions are situated in a small cave complex along with other cult inscriptions and symbols associated with a non-Buddhist Rdzogs-chen tradition (also see fig. xxv). This execution of this inscription was informed by rudimentary dbu-can scripts of the Gnam-mtsho region.

Fig. xxv. Prayer to the Primordial Buddha, Gnam-mtsho

This inscription is a prayer to Kun-tu bzang-po, the Buddha of the Rdzogs-chen tradition (see fig. 53). The array of symbols and unusual paleography of other cult inscriptions in this cave complex indicate that it was occupied by those practicing a non-Buddhist brand of Rdzogs-chen (also see figs. xxiii, xxiv). The rudimentary form of the dbu-can script, including the almost horizontal gi-gu vowel sign and zhabs-kyu written as a single-line ligature, together with a prefixed (sngon-’jug) a-chung in tu, point to the Early Historic period.

Vestigial period

With the bstan-pa phyi-dar and the flowering of Classical Tibetan literature, there was a significant expansion of epigraphic activities in Upper Tibet. The bulk of this activity took place in Buddhist temples and monasteries, but there was also an increase in the fund of Upper Tibetan rock inscriptions. Adhering to ancient techniques of painting and carving, rock inscriptions were the more folkish or unlettered counterpart to the institutional production of religious epigraphs. Although the quantity of rock inscriptions rose, overtly religious themes came to dominate the corpus, displacing ‘secular’ content as the focus of the epigrapher. There are many unanswered questions about the literary development of the Tibetan language, but it is generally agreed that the transition from Old Tibetan to Classical Tibet took place in the 10th to 12th centuries CE. This is reflected in the adoption of a grammatical, orthographic and calligraphic system that with relatively few changes has continued to the present day.

Scripts, non-cursive and cursive, associated with Late Old Tibetan and Classical Tibetan texts serve as a paleographic benchmark for identifying rock inscriptions of the Vestigial period, the inception of which is dated nominally in this work to ca. 1000 CE. It is unlikely however that in such a large and religiously diverse place as Upper Tibet its paleography can be neatly categorized into pre- and post-1000 CE groups. We might expect that there was a period of transition in which both earlier and later scripts were exploited or which was marked by hybrid forms of writing. Nowhere is the ambiguity between the Old Tibetan and Classical Tibetan paleographic divide greater than in rock inscriptions comprised of only one syllable. Grammatical and semantical tools of analysis aiding in the dating of inscriptions are barely applicable at this level of brevity.

Non-Buddhists and Buddhists possessed customary scripts, which can aid in identifying the sectarian affiliations of rock inscriptions dating to the Vestigial period. For example, wispier and scrawlier dbu-can scripts appear to have been almost exclusively the province of non-Buddhists. Buddhist inscriptions tend to be of larger dimensions. Non-Buddhists at Gnam-mtsho often applied an ochre pigment that is orange red in hue, as opposed to the deeper red pigments favored by Buddhists for their epigraphy. Sometimes the opposite is true, and it is the non-Buddhist inscription painted in a darker color, but in any case, Buddhist and non-Buddhist epigraphs in close proximity commonly exhibit contrasting shades of red ochre. Pigments used in writing Buddhist mantras were frequently more robustly applied than the applications of non-Buddhists.

Fig. xxvi. The four types of nasalization diacritics represented in the ancient rock inscriptions of Upper Tibet (from upper left in clockwise direction): simple sna-ldan mark (dot), bipartite sna-ldan mark (dot and crescent), hooked sna-ldan mark (volute), and another bipartite sna-ldan mark (hook and crescent)

It is important to distinguish the four main styles of nasalization diacritics in the rock inscriptions of Upper Tibet, for they constitute a key paleographic feature. The Tibetan transcription of Sanskritic mantras (Buddhist and non-Buddhist) includes the use of the anusvāra, marking the nasalization of vowels.* In Upper Tibetan rock inscriptions it occurs with the syllable O, nasalizing it to produce Oṁ. In its most simple form, the anusvāra consists of a dot or teardrop inscribed above the O vowel sign (na-ro/nga-ro). Tibetan Buddhism refers to this diacritic as klad-kor, thig-le and rjes-su na-ro (literally ‘after the vowel sign O’). In the vernacular, it is frequently referred to by its phonological value: ma phyed-ka (‘half of m’).

For an exposition of the Indic tradition of Buddhist mantra writing, see Jayarava 2011.

In the Tibetan transcription of Sanskritic mantras a more elaborate nasalizing mark called anunāsika is also used. It consists of a dot or teardrop cradled in a crescent moon (candrabindu). In tantric forms of Tibetan religion this is equivalent to the conjoined sun and moon (nyi-zla). In Upper Tibetan rock inscriptions, the diacritic nyi-zla only occurs with the syllables O/Ō and hu/, rendering them /Ōϻ and huṁ/hūṁ. The use of the nyi-zla diacritic with the letter O is peculiar to Tibetan scripts, thus the employment of a special character for it in this work. Following Indic convention, the use of the dot and crescent diacritic with the syllable hu/ is standard in Tibetan mantra construction (as the vowel sign is positioned below the cap line or ‘head’ (mgo) of the letter). There are also examples in Upper Tibetan epigraphy of the sna-ldan mark with the syllable hu/hū being reduced to just a dot or crescent.

A variation of the nyi-zla in the ancient rock epigraphy of Upper Tibet consists of a volute to represent the dot and crescent with O/Ō and hu/. I refer to this diacritic as the ‘hooked sna-ldan mark’ (rendered Oḿ, huḿ/hūḿ). The hooked sna-ldan mark is intermediate in form between the simple and bipartite nasalization diacritics. Occasionally, a hook or volute is combined with the crescent in Ō and (rendered Ōɱ and hūɱ). This is the most elaborate sna-ldan mark in Upper Tibetan epigraphy. I have chosen special characters to designate the three varieties of Tibetan anunāsika with the letter O because they are atypical of Indian scripts. Even in Tibetan Buddhist calligraphy they have been superseded by the simplified anusvāra. In this work, I refer to the dot and crescent and hook and crescent diacritics as ‘bipartite sna-ldan marks’ (sna-ldan means ‘nasal [sound]’).

The lineaments of the following examples of monosyllabic inscriptions of A and Oṁ generally conform with Classical Tibetan calligraphic standards in literary works. Thus, they are best dated to the Vestigial period. However, a somewhat earlier date for their creation cannot be excluded.

Fig. xxvii. The mystic letter A worshipped in modern times as a self-formed (rang-’byung) religious symbol with paper rlung-rta and dabs of butter, Gnam-mtsho

This inscription represents the one of most common varieties of dbu-can script in Classical Tibetan works (see fig. 66). It was superimposed on a smaller, more elementary style letter A. Note the differences in the pigment coloration and form of the two scripts.

Fig. xxviii. The syllable Ōḿ with a diacritic mark consisting of a hooked sna-ldan mark, Gnam-mtsho. Note the figure of a bird superimposed on the inscription

The style of nasalization diacritic illustrated here is common in the Buddhist epigraphy of Upper Tibet dating to the Vestigial period (and possibly earlier as well), but it was also used to a lesser extent by non-Buddhists in their rock inscriptions. As in this inscription, the hook usually faces towards the left (also see figs. xvii, xviii, supra), but not always (see figs. 153, 366).*

For another occurrence of a hooked sna-ldan mark oriented to the right in Lo Mustang, see August 2017 Flight of the Khyung, fig. 26.

Fig. xxix. The Dpa’-bo ’bru-lnga mantra of non-Buddhists, Gnam-mtsho

Here is an occurrence of the hooked sna-ldan mark in a mantra belonging to Yungdrung Bon or another non-Buddhist group that may have preceded this religion in the Gnam-mtsho region (see fig. 105). A hooked sna-ldan mark is also seen in the hūḿ syllable, as is found in Buddhist variants of the same period.

Fig. xxx. The syllable Oṁ in the lan-tsha script, Gnam-mtsho

This is the only painted rock inscription in the lan-tsha script discovered in Upper Tibet (see fig. 63). Highly degraded, it presumably predates the 15th century CE, like most other red ochre inscriptions in the region.

Fig. xxxi. An unusually configured syllable Ōϻ, Gnam-mtsho

This inscription has the two elements of the bipartite sna-ldan mark spread widely apart (see fig. 37). It appears to be an idiosyncratic rendering of the sacred syllable dating to the Vestigial period, but an earlier date for its creation cannot be entirely discounted.

Fig. xxxii. The inscription Oϻ ma hūṁ with accompanying symbols, Gnam-mtsho

The bipartite sna-ldan mark over the vowel O is a grammatical form commonly found in the rock inscriptions of Upper Tibet, Buddhist and non-Buddhist. This mantra, with its particular sequence of syllables and counterclockwise swastika, belongs to a non-Buddhist tradition (see fig. 238). The conjoined sun and moon subject above the accents the sna-ldan mark, underlining its mystical association with these prime tantric reference points.

Fig. xxxiii. The inscriptions Ōϻ A hūṁ and A Oṁ, Central Byang-thang

The three seed syllables (sa-bon) written vertically on the left are Buddhist, while the two seed syllables on the right, given their sequence, appear to be non-Buddhist (see fig. 275). The Ōϻ features a bipartite sna-ldan mark. Buddhist mantras with a sna-ldan mark of two elements are found in a variety of other media including religious objects made of metal and wood.† The inscribing of Buddhist mantras with this style of diacritic fell into disuse around 1400 CE.* There do not appear to be any examples of the bipartite sna-ldan mark in the rock inscriptions of Upper Tibet predating the late 10th century CE, indicating an approximately four-century-long period of popularity for this calligraphic embellishment. Nonetheless, as a stylistic gesture, Ma-ṇi mantras were occasionally engraved and written using the old-fashioned bipartite sna-ldan mark in later periods and even today.

An engraving of a Ma-ṇi mantra in four different languages and scripts using the old technique of carving letters directly into stone is found in Tyr (Yongning Temple), on the mouth of the Amur River, and dated to 1413 CE. The Tibetan example of the mantra exhibits modern paleographic features. On this multilingual inscription, see Ligeti 1961. In the photograph provided (ibid.) the first syllable of the mantra in dbu-can is not visible). The so-called Stele of Sulaiman (1348 CE) has the Ma-ṇi mantra engraved in six different scripts (dbu-can, lan-tsha, ’phags-pa, Old Uyghur, Chinese, and Tangut). The first syllable of the dbu-can version has a bipartite ma-slad-skor. For an image of this stele, see Wikipedia:
https://en.wikipedia.org/wiki/Stele_of_Sulaiman

The bipartite sna-ldan mark is also seen in temple inscriptions of Western Tibet, as is the hooked sna-ldan diacritic. For example, both marks are featured in Ma-ṇi and other mantras accompanying the frescos of Mang-brag in Gu-ge, a cave temple dated on art historical grounds to ca. 1250–1400 CE. There are also a few occurrences of the reverse letter i (gi-gu log) at Mang-brag, as there are in Ma-ṇi mantras in the rock inscriptions of Upper Tibet. On Mang-brag, see January 2015 Flight of the Khyung. For the hooked sna-ldan mark over the syllables Ō and written multiple times on a piece of birchbark and sealed in a copper alloy mchod-rten, which radiocarbon analysis suggests took place in the late 12th or 13th centuries CE, see Hatt 1980, p. 187 (fig. 15). Nevertheless, this author holds (ibid. 206, 207) that some of the folded birch bark sheets enshrined in the mchod-rten may be older than the date of its consecration. For a red ochre Ma-ṇi mantra with a bipartite sna-ldan mark over the Ō and a gi-gu log (reverse vowel i mark) in Lo Mustang, see August 2017 Flight of the Khyung, fig. 23. For a carved Ma-ṇi mantra with the hooked sna-ldan mark over the O in Lower Ladakh, see Denwood 1980, p. 160 (fig. B); Linrothe 2003, item no. 83116. On a thang-kha of the tutelary god (yi-dam) Acala with inscriptions of mantras carrying the bipartite sna-ldan mark over the O and the hooked sna-ldan over the h in hūṁ, see Zla-ba tshe-ring et al.-b, p. 85. Henss (2014: 184) dates this thang-kha to ca. 1300 CE.

Fig. xxxiv. Buddhist mantra for the deity Rdo-rje sems-dpa’, Gnam-mtsho

This inscription presents another variant of the syllable Oṁ (see fig. 92). In this example, the sna-ldan mark is written as a teardrop-shaped form half as large as the letter radical (ming-gzhi).

Fig. xxxv. The Buddhist Rigs-gsum mgon-po mantras, Gnam-mtsho

Here we find use of the simplest form of the Oṁ syllable sans the a-chung (see fig. 61).* Examples with the a-chung are more common in ancient rock inscriptions of Upper Tibet.

On Rigs-gsum mgon-po inscriptions and mchod-rten in Ladakh, see Francke 1915.

Fig. xxxvi. A bichrome Ma-ṇi mantra, Central Byang-thang

This well written epigraph displays a bipartite sna-ldan mark consisting of a crescent and a hook and the gi-gu log (see fig. 260). As with other inscriptions in this section of the paleographic typology, attribution to the Vestigial period is indicated.

Fig. xxxvii. The non-Buddhist Dpa’-bo ’bru-lnga mantra, Gnam-mtsho

This is a non-Buddhist example of the use of a bipartite sna-ldan diacritic. While this calligraphic motif has long fell out of fashion with the Buddhists, Yungdrung Bon continues to fully embrace it in their presentation of mantras. This faded red ochre inscription bears many features of more modern dbu-can scripts.

Fig. xxxviii. A modern version of the Dpa’-bo ’bru-lnga mantra

For comparative purposes, this is a modern representation of the mantra. End marks (shad versus the gter-tsheg) notwithstanding, it exhibits a similar calligraphic style to its counterpart in fig. xxxvii.

An examination of the letter a-chung, prefixed and subjoined in the inscriptions of the Early Historic period and Vestigial period illustrated in this work reveals many different forms of the letter. These range from attenuated marks and inverted U-shaped forms to standard types inscribed in two strokes.*

For a table of different types of a-chung letters taken from a wide range of sources, see Uray 1955, Table II, Table II (Suite).

Fig. xxxix. A prayer to the Karmapa, Central Byang-thang

While the mantras in figs. xxxvi and xxxvii were well executed, here is an example of an inscription in dbu-can that was more crudely written (see fig. 281). The lineaments of the letters and the orthography of the inscription suggest attribution to the Vestigial period.

Fig. xl. Buddhist prayer to the savioress, Gnam-mtsho

This inscription represents the next stage in cursivization after the process of streamlining Tibetan scripts in the post-Imperial period (see fig. 13). This kind of dbu-med script was well used in textual productions of the Vestigial period.

Fig. xli. A prayer to the Stag-lung bka’-brgyud hierarch

This is an example of a more freely scrawled style of dbu-med calligraphy. The contents of the inscription demonstrate that it was not written before the late 12th century CE.

Fig. xlii. A prayer to the of water-spirits of Gnam-mtsho

This inscription was rendered in a fully developed dbu-med script and in much smaller letters than most other inscriptions at Gnam-mtsho. Written in a black pigment, is it known whether it was indited by Buddhists or non-Buddhists. Nevertheless, inscriptions with comparable characterictics, where identifiable, are non-Buddhist.

The paleographic traditions in the rock inscriptions of Upper Tibet quite commonly include head marks. Head marks (yig-mgo; older usage: mgo-yig/dbu-yig) typically signal the beginning of a text, be it a mantra, sentence or passage. In literature, head marks are also used at the beginning of a folio or book chapter. In the epigraphy of Upper Tibet, the use of these grammatical marks is mostly linked to inscriptions of the Vestigial period. There are however several instances of head marks in scripts assigned to the Early Historic period.*

For the use of a yig-mgo in an inscription engraved at the bottom of a silver cup dated to the Imperial period, see Carter 1998, p. 23 (fig. 2). For a survey of headmarks in early Tibetan documents, see Scherrer-Schaub and Bonani 2002, pp. 191, 192. According to Schaub and Bonani (ibid.), the ‘curl-like symbol’ grew more elaborate over time.

Head marks occurring in the rock inscriptions of Upper Tibet are mostly standard calligraphic forms found more universally in Tibetan writings. The most common variants in the epigraphy of Upper Tibet consist of a simple volute, the so-called ‘old grammatical head mark in front’ (brda-rnying yig-mgo mdun-ma). There are also instances of two simple volutes being used as a head mark, the second one called the ‘old grammatical head mark behind’ (brda-rnying yig-mgo sgab-ma). The more elaborate head marks commonly associated with Classical Tibetan texts are also found in the rock inscriptions of Upper Tibet, but are uncommon. Known simply as a yig-mgo (mdun-ma and sgab-ma), they consist of one or two curved lines underscoring each of two or sometimes three volutes. In the epigraphy of the region it was popular to add one or two vertical strokes after the volute(s) as part of the yig-mgo. This defining regional characteristic also occurs in other ancient Tibetan literature and inscriptions. In this work, I refer to one vertical line appearing as part of the head mark as ‘initial shad’ and two parallel vertical lines as ‘initial nyis-shad’.

Although a chronological progression from less to more complex head marks appears to be reflected in the epigraphy of Upper Tibet, the tradition of making rock inscriptions in the region was open to a wide spectrum of society. This surely included local herders and farmers who were not bound by the stricter conventions of Tibetan literary centers. As with other paleographic and grammatical traits, older and newer types of head marks may have been used interchangeably depending on the individual proclivities of epigraphers.

Fig. xliii. Brda-rnying yig-mgo mdun-ma and initial shad (see fig. 91)

Fig. xliv. Brda-rnying yig-mgo, mdun-ma and sgab-ma, with appended initial nyis-shad (see fig. 207). This head mark occurs in an inscription that may possibly date to the post-Imperial period

Fig. xlv. An underlined yig-mgo mdun-ma forming a circle (see fig. 195). Two sets of initial nyis-shad are appended to it.

Fig. xlvi. Another ancient type of head mark consisting of a yig-mgo with three volutes (see fig. 197)

Fig. xlvii. A yig-mgo mdun-ma and sgab-ma, each surmounted by a hook motif (see fig. 352)

Fig. xlviii. A rare occurrence of a yig-mgo comprised of three interconnected volutes without underlining (see fig. 97). Above each volute the conjoined sun and moon symbol (nyi-zla) was inscribed

As mentioned, there are also red ochre rock inscriptions at Gnam-mtsho that postdate the Vestigial period. In a small cave adjacent to the Sangs-rgyas yar-byon sgrub-phug, Bkra-shis do-chen, the personal (ordination) names of three Stag-lung throne holders were written by three different individuals as a devotional act (Bellezza 1997a: 212).* The three personalities are Rje kun-dga’ bkris (bkra-shis) (1536–1605 CE), Ngag-dbang rnams-rgyal (1574–1621 CE) and Ngag-dbang bkris (bkra-shis) dpal-[grub] (1600–1671).† These hierarchs were the 16th, 17th and 20th Stag-lung khri-pa respectively. As averred in the Stag-lung tradition, this epigraphic evidence (as well as local toponyms) demonstrates that the subsect maintained a long-term presence at Gnam-mtsho. It also establishes that the making of rock inscriptions in red ochre continued in isolated cases at Gnam-mtsho well into the 17th century CE. Interestingly, there are no instances of ordination names of Buddhist lamas in any of the rock inscriptions of Upper Tibet attributed to the Vestigial period.

In another small cave adjacent to Sangs-rgyas yar-byon sgrub-phug there is an inscription written in dbu-med announcing the presence of a footprint (zhabs-rjes) in the cave wall (Bellezza 1997a: 212). Also, in a small cave near the Brag-dkar formation of Bkra-shis do-chen there is an inscription that reads; Stag lung pa’i phug (Cave of the Stag-lung-pa), located near the entrance (Bellezza 1997a: 211, 212). Photographic documentation of these inscriptions is unavailable.

On the succession history of the Stag-lung hierarchs, see Schwieger 1996.

Fig. xlix. “Ngag dbang mam rgyal mkhyeno” (upper), “Chos rgyal mo gus grub mkhyeno” (lower), Gnam-mtsho

Both inscriptions finish with mkhyeno (conjunct form of mkhyen no), showing that they are prayers to lamas to look favorably upon the inscribers. The inclusion of a personal name in the upper inscription indicates that it was made while the lama was still alive. The syllable rnam was written using the sna-ldan diacritic rather than with the full letter ma, a more formal gesture on the part of the epigrapher. The lower inscription includes a common epithet for high lamas, Chos-rgyal (Sovereign of the Dharma), which probably refers to one of the Stag-lung-pa hierarchs. Unusually, this title is written with a feminizing agent (mo). The next portion of the inscription (gus-grub) seems to mean ‘accomplished with devotion’. The two inscriptions use an old grammatical style (brda-rnying) head mark. The head mark and first syllable of the lower inscription are partially superimposed on an older inscription that appears to read Oṁ (probably dating to the Vestigial period; head mark partly cut in the photograph). The last syllable of the lower inscription is superimposed on the finial of a Buddhist mchod-rten (for this mchod-rten see Bellezza 2017-2018, fig. 20b).

Fig. l. “Ngag dbang bkris dpal”

Although only the proper the name only is visible in this photograph, the inscription may possibly continue with the words: mkhyen no, making it a prayer.

Fig. li. “Rje Kun dga’ bkris mkhyen no”

The inclusion of a proper name in this prayer indicates that is was made during the lifetime of this lama. The vowel sign (’greng-bu) in Rje, an honorific title, is now mostly missing.

The rock inscriptions in figs. xxxxix and li were written in fully modern forms of the dbu-can script, setting them apart from other epigraphs presented in this paleographic analysis. Fig. l, an epigraph from the 17th century CE, is more roughly inscribed in a chirography producing thicker strokes and squatter letters. Most red ochre epigraphs attributed to the Vestigial period in this work exhibit more wear, more darkening or browning of the pigment or more fading than the specimen in fig. l, justifying assignment to an earlier period. Also, there is no other handwriting (yi-ge’i rnam-pa) in this survey closely matching the example pictured in fig. li. Nevertheless, this inscription raises the question of which other specimens attributed to the Vestigial period in this work might in fact be more recent. It is possible that there are late rock inscriptions mimicking scripts used in the Vestigial period that have escaped detection.

Another group of potentially later inscriptions is found at a site called Lha-khang-dmar-chags. These were written as a kind of sacred graffiti on the walls of a defunct non-Buddhist temple site (see figs 292–315). Similarly, there are a few other inscriptions included in this survey that on paleographic grounds are probably best assigned to the post-14th century CE period (e.g., figs. 166, 168, 239).

Physical qualities of the inscriptions

Here we review the physical qualities of painted and carved rock inscriptions in Upper Tibet and the tools that may have been used to create them. I am not aware of any Tibetan work that details how ancient rock inscriptions were made. Moreover, it is not clear how relevant in practical terms extant Tibetan manuals on carving and painting are to the rock inscriptions of this study. As part of the rig-gnas chen-po lnga (five major traditional fields of study) there is the bzo rig-pa (study of craftsmanship) in which the carving of various materials and painting are included, as well as the sgra rig-pa (study of grammar) covering literary composition. As rock inscriptions are closely aligned technically with older rock art, more archaic methods of production than those recorded in bzo rig-pa texts may be indicated. Reconstructing how rock inscriptions were crafted in Upper Tibet will require a regimen of scientific testing.

Snellgrove and Richardson (1968: 54, 146) observe that earlier Tibetan crafts were replaced by arts and crafts coming from India. The fabrication of ancient rock inscriptions in Upper Tibet belongs to the older tradition of arts and crafts.

Paper is a smooth and absorbent medium on which a variety of writing instruments can be used. Conversely, rock is a hard, rough and often irregular surface with minimal capabilities to absorb and fix ink. To cope with these inherent drawbacks of inditing on stone, epigraphers relied on mineral pigments or cutting tools.

Pigments such as red ochre (oxides of iron) are extremely durable forming a strong bond with stone. Their imperviousness to wear and weathering, geochemical stability, and ability to fix to stone surfaces are properties sought after by artists and epigraphers worldwide for millennia. In addition to red ochre (ranging in color from reddish orange, to scarlet and magenta), yellow ochre, and white and black mineral pigments were used, albeit less commonly, to inscribe rock inscriptions in Upper Tibet. To my knowledge, the chemical composition of mineral pigments used in Upper Tibetan rock art and rock inscriptions has not been analyzed. The most common white pigments used in rock art in other locations, suggests that either calcium sulfate, calcium carbonate or magnesium calcium carbonate was employed in Upper Tibet to produce a white color. Similarly, oxides of manganese would have produced a black pigment for rock inscription. There are several finely written epigraphs in a dbu-med script at Gnam-mtsho that may have been written in lampblack (paracrystalline carbon), an ink traditionally employed in Tibet for creating texts.* Orpiment (arsenic sulfide) could have been exploited for its bright yellow color in one or two inscriptions.*†

The instruments used to write Old Tibetan documents included the bamboo or cane pen (snyu-gu) and possibly other types of wooden pens (smyig-ma). The way the nib was cut regulated ink flow and the consequent thickness of strokes. Ink (snag-tsha) was produced from carbonized substances mixed with oil, gum or glue. For more on traditional writing technology in Tibet, see Jamgӧn Kongtrul Lodrӧ. 2012, pp. 264–269; Scherrer-Schaub 1999, pp. 6, 7.

On the preparation and use of mineral pigments for religious paintings in Tibet, see Jackson 1988.

The methods followed for preparing mineral pigments by rock artists and epigraphers in Upper Tibet is unknown. We can surmise that pigments were ground into a fine powder and possibly mixed with other ingredients such as binding and fixative agents. Mineral pigments may have been heat treated to enhance their color and friability.* Once prepared, water and/or oils (such as butter or animal fat) were added to powdered minerals, which may then have been filtered and decanted to produce the finished pigment for inscription.

For comparative purposes, on the mining, heating and use of admixtures in making ochre pigments by Australian aborigines, see, for example, Dobrez 2014, pp. 377, 378, 380. For the preparation of ochre pigment for the painting of contemporary murals in Kerala, see Mini 2010, pp. 637, 638.

It is not certain what kinds of writing instruments (yig-chas) were used to make rock inscriptions in Upper Tibet. The diversity in the execution of inscriptions suggests that a variety of different instruments for inditement were exploited. Pigments were variously lightly, thickly, roughly or smoothly applied, reflecting reliance on a selection of writing tools (technique notwithstanding). Quills may have been used as a kind of disposable stylus. Inscribing on stone would have quickly degraded the writing tip but this would not have been a big concern, for bird feathers are widely available in Upper Tibet, especially near lakes with their large aquatic bird populations. Conceivably, the use of nibs made from other types of materials (bone, horn, wood, etc.) could have been used as well. Brushes (spu-snyug, pir) fashioned from animal hair or wool are another obvious choice as a writing apparatus. However, criteria for discriminating between inscriptions made with a brush or stylus have not been established for the rock inscriptions of Upper Tibet.

Uncertainty also surrounds methods used to engrave rock inscriptions in Upper Tibet. Like the painted inscriptions, there is a good deal of variety in their physical make-up, strongly suggesting that different kinds of tools and techniques were involved in their production. Engravings of finer and deeper lines were made with sharp-edged ferrous tools such as chisels or burins. Other inscriptions with thicker but shallower and less defined lines appear to have been created through abrading or pecking, necessitating the application of implements like grinders and hammer and punch.

Part Three: A statistical and religious profile of rock inscriptions in Upper Tibet

Statistical analysis

Much can be inferred about the religious complexion of Upper Tibet in the Early Historic and Vestigial periods by examining the number and types of rock inscriptions. Although this survey is not exhaustive, it is sufficiently comprehensive to furnish statistically relevant information. The statistical data provided has a margin of error of up to 2%, as the tally of various categories is not a strict count.

A total of 525 rock inscriptions painted or carved in the old tradition were used in this statistical analysis* Of this total, 340 or 65% of the rock inscriptions are located around Gnam-mtsho. In the rest of Upper Tibet only 185 rock epigraphs were documented. The disproportionately high number of inscriptions at Gnam-mtsho demonstrates that this was a prime center of epigraphic activity in Upper Tibet. The historical factors accounting for the concentration of two-thirds of known rock inscriptions at this second largest lake on the Tibetan Plateau is unclear. Its proximity to Central Tibet, the heart of the Tibetan world, is likely to have played a role. Religious pressures coming from Central Tibet probably acted upon Gnam-mtsho more strongly than in areas of the Byang-thang farther removed from this hothouse of religious activity. In the Vestigial period sectarian tensions and a consolidation of ecclesiastic territories were rife in Central Tibet, with Buddhism and Yungdrung Bon carving out religious, social and political identities, which in some cases are still in place today. The religious and political turmoil of that time spilled over into other regions of Tibet, as for instance, the history of Western Tibet illustrates.

Around 30 inscriptions documented in Upper Tibet are not photographically represented in this work but are noted in the catalogue below.

In the Eastern Byang-thang (excluding Gnam-mtsho) and Central Byang-thang 60 rock inscriptions were recorded (11% of the total), and 120 in the Western Byang-thang and Far Western Tibet (23% of the total). All epigraphs at Gnam-mtsho are painted, as are an additional 20 in the Eastern Byang-thang, five in the Central Byang-thang, 33 in the Western Byang-thang, one in the Upper Gtsang-po valley, and six in Far Western Tibet, 77% of the total. There appears to be no other region in Tibet with such a high number of ancient epigraphs written on stone with mineral pigments. The other 120 (23%) inscriptions documented in Upper Tibet were produced by directly cutting letters into rock surfaces.

Of the 340 rock inscriptions recorded at Gnam-mtsho, 16 are attributed to the Early Historic period, 289 to the Vestigial period (amongst which are a few later specimens), with an additional 35 potentially predating the beginning of the Vestigial period (ca. 10th century CE). Thus, 86% of the inscriptions at Gnam-mtsho appear to date to the Vestigial period, while only 5% are unambiguously attributed to the Early Historic period and 9% could belong to either period. As already explained, the higher incidence of epigraphic activity in the Vestigial period appears to have been spurred on by religious rivalries between Buddhists and non-Buddhists.

In the Western Byang-thang (excluding Lha-khang dmar-chags and Do dril-bu) and Far Western Tibet, 12 inscriptions are assigned to the Early Historic period, 15 to the Vestigial period, plus 56 that may possibly predate the Vestigial period. The factors accounting for a much higher percentage (80%) of inscriptions predating or potentially predating 1000 CE in more western regions are not entirely clear. As observed, close proximity to Ladakh, an early center of Tibetan epigraphy, appears to be a major factor. The types of rock art mchod-rten and religious dedications in Far Western Tibet point to cultural influences from more northern and western regions. The ruined temple of Lha-khang dmar-chags in the Western Byang-thang is a special case, with many of its 33 mineral pigment inscriptions possibly made after the Vestigial period. A few epigraphs written on the interior walls of an ancient residence at Do-dril-bu may also postdate the Vestigial period.

As discussed, rock inscriptions containing explicit historical material are uncommon in Upper Tibet. There are only ten or 11 instances of clan names in the entire region. There are just nine personal names and epithets, and four indications of non-religious professional status. Additionally, three different toponyms are mentioned in five inscriptions.

On a religious criterion, the 525 inscriptions of this statistical survey can be divided as follows: 166 are patently Buddhist, 172 non-Buddhist, and 187 inscriptions have an indeterminate identity (Buddhist or non-Buddhist). The inscriptions whose religious affiliations are in question typically include monosyllabic specimens and those containing no explicit religious indications. In this latter category are most inscriptions dating to the Early Historic period and many other specimens potentially predating 1000 CE. From what we know of the religious complexion of Upper Tibet in that period, it can be inferred that most of these older epigraphs were produced by non-Buddhists. In fact, only four inscriptions positively assigned to the Early Historic period can be identified confidently as Buddhist. As discussed in Part One and Part Two, the epigraphic record mirrors the important place of non-Buddhist traditions in the Upper Tibetan religious scene of the Early Historic period and Vestigial period.

The explicitly religious rock inscriptions of Upper Tibet are overwhelmingly composed of just seven different mantras. The single most common inscription carved and painted using the old techniques is the Ma-ṇi mantra. A total of 110 specimens were documented, making up 22% of the total number of ancient rock epigraphs surveyed in Upper Tibet. Seventy-two Ma-ṇi mantras were documented at Gnam-mtsho alone, three in the Central Byang-thang and 36 in the Western Byang-thang and Far Western Tibet. Ma-ṇi mantras are not found at other sites in the Eastern Byang-thang or at Lha-khang-dmar-chags. Other Buddhist mantras represented include Oṁ A huṁ (18 examples) and the Rigs-gsum-mgon-po mantras (eight examples).

Surprisingly, only two rock inscriptions consisting of the Vajra Guru mantra was surveyed in Upper Tibet. In fact, no other references to the highly important Buddhist master Gu-ru rin-po-che (Padma ’byung-gnas) were discovered in the rock epigraphs. This indicates that the very popular Vajra Guru mantra of later times was not well known or was not favored as an epigraphic subject in the epigraphy of the Vestigial period. The lack of this mantra in Upper Tibet may have been influenced by the Buddhist conversion of most of the region mostly by the Bka’-gdams-pa and Bka’-brgyud-pa sects, rather than by the Rnying-ma-pa (which did not have a large institutional presence in much of the region).

Common mantric syllables standing alone in the rock inscriptions of Upper Tibet are A (81 examples) and Oṁ (43 examples). In some epigraphic and rock art contexts, these sacred syllables were created by non-Buddhists, but others belong to Buddhism. The religious identity of certain examples can be ascertained through an analysis of their paleographic traits or through accompanying rock art. However, the religious affiliation of many monosyllabic inscriptions in the region has not been determined.

The most common non-Buddhist mantra in the rock inscriptions of Upper Tibet is what Yungdrung Bon calls Sale-‘od. Forty-one of these mantras have been documented: 31 at Gnam-mtsho, three in the Central Byang-thang and seven at Lha-khang dmar-chags. However, no Sale-’od or other non-Buddhist multisyllabic mantras were discovered in Far Western Tibet. Potential non-Buddhist mantras in Far Western Tibet are monosyllabic; e.g., Oṁ and A. Another likely non-Buddhist mantra in the west reads: Oṁ ma.

The single most popular mantra of Yungdrung Bon is the Ma-tri. Nevertheless, among ancient rock inscriptions just five examples were noted. Another common Yungdrung Bon mantra is called the Du-tri-su. Four of these occur in rock inscriptions, one at Bkra-shis do, Gnam-mtsho, and three at Lha-khang dmar-chags. A more common non-Buddhist mantra in the rock epigraphy is A Oṁ huṁ (11 examples), with nine of these being found at Bkra-shis do, one at Dang-ra g.yu-mtsho and one at Lha-khang dmar-chags. There are also nine non-Buddhist religious inscriptions at Bkra-shis do reading: Bso A phaṭ /A phaṭ. Most of these are accompanied by counterclockwise swatikas drawn in the same hand. A few exceptions notwithstanding, non-Buddhist multisyllabic inscriptions can be dated to the Vestigial period.

Doctrinal significance of mantras

As approximately 85% of all ancient rock inscriptions included in this survey is mantric in nature, understanding the religious functions of mantras is essential. The preponderance of these types of inscriptions in Upper Tibet discloses that religious beliefs and interests constituted the chief motivation of epigraphers. Moreover, many non-mantric inscriptions are religious in nature, overtly or tacitly. Religion, in a word, was the overriding concern, at least for those who made the effort to leave rock inscriptions in the region. Thus, it can be deduced that the main media of communications for non-religious affairs in Upper Tibet in the Early Historic period and Vestigial period were textual or non-literary in nature. Until the modern period, the oral tradition (songs, myths, genealogies, ecstatic recitations, etc.) remained the most important medium for transmission of so-called secular or popular traditions in Upper Tibet.

The mantric syllables A and Oṁ are laden with religious meaning in Buddhism and Yungdrung Bon, with much of their significance shared by both religions. Here only a résumé of a few main functions is given. A and Oṁ are commonly called seed syllables (sa-bon) or heart syllables (snying-po yig-ge), conveying their fundamental doctrinal status. Most broadly, A and Oṁ symbolize or embody the full spectrum of Buddhist or Yungdrung Bon teachings. More specifically, A in Buddhism is the seed syllable visualized at the mystic center of the throat in various rituals and meditations, while Oṁ is the seed syllable at the forehead. These two syllables are the sound equivalent of spiritual realization and the form of the Buddha. Similarly, in Yungdrung Bon, A represents the primordial nature, essence of wisdom, purity, bliss, boundlessness, eternity, and other such profound qualities. Oṁ represents their Buddha (Sangs-rgyas), Ston-pa gshen-rab, and the methods leading to enlightenment, as well as the five types of wisdom (ye-shes lnga) and five aspects of the Buddha (sku-lnga). This symbolic pentad is reflected in the five calligraphic elements of the Yungdrung Bon Oṁ.

Oṁ A huṁ and A Oṁ huṁ are seed syllable mantras of the Buddhists and Yungdrung Bonpo respectively. They mystically represent the three bodies of the Buddha (sku-gsum), the purified body, speech and mind (sku gsung thugs), and the enlightenment form present in the three worlds (khams-gsum), etc.

The Ma-ṇi mantra of Buddhists and the Sale-’od mantra of Yungdrung Bon have an elaborate range of meanings. By far, the most common mantra in the Tibetan world is the Ma-ṇi. A key source for its history, doctrinal significance and ritual usage is the collection of texts known as the Ma ṇi bka’ ’bum, attributed to King Srong-btsan sgam-po (7th century CE). These treasure texts (gter-ma) were ‘revealed’ between the 11th and the middle of the 13th century CE.* Nonetheless, the origins of the cult of Spyan-ras-gzigs in Tibet can be traced back to the Imperial period. Canonical texts pertaining to this god are listed in the Ldan kar ma catalogue, compiled in 812 CE (Kapstein 2003: 527). The Ma-ṇi mantra is centered around the ‘Possessor of the Jewel and Lotus’ (Ma-ṇi Padme), the deity Spyan-ras-gzigs (Avalokiteśvara) (cf. Kapstein 1997: 71). Spyan-ras-gzigs is seen by Buddhists as being the patron god of Tibet. Most commonly, the six syllables represent the six transmigratory realms of existence (’gro-ba rigs-drug) and the purification of the prevailing emotional affliction associated with each one.†

On the history of this collection of texts, see Melnick and Bell n.d.; Kapstein 2003; Dargyay 1989.

For a traditional explanation of the significance of the Ma-ṇi mantra based on the Ma ṇi bka’ ’bum, see Kapstein 1997, pp. 74–76; Kyabje Yonzin Trijang Dorje Chang Losang Yeshe Tenzin Gyatso Pal Zangpo 1982.

Sale-’od along with the Ma-tri and Du-tri-su are the three most important mantras of Yungdrung Bon and are known collectively as Snying-po rnams-gsum. They are reputed to contain syllables from the Zhang-zhung language, but some of these have a Sanskrit etymology. According to Yungdrung Bon doctrine, the three main mantras were bestowed upon humanity by the Buddha Gshen-rab in the remote past.

The Sale-’od mantra is most commonly associated with the deity of effulgence, Gshen-lha ’od-dkar. Each of the ten syllables of this mantra represents a major aspect of Yungdrung Bon teachings. For example, the two syllables sa and le are associated with the luminosity (gsal-ba) of the natural state (gnas-lugs), which is non-material and free from defects. Although hardly if ever seen in the ancient epigraphic record of Upper Tibet, two syllables of the Sale-’od mantra, sa and le, are now often combined in conformance with Sanskritized Tibetan. The Ma-tri mantra is dedicated to the Buddha Ston-pa gshen-rab. The second syllable of this mantra, ma, is connected to the wisdom goddess of compassion, Shes-rab ’byams-ma. Like the Ma-ṇi mantra, the final six syllables are linked to the six realms of transmigratory existence. Each of the syllables of the six realms represents a Buddha that guides sentient beings towards enlightenment. The 15 syllables of the Du-tri-su mantra recapitulate the qualities of the primordial mind and the purification of the suffering of the six realms of transmigratory existence.

The cult of Spyan-ras-gzigs can be traced from ancient India to Tibet of the Imperial period thence to the latter diffusion of Buddhism, strongly suggesting that the doctrinal significance and ritual and meditative functions assigned the Ma-ṇi mantra have remained relatively stable over time. The import of other Buddhist mantras in the rock inscriptions of Upper Tibet is not liable to have changed much either in the last millennium.

The doctrinal and ritualistic continuity exhibited by non-Buddhist mantras in the epigraphy of Upper Tibet is somewhat in question. Non-Buddhist mantras and religious content of the Early Historic period are likely to have been the product of groups retaining archaic practices and beliefs. Thus, early mantric syllables (like A and Oṁ) may have carried distinctions in meaning and usage from the value placed upon them subsequently by Yungdrung Bon adherents. As discussed in Part Two, the sectarian affiliations of non-Buddhist sects in Upper Tibet in the Vestigial period is unclear. Whether they saw themselves as autonomous or belonging to Yungdrung Bon, non-Buddhists in the region widely adopted mantras that became the mainstay of Yungdrung Bon. If in fact there were localized non-Buddhist cult groups in Upper Tibet in the Vestigial period, their understanding of these mantras may have varied somewhat from mainstream religious concepts reviewed above. It is not my intention here to minimize continuity in the conception and deployment of mantras between any local non-Buddhist traditions and Yungdrung Bon, but only to point out that these links may have been tempered by a process of sectarian and doctrinal evolution.

In the Yundrung Bon tradition, the recitation of the three primary mantras is used as a method of purification and realization. These mantras are applicable to the three classes of doctrine: sutra, tantra and Rdzogs-chen.* That Rdzogs-chen (Great Perfection), a penetrating philosophical view and praxis, was practiced at Gnam-mtsho is evidenced in the early epigraphy of the region.† There are several rock inscriptions that refer to Kun-tu bzang-po (All Goodness), the primordial Buddha of the Rdzogs-chen tradition (figs. 53, 143, 221). Some, if not all references to the primordial Buddha in the epigraphy of Gnam-mtsho were made by non-Buddhists, as indicated by the scripts, proximate rock art and the positioning of Buddhist inscriptions. Textual evidence corroborates that non-Buddhist Rdzogs-chen practitioners were active in various parts of Far Western Tibet and the Western and Central Byang-thang prior to the 12th century CE (Reynolds 2005). Moreover, the Yungdrung Bon master Ldong-sgom zhig-po is recorded as having made pilgrimage to Gnyan-chen thang-lha in the Vestigial period (Achard 1998: 36, 37). As this Rdzogs-chen saint is said to have actively sought out many masters during his pilgrimages (ibid.), it would be natural for him to have interacted with fellow practitioners in the Gnam-mtsho region.

According to the biography of a Yundrung Bon Rdzogs-chen practitioner named Bon-zhig khyung-nag (12th century CE), at the moment of his enlightenment he heard the Sale-’od mantra resounding in all directions. See Achard 1998, p. 33.

On historical aspects of non-Buddhist Rdzogs-chen traditions in the Early Historic period, see Gibson 1998; Kapstein 2009; O’Donovan 2004.

Buddhist and non-Buddhist contestation

Aside from reflecting the beliefs and devotion of Buddhists and non-Buddhists in Upper Tibet, mantric rock inscriptions served as sectarian markers, telling us much about the relationship between the two basic groups of religious adherents in the region. The relative positions of the inscriptions of these religious groups is characterized by a good deal of serial activity involving the creation and recreation of mantras in close quarters. Three interrelated epigraphic-related actions are evident in this jostling of Buddhist and non-Buddhist mantras on rock surfaces: erasure or defacement, juxtaposition and superimposition. Nowhere is this more evident in Upper Tibet than at Gnam-mtsho.

The primarily red ochre inscriptions of Gnam-mtsho are found in more than 50 limestone caves and rock faces on the southeast and north sides of the lake. At many of these sites the same pattern of epigraphic activity is observed. It consists of the initial inscribing of a non-Buddhist mantra to which a Buddhist mantra was added. In some cases, this process of inscription was repeated until cave walls were filled with mantras (e.g., figs. 28, 64, 172, 225). Most subsequent inscription on crowded rock walls at Gnam-mtsho was done by Buddhists. In other places, the mantras of both religions stand alone, the available space sufficient or of little appeal, discouraging subsequent writing in the same spot.

The inscribing of mantras here and there on stone surfaces shows that the process of epigraphic amendment was by no means an organized or orderly undertaking. The effacing or disfiguring of mantras, typically non-Buddhist ones, and the squeezing of Buddhist mantras next to or over those of the other is emblematic of a dynamic interplay. The destructive element in the epigraphic record indicates that the encounter between the two religious circles involved certain tensions. These appear to have been the result of sectarian competition, the political and economic context of which is difficult to assess beyond the generalities of a Buddhist victory. The confrontational aspect is brought further into focus by the erasure, defacement and the painting and carving over of rock art with a non-Buddhist dimension (such as counterclockwise swastikas, tiered shrines and hunting scenes) all over Upper Tibet.

Examples of the scratching or rubbing out of Sale-’od mantras are evident at various locations around the Bkra-shis do headland (e.g., figs. 11, 15, 147). An example of a carved Ma-tri mantra being scratched out is also known (fig. 288). The wear upon these erasures indicates that they were not done recently. There are many examples at Gnam-mtsho of the juxtaposition or superimposition of Buddhist mantras upon earlier non-Buddhist inscriptions (e.g., figs. 66, 73, 99, 103, 105, 110, 146, 189).

Inscriptions announcing that the Stag-lung-pa had arrived in the Klu-khang grotto of Bkra-shis do are another example of sectarian differences (e.g., figs. 43, 51). These hastily written inscriptions may convey that the inscribers were working under less than ideal conditions. That the work of epigraphers may have been viewed by other sects as a hostile action is in line with instances of sectarian troubles at Gnam-mtsho in the Vestigial period, as recorded in the Stag lung chos ’byung.

The historical perspective afforded by the ancient epigraphy and rock art of Gnam-mtsho and other areas of Upper Tibet indicates that the writing of mantras had become a sectarian tool of affirmation and domination. They were not merely religious expressions but political instruments as well. The epigraphic record squares nicely with what we know about conflicts between Buddhists and non-Buddhists in the Tibetan historical record more generally (see Part One). That Buddhist epigraphy often came later, in greater quantity and in larger letters reflects the eventual superior political status of Buddhism, which led to its spread in most quarters of Upper Tibet.

The epigraphic record indicates that non-Buddhists did not give up their erstwhile domination without some resistance. There is an example of a Sale-’od mantra superimposed on a Ma-ṇi mantra (fig. 11). The deployment of the mantra Bso A phaṭ in caves of Bkra-shis do (e.g., figs. 131, 132, 146, 157), a stronghold of non-Buddhist religious tradition in Imperial times and before, was designed to purge obstacles and other unwanted forces. Seen in the greater scheme of epigraphy and rock art on the Byang-thang, the object of expelling may well have been the Buddhists themselves.

While confrontation between Buddhists and non-Buddhists is attested in the Upper Tibetan epigraphic record, any resort to violence may have been of short duration or sporadic in nature. Had the imposition of Buddhism in the region involved sustained warfare, non-Buddhist rock inscriptions might be expected to have suffered far more damage. Also, had lengthy battles ensued, Buddhist mantras are likely to have been placed in an even more strident or aggressive manner. Although there are non-Buddhist exorcistic formulae, invectives and bellicose or slanderous speech are little represented in the rock inscriptions of Upper Tibet. Thus, the epigraphic evidence suggests that competition between Buddhists and non-Buddhists was mostly played out in the intellectual and affective arenas. While bloodshed may have infected Upper Tibet’s conversion to Buddhism (as the Stag lung chos ’byung and Mnga’ ris rgyal rabs aver), the region does not appear to have been engulfed in a protracted military tussle.

Ironically, most Tibetan pilgrims to Bkra-shis do these days are unconcerned with amcient sectarian preoccupations revealed by the epigraphy of the sacred site. They add their frayed blessed cords (srung-mdud) and butter offerings (mchod-pa) to non-Buddhist and Buddhist inscriptions alike, seeing them as magical manifestations of the power of Tibetan religion.

Catalogue of Upper Tibetan rock inscriptions

Introduction

This second half of the monographs features 525 inscriptions in 394 photographs. There are also references to other inscriptions for which there is no photographic coverage. The epigraphy is set out in geographic sequence beginning at Gnam-mtsho in the east and continuing westwards across the Byang-thang to Ru-thog. Each inscription is presented using standard categories of information including:

  1. Photograph and caption
  2. Reading of the inscription(s)
  3. Type of script
  4. Grammatical marks, if any
  5. Translation, if provided
  6. Technique of production
  7. Estimated date of production
  8. Proximate rock art, if applicable
  9. Location
  10. Additional remarks, if appropriate

The transliteration of Tibetan inscriptions follows the modern system (Wylie) with the addition of the following non-standard notations:

Oḿ: Oṁ with hooked sna-ldan mark
Ōḿ: Oṁ with hooked sna-ldan mark and subjoined a-chung
: Oṁ with sna-ldan mark consisting of dot and crescent
Ōϻ: Oṁ with sna-ldan mark consisting of dot and crescent and subjoined a-chung
Ōɱ: Oṁ with sna-ldan mark consisting of hook and crescent and subjoined a-chung

Also:

Ā: A with subjoined a-chung
hūṁ: huṁ with subjoined a-chung
hūḿ: huṁ with hooked sna-ldan mark and subjoined a-chung
ǐ: reversed vowel i (gi-gu log)

Other notations used in this work are as follows:

#: illegible syllable
{}: reading of bracketed syllables and words uncertain
+: joining of grammatical mark immediately following one another
…: indeterminate number of illegible or missing syllables in an inscription; a lacuna in a translation
(=): standard spelling equivalent
(C.T. =): Classical Tibetan word equivalent

Some of the inscriptions in the catalogue are designated as illegible or partially illegible, signifying that according to the standards of this study, a reading could not be settled upon. Subjecting unread inscriptions to further examination, additional field inspection, augmented photography and/or greater use of digital enhancement software may aid in obtaining more valid readings, partial or complete.

Bkra-shis-do, Gnam-mtsho

The largest concentration of rock inscriptions in Upper Tibet are located on rocky headlands along the southeast and northern shores of Gnam-mtsho. Gnam-mtsho is the largest lake in Inner Tibet, that portion of the Plateau comprising the Tibet Autonomous region (TAR). This brackish lake of nearly 2000 m² stretches across the southeastern extremity of the Byang-thang.*

For more information on the geography, history and culture of Gnam-mtsho, see Bellezza 1997a.

Fig. 1. Bkra-shis do (middle) as seen from an outlier of the Gnyan-chen thang-lha range. Bkra-shis do-chung is on the left and the larger Bkra-shis do-chen formation on the right. In the background is the north shore of Gnam-mtsho.

Bkra-shis do (Auspicious Headland) is a wedge-shaped headland projecting into the southeast side of Gnam-mtsho.* This famous promontory is home to around 350 ancient rock inscriptions written with red ochre and black mineral pigments. Bkra-shis do is one of the most important cultural and archaeological sites on the Byang-thang. According to Yungdrung Bon tradition, this headland was called Gnam-mtsho do-ring (Sky Lake Long Headland) and was one of the 37 assembly centers (’du-gnas) of the bon-po before and during the Imperial period.† The historical significance of Bkra-shis do is partly attributable to its geographic position. It lies near the foot of two well-used passes (La-rgan la and Rkyang la), linking the Byang-thang with ’Dam-gzhung and Central Tibet. Also, the many caves of the headland furnished shelter and served as nuclei for the construction of residential structures along the base of the escarpments.

On the caves, rock art, folklore, and religious history of Bkra-shis do, see Bellezza 1997a, pp. 173–218, 221–230. Also see, Bellezza 2008, passim; 2001, pp. 200–211; 2002a, pp. 126–130; 2002b; Suolang Wangdui 1994, pp. 141–156.

On the ’Du-gnas so-bdun, see Karmay 1972, pp. 38, 40, 41; Uebach 1999; Bellezza 2008, pp. 284, 290–292.

In addition to a wealth of rock art and rock inscriptions, there are structural remains at Bkra-shis do that appear to predate the 11th century CE. Ancient residential ruins are concentrated in the central portion of the South side of Bkra-shis do-chung. These ruins have morphological characteristics associated with archaic rock shelter complexes found in other parts of the Byang-thang. The headland is also notable for its stands of dwarf willow and juniper, creating special ecological niches in which animals such as the musk deer (gla) once thrived. There is a small Buddhist Rdzogs-chen monastery, and a cave temple named Rgwa-lo gzims-phug at Bkra-shis do-chung, both of which belong to the Rnying-ma sect.

Bkra-shis do terminates in two prominent limestone formations: Bkra-shis do-chung and Bkra-shis do chen. These towering red and white projections are separated from each other by a gap of about 250 m. The circumambulatory pilgrims’ route around Bkra-shis do-chung is approximately 3 km in length and the circuit around Bkra-shis do-chen is about 9 km in length. There are no less than 60 shallow caves and passageways in the two formations of Bkra-shis do. Among these, seven caves and surrounding areas at Bkra-shis do-chen and 21 caves and surrounding areas at Bkra-shis do-chung boast ancient rock inscriptions.

Most inscriptions at Bkra-shis do were made using red ochre pigments. A famous local source of this mineral is a small pit mine situated near the east end of Bkra-shis do-chen (Bellezza 1997a: 210, 211). This mine is called Rdo-rje phag-mo rakta and is said to be the blood of this important Buddhist tantric goddess and guardian of Gnam-mtsho. The iron oxide from this mine is still prized by Tibetans as a sacred and medicinal substance.

Since circa 2000, Bkra-shis do has been converted into an extremely popular tourist attraction. In the height of the season (June-October) several hundred buses and cars visit the site daily. This huge influx of mostly Chinese visitors (but also Tibetans and foreigners) has had a negative impact on the environmenal and cultural integrity of the headland, a place that until the 1990s was home to a coterie of local herders (’brog-pa), meditators and Tibetan pilgrims on a seasonal or permanent basis. Unfortunately, as of 2015, no effort has been made to protect and conserve the rock inscriptions, pictographs or other archaeological resources of Bkra-shis do. Increasingly, rock inscriptions and rock art are being destroyed by inappropriate construction projects, particularly on the south sides of Bkra-shis do-chung and Bkra-shis do-chen. Other rock inscriptions and art have been damaged by pilgrims who commonly believe that they are miraculously self-formed. Although this gives the ancient mineral pigment applications a sacred standing, it leads to the dabbing of butter on the inscriptions and art as an act of offering and sign of devotion. However, the fats may have a deleterious effect on the stability of pigments employed to make the ancient epigraphs and pictographs. Another growing problem is the use of synthetic paints to create new inscriptions and art, which is obscuring and defacing ancient mineral pigment applications.

Bkra-shis do-chen, Gnam-mtsho

Fig. 2. Sangs-rgyas yar-byon phug, Bkra-shis do-chen, Gnam-mtsho

Fig. 3. Fragmentary or incomplete Ma-ṇi mantra and mantric syllable

ma ṇi pad (upper)

Ōm (lower left)
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Sangs-rgyas yar-byon phug, Bkra-shis do-chen

Fig. 4. Fragmentary inscription

{thed rang do}
lcam za
Script: rudimentary dbu-can
Grammatical marks: intersyllabic tsheg
Translation: “Female of the clan.” (ln. 2)
Technique: red ochre handwriting
Estimated date: post-Imperial period
Location: Sangs-rgyas yar-byon phug, Bkra-shis do-chen

Fig. 5. Buddhist mantra

Oṁ # {ha}
Script: dbu-can
Grammatical marks: None
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Sangs-rgyas yar-byon phug, Bkra-shis do-chen
Note: This is a much larger inscription to the right of fig. 4. Below the inscription in this photograph are pictographs of a bird (visible) and a hornless deer appropriately labeled bya and yu-pa. These inscriptions date no earlier than the Vestigial period. Also see Suolang Wangdui 1994, p. 152 (fig. 196).

Fig. 6. Sale-’od mantra

Ā {A} dkar sale ’od A yang {Ōm}
’du
Script: dbu-can
Grammatical marks: brda-rnying yig-mdo mdun-ma+initial nyis-shad, intersyllabic tsheg
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Sangs-rgyas yar-byon phug, Bkra-shis do-chen
Note: Situated above inscriptions in fig. 3

Fig. 7. Buddhist prayer

sa pa kun las rnams (= rnam) rgyal
grol zhing
Script: rudimentary dbu-med
Grammatical marks: intersyllabic tsheg
Translation: “…entirely liberated” (ln. 2)
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Sangs-rgyas yar-byon phug, Bkra-shis do-chen
Note: The first line includes a personal name. The last two syllables of the first line are written is an abridged form (bsdus-yig). The left and right sides of inscription are partly cut in this photograph.

Fig. 8. Mantric syllable and prayer

hām (upper)

{rgan gyi rab} (lower)
me nub rgyal mtshan ’dzin
# {la} lha phyag ’tshal
…{la pra dzanya gu dhu hru}
Script: rudimentary dbu-med
Grammatical marks: intersyllabic tsheg, shad-gnyis (lns. 2–4)
Translation: “…Prostrations {to}…the deity holder of the undiminished victory banner.”
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Sangs-rgyas yar-byon phug, Bkra-shis do-chen
Note: The sna-ldan mark in the upper inscription is partly cut. Also the first line of the long inscription is not presented in full in this image. The final (fourth) line contains a series of Sankritized syllables serving as a eulogy or salutation for the deity noted. Situated below inscription in fig. 7.

Fig. 9. The Twin Caves, Bkra-shis do-chen, Gnam-mtsho

Fig. 10. A view of one of the walls of the Twin Caves with Ma-ṇi mantra (middle left), two endless knot symbols (right), three A between two flaming jewels symbol (bottom left), and the Sale ’od mantra below the flaming jewels on the right.

Oϻ ma ṇĭ pad me {huṁ}
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period
Proximate rock art: two endless knots (pa-tra), two flaming jewels (nor-bu me-’bar), standing anthropomorph, and other red ochre applications
Location: Twin Caves, Bkra-shis do-chen
Note: The deep red hue and elementary form of the upper endless knot suggests that it was painted by non-Buddhists. Conversely, the complexity of the design of the lower endless knot and the way the coloration and ablation of the pigment matches the Ma-ṇi mantra indicates that it has a Buddhist identity. The two flaming jewels at the bottom of the image were most probably produced by non-Buddhists.

Fig. 11. Fragmentary Sale-’od mantra partially superimposed on Ma-ṇi mantra

# # dkar sa le {’od} # yang Oϻ… (upper)

Oṁ ma ṇi padme hūṁ (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (upper)
Technique: red ochre handwriting
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika superimposed on the Ma-ṇi mantra
Location: Twin Caves, Bkra-shis do-chen

Fig. 12. Ma-ṇi mantra

{Oṁ} ma ṇĭ pad me hūṁ
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre handwriting
Estimated date: Vestigial period
Proximate rock art: standing anthropomorph
Location: Twin Caves, Bkra-shis do-chen

Fig. 13. Buddhist prayer

grol (= sgrol) ma nyer 1
rang byon g.yas {g.yon} #
{nyid la gsal} #
spor yod lags
# ge’o (= dge’o)
Script: rudimentary dbu-med
Grammatical marks: intersyllabic tsheg
Translation: “The self-formed twenty-one Savioresses (Sgrol-ma) [to the] right {and left}…clearly visible…May it be auspicious”
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Twin Caves, Bkra-shis do-chen
Note: The number in first line is partly written out and in numeral form. Lags is written in an abbreviated form. Situated below inscription pictured in fig. 12.

Fig. 14. Letter A written three times

A
A
A
Script: rudimentary dbu-can (upper two) and dbu-can (lower)
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Possibly Early Historic period
Location: Twin Caves, Bkra-shis do-chen
Note: The line below the inscription belongs to a pictographic composition consisting of a large oval outline. Inside it is the dbu-can syllable na finely inscribed by another hand. Below the oval is the inscription in fig. 11.

Fig. 15. Partially effaced Sale-’od mantra

A A dkar sa le A yang Oɱ
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Twin Caves, Bkra-shis do-chen
Note: Situated below flaming jewels depicted in fig. 10.

Fig. 16. Non-Buddhist mantric syllable

{Oṁ}
Script: Rudimentary dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Early Historic period
Proximate rock art: curvilinear design below inscription
Location: Twin Caves, Bkra-shis do-chen

Fig. 17. Two fragmentary mantras

A A (upper)

Oṁ ma ṇǐ pa (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: Twin Caves, Bkra-shis do-chen
Note: The upper inscription appears to be the first two syllables of the Sale-’od mantra. These inscriptions are situated below curvilinear subject in fig. 14.

Fig. 18. Mantric syllable

Oṁ
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period or somewhat earlier
Location: Twin Caves, Bkra-shis do-chen
Note: This syllable may have a highly worn subjoined a-chung. This inscription is situated near the floor of the cave on the same wall but at some distance from as fig. 13.

Fig. 19. Ma-ṇi mantra

Oṁ ma ṇi padme hūṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Twin Caves, Bkra-shis do-chen

Fig. 20. Mantric syllable

Oḿ
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Proximate rock art: triple gems (nor-bu gsum), clockwise swastika (partly cut in photograph) and unidentified subject (flower?)
Estimated date: Vestigial period or somewhat earlier
Location: Twin Caves, Bkra-shis do-chen
Note: The inscription may have been written in conjunction with one or more of the pictographs.

Fig. 21. Non-Buddhist mantra

Oḿ ’du
Script: dbu-can
Grammatical marks: intersyllabic tsheg, shad-gnyis
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Twin Caves, Bkra-shis do-chen
Note: This appears to be an abbreviated form of the Sale-’od mantra. The anterior part of the first syllable is cut in the photo.

Fig. 22. Incomplete Ma-ṇi mantra

me {huṁ} hri
Script: dbu-can
Grammatical marks: none
Technique: black pigment handwriting
Estimated date: Vestigial period
Location: Twin Caves, Bkra-shis do-chen

Fig. 23. The twin chambers of Brag phying-gur phug, Bkra-shis do-chen, Gnam-mtsho

Fig. 24. Ma-tri and another mantra

Ōḿ ma tri mu ye sale ’du’ (upper)

Oṁ ma (lower)
Script: dbu-can
Grammatical marks: yig-mgo mdun-ma+initial nyis-shad, intersyllabic tsheg, nyis-shad (upper)
Technique: red ochre handwriting
Proximate rock art: inscriptions are bracketed by auspicious symbols, including (from top left in clockwise direction): victory banner (rgyal-mtshan), five-pointed star (skar-ma rtse-lnga), conch (dung-dkar), triple gems, parasol (dbu-gdugs) superimposed on a pair of deer, fish (nya, partially cut in photograph), endless knot, and vase (bum-pa); also, anthropomorphic figures on the bottom left
Estimated date: Vestigial period
Location: Brag-phying gur-phug, Bkra-shis do-chen
Note: The lower inscription may be an incomplete Ma-ṇi mantra. The eight sacred symbols have a non-Buddhist identity and some of them may have been made by the same hand as the Ma-tri mantra. Except for the triple gems and star, these symbols are part of the standard set of eight auspicious symbols known as bkra-shis rtags-brgyad; however, they are rendered in a more archaic in style. Also see Bellezza 2001, pp. 208, 338 (fig. 10.38); January 2013 Flight of the Khyung, fig. 10.

Fig. 25. Sale-’od mantra

A A dkar sale ’od
A yang Oṁ ’du
Script: dbu-can
Grammatical marks: shad+nyis-shad
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Brag-phying gur-phug, Bkra-shis do-chen

Fig. 26. Fragmentary Sale-’od mantra

A A dkar sale ’od
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Brag-phying gur-phug, Bkra-shis do-chen

Fig. 27. Sale-’od mantra

A A dkar sale ’od A yang Oṁ du
Script: dbu-can
Grammatical marks: shad+nyis-shad
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Brag-phying gur-phug, Bkra-shis do-chen
Note: The inscription and some rock art in this cave have been defaced by the application of butter done by pilgrims. The strings seen in the image are discarded srung-mdud (protection cords).

Fig. 28. One of the walls of Brag-phying gur-phug, Bkra-shis do-chen. Several Ma-ṇi mantras and other mantric syllables are discernable in the image.

Note: See below for details.

Fig. 29. Incomplete Ma-ṇi mantra with letter A superimposed on right side.

Ōṁ ma ni pad hūṁ

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Brag-phying gur-phug, Bkra-shis do-chen
Note: Located on lower left side of fig. 28.

Fig. 30. Three Ma-ṇi mantras and partial Tibetan alphabet

Ōḿ {ma ni} pad me huṁ (upper)

ka ga kha {nga} cha ca ja {nya} ta tha da (upper middle)

Ōḿ ma ni pad me huṁ (lower middle)

Oṁ ma ni pad me huṁ (lower right)

mi (lower left)
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period
Proximate rock art: The last syllable of the upper inscription is superimposed on a mchod-rten painted above the upper middle inscription, mchod-rten on left side of image (partially visible), and mchod-rten partially superimposed on the lower middle and lower right inscriptions
Location: Brag-phying gur-phug, Bkra-shis do-chen
Note: These inscriptions are seen in the middle of fig. 28. There appear to be the traces of at least one letter to the right of the uppermost mchod-rten and traces of letters to the left and to the right of the lowermost inscription. The first 11 letters of the Tibetan alphabet in the upper middle inscription appears to be a practice exercise that may possibly have been written by the same hand as one of the Ma-ṇi mantras.

Fig. 31. Mantric syllable

Ōṁ
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial shad
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Brag-phying gur-phug, Bkra-shis do-chen
Note: Located on upper middle side of fig. 28.

Fig. 32. Mantric syllable

Ōḿ
Script: dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: Vestigial period
Location: Brag-phying gur-phug, Bkra-shis do-chen
Note: An aquatic bird was partially superimposed on the inscription at a much later date. Situated on middle left side of fig. 28.

Fig. 33. Mantric syllable

A
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre handwriting
Estimated date: possibly Early Historic period
Location: Brag-phying gur-phug, Bkra-shis do-chen
Note: Situated in lower middle part of fig. 28.

Fig. 34. Mantric Inscription

ga # ma hi ru ngo
sho # # # ba bya {chi}
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial shad, shad
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: several pigment applications
Location: Brag-phying gur-phug, Bkra-shis do-chen

Fig. 35. Ma-ṇi mantra

Oϻ ma ṇi padme hūṁ hri
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Brag-phying gur-phug, Bkra-shis do-chen

Fig. 36. Buddhist Seed syllables

Ōϻ Ā huṁ
Script: dbu-can
Grammatical marks: none.
Technique: red ochre writing
Estimated date: Vestigial period
Location: Brag-phying gur-phug, Bkra-shis do-chen

Fig. 37. Mantric syllable


Script: dbu-can
Grammatical marks: none.
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Brag-phying gur-phug, Bkra-shis do-chen

Fig. 38. Incomplete Ma-ṇi mantra

Oϻ ma ṇi padme
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: battle scene consisting of five figures (one arrayed against four)
Location: Brag-phying gur-phug, Bkra-shis do-chen
Note: The inscription was partially superimposed over the battle scene.

Fig. 39. Buddhist inscription labelling proximate rock art

dkon chog gsum
{nor} bu {lo}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: clockwise swastika, two flaming jewels symbols, linear motif consisting of parallel lines and crossmarks, endless knot, and vase (lower left corner)
Location: Unnamed cave, situated near eastern extremity of Bkra-shis do-chen
Note: Dkon-mchog gsum (sic) are the Triple Gems (Sangs-rgyas, Chos and Dge-’dun) represented by the larger flaming jewels symbol. The smaller flaming jewels, endless knot and vase were painted in a red ochre of different hues evidently by different artists. The style of the two former symbols indicates that they were probably made by non-Buddhists. There are three Ma-ṇi inscriptions nearby, one of which may have been written by the same hand that made the Ma-ṇi inscription directly below it. There is also an inscription that may read: phyug # la lo.

Fig. 40. Counterclockwise swastika with mantric syllables inside its arms

ma {ya A khram} (from top left in a counterclockwise direction)
Script: rudimentary dbu-can, dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Stag-lung phug, Bkra-shis do-chen
Note: The initial letter and direction in which this mantra is to be read is unclear. This mantra and the swastika were made by non-Buddhists. They represent a magical diagram of some kind. On this inscription and rock art, also see Bellezza 2002a, pp. 130, 204.

Fig. 41. One of the ancillary chambers of the Klu-khang, Bkra-shis do-chen

Fig. 42. Another ancillary chamber of the Klu-khang, Bkra-shis do-chen

Fig. 43. Religious inscription

stag {lang la} phebs
Script: dbu-med
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: Klu-khang, Bkra-shis do-chen
Note: This incomplete inscription appears to hail the arrival of the Stag-lung-pa subsect to the area.

Fig. 44. Mantric syllable

Ōϻ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Klu-khang, Bkra-shis do-chen
Note: Although the larger figure partially surrounding the syllable resembles the letter l, other linear markings situated nearby in the same ochre hue and bearing comparable wear characteristics do not appear to be letters.

Fig. 45. Several non-Buddhist inscriptions, flaming jewels symbol (upper middle), mchod-rten (middle right) and three swastikas (middle left and bottom)

Note: See below for details. On the mchod-rten, see Bellezza 2017-2018, fig. 14c.

Fig. 46. Non-Buddhist seed syllables

Ā Ōϻ hūṁ
Script: dbu-can
Grammatical marks: initial nyis-shad
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika below (partially visible)
Location: Klu-khang, Bkra-shis do-chen
Note: Situated in lower middle part of fig. 45.

Fig. 47. Various letters probably inscribed by non-Buddhists

A

ga

ka

Ōm huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: wild yak on left side (anterior portion visible)
Location: Klu-khang, Bkra-shis do-chen
Note: The vowel sign in huṁ has been reduced to a curved ligature. Situated on lower left of fig. 45.

Fig. 48. Fragmentary Ma-ṇi mantra

A ma ṇa pa
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: quadruped above mantra
Location: Klu-khang, Bkra-shis do-chen
Note: It is possible that the mark over the letter na is the vowel sign gi-gu. The animal may have been made at same time as mantra.

Fig. 49. Ma-ṇi mantra

{Oṁ} ma ṇi padme hūṁ {hri}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Klu-khang, Bkra-shis do-chen

Fig. 50. Fragmentary Sale-’od mantras

A A dka (upper)

A Ā (lower)
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika above inscription, linear subject
Location: Klu-khang, Bkra-shis do-chen
Note: The subjoined a-chung of the final A is partially cut in the photograph.

Fig. 51. Buddhist inscription

stag lung la phebs (lower)
Script: dbu-med
Grammatical marks: intersyllabic tsheg
Translation: “The Stag-lung have arrived”.
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: various red ochre applications
Location: Klu-khang, Bkra-shis do-chen
Note: The upper inscription is illegible. There are three main layers of pigment applications on the pictured cave wall. The first (oldest) layer consists of orange-red applications that appears to have been effaced and smudged. The second layer is made up of the Buddhist inscription. The third and most recent layer is comprised of ochre crayoned scribbles.

Fig. 52. Five non-Buddhist mantric letters

A A {cha} # #
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: post-Imperial period
Proximate rock art: each syllable is boxed in by red ochre lines
Location: Klu-khang, Bkra-shis do-chen
Note: Below the inscription is the syllable ku (not pictured). The compartmented mantric syllables are part of a larger rectangle subdivided into various parts above which stands an anthropomorphic figure. Also see Bellezza 2000, p. 43 (fig. 10a).

Fig. 53. Non-Buddhist inscription

kun ’tu bzang yin
Script: rudimentary dbu-can
Grammatical marks: two intersyllabic tsheg
Translation: “I am Kun-’tu bzang-[ po]”/“[This] is Kun-’tu bzang-[ po]”.
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Klu-khang, Bkra-shis do-chen
Note: Inscription situated immediately right of anthropomorphic figure noted in description of fig. 52. This inscription confirms the practice of religious traditions associated with the primordial buddha Kun-tu bzang-po (sic). It either refers to the inscriber as having mystically achieved a state tantamount to the deity or to the rock formation itself being an emanation of the deity. On all material phenomena being of the essence (ngo-bo) of Kun-tu bzang-po in the Yungdrung Bon tradition, see Achard 1998, p. 39.

Fig. 54. Incomplete or inchoate non-Buddhist mantra

A A A dkar sale ’od A
Script: rudimentary dbu-can, dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: post-Imperial period
Location: Klu-khang, Bkra-shis do-chen

Fig. 55. Non-Buddhist mantric syllable

A
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Proximate rock art: minor markings
Estimated date: Vestigial period or somewhat earlier
Location: Klu-khang, Bkra-shis do-chen
Note: Located immediately above mantra in fig. 54.

Fig. 56. Non-Buddhist mantric syllable

A
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: minor markings
Location: Klu-khang, Bkra-shis do-chen

Fig. 57. Mantra

Oϻ #
Script: dbu-can
Grammatical marks: none
Technique: red pigment writing
Estimated date: Vestigial period or later
Location: Klu-khang, Bkra-shis do-chen
Note: It is not clear if these letters are part of a longer mantra because of a lack of photographic evidence.

Fig. 58. Partially obliterated mantric syllable

hūṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Thang-lha’i mdzod-khang, Bkra-shis do-chen
Note: In a nearby cave is an inscription written in a black pigment that reads: grub thob dpal na mo. See Bellezza 1997a, p. 217. No further assessment of this inscription is available.

Bkra-shis- do-chung, Gnam-mtsho

Fig. 59. Ma-ṇi mantras

{} ma ni padme hūṁ hri (upper)

{ma ni} pad {me} hūṁ (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: near O-rgyan phug, Bkra-shis do-chung
Note: A Ma-ṇi mantra was recently carved over the lower painted one.

Fig. 60. Ma-ṇi mantra

Ōṁ ma nĭ pad me {huṁ}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: near O-rgyan phug, Bkra-shis do-chung
Note: Image enhancement software reveals that there is also a fragmentary lower Ma-ṇi mantra that appears to have been made by the same hand, upon which another inscription (incomplete in the photo) was superimposed. The first three syllables of the superimpoised inscription read: ’der skyag sa.

Fig. 61. Ma-ṇi and Phyag-na rdo-rje mantras

Oṁ ma ṇi padme huṁ
Oṁ bardzra ba
Oṁ bardzra pa ṇi huṁ
huṁ
Script: dbu-can
Grammatical marks: none
Estimated date: Vestigial period
Technique: red ochre writing
Location: near O-rgyan phug, Bkra-shis do-chung

Fig. 62. Mantra

ma
Oṁ
huṁ
ma la swa ha
Script: lan-tsha related and dbu-can
Grammatical marks: shad
Estimated date: Vestigial period
Technique: red ochre writing
Location: Gdung-rten mkha’-’gro gter-yig, Bkra-shis do-chung

Fig. 63. Mantric syllable

Oṁ
Script: lan-tsha
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Gdung-rten mkha’-’gro gter-yig, Bkra-shis do-chung

Fig. 64. Rock face and tiny cave near Gu-ru sgrub-phug

Note: See below for details.

Fig. 65. Two mantras

hūṁ ba-dzra (left)

phaṭ (right)
Script: dbu-can
Grammatical marks: shad (right)
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: Situated in upper part of fig. 64.

Fig. 66. Mantric syllables

A (upper)

{A} (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: The upper inscription was partially superimposed on the lower inscription. Also see Bellezza 2001, pp. 207, 208, 336 (fig. 10.35). Situated on the middle left side of fig. 64.

Fig. 67. Defaced Sa-le-’od inscription

# # # # le ’od # yang ’du #
Script: dbu-can
Grammatical marks: shad
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: Situated immediately above syllable in fig. 66.

Fig. 68. Inscription

dran pa
dran pa rten
Script: dbu-can
Grammatical marks:
Translation: “Memory, receptacle of memory.”
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: lines below inscription and flaming jewels symbol to the left
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: The style of the symbolic representation suggests that the rock art and inscription were created by a non-Buddhist. Situated to left of fig. 66.

Fig. 69. Rigs-gsum mgon-po mantras

Oṁ ma ṇi pad me hūṁ
Oṁ ba ki sha ri muṁ
Oṁ badzra pa ṇi hūṁ
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial nyisshad, intersyllabic tsheg, each line terminates in nyis-shad
Technique: red ochre writing
Estimated date: Vestigial period or later
Proximate rock art: geometric subject (consisting of perpendicular lines bracketing two triangles) above mantras
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: Situated immediately below fig. 68.

Fig. 70. Non-Buddhist mantras

A (left)
A
A {da}

A (right)
Ōḿ {huṁ}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: pa tra
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: Mantra on the right is partially cut in photograph. Situated to the right of fig. 69.

Fig. 71.

dran
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: Situated to right fig. 70.

Fig. 72. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: Situated to right of fig. 66.

Fig. 73. Ma-ṇi and fragmentary Sale-’od mantras, mantric syllable, long inscription, and rock art

Ōϻ ma ṇi padme hūṁ (upper)

A # sale ’od # (lower right)
ya

huṁ (bottom left)
Script: dbu-can
Grammatical marks:
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: two counterclockwise swastikas above Ma-ṇi mantra and horse (lower left) and minor pigment applications
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: For inscriptions situated in middle of photograph, see fig. 74.

Fig. 74. Long inscription pictured in fig. 73

A A A A A A
A A A A A A A A A A A
A A A {A} A A A A A A
A A A A A A A {A A}
A sum bcu rtsa
gsum ’di
ra dzi bum {pa}
{bdu nang} # # # #
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+ initial nyisshad (upper four lines), intersyllabic tsheg (upper five lines)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: horse
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: The letter A appears 33 times in the first four lines not including the three examples that have been crossed out in the second and third lines using red ochre (possibly by the inscriber himself). The letter A in the fifth line appears to be part of the name of the inscription: “Thirty-Three A” (written in a non-standard manner with the letter r added after the third syllable). The first two syllables of the eighth (final) line may possibly be part of the seventh line. The final line of the inscription was partially superimposed on the horse below. The significance of the inscription is unclear. It appears to have been written by a non-Buddhist. The letter A may possibly represent a seed syllable (sa-bon) for a non-Buddhist tutelary deity or buddha. Alternatively, the inscription may be dedicated to the 33 dimensions of gods situated above the cosmic mountain (G.yen-khams sum-bcu rtsa-gsum). Also see Bellezza 2001, pp. 208, 337 (fig. 10.37).

Fig. 75.

Oṁ
A
{huṁ}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: The lowest syllable is partially obliterated and cut in the photograph.

Fig. 76. Non-Buddhist mantric syllables

A
Ōṁ
Script: dbu-can
Grammatical marks: yig-mgo mdun-ma+sgab-ma+ initial nyis-shad
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung

Fig. 77. Mantric syllables

Ā (upper)

Oṁ (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: significant pigment application below lower inscription that appears to include letters
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung

Fig. 78. Non-Buddhist mantric syllable

A
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma+initial nyis-shad
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung

Fig. 79. Non-Buddhist mantric syllables

A
Oṁ
huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: Except for the superscribed vowel symbol of the middle syllable, the lower two syllables are not pictured.

Fig. 80. Mantric syllables

Oṁ
A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung

Fig. 81. Two fragmentary inscriptions, Dpa’-bo ’bru-lnga mantra, and Buddhist inscription

ma # sa # (upper)

A Ōṁ huṁ raṁ {dza} (upper middle)

thal ru rin po {che} la na mo (lower middle)
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial shad, nyis-shad (lower middle)
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: The lower middle inscription is in praise of a Buddhist lama who appears to belong to a clan (sic: rus) or camp known as Thal. The lower inscription of three lines is largely illegible. A Dpa’-bo bru-lnga mantra was written in a black pigment upon a boulder in an ancient rock shelter on the headland of Rdo-ring, Gnam-mtsho. On this site and inscription, see Bellezza 1997a, pp. 271, 272.

Fig. 82. Mantric syllables

Oṁ # # (upper)

A (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: Located above inscriptions in fig. 81. Also see Bellezza 2001, pp. 207, 208, 337 (fig. 10.36).

Fig. 83. Ma-ṇi inscription

Oṁ ma ṇi pad me huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: Located below inscriptions in fig. 82.

Fig. 84. Non-Buddhist mantric syllables and inscription

A (upper)
Oṁ
{huṁ}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: Ochre pigment applications between the two inscriptions
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: The lower inscription of at least three lines has been largely destroyed.

Fig. 85. Sri-gcod ’bum-pa (white structure at base of rock formation) with inscription set beside a large pictographic mchod-rten.

Note: On the base of the mchod-rten located to the right of the inscription, there is a large brda-rnying yig-mgo mdun-ma with an appended nyi-shad and what remains of the syllable Oṁ or A written in very large characters. The erosion exhibited by this is inscriptions indicates that it is of considerable age. See fig. 86 for more details. For the adjacent pictographic mchod-rten (one of five in a row), see Bellezza 2001, pp. 209, 210, 339 (fig. 10.40); March 2011 Flight of the Khyung, fig. 17; 2017-2018, mchod-rten 15a.

Fig. 86. Non-Buddhist mantric syllables

A A
A
A
A
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Early Historic period
Proximate rock art: mchod-rten (tiny portion of which is visible on the right)
Location: Sri-gcod ’bum-pa, Bkra-shis do-chung
Note: Also see Bellezza 2001, pp. 209, 338 (fig. 10.39).

Fig. 87. Mantric syllables

A (upper)

A (lower)
Script: dbu-can (upper) and rudimentary dbu-can (lower)
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: Red ochre application that may be obscuring additional letters
Location: Sri-gcod ’bum-pa, Bkra-shis do-chung

Fig. 88. Four Ma-ṇi mantras and other mantras, Cave I near Bka’-brgyud phug-chen, Bkra-shis do-chung

Note: See below for details.

Fig. 89. Two Ma-ṇi mantras and sacred syllables

A (upper)

Oϻ ma ṇi pad me hūṁ hri (middle)

Oṁ ma ṇi pad me (lower left)

huṁ hri (lower right)
Script: dbu-can
Grammatical marks: intersyllabic tsheg, shad (middle); initial shad-nyis (lower left)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: pair of fish and other sacred symbols partly visible on lower right
Location: Cave I near Bka’-brgyud phug-chen, Bkra-shis do-chung
Note: Inscriptions situated on middle left side of fig. 88. On these sacred symbols, see January 2013 Flight of the Khyung, figs. 6–9.

Fig. 90. Fragmentary Ma-ṇi mantra

Oṁ ma ṇi
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: parasol-like subject and three or four counterclockwise swastikas above inscription
Location: Cave 1 near Bka’-brgyud phug-chen, Locus 1, Bkra-shis do-chung
Note: Situated in middle of fig. 88.

Fig. 91. Fragmentary Ma-ṇi mantra

Ōṁ ma ṇǐ
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial shad, intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: Cave I near Bka’-brgyud phug-chen, Bkra-shis do-chung
Note: Located below inscriptions in fig. 89.

Fig. 92. Buddhist purification mantra

Oṁ badzra sa{ṭva} huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Cave I near Bka’-brgyud phug-chen, Bkra-shis do-chung
Note: Situated on right side of fig. 88.

Fig. 93. Ma-ṇi mantra

Ōϻ ma ni pad me huṁ hri
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: Cave I near Bka’-brgyud phug-chen, Bkra-shis do-chung

Fig. 94. Ma-ṇi mantra

Ōḿ ma ni pad me hūḿ
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma, intersyllabic tsheg, nyis-shad
Technique: red ochre writing
Estimated date: Vestigial period
Rock art: clockwise swastika
Location: Cave II near Bka’-brgyud phug-chen, Bkra-shis do-chung

Fig. 95. Mantra for tutelary deity

Oṁ hri ha ha hūṁ hūṁ phaṭ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Cave II near Bka’-brgyud phug-chen, Bkra-shis do-chung

Fig. 96. Fragmentary mantra

{A} #
#
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika (left) and circle (middle right)
Location: Cave II near Bka’-brgyud phug-chen, Bkra-shis do-chung
Note: The swastika and inscription may possibly form an integral composition. Situated above mantra in fig. 95.

Fig. 97. Initial mark

Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma+sgab-ma+nyi-zla rtags+initial nyis-shad
Technique: red ochre writing
Estimated date: Vestigial period
Location: Cave II near Bka’-brgyud phug-chen, Bkra-shis do-chung

Fig. 98. Fragmentary Ma-ṇi mantra

Oṁ ma
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Cave II near Bka’-brgyud phug-chen, Bkra-shis do-chung
Note: A Ma-ṇi mantra was engraved over the inscription in recent times.

Fig. 99. Ma-ṇi mantra and non-Buddhist mantric syllable

{Oṁ} ma nǐ pad {me hūṁ} (upper)

A (lower)
Script: dbu-can (upper), rudimentary dbu-can (lower)
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period (upper), Early Historic period (lower)
Location: Cave III near Bka’-brgyud phug-chen, Bkra-shis do-chung

Fig. 100. Mantric syllable and Vajra Guru mantra

A (upper)

Ōṁ A hūṁ badzra gu ru pad{me} siddhi hūṁ (lower)
Script: rudimentary dbu-can (upper), dbu-can (lower)
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier (upper), Vestigial period or later (lower)
Location: Cave III near Bka’-brgyud phug-chen, Bkra-shis do-chung
Note: The right side of th lower inscription is partly cut in the photograh.

Fig. 101. Mantric syllable

Oṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Cave III near Bka’-brgyud phug-chen, Bkra-shis do-chung

Fig. 102. Three large Ma-ṇi mantras and other mantras, base of escarpment near Bka’-brgyud phug-chung, Bkra-shis do-chung. The largest Ma-ṇi mantra (lower right) was made much more recently.

Note: See details below.

Fig. 103. Ma-ṇi and Sale ’od mantras

A A dkar sa le ’od A yang Oṁ ’du (upper)

Ōϻ ma ṇi pad me hūṁ (lower)
Script: rudimentary dbu-med (upper), dbu-can (lower)
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: escarpment near Bka’-brgyud phug-chung, Bkra-shis do-chung
Note: The last syllable of the upper inscription has an inverted vowel sign. Situated on left side of fig. 102.

Fig. 104. Non-Buddhist mantric syllables

{A} Ōϻ huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: escarpment near Bka’-brgyud phug-chung, Bkra-shis do-chung
Note: Very large inscription situated in lower left half of fig. 103. The Ma-ṇi mantra in fig. 103 was superimposed on the inscription.

Fig. 105.

A Oṁ hūḿ raṁ dza (upper)

Ōϻ ma ni pad me huṁ hri (lower)
Script: dbu-can
Grammatical marks: shad (upper); intersyllabic tsheg, gter-tsheg (lower)
Technique: red ochre writing
Estimated date: Vestigial period
Location: escarpment near Bka’-brgyud phug-chung, Bkra-shis do-chung
Note: Situated in right half of fig. 102.

Fig. 106. Buddhist mantric syllables

Oṁ A huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: escarpment near Bka’-brgyud phug-chung, Bkra-shis do-chung
Note: Located immediately above fig. 105.

Fig. 107. Non-Buddhist mantric syllables

A Oṁ huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: escarpment near Bka’-brgyud phug-chung, Bkra-shis do-chung
Note: Situated immediately to right of inscription in fig. 105.

Fig. 108. Inscription

{ni} pa
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: escarpment near Bka’-brgyud phug-chung, Bkra-shis do-chung
Note: Probably an incomplete Ma-ṇi mantra. Situated on far-left side of fig. 102.

Fig. 109. Sale-’od and Ma-ṇi mantras and mantric letters and syllables

A A dkar sa le ’od (upper)
A yang Ōϻ ’du

Ōϻ ma ṇi pad me hūṁ (middle)

A ’ca (lower left)

A (right side)
Oṁ
huṁ
Script: dbu-can
Grammatical marks: intersyllabic tsheg (upper and middle)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: endless knot
Location: Mi-tshad phug, Bkra-shis do-chung
Note: These inscriptions were made by different hands. The initial letter A in the Sale-’od mantra was written or rewritten by a different person than the rest of the inscription.

Fig. 110. Ma-ṇi and fragmentary Sa-le-’od mantras

Ōϻ ma ṇi pad me hūṁ (upper)

A A {dkar} (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (upper)
Technique: red ochre writing
Proximate rock art: flaming jewels and conjoined sun and moon
Estimated date: Vestigial period (upper), possibly post-Imperial period (lower)
Location: Mi-tshad phug, Bkra-shis do-chung
Note: The two sacred symbols appear to have been made by the same hand as the incomplete Sale-’od mantra. The Ma-ṇi mantra seen at the top of the image was written much more recently.

Fig. 111. Two caves before the Bde-mchog yab-yum formation (proceeding in a clockwise direction, Bkra-shis do-chung

Fig. 112. Fragmentary Sale-’od mantra

# # # sa le ’od # #

Oϻ ’du
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: black pigment applications
Location: west cave of two caves before the Bde-mchog yab-yum formation, Bkra-shis do-chung.
Note: The inscription is cut in the photograph.

Fig. 113. Mantric syllable

Ōṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Right cave of two caves before the Bde-mchog yab-yum formation (proceeding in a clockwise direction, Bkra-shis do-chung,
Note: Situated above inscription in fig. 112.

Fig. 114. Inscription and fragmentary Ma-ṇi mantra

{bon} po (upper)
do

Ōϻ ma (lower)
Script: dbu-med (upper), dbu-can (lower)
Grammatical marks: shad (upper)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: various red ochre applications.
Location: Near Rdo-zhun phug, Bkra-shis do-chung

Fig. 115. Buddhist seed syllables and Ma-ṇi mantra

{Oṁ} (upper and right side)
A
huṁ

Ōϻ ma ṇi pad me huṁ (middle)
Script: dbu-can
Grammatical marks: one intersyllabic tsheg (middle)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: circle above the syllable me
Location: Near Rdo-zhun phug, Bkra-shis do-chung
Note: The better condition, calligraphy and placement of the first two syllables of the middle mantra indicate that they were made by a different hand than the rest of the mantra. The uppermost syllable of the upper mantra is cut in the photograph impeding a full assessment.

Fig. 116. Pad-sdong phug, Bkra-shis do-chung

Fig. 117. Fragmentary Ma-ṇi mantra

Oṁ ma ni pad me
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Pad-sdong phug, Bkra-shis do-chung
Note: Situated on upper right side of fig. 116.

Fig. 118. Inscription

dran sod
kya
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Translation: “Remember!”
Technique: red ochre writing
Estimated date: Vestigial period
Location: Pad-sdong phug, Bkra-shis do-chung
Note: This inscription is written in the vernacular of the Gnam-mtsho and ’Dam-gzhung region.

Fig. 119. Inscriptions and non-Buddhist mantric syllables

dpa’ # (left)

A Oṁ huṁ (middle)

A ma Oṁ (right)

A (lower right)
Script: dbu-med (left), dbu-can
Grammatical marks: intersyllabic tsheg (left)
Technique: red ochre writing
Estimated date: Vestigial period (left), Vestigial Period or earlier (middle, right, lower right)
Proximate rock art: flaming jewels (far right)
Location: Pad-sdong phug, Bkra-shis do-chung
Note: The right inscription may have been made in conjunction with the flaming jewels. There appears to be more to the dbu-med inscription on the left but photographic coverage is lacking. See fig. 120 for the dbu-med inscription near the bottom of the photograph.

Fig. 120. Prayer to the female water spirits

rnam mtsho ’phyug mo ’i mkhro’ klu gdong ma rang
byon mkhyen no / bkris
Script: dbu-med
Grammatical marks: initial tsheg-shad, some intersyllabic tsheg, gter-tsheg after both lines
Translation: “It is written: be aware of us self-formed serpent-faced dakinis of Rnam-mtsho (= Gnam-mtsho) phyug-mo”.
Technique: black pigment writing
Estimated date: Vestigial period
Location: Pad-sdong phug, Bkra-shis do-chung
Note: The word mkha’-’gro is written in a conjunct form (bskungs-yig). A quincunx of dakinis with the serpent face of the water spirits is derived from the Ma-rgyud tradition of Yungdrung Bon (mkha’-’gro klu-gdong lnga). See Bellezza 1997a, p. 110. Given this inscription, it may be that an older name for Pad-sdong phug was Klu-gdong phug. Ribbed surfaces in this cave are associated with the serpent faced goddesses (ibid., 180, 181).

Fig. 121. Mantric syllables

A (upper)

Oṁ (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Pad-sdong phug, Bkra-shis do-chung
Note: For dbu-med inscription at bottom of photograph, see fig. 122.

Fig. 122. Buddhist prayer

Stag lung pa ’khyeno (= mkhyen)
Script: dbu-med
Grammatical marks: none
Technique: red ochre writing
Translation: “Stag-lung-pa be aware of us.”
Estimated date: Vestigial period
Location: Pad-sdong phug, Bkra-shis do-chung

Fig. 123. Buddhist memorial prayer

tshe ’das chag nga ma
# # #
mdo yu ma nu mo
ta sren rdoe
A A pa mgrol ma
Script: dbu-can
Grammatical marks: none
Technique: black pigment writing
Translation: “The departed Chag-nga-ma, # # #, Mdo yu-ma, Ta-sren, mother, father, Rdo-rje, Mgrol-ma (= Sgrol-ma).”
Estimated date: Vestigial period or later
Proximate rock art: two clockwise swastikas
Location: Pad-sdong phug, Bkra-shis do-chung
Note: The name Rdo-rje is written in abbreviated form. The names of the departed must have been registered by their relatives.

Fig. 124. Fragmentary Ma-ṇi mantra

{ma} ṇi padme huṁ hri
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Pad-sdong phug, Bkra-shis do-chung
Note: The left side of the inscription is cut in the photograph.

Fig. 125. A cave with a large red ochre pictograph of a khyung with pictograph of Phyag-na rdo-rje below (middle of photograph). Note the two counterclockwise swastikas on the upper left side of the photograph

Note: The name of this cave was coined by the author. See below for details.

Fig. 126. Ma-ṇi mantras and mantric syllable

Oḿ ma ṇi pad me hūḿ (upper left)

A (upper right)

Oḿ ma ṇǐ pad me hūḿ (middle)
hri

{A ma}… (lower)
Script: dbu-can
Grammatical marks: shad (upper left); brda-rnying yig-mgo mdun-ma+initial nyis-shad, shad (middle)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika (far right)
Location: Big Khyung Cave, Bkra-shis do-chung
Note: The lower inscription appears to be an incomplete Ma-ṇi mantra. There are a several letters to the right of the upper right inscription but they are not legible in the photographs available for study. The lower Ma-tri mantra and counterclockwise swastika just above were made recently with synthetic paint. The Oṁ below the Ma-tri mantra carries a date of 2000 CE. This modern painted inscription is already fading. Situated in left half of fig. 125.

Fig. 127. Non-Buddhist mantric syllables

bso (upper)

A Ōϻ hūṁ (lower)
Script: dbu-can
Grammatical marks: shad (lower)
Technique: red ochre writing
Estimated date: Vestigial period
Location: Big Khyung Cave, Bkra-shis do-chung
Note: Situated to right of upper Ma-ṇi mantra in fig. 126.

Fig. 128. Ma-ṇi mantra

me {hri}

Oḿ ma na pad
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: large pictographs of khyung and Phyag-na rdo-rje to right (partially visible in photograph)
Location: Big Khyung Cave, Bkra-shis do-chung
Note: Situated below and to right of inscription in fig. 127.

Fig. 129. Mantra of Phyag-na rdo-rje

Ōḿ badzra pa ni huḿ
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+shad-nyis
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: image of Phyag-na rdo-rje standing on corpse above inscription (only small portion of which is visible in the photograph)
Location: Big Khyung Cave, Bkra-shis do-chung
Note: Situated on lower right side of fig. 125.

Fig. 130. Buddhist Prayer

lha {chang} # #
bar snang # #
# # # # # #
dgyes pa rdo rje #
Script: dbu-can
Grammatical marks: some intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Translation: “Deity…Intermediate Space…Joyful Adamantine One.”
Location: Big Khyung Cave, Bkra-shis do-chung
Note: The photographic record of this inscription is inadequate for a full assessment. Situated well above inscription in fig. 129.

Fig. 131. Non-Buddhist mantra

bso A phaṭ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: partially superimposed deer in black pigment
Location: Big Khyung Cave, Bkra-shis do-chung

Fig. 132. Non-Buddhist mantra

bso
A phaṭ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika below inscription that appears to have been made by same individual who wrote the mantra
Location: Big Khyung Cave, Bkra-shis do-chung
Note: Situated directly above inscription in fig. 131.

Fig. 133. Mantric syllable

Ōϻ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Big Khyung Cave, Bkra-shis do-chung

Fig. 134. Sale-’od mantra and other inscription

A A {dkar sale} ’od A yang
Ōṁ ’du (upper)

bod bla {ma dpang} (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (lower)
Technique: red ochre writing
Translation: “Tibetan Lama…”
Estimated date: Vestigial period
Proximate rock art: deer, anthropomorph and bird superimposed on upper inscription
Location: Big Khyung Cave, Bkra-shis do-chung
Note: Situated to right of inscriptions in fig. 126.

Fig. 135. Fragmentary religious inscription of at least eight lines

kyi gling…’dzam {bu} gling
lha la khyad {par ’phags pa}…{phyug}
# # la khyad {par ’phags} # # gsum
{bsags} khyad {par}…sa bon
…thog
Script: rudimentary dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+ initial nyisshad, nyis-shad
Technique: red ochre writing
Estimated date: Early Historic period
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: Only a few syllables in the first five lines of the inscription are discernable, precluding a coherent reading.

Fig. 136. Inscription

{rgyal} #
{kyid pas mchod}
bul
Script: dbu-can
Grammatical marks: unclear
Technique: red ochre writing
Estimated date: Imperial period
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: This inscription appears to be a religious offering.

Fig. 137. Sale-’od mantra and other inscription

A A dkar sa le ’od A yang Oṁ ’du (upper)

chag sdang la (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Translation: “Lust, hatred…”
Technique: red ochre writing
Estimated date: Vestigial period
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: The lower inscription with its mention of two of the three afflictive mental conditions or poisons (dug-gsum), may possibly have been written to depreciate the non-Buddhist mantra above.

Fig. 138. Illegible inscription

# # # # #
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+ initial nyisshad
Technique: red ochre writing
Estimated date: unclear
Proximate rock art: non-Buddhist mchod-rten below (only finial visible in photograph)
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: The inscription may have been made by the same individual as the mchod-rten. Above a nearby mchod-rten is a red ochre dbu-can inscription that appears to read: nang. Superimposed on this mchod-rten is an inscription in tiny black letters of at least ten lines in what appears to be dbu-med (for this mchod-rten, see Bellezza 2017-2018, fig. 10l). Almost nothing of this extremely degraded inscription is readable despite adequate photgraphic documentation.

Fig. 139. Phyag-na rdo-rje and Ma-ṇi mantras

Ōṁ badzra pa ni hūṁ (upper)

{Oṁ} ma {ni} pad me {huṁ} (middle)

Oṁ ma ṇi pad me huṁ hri (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: lower inscription superimposed on what appear to be non-Buddhist tiered shrines
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: There is another inscription of a few letters (possibly an incomplete Ma-ṇi mantra) above and to the right of the lower inscription.

Fig. 140. Fragmentary inscription(s)

Script: dbu-can
Grammatical marks: unclear
Technique: black pigment writing
Estimated date: possibly post-Imperial period
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: The one or more inscriptions in the image is virtually illegible. The word zlog (repulse, reverse) can be made out in the third line using image enhancing software.

Fig. 141. Fragmentary inscription

# # lnga’o # # mi ’jig pa # (ln. 3)
khyed rnams… (ln. 4)
yig ge ’di # # # zhig dbang… (ln. 5)
chod lo
Script: dbu-can
Grammatical marks: none
Technique: black pigment writing
Translation: “The fifth…without fear…You all…this writing…power…”.
Estimated date: possibly post-Imperial period
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: Most of this inscription of at least six lines is illegible. The two syllables yig ge were written is an abridged form (bsdus-yig).

Fig. 142. Fragmentary inscription

’di la (ln. 1)
Script: dbu-can
Grammatical marks: unclear
Technique: black pigment writing
Estimated date: Vestigial period
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: Almost all this inscription remains illegible.

Fig. 143. Inscription and non-Buddhist mantra

…kun {bzang} rang byon (upper)
… ’dug bzhug pa red
…gter mang po yod pa red
…{kun gyi} dad mos mdzod

…mal mal sa ha (lower)
Script: dbu-med (upper), dbu-can (lower)
Grammatical marks: intersyllabic tsheg, shad, gter-tsheg (upper); nyis-shad (lower)
Technique: black pigment writing
Translation: “…There is the self-formed Kun-[tu] bzang-[po]… residing…There are many religious treasures. Keep the faith in…”
Estimated date: Vestigial period or later
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: The lower inscription appears to be the last three syllables of a Du-tri-su mantra.

Fig. 144. Fragmentary non-Buddhist inscription

…gyis sya bar zhu

…sha nye ring kha lag
…{’brel}
Script: rudimentary dbu-med
Grammatical marks: intersyllabic tsheg, nyis-shad
Estimated date: Vestigial period
Technique: black pigment writing
Proximate rock art: counterclockwise swastika
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: The counterclockwise swastika appears to have been made by the same individual who wrote the inscription. The swastika and the word sya, an ancient form of gsung (speech), indicate that this inscription was made by a non-Buddhist. The legible words might suggest an offering to the dead, attesting that the relatives and all material things needed have been provided. Situated below inscriptions in fig. 143.

Fig. 145. Fragmentary Sale-’od mantra

A dkar sa le ’od A yang
Script: dbu-can
Grammatical marks: unclear
Technique: black pigment writing
Estimated date: Vestigial period
Proximate rock art: two counterclockwise swastikas
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: The swastikas and mantra appear to have been made by the same hand.

Fig. 146. Non-Buddhist mantra and Ma-ṇi mantra.

bso A phaṭ (upper)

Ōṁ ma ṇi pad me hūṁ hri (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg, gter-tsheg (lower)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika above
Location: In or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: The upper mantra and swastika appear to have been made by the same individual.

Fig. 147. Effaced Sa-le-’od mantra

A A dkar sal le ’od A yang Oḿ ’du
Script: dbu-can
Grammatical marks: shad
Technique: red ochre writing
Estimated date: Vestigial period
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung

Fig. 148. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Proximate rock art: bird-like subject and row of five subjects resembling sacrificial cakes (gtor-ma) or tiered shrines
Estimated date: Vestigial period
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: The inscription and rock art appear to have been made by the same hand.

Fig. 149. Non-Buddhist mantra

A Oṁ (upper)
huṁ

{ca} (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: shrine-like subject, five-pointed star and two counterclockwise swastikas
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung
Note: The upper inscription and rock art appear to have been made by the same hand.

Fig. 150. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: in or near Sdig-pa’i ’dzul-khung, Bkra-shis do-chung

Fig. 151. Non-Buddhist mantra and possible Ma-ṇi mantra

A phaṭ (upper)
Script: dbu-can
Grammatical marks: unclear
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: hourglass-like subject and counterclockwise swastika
Location: Dbu-ma phug, Bkra-shis do-chung
Note: The swastika, other pictograph and inscription appear to have been produced by the same individual. Only the superscript above the letter A of a Ma-ṇi mantra is visible at the bottom of the photograph, insufficient evidence to properly appraise the inscription.

Fig. 152. Bon gyi phug, Bkra-shis do-chung

Fig. 153. Buddhist seed mantra and Sale-’od mantras

Oḿ (upper)
A
huḿ

A A dkar sa le ’od A yang Oṁ ’du (middle)

A A {dkar sa le ’od A yang Oṁ} ’du (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (lower)
Technique: red ochre writing
Estimated date: Vestigial period
Location: Bon gyi phug, Bkra-shis do-chung

Fig. 154. Non-Buddhist mantras

bso A
bso A phaṭ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: rectilinear subject above, counterclockwise swastika below
Location: Bon gyi phug, Bkra-shis do-chung
Note: The syllable phaṭ appears to be linked to both lines. The first syllable of the second line was partially effaced. The swastika appears to have been made by the same individual as the inscription. Located to left of fig. 159.

Fig. 155. Inscription and non-Buddhist mantra

{du dod} rgyud pa (upper)

zha # (upper right)

bso (lower right)
Script: rudimentary dbu-med (upper), dbu-can
Grammatical marks: intersyllabic tsheg, nyis-shad (upper)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: two counterclockwise swastikas
Location: Bon gyi phug, Bkra-shis do-chung
Note: The swastikas appear to be associated with the non-Buddhist inscription on the lower right side of the photograph. The letters zha # are partially superimposed on the lower inscription.

Fig. 156. Inscription

bon/ban kyi phag pa
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Translation: “The pig of Buddhists.”/ “The pig of Bon.”
Location: in or near Bon gyi phug, Bkra-shis do-chung
Note: This inscription as it stands may not be the original wording because the meaning it conveys is highly irregular. The reading given varies according to whether the faint superscribed na-ro is incorporated or disregarded. It may be that the word phag originally had a zhabs-kyu vowel sign; ochre pigment traces can be seen. If so, it renders the word phug (cave). Perhaps, therefore, the inscription should be read as: “Cave of Bon”.

Fig. 157. Non-Buddhist mantra

bso A phaṭ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: in or near Bon gyi phug, Bkra-shis do-chung
Note: The swastika appears to have been made by the same individual as the inscription.

Fig. 158. Fragmentary Sale-’od mantra

A A
sa le A
{’du}
Script: dbu-can
Grammatical marks: some intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: in or near Bon gyi phug, Bkra-shis do-chung

Fig. 159. Fragmentary Du-tri-su mantra

A rmad du tri su nag po shi shi {mal}
gsung bar {zhu}
Script: dbu-can (upper line), dbu-med (lower line)
Grammatical marks: intersyllabic tsheg (lower line)
Technique: red ochre writing
Translation: “The speech that is given: A rmad du tri su nag po shi shi {mal}.”
Estimated date: Vestigial period
Location: in or near Bon gyi phug, Bkra-shis do-chung

Fig. 160. Ma-ṇi mantra

Oḿ
ma
ṇi
padme
huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: in or near Bon gyi phug, Bkra-shis do-chung
Note: The very top and bottom of the inscription are cut in the photograph.

Fig. 161. Buddhist seed syllables

Oṁ
A
huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: in or near Bon gyi phug, Bkra-shis do-chung
Note: The top of the inscription is cut in the photograph.

Fig. 162. Mna’-bshag phug, Bkra-shis do-chung

Note: See details below.

Fig. 163. Ma-ṇi mantra (upper), Ma-ṇi mantra (middle right), inscription (middle left) Ōṁ (far right), Sale-’od mantra (left), and counterclockwise swastika (lower left), Mna’-bshag phug

Ōḿ {ma} ni {pad} me {hūṁ} (upper)

Oṁ ma ni pad me hūṁ (middle right)

{ra da} (middle left)

Ōṁ (far right)
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial shad (upper)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: Mna’-bshag phug, Bkra-shis do-chung
Note: The Ōṁ on the far right appears to be part of a longer sequence of syllables but the photographic record is insufficient for a full assessment. For the Sale-’od mantra (lower left), see fig. 164.

Fig. 164. Sale-’od mantra

A A dkar sa le ’od A yang {Oṁ} ’du
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: Mna’-bshag phug, Bkra-shis do-chung
Note: The first letter A of the inscription is damaged and not visible in the photograph. Located on left side of fig. 163.

Fig. 165. Non-Buddhist mantric syllables

A A
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Early Historic period
Proximate rock art: unidentified figure
Location: Mna’-bshag phug, Bkra-shis do-chung
Note: Located below and to left of inscription in fig. 163.

Fig. 166. Probable sutra inscription

Script: dbu-can
Grammatical marks: unclear
Technique: black pigment writing
Estimated date: Vestigial period or later
Location: Mna’-bshag phug, Bkra-shis do-chung
Note: This fragmentary but extensive inscription has not been examined closely.

Fig. 167. Zhwa-dmar phug, Bkra-shis do-chen

Fig. 168. Buddhist mantras

Oṁ ma ṇi padme hūṁ

Oṁ ma ṇi padme hūṁ

Oṁ ma ṇi pad me hūṁ

Oṁ bdzara twasa hūṁ

Oṁ ma ṇi padme hūṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or later
Location: Zhwa-dmar phug, Bkra-shis do-chung
Note: These mantras while crudely written, exhibit modern calligraphic features suggesting that they may postdate the 14th century CE. In the fourth mantra the syllabes twa and sa have been transposed.

Fig. 169. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Zhwa-dmar phug, Bkra-shis do-chung

Fig. 170. Ma-ṇi mantra

Ōṁ ma ṇi pad huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Zhwa-dmar phug, Bkra-shis do-chung
Note: The superimposed mantra was written recently.

Fig. 171. Sgrol-ma phug, Bkra-shis do-chung

Fig. 172. A portion of the rear wall of Sgrol-ma phug with various inscriptions and rock art.

Note: See below for details.

Fig. 173. Ma-ṇi mantras and mantric syllables

(upper right)

Oṁ ma ṇi pad me huṁ (left)

Oṁ ma ṇi pad me huṁ (right)

Oṁ (lower left)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (right)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: sun, tiered shrine (only bottom part visible in photograph)
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: Inscriptions situated on upper right side of fig. 172.

Fig. 174. Buddhist seed syllables and another inscription

Oḿ (right)
A
hūṁ

ṇǐ (left)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: Below is what appears to be an incomplete letter A. Situated to right of fig. 173 and to left of fig. 174.

Fig. 175. Desultory syllables and letters some of which probably formed Ma-ṇi mantras

Oṁ (top)

A (lower middle left)

A pa (lower left)

dme (lower right)

hūṁ (lower right)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: various red ochre applications
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: For the large vertical inscription on the left, see fig. 174. The inscription in the lower left corner is liable to read huṁ but photographic documentation is inadequate for a complete reading. What may be an incomplete huṁ is situated to left of lower right inscription.

Fig. 176. Buddhist seed syllables

Oṁ A hūṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: Situated on upper right side of fig. 172.

Fig. 177. Seed syllables and Ma-ṇi mantra

Oṁ Oṁ (upper)

Oṁ ma ṇi pad me hūṁ hri (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: The first inscription appears to represent an incomplete Ma-ṇi mantra. The vowel sign of the second Oṁ of the upper inscription was not completed. Situated immediately to the right of fig. 176.

Fig. 178. Ma-ṇi mantra and another inscription

Ōϻ ma ṇi pad me hūṁ (upper)

ma cig rang byon # (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (upper)
Technique: red ochre writing
Translation: “Self-formed Only Mother…”
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika between the two inscriptions
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: Situated to left of fig. 173. The lower inscription refers to a female deity.

Fig. 179. Mantric syllable


Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: Situated in the middle of fig. 172.

Fig. 180. Non-Buddhist mantra

bso A phaṭ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: The swastika appears to have been made by same hand as the inscription. Situated above fig. 179.

Fig. 181. Dpa’-bo ’bru-lnga mantra

A Ōϻ hūṁ raṁ dza
Script: dbu-can
Grammatical marks: gter-tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: Situated immediately to left of fig. 180.

Fig. 182. Inscription

skyo rten po do phug du gtad do
do phug gĭ lha ’dre rnams kyĭs
gtad {tshĭg} ’dĭs dam nyams
{d/sgrol}
Script: rudimentary dbu-med
Grammatical marks: intersyllabic tsheg, shad
Translation: At the headland cave I handover Skyo rten-po. Lha and ’dre of the headland cave, {slay} the oath-breaker through the {words} of this malediction.”
Technique: red ochre writing
Estimated date: post-Imperial period
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: Translation is provisional as two words are not fully legible. Situated below figs. 180 and 181. On this inscription, see Bellezza 2008, p. 187 (fn.193).

Fig. 183. Inscription

Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Sgrol-ma phug, Bkra-shis do-chung
Note: This inscription is illegible. It appears to have been purposely erased.

Fig. 184. Non-Buddhist mantric syllables

A (upper)

A (lower)
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Early Historic period
Location: Unclassified, Bkra-shis do-chung

Fig. 185. Sale-’od mantras

A A dkar sa le ’od A yang Oṁ du
A A dkar sa le ’od A yang
Script: rudimentary dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: nyi-zla and nor-bu me-‘bar below inscription
Location: Unclassified, Bkra-shis do-chung
Note: The two sacred symbols may have been made by the same hand as the inscription. The last portion of the inscription is cut in the photograph.

Other sites at Gnam-mtsho
Rock inscriptions in red ochre and other mineral pigments are found along the western half of the rugged north shore of Gnam-mtsho. These ancient epigraphs occur in a series of clefts and caves in limestone formations facing directly onto the lake. The epigraphs are comparable in contents, paleography and techniques of production to those of Bkra-shis do. Likewise, an overarching historical theme is the encounter between the non-Buddhist and Buddhist religions, as seen in the relative placement of inscriptions.

Rta-mchog ngang-pa do, Gnam-mtsho

Rock art and inscriptions are found in most of the approximately one dozen caves around Rta-mchog ngang-pa do (Excellent Horse Goose Headland), the third longest headland at the great lake. This rugged headland is situated in the middle of the north shore of Gnam-mtsho. The inscriptions are presented beginning at the base of the east side of the headland and proceeding in a clockwise direction. On Rta-mchog ngang-pa do, see Bellezza 1997a, pp. 262, 263; December 2008 Flight of the Khyung.

Fig. 186. The headland of Rta-mchog ngang-pa do, Gnam-mtsho

Fig. 187. Non-Buddhist mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: white pigment writing
Estimated date: Vestigial period
Proximate rock art: two counterclockwise swastikas
Location: Rta-mchog ngang-pa do
Note: The inscription and rock art may have been made by the same individual. On this rock art, see June 2016 Flight of the Khyung, fig. 46.

Fig. 188. Buddhist prayer and Buddhist mantras

bla ma sa skya {pa} mkhyeno (upper)

# ba’i bla ma sangs lhun
{grub}
mkhyen

Oṁ (left)
A
hūṁ

hūṁ badzra phaṭ (middle)
hūṁ badzra phaṭ

hūṁ (lower right)
bra
dza
Script: dbu-can
Grammatical marks: some tsheg (upper)
Technique: red ochre writing
Translation: “Be aware of us Sa-skya-pa lama. Lama Sangs-{rgyas] lhun-{grub} be aware of us.”
Estimated date: Vestigial period
Location: Rta-mchog ngang-pa do
Note: The upper line of the upper inscription is mostly cut in the photograph. The last two syllables of the lower right mantra (corrupted form of huṁ badzra) can be seen in fig. 189. The two middle mantras are for affirming a Buddhist identity and for expelling negative influences (the object probably being non-Buddhists).

Fig. 189. Non-Buddhist prayer and Buddhist mantras

slob dpon stong rgyung mthu (upper)
chen sku la skyabsu mchi’o

phaṭ (middle left)

Oṁ ma ṇǐ pad me hūṁ (middle right)

Ōṁ A hūṁ badzra {gu ru} pad{me} siddhi hūṁ (lower)

Ā (lower right)
Script: dbu-med (upper), dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial shad, intersyllabic tsheg (upper)
Technique: red ochre writing
Translation: “I take refuge in the personage Master Stong-rgyung mthu-chen.”
Estimated date: Vestigial period
Proximate rock art: crescent moon (lower left)
Location: Rta-mchog ngang-pa do
Note: The phaṭ of the middle left inscription may be preceded by other syllables, but photographic evidence for assessment is lacking. Located directly below fig. 187.

Fig. 190. Buddhist mantras

Ōṁ ma ṇi pad me hūṁ (upper)

huṁ badzra {phaṭ} bla ma ma # # (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: crescent moon and clockwise swastika
Location: Rta-mchog ngang-pa do
Note: There are three more lines of fragmentary mantras visible in the lower half of the photograph Located directly below fig. 188.

Fig. 191. Incomplete mantra

A A d r
k
Script: dbu-can
Grammatical marks: none
Technique: yellow ochre writing
Estimated date: Vestigial period
Location: Rta-mchog ngang-pa do
Note: This appears to be an incomplete Sale-’od mantra. The left side of the inscription is partly cut in the photograph.

Fig. 192. Mantric syllables

A A (upper)

A (upper middle)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Rta-mchog ngang-pa do
Note: See fig. 193 for lower inscriptions.

Fig. 193. Inscription and incomplete Sale-’od mantra

{mchog ’gyur} # # # # # # (upper)

A (middle right)

A A dkar sa le ’od A (lower)
Script: rudimentary dbu-med (upper), dbu-med
Grammatical marks: unclear
Technique: red ochre writing
Estimated date: Vestigial period
Location: Rta-mchog ngang-pa do
Note: The reading of the upper inscription is complicated by the unusual combination of vowel signs. There is a highly fragmentary fourth inscription immediately below the Sale-’od mantra that is hardly visible in the photograph. This fragmentary inscription is also seen in lower part of fig. 192.

Fig. 194. Fragmentary inscription

Script: dbu-can
Grammatical marks: brda-rnying yig-mdo mdun-ma (each line)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: unidentified subject below inscription
Location: Rta-mchog ngang-pa do
Note: This inscription of three lines defies a coherent reading even with the higher resolution photographs available. Inscription partially visible on upper right side of fig. 192.

Fig. 195. Non-Buddhist and Buddhist mantras and accompanying dedication

A A dkar sa le ’od A yang Oṁ ’du (upper)

Oϻ ma ni pad me huṁ (middle)

kun gyis syang bar zhu (lower)
Script: dbu-can, dbu-med (lower)
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial nyis-shad+nyis-shad (upper); intersyllabic tsheg (middle and lower)
Technique: white pigment writing (upper right), red ochre writing
Translation: “All offered as the speech.”
Estimated date: Vestigial period
Proximate rock art: white pigment curvilinear design above and to left of upper inscription
Location: Rta-mchog ngang-pa do
Note: The white pigment subject appears to be an elaborate kind of yig-mgo bdun-ma. The sophisticated execution of the design matches a non-Buddhist mchod-rten depicted above (not pictured). It appears that the lower line was added to join the significance of the two mantras belonging to different religious traditions. Syang is an old form of gsung used particularly by non-Buddhists.

Fig. 196. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or later
Proximate rock art: geometric design and anthropomorphic figures
Location: Rta-mchog ngang-pa do
Note: The letter A was framed by rock art at a much later date. Situated below fig. 193.

Fig. 197. Non-Buddhist mantra and mantric syllable

A A dkar Oṁ # (upper)

A (lower)
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma+sgab-ma+initial nyis-shad (upper); brda-rnying yig-mgo mdun-ma+initial nyis-shad (lower)
Technique: red ochre writing
Estimated date: Vestigial period
Location: Rta-mchog ngang-pa do
Note: The first part of the head mark of the lower inscription is partly cut in the photograph.

Fig. 198. Inscription

se ghing bdar pa ma {da}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: straight lines and wavy line
Location: Rta-mchog ngang-pa do
Note: Sengge (lion) may possibly be intended as the first two syllables. The lines appear to accompany the inscription.

Fig. 199. Mantric syllable and Non-Buddhist seed syllables

Ā (left)

Ā Ōṁ huṁ (right)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: Rta-mchog ngang-pa do
Note: The photographic record is inadequate to properly assess the inscription on the left and any others in the proximity.

Fig. 200.

Ōṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: A little west of Rta-mchog ngang-pa do

Fig. 201. Fragmentary Ma-ṇi mantra

Oṁ ma ṇi pad
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: clockwise swastika, crossed thunderbolt (rdo-rje rgya-gram), and lines
Location: A little west of Rta-mchog ngang-pa do

Fig. 202. Rigs-gsum mgon-po mantras

Oṁ wa gi shwa ri hūṁ
Oṁ ma ṇi pad me hūṁ
Oṁ badza (= badzra) pa {ni} huṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: A little west of Rta-mchog ngang-pa do
Note: Note the unusual letter wa in the upper line.

Khyi-rgan gag-pa do, Gnam-mtsho

Khyi-rgan gag-pa do (Old Barking Dog Headland) is located west of Rta-mchog ngang-pa do. This small headland is most noteworthy for its red ochre pictographs.

Fig. 203. Inscription

bsan ra
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Early Historic period
Location: Khyi-rgan gag-pa do

Fig. 204. Non-Buddhist mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: Khyi-rgan gag-pa do
Note: The letter and swastika appear to have been made by the same hand.

Lug-do, Gnam-mtsho

Lug-do (Sheep Headland) is a smaller limestone headland on the western half of the north side of Gnam-mtsho. On this site, see Bellezza 1997a, pp. 260–262.

Fig. 205. The main rock face at Lug-do, Gnam-mtsho

Fig. 206. Fragmentary Sa-le-’od mantra

A A dkar {sa}
’od A yang Oṁ
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: Lug do

Fig. 207. Fragmentary inscription

lug do brag khung
rdza ’di # # pa
Script: rudimentary dbu-med
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma+initial nyis-shad, intersyllabic tsheg
Technique: red ochre writing
Translation: “Cave of Sheep Headland rock formation.” This rock…”
Estimated date: post-Imperial period
Location: Lug do
Note: This inscription establishes the antiquity of the name of the headland. The style of paleography indicates that it was written by an adherent of a non-Buddhist tradition at Gnam-mtsho. There may have been a third line (and possibly others) to the inscription, which were effaced by the superimposition of carved lines of an extensive Buddhist prayer added on a ground of red ochre. Situated on middle left side of fig. 205.

Fig. 208. Ma-ṇi mantras

Oṁ ma ṇi padme hūṁ (left)

Oṁ ma ni (right)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: lotus petal-like design (right half), counterclockwise swastika (lower left corner)
Location: Lug do
Note: Situated in middle of fig. 209.

Fig. 209. Non-Buddhist mantric syllables

# {Oṁ chen}
Script: dbu-can
Grammatical marks: possible yig-mgo mdun-ma (upper row of marks)
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: Above the inscription appear to be symbols resembling the letter ba.
Location: Lug do

Fig. 210. Inscription

lug do yi brag {’di yi}
{yin no} kha {dron} gi #
Script: rudimentary dbu-med
Grammatical marks: intersyllabic tsheg, {nyis-shad} (upper line)
Translation: “The rock formation of Sheep Headland…”
Technique: red ochre writing
Estimated date: post-Imperial period
Location: Lug do
Note: The last syllable in the second line is cut in the photograph. Like the inscription in fig. 207, this one corroborates the antiquity of the local place name.

Fig. 211. Inscription

kye lug do’i
Script: dbu-med
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma+initial nyis-shad, intersyllabic tsheg
Technique: black pigment writing
Translation: “Listen! Of Sheep Headland.”
Estimated date: Vestigial period
Location: Lug do
Note: Superimposed on inscription in fig. 210 (middle right side).

Fig. 212. Fragmentary inscription

# # # # kyi
Script: rudimentary dbu-med
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: Lug do

Fig. 213. Inscription with title and personal name

{dpon} jo ldan kyi
Script: rudimentary dbu-med
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Translation: “Of {leader} Jo-ldan.”
Estimated date: Vestigial period
Location: Lug do

Fig. 214. Mantric syllables

Oṁ {ma ni} (top left)

Oḿ # (bottom left)

Oṁ (upper right)

A kha (lower right)
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Lug do
Note: The final letter kha in the lower right inscription is partially cut in the photograph. Above the upper right syllable Oṁ is what may be an inscription with highly unusual letter formations.

Ra-ma do, Gnam-mtsho

Ra-ma do (Female Goat Headland) is a towering headland with a large cave by the lakeshore. On this site, see Bellezza 1997a, pp. 260–262.

Fig. 215. Fragmentary Sale-’od mantra

A dkar sa le
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: unidentified red ochre subject above inscription (partially visible in photograph), unidentified subject
Location: Ra-ma do

Fig. 216. Letters

#
# # #
Script: dbu-can
Grammatical marks: nyis-shad
Technique: red ochre writing
Estimated date: Vestigial period
Location: Ra-ma do
Note: Situated below fig. 204.

Fig. 217. Inscription

# rin chen # # yin no
brag yang (= kyang) {bzang por} gda’
Script: rudimentary dbu-med
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial nyis-shad (each line), intersyllabic tsheg, nyis-shad+nyis-shad (both lines)
Technique: red ochre writing
Translation: “I am… Rin-chen… Here too is the {good} rock formation.”
Estimated date: Vestigial period
Proximate rock art: three horse riders, concentric circle, minor red ochre applications, and much more recent black pigment applications
Location: Ra-ma do
Note: Due to the peculiar formation of the letters and superimposition of rock art, this inscription is difficult to decipher.

Fig. 218. Mantric syllables and letters

{Oṁ}
{nyer}
#
{Oṁ}
#
Script: rudimentary dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: red ochre pigment applications (possible letters among them)
Location: Ra-ma do

Note: A red ochre six-syllable Ma-ṇi mantra is also found at Ra-ma do.

Stong-shong phug, Gnam-mtsho

Stong-shong phug is a sizable cave with a roundish floor plan and a natural hole in the roof. Once a non-Buddhist religious sanctuary, the cave was more recently used to shelter sheep and goats. Recently, the Chinese government has requisitioned the site for the burgeoning tourist industry and built walkways and paved over the floor of the main cave. There are also several smaller caves at the site. On this site, see Bellezza 1997a, pp. 248–250.

Fig. 219. Stong-shong phug. Gnam-mtsho

Note: The inscriptions and rock art are concentrated in the niche situated in the middle of the photograph.

Fig. 220. Main concentration of inscriptions and rock art, Stong-shong phug

Note: See below for details below. The inscription written on the left side of the photograph does not appear to be of great age. It reads: {slob stong} # chen # # # # # #.

Fig. 221. Non-Buddhist mantric syllable and prayer, and syllable

A (upper)

kun ’tu (C.T. = tu) bzang mkhyen (middle)

rgyo (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (middle)
Technique: red ochre writing
Translation: Middle: “Be aware of us Kun-’tu bzang-[po].” Lower: “Sex.”
Estimated date: Early Historic period (upper, middle), Vestigial period (lower)
Proximate rock art: anthropomorphic figure with peaked headgear (partially visible in photograph)
Location: Stong-shong phug
Note: The lower inscription appears to have been written by the same individual who painted the anthropomorphic figure. The unusual ligature on the right side of the syllable is unidentified. It is not clear if the inscription is a disparaging reference to the anthropomorph or other inscriptions/rock art at the site, or alternatively, a reference to a mystic practice. Situated on middle left side of fig. 220.

Fig. 222. Mantric syllables

A A
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: possibly Early Historic period
Location: Stong-shong phug
Note: Other letters were possibly obliterated by the application of red ochre above and to the right of the legible inscription. Situated on lower middle left side of fig. 220.

Fig. 223. Inscription

{sod} # #
da
Script: dbu-can
Grammatical marks: possibly intersyllabic tsheg
Technique: red ochre writing
Translation: “{Kill!}…”
Estimated date: Vestigial period
Proximate rock art: stepped shrine above and anthropomorphic figure to right (only tiny portions of which are visible in photograph)
Location: Stong-shong phug

Fig. 224. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: conjoined sun and moon (middle), swastika (lower right), white pigment swastika (middle right)
Location: Stong-shong phug

Fig. 225. Mantric syllables and extensive but fragmentary inscriptions

A (upper)
{hum}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: stepped shrine to left (partly visible) and minor pigment applications
Location: Stong-shong phug
Note: Six or seven irregular lines are discernable in the photograph. Despite having detailed images, only some individual letters are recognizable but not coherent sentences. It is not clear if this writing constitutes one inscription with various clauses (don-tshan), or multiple inscriptions. For example, note jo {bu} (“{son} of the lord”; top right), sa thebs (to plant/establish in the ground; directly below), sde (section, division; middle of photograph), nye ma nye pha ({maternal and paternal relatives; below). From what can be read, this inscription(s) it is of considerable historical importance. Situated on right half of fig. 220. See fig. 226 for more details.

Fig. 226. View of upper right portion of inscription in fig. 225

rgyags na rgyag
jo mo {rgol}
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial nyis-shad
Technique: red ochre writing
Estimated date: Vestigial period
Location: Stong-shong phug
Note: There is not enough of the inscription visible to provide a reliable a translation. The upper line seems to convey attainment of satisfaction, as in after a full meal. The lower line seems to refer to a mistress, female leader or goddess.

Fig. 227. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: Stong-shong phug
Note: The letter and swastika may have been made by the same hand.

Fig. 228. Possible invective

{rgyon} #
{dang khĭ} sa {rgya}
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial nyis-shad, intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: Stong-shong phug

Fig. 229. Fragmentary inscription

# ma #
# ngo mo
# sya
# ni
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Stong-shong phug

Fig. 230. Mantric syllable

Ā
Script: dbu-can
Grammatical marks: none
Technique: white mineral pigment writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: Stong-shong phug
Note: The syllable and swastika appear to have been made as an integral composition. The letter ma written in red ochre is situated lower and to the right (not pictured).

Fig. 231. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Stong-shong phug

Fig. 232. Inscription

pha spun kyǐ sto A ma dgro snyags
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Translation: “Lineage mother of the parents and siblings, the Dgro [and] Snyag.”
Estimated date: possibly post-Imperial period
Proximate rock art: tiered shrine
Location: Stong-shong phug
Note: The word sto appears to designate a tribe or lineage in either a proper or generic sense. Dgro and Snyag can be identified as clan names. The ’Gro/Sgro and Snyags/Gnyags (sic) were prominent clans of considerable political importance in the Imperial period, according to Tibetan historical documents and catalogs of principalities (Dotson 2012). The tiered shrine and inscription appear to have been made as an integral composition. This seems to augment the significance of the inscription as registering a feast of special proportions. Located in small cave near the main Stong-shong phug.

Se-mo do, Gnam-mtsho

Se-mo do is the largest of the three main islands in Gnam-mtsho. It is situated opposite of Rigs-lnga do about 4 km offshore. There are a series of caves and the foundations of archaic residential structures on the south side of the island. On this site, see Bellezza 2014a, pp. 450–460; Bellezza 1997a, pp. 159–165.

Fig. 233. The island of Se-mo do, Gnam-mtsho, with Gnyan-chen thang-lha (upper left)

Fig. 234. Buddhist seed syllables and Sale-’od mantras

Oṁ (right)
A
hūṁ

{A A dkar sale ’od A yang Ōṁ ’du} (left)
A A dkar sale ’od A yang Ōϻ ’du
Script: dbu-can
Grammatical marks: gter-tsheg (left)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: two flaming jewels, counterclockwise swastika (between first two syllables of inscription on the right)
Location: Klu-khang, Se-mo do
Note: The mass of letters at the bottom of the photograph are illegible because photographic coverage is inadequate. For the same reason, it is not clear if there is a third Sa-le-’od mantra or some other inscription on the left.

Fig. 235. Buddhist seed syllables

Ōṁ
Ā
hūṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Klu-khang, Se-mo do
Note: Situated to right of fig. 234.

Fig. 236. Fragmentary inscription with Snying-po rnams-gsum mantras

# # # tshang mkhyeno
{A rmad} du tri su nag po zhi zhi mal mal
# A dkar sa le ’od A {yang Ōϻ} ’du
le ’du {Ōṁ} ma tri mu ye sa le ’du
Script: dbu-can
Grammatical marks: none
Estimated date: Vestigial period
Proximate rock art: conjoined sun and moon superimposed on third line of inscription, counterclockwise swastika superimposed on fourth line of inscription
Technique: red ochre writing
Location: Klu-khang, Se-mo do
Note: All four lines may have been written by the same individual. The first two visible syllables of the fourth line appear to be the final two syllables of another Ma-tri mantra.

Rigs-lnga do, Gnam-mtsho

Rigs-lnga do (Crown of Five Diadems Headland) is so named for the pinnacles on the formation. The inscriptions are found in several shallow caves. According to a local oral tradition, a minor rocky dispersion at the headland is the remains of a Tibetan fishing camp from the 11th and 12th centuries CE.

Fig. 237. Rigs-lnga do, Gnam-mtsho. The pinnacles on the summit of the limestone headland are envisioned as diadems of a tantric crown

Fig. 238. Non-Buddhist seed syllables

Oϻ ma hūṁ
Script: dbu-can
Grammatical marks:
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: conjoined sun and moon above inscription, counterclockwise swastika and two unidentified subjects below inscription
Location: Rig-lnga do

Fig. 239. Mantric syllable, Buddhist prayer and Ma-ṇi mantra

A (upper)

karma pa mkhyen (middle)

Oṁ ma ṇi padme hūṁ (lower)
Script: dbu-can
Grammatical marks: yig-mgo mdun-ma+sgab-ma+initial nyis-shad, nyis-shad (lower)
Technique: yellow ochre (upper), red ochre (middle) and black mineral pigment (lower) writing
Estimated date: Vestigial period (upper, middle), post-Vestigial period (lower)
Proximate rock art: flaming jewels in red and yellow ochre
Location: Rig-lnga do
Note: The upper inscription and flaming jewels symbol appear to have been made by the same individual. The way in which the middle inscription was superimposed on the flaming jewels suggests that this symbol was painted by non-Buddhists. A lower-most inscription visible in the photograph, a personal name, was made at a much later date. It reads: blo gros nor ldan ma. There is also a fragmentary inscription (illegible) visible on the top right side of the photograph.

Fig. 240. Mantric syllable

Ōṁ
Script: dbu-can
Grammatical marks: none
Technique: yellow ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Rig-lnga do

Fig. 241. Buddist seed syllables, Ma-ṇi mantra and Rigs-gsum mgon-po mantras

Ōḿ ma (upper)
Ā
hri
#

Ōḿ ma ṇi padme hūṁ hri (middle)

Ōḿ ma ṇǐ pad me huṁ hri (lower)
Ōḿ wa gi sha ri muṁ
Oḿ’ badzra pa ni huṁ
do pa bla ma mkhyen
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma+initial nyis-shad, intersyllabic tsheg, nyis-shad (middle); yig-mgo mdun-ma+sgab-ma+initial nyis-shad, nyi-shad (lower, third and fourth lines)
Technique: yellow ochre (upper) and red ochre writing
Translation: (last line of lower inscription): “Lama of the headland be aware of us.”
Estimated date: Vestigial period
Proximate rock art: red ochre pigment applications
Location: Rig-lnga do

Fig. 242. Mantric syllables

Ā

Oṁ

{A}
Script: dbu-can
Grammatical marks: none
Technique: yellow ochre writing
Estimated date: Vestigial period
Location: Rig-lnga do
Note: The lower letter is partially cut in the photograph.

Fig. 243. Ma-ṇi mantra, Buddhist refuge prayer and mantric syllables

Ōϻ ma ṇi padme hūṁ hri Ōϻ
(bla ma kyabsu (= skyabsu) che’o
sang (= sangs) rgyas la kyabsu (= skyabsu) che’o
Ōϻ
Ā
hūṁ
Script: dbu-can
Grammatical marks: yig-mgo mdun-ma+sgab-ma
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: mantras boxed in and with what appears to be an enthroned lama on the left
Location: Rig-lnga do

Fig. 244. Ma-ṇi mantra

Oṁ ma ṇi pa dme hūṁ
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: row of lotus petals below inscription
Location: Rig-lnga do
Note: The two syllables pad and me were incorrectly split.

Fig. 245. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: surrounding red ochre pigment applications
Location: Rig-lnga do

Lce do, Gnam-mtsho

Lce-do is a tiny cave on the west side of Gnam-mtsho pullulating with early pictographic art. On this site, see Bellezza 1997a, pp. 238–246. The rock art of this cave will form the subject of another article for Flight of Khyung.

Fig. 246. The headland of Lce do, Gnam-mtsho. The small cave with the bulk of inscriptions and pictographs is situated at the base of the rock formation in the middle of the photograph

Fig. 247. First four letters of the Tibetan alphabet

ka ga kha nga
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: surrounding red ochre pigment applications
Location: Lce do
Note: The order of the middle two letters has been reversed.

Fig. 248. Syllable

rno
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: surrounding red ochre pigment applications
Location: Lce do

Fig. 249. Two syllables

Oṁ na
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: surrounding red ochre pigment applications
Location: Lce do
Note: The sna-ldan mark is missing. Also see Suolang Wangdui 1994, p. 139 (fig. 170).

Fig. 250. Mantric syllables

# # ma {ni ni}
{hūṁ}
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Lce do
Note: Probably an incomplete Ma-ṇi mantra.

Fig. 251. Mantric syllables

# (left)
Oḿ Ā huṁ (right)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Lce do
Note: Situated immediately to right of the inscription in fig. 250.

Fig. 252. Fragmentary mantras and prayer

{Oṁ} # ma dza ni dzra
# # # # # {hūṁ}
# # mkhyen no
# #
# # {hā huṁ}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or later
Location: Lce do
Note: Photographic documentation is not adequate to fully assess this inscription.

Fig. 253. Historical record

# # # # # na ston ’brug rgya gyi… (upper)

# # #

# sgre gseng res ’go byas (middle)
pas mi ’go lnga gser srang
rgya (= brgya) tham ba (= pa) khyer nas bod du song

yo ser bas ’go byas pas {mi}’go lnga {bcu} (right)
gser srang stong khyer nas # #…
song
# # #
# #
# # rgyal
# # #
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma, intersyllabic tsheg (middle); nyis-shad (middle, ln. 1)
Technique: red ochre writing
Translation: Middle: “Sgre-gseng-re led five headmen and took all one hundred srang of gold and went to Tibet.” Right: “Yo-ser-ba led fifty headmen and took a thousand srang of gold and went to…”
Estimated date: post-imperial period
Proximate Rock art: bird with long neck and folded wings to right of middle inscription
Location: Lce do
Note: Srang is a unit of currency in gold, the value of which is unclear in this historical context (on srang in the 18th to 20th centuries, see Bertsch 2002). At Gnam-mtsho and in other regions of Upper Tibet, the word Bod commonly denotes Central Tibet. The last four lines of the inscription on the right appear to be proper names of the individuals involved in the travel and transaction mentioned. There are also two small-sized letters (ha A) bracketed by curvilinear marks between the second and third lines of the middle inscription. This group of interrelated inscriptions documents the transfer of wealth from prominent figures in the Gnam-mtsho region to unnamed beneficiaries in Central Tibet. The reason for the travel to Central Tibet and payment is unclear. One possibility is that the inscriptions record the invitation of a religious teacher(s) to the Byang-thang. The amounts of gold noted in the inscriptions vary ten-fold, as if there was a competition or escalation involved in the payment being offered, possibly as an expression of piety or commitment on the part of the paying parties. Unfortunately, the available photographic evidence is inadequate to fully assess the inscriptions. They are found away from the main cave at Lce-do.

Fig. 254. Inscription

Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma, intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period or possibly earlier
Proximate rock art: curly lines and other pigment applications
Location: Lce do
Note: Due to the unusual formation of some letters and non-standard orthography, no reading of this inscription is attempted. It begins with the formal word of address kye, immediately followed by shǐ-po denoting a dead person. The first word in the fourth complete line is probably rgyag-pa (to cast aside), rather than rkyag-pa (excrement), following the chirography of the last word of the first line (tshag, not tshak). This inscription may refer to the discharge of a funerary ritual. Situated below fig. 253.

Fig. 255. Fragmentary Buddhist prayer

…srung nas gsol ba…(ln. 3)
…khyung # mkhyen
{dbang} grags pa mkhyeno

hūṁ (left)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (right)
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: surrounding red ochre pigment applications
Location: Unclassified, Gnam-mtsho

Other sites on the Byang-thang

Rock inscriptions both carved and painted are located at various sites across the Byang-thang. They are found in Smad-pa, Gzhung-med, Dang-ra g.yu-mtsho, Bkra-ris gnam-mtsho, Sger-rtse, and one in the upper Yar-lung gtsang-po valley. The bulk of painted specimens are situated in the Eastern Byang-thang. Like Gnam-mtsho, the ancient epigraphs of other locations on the Eastern Byang-thang often bear physical signs of the encounter between Buddhist and non-Buddhist religions, including erasure, juxtaposition and superimposition. The epigraphs of the Central Byang-thang and Western Byang-thang are mostly carved and contain a higher percentage of Buddhist examples. Except at Lha-khang dmar-chags, Sa-le-’od, Ma-tri and Dpa’-bu ‘bru-lnga mantras are rare in the Central Byang-thang and Western Byang-thang.

Lha-ris sgrub-phug, Smad-pa

Lha-ris sgrub-phug is an ancient cave shelter situated in a side valley of the Smad-pa basin, in what is now Shan-rtsa county. The cave is suspended approximately 35 m above the valley floor and faces east.

Fig. 256. Mantric syllable

A
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: probably Vestigial period
Location: Lha-ris sgrub-phug, Smad-pa
Note: For the group of older pictographs below the inscription, see Bellezza 2017a, p. 12 (fig. 5); 2008, p. 165 (fig. 274); Suolang Wangdui 1994, p. 134 (fig. 161).

Slob-dpon phug, Gzhung-smad

Slob-dpon phug is a small cave perched high above Ze-ling bdud-mtsho/Gser-gling mtsho in Shan-rtsa county. According to local legend, Gu-ru rin-po-che spent time here in meditation.

Fig. 257. Mantric syllable

{Ōϻ}
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Slob-dpon phug, Gzhung-smad
Note: There are a few earlier pictographs in this cave. For two tiered shrines among them, see Bellezza 2008, p. 184 (fig. 334); 2017-2018, figs. 6a, 6b.

Fig. 258. Buddhist mantric syllable

hūṁ
Script: dbu-can
Grammatical marks: gter-tsheg
Technique: yellow ochre writing
Estimated date: Vestigial period
Proximate rock art: five or more red ochre counterclockwise swastikas
Location: Slob-dpon phug, Gzhung-smad
Note: The inscription was superimposed on a red ochre counterclockwise swastika. Unfortunately, the color calibration scale in the photograph is obstructing part of the inscription.

Chos-lung O-rgyan bsam-gtan-gling, Gzims-phug

Chos-lung O-rgyan bsam-gtan-gling is a cave nestled in an eponymous gorge high above the Rnying-ma monastery of Gzims-phug, in Shan-rtsa county. According to local legend, Gu-ru rin-po-che mediated in this cave.

Fig. 259. Non-Buddhist mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika in box, minor red ochre applications
Location: Chos-lung O-rgyan bsam-gtan-gling, Gzims-phug
Note: The counterclockwise swastika in a matching box establishes the non-Buddhist identity of the inscription. Also see Bellezza 2014a, p. 446 (fig. THE7.4); figs. 374, 375 infra. Also in the same cave are two examples of the non-Buddhist mantra A Om huṃ and another specimen of the syllable A. See ibid., p. 450.

Rta-ra dmar-lding, Gzims-phug

Rta-ra dmar-lding is the name of a defile near the Buddhist monastery of Gzims-phug, in Shan-rtsa county. There are ancient cliff dwellings and what appear to be the ruins of non-Buddhist religious structures at the site. On this loction, see Bellezza 2014a, pp. 444–450.

Fig. 260. Ma-ṇi mantra

Ōɱ ma ṇǐ pad me hūḿ
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial nyis-shad, intersyllabic tsheg, nyis-shad
Technique: white pigment (oxides of calcium?) writing on a red ochre ground
Estimated date: Vestigial period
Proximate rock art: partial white pigment counterclockwise swastika (below the yig-mgo), one or two purposely obscured white pigment counterclockwise swastikas (bottom left), red ochre tiered shrine (bottom middle right)
Location: Rta-ra dmar-lding, Gzims-phug
Note: This inscription documents the Buddhist occupation or early tenure at the regionally important site of Gzims-phug. See Bellezza 2014a, pp. 446, 447; also see ibid., p. 445 (fig. THE7.1); 2008, pp. 189, 190 (fig. 356).

Fig. 261. Inscription

sgon (= dgon) pa
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma, intersyllabic tsheg
Translation: “Monastery.”
Technique: red ochre writing
Estimated date: Vestigial period
Location: Rta-ra dmar-lding, Gzims-phug

Fig. 262. Fragmentary Sale-’od mantra

A # {dkar} sa le ’od A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: two counterclockwise swastikas below inscription
Location: Rta-ra dmar-lding, Gzims-phug
Note: The swastikas may have been made by same individual who wrote the mantra. These figures and the inscription were scrawled on a prepared stucco surface fixed to the rock formation, which formed the rear wall of a now ruined residential structure.

Fig. 263. Rigs-gsum mgon-po mantras

Ōϻ wa gi shwa ri muṁ Ōϻ ma ṇi padme hūṁ Ōϻ badzra pā ni hūṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing and carving
Estimated date: Vestigial period
Location: Rta-ra dmar-lding, Gzims-phug
Proximate rock art: surrounding the inscription is a thick band of white pigment topped by two conjoined sun and moon symbols
Note: Both carved in relief and painted, this inscription was made using an unusual technique. The addition of an a-chung to the final pa in the inscription is an old-fashioned orthographic feature.

Chu-ro, Gzhung-smad

Chu-ro (Water Remains) is situated in a large overhang in the middle of a towering blue and red limestone formation, in what is now Shan-rtsa county. In addition to inscriptions and a few pictographs, this ancient religious sanctuary preserves three early stone and mortar mchod-rten. Epigraphic, rock art and structural evidence at the site documents both non-Buddhist and Buddhist phases at Chu-ro. On this cave sanctuary and its structural remains and rock art, see October 2013 Flight of the Khyung.

Fig. 264. Fragmentary inscription on mchod-rten

sa kha la {gzug}
# # {bar} ghar {ci}
Script: dbu-can
Grammatical marks: none
Translation: “Established on the earth…”.
Technique: red ochre writing
Estimated date: Vestigial period
Location: Chu-ro, Gzhung-smad
Note: The photographic documentation of this inscription is inadequate. Like other inscriptions on the mchod-rten of Chu-ro, this example appeared ad hoc, strongly suggesting that it was made subsequent to the establishment of the religious monument. Like several other cave sanctuaries in Shan-rtsa, Chu-ro was established by non-Buddhists.

Fig. 265. Ma-ṇi mantra

Oṁ ma ṇǐ pad me hūṁ hri
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma+initial nyis-shad, intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period
Location: Chu-ro, Gzhung-smad
Note: There is also a fragmentary inscription below the Ma-ṇi mantra that is hardly readable in the photograph. There are no archaeological or epigraphic signs of an early Buddhist occupation of the site, strongly suggesting that the makers of the mantra and the mchod-rten belonged to different religious traditions.

Fig. 266. Fragmentary Ma-ṇi mantra and another Buddhist mantra on mchod-rten

Oḿ {ma ṇǐ pad} me hūṁ (left)

Ōm hi swaha (right)
Script: dbu-can
Grammatical marks: none visible
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: conjoined sun and moon matching in terms of pigment quality and wear the inscription on the right
Location: Chu-ro, Gzhung-smad
Note: The mantra on the right is cut in the photograph.

Fig. 267. Illegible inscription of two lines on mchod-rten

Script: dbu-can
Grammatical marks: none visible
Technique: red ochre writing
Estimated date: Vestigial period
Location: Chu-ro, Gzhung-smad

Fig. 268. Ma-ṇi and other mantra

Oṁ ma ṇǐ pad me hūṁ hri (upper)

sa ru lu {ru lu} # hūṁ (lower)
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial nyis-shad, intersyllabic tsheg (upper); intersyllabic tsheg (lower)
Technique: red ochre writing
Estimated date: Vestigial period
Location: Chu-ro, Gzhung-smad
Note: The upper inscription is not fully visible in this photograph but there is further photographic documentation. The vowel sign (’greng-bu) of the syllable me in the Ma-ṇi inscription was written in the opposite direction (top of stroke to the right).

Fig. 269. Mantric syllable on mchod-rten

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Chu-ro, Gzhung-smad

Dar-lung phug-pa, Gzhung-smad

Dar-lung phug-pa is an ancient residential complex consisting of two small rock shelters perched above a rocky ravine, in Shan-rtsa county. On this site, see November 2013 Flight of the Khyung.

Fig. 270. Mantric syllable

Ōṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Dar-lung phug-pa, Gzhung-smad

Fig. 271. Non-Buddhist mantric syllables

A (upper)

A (middle)

A (lower)
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Early Historic period
Location: Dar-lung phug-pa, Gzhung-smad
Note: Also see November 2013 Flight of the Khyung, figs. 4–6.

Fig. 272. Non-Buddhist mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: beige pigment (clay?) writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: Dar-lung phug-pa, Gzhung-smad
Note: The syllable and swastika appear to have been made by the same individual. Also see November 2013 Flight of the Khyung, fig. 12.

Bshag-bsangs, Gro-ba

Bshag-bsangs (also called Shar-tshang) consists of a series of outcrops stretching for 320 m along a valley floor, and is situated approximately 15 km northwest of the township headquarters, Gro-ba, Nyi-ma county. This site is located within the district traditionally known as Nag-tshang. This narrow north-south oriented valley links the basin of Ri-khrod with the lake basins of Phug-pa and Dmar-zhal. On rock art at this site, see Suolang Wangdui 1994, pp. 111–117.

Fig. 273. Mantric syllables

A Oṁ (upper)

Oṁ ma (upper middle)

Oṁ ma (lower middle)

A (lower left)

A (lower right)
Script: dbu-can
Grammatical marks: sna-ldan diacritic absent from the lower two Oṁ syllables
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: wild ungulate
Location: Bshag-bsangs, Gro-ba
Note: These appear to be non-Buddhist mantric syllables. The inscriptions and animal figure may have been carved in the same period.

Fig. 274. Ma-ṇi mantra and other mantric syllables

Ōϻ ma ṇi pad me hūḿ (upper)

A (middle left)

ka (middle right)

Oṁ (lower left)
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma, intersyllabic tsheg (upper)
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: two deer, another wild ungulate and other carvings
Location: Bshag-bsangs, Gro-ba
Note: The vowel sign of the Ōϻ in the upper inscription was superimposed on a deer petroglyph. Some retouching of the carvings has occurred.

Fig. 275. Various mantric syllables

Ōϻ (left)
A
hūṁ

A (lower middle)

A (right)
Oṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: horse rider (lower right) of same general period
Location: Bshag-bsangs, Gro-ba
Note: The inscription on the left is Buddhist while the one on the right appears to have been made by non-Buddhists. The Ōm syllable visible in the middle of the photograph was carved much more recently.

Fig. 276. Buddhist seed syllables

Oṁ
A
hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Location: Bshag-bsangs, Gro-ba
Note: The lower syllable is partly cut in the photograph.

Fig. 277. Buddhist mantric syllables

Oṁ A
hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: surrounding older carvings, including raptor with spread wings (upper right side) and linear form in between the syllables
Location: Bshag-bsangs, Gro-ba
Note: The lower syllable was not completed.

Fig. 278. Buddhist seed syllables

Ōϻ Ā hūṁ
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial shad, intersyllabic tsheg
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: more recent horse rider
Location: Bshag-bsangs, Gro-ba
Note: The lighter re-patination of the Ōϻ syllable appears to be the result of retouching. The huṁ syllable has an elaborate sna-ldan mark.

Fig. 279. Mantric syllables

A

A
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Location: Bshag-bsangs, Gro-ba

Fig. 280. Inscription

# {sra} # mo…
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Location: Bshag-bsangs, Gro-ba
Note: The right side of the inscription is cut in the photograph.

Fig. 281. Buddhist prayer

karma pa mkhyeno
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika
Location: Bshag-bsangs, Gro-ba
Note: The inscription and swastika may have been made by the same hand. Even today Upper Tibetan pastoralists are sometimes indiscriminate in their orientation of swastikas.

Fig. 282. Probable non-Buddhist mantric syllables

A (upper)

A (lower)
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Bshag-bsangs, Gro-ba

Fig. 283. Mantric syllable

Oṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Location: Bshag-bsangs, Gro-ba
Note: For another solitary Oṁ superimposed on the carving of a camel at the same site, see Suolang Wangdui 1994, p. 116 (fig. 121).

Fig. 284. Buddhist inscription

karma
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: line and three horse riders
Location: Bshag-bsangs, Gro-ba
Note: The rock art appears to belong to a later period. Also see Suolang Wangdui 1994, p. 112 (fig. 111).

Fig. 285. Inscription

sta
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Bshag-bsangs, Gro-ba
Note: Other readings are possible (e.g., sa ta, sa.ha, etc.).

Fig. 286. Vajra Guru mantra

Oṁ A hūṁ ba dzra gu ru padma siddhi hūṁ (upper)

hūṁ (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: carved writing
Estimated date: Vestigial period or later
Proximate rock art: two wild ungulates (left), horse rider (upper right), standing anthropomorph (lower right), counterclockwise swastika, and bow and arrow (lower left)
Location: Bshag-bsangs, Gro-ba
Note: The inscription, horse rider, anthropomorph and bow and arrow were made in the same period and possibly by the same artist. Despite it orientation, the swastika belongs to the same Buddhist composition as the other figures. Situated on same rock face as fig. 284. Also see Suolang Wangdui 1994, p. 117 (fig. 122).

Fig. 287. Mantric syllables

A A hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: standing figure brandishing a sword standing above a row of seven conical objects with three-pronged tops, and horse rider
Location: Bshag-bsangs, Gro-ba
Note: This inscription and composition is situated immediately above the one in fig. 286. This also appears to be an integrated composition, forming a unified scene with the composite figures below. The various subjects (sword, bow and arrow, crown of three diadems, ostensible offering cakes (gtor-ma), etc.) strongly suggest that this is a portrayal of a Buddhist ritual performance. The letter A flanking the central figure with sword is an unusual feature that may have been inspired by a non-Buddhist epigraphic tradition. Also see Suolang Wangdui 1994, p. 117 (figs. 122, 123).

Fig. 288. Ma-tri and another mantra

Oṁ ma tri mu ye sa le ’du (upper)

A (lower left)

Oṁ ma (lower right)
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial shad, intersyllabic tsheg (upper)
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: branched linear subject above mantra
Location: Bshag-bsangs, Gro-ba
Note: The upper mantra was defaced. Note the isolated head mark on the lower left corner of the photograph.

Am-nag, Dang-ra g.yu-mtsho

Am-nag (Black Rock Formation) is the name of a non-Buddhist cave retreat on the eastern shore of Dang-ra g.yu-mtsho in the traditional district known as Nag-tshang. The one possible inscription in this location is found on the mud plaster surface of a ruined masonry façade.

Fig. 289. An apparent fragmentary inscription (upper left side)

Script: dbu-can
Grammatical marks: none visible
Technique: red ochre writing
Proximate rock art: parasol (upper middle), vase (upper right), counterclockwise swastika (middle), and scroll and lotus designs (lower)
Estimated date: Vestigial period or later
Location: Am-nag, Dang-ra g.yu-mtsho
Note: None of the ostensible inscription is readable in this image.

Do dril-bu, Bkra-ri gnam-mtsho

Do-dril-bu (Bell Island), an erstwhile island, is located on the east side of Bkra-ri gnam-mtsho. The site now falls in Mtsho-chen county, part of the district traditionally known as Phyogs-bcu. It hosts an important array of early all-stone corbelled residences. It appears that the single inscription recorded at Do-dril-bu was written to signal Buddhist control of the locale. Buddhist occupation of Do-dril-bu, however, was nominal and does not appear to have involved permanent settlement. On Do-dril-bu, see Bellezza 2001, pp. 237–243; June 2014 Flight of the Khyung.

Fig. 290. Rigs-gsum mgon-po mantras

Oṁ ma ṇi pad me hūṁ
hri
Oṁ bha ki sha ri muṁ
Oṁ badzra pa ṇǐ hūṁ
Script: dbu-can
Grammatical marks: initial nyis-shad, intersyllabic tsheg, nyis-shad
Technique: white pigment writing on red ochre ground and red ochre writing on white pigment ground
Estimated date: Vestigial period
Location: Do dril-bu, Bkra-ri gnam-mtsho
Note: Also see Bellezza 2001, p. 379 (fig. A-12).

Gu-ru rin-po-che phug, Bkra-ri gnam-mtsho

Gu-ru rin-po-che sgrub-phug is situated in a sacred bright white limestone formation known as Mu-ro ri. Located on the north lakeshore, this cave overlooks Bkra-ri gnam-mtsho.

Fig. 291. Inscription

Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: unclear
Location: Gu-ru rin-po-che phug, Mu-ro ri, Bkra-ri gnam-mtsho
Note: Available photographic documentation is inadequate to read this inscription.

Lha-khang dmar-chags, Da-rog mtsho

Fig. 292. The ruined Yungdrung Bon retreat of Lha-khang dmar-chags, Da-rog mtsho, ’Brong-pa (’Brong-pa tsho-dgu). Many of the inscriptions are located on the central structure which is covered in clay plaster (middle)

This ruined cave temple complex is nestled in a towering limestone formation on the north side of Da-rog mtsho (also spelled Da-rogs mtsho, Dwa-rog mtsho). Locally known as Lha-khang dmar-chags (Red-painted Temple), epigraphic and artistic evidence indicates that this facility functioned as a Rdzogs-chen retreat center during the Vestigial period for non-Buddhist practitioners.* Some of the painted inscriptions and artwork at the site form an integral part of Lha-khang dmar-chags, indicating that at least some of the structures at the site date to the Vestigial period. Many of the inscriptions and pictographs however were evidently added to the masonry walls of the site after its decline, as suggested by their ad hoc placement and the superimposition of images and letters. These non-integral inscriptions appear to have been added at various times. The proliferation of mantras and symbols at the site now associated with Yungdrung Bon strongly indicate that this religion and/or its localized cult predecessor occupied Lha-khang dmar-chags exclusively and were responsible for its establishment. The most famous non-Buddhist religious sites at Da-rog mtsho are Sha-ba gdong lha-khang and Mtsho gling, which are said to have been active in the Imperial period.† These places and the personalities connected to them became incorporated into the Yungdrung Bon religion.

On this site, see Bellezza 1999; June 2010 Flight of the Khyung.

On these two sites and the exploits of the celebrated Rdzogs-chen practitioner Snang-bzher lod-po (8th century CE) taken from Rdzogs pa chen po zhang zhung snyan rgyud biographical materials, see Karmay 1972, pp. 97–99, Norbu 1995, pp. 214–216; Bellezza 1999; Reynolds 2005, pp. 100–103. For a field survey of Sha-ba gdong lha-khang and Mtsho Gling and other archaeological sites at Da-rog mtsho, see Bellezza ibid.; 2014a, pp. 462–478.

Da-rog mtsho is also linked to non-Buddhist Rdzogs-chen practitioners of ca. 950–1100 CE period, successors in a Rdzogs-chen lineage issuing from the great saints Ta-pi hri-tsa and Snang-bzher lod-po. The later masters include Dpon-chen btsan-po who is recorded as meditating at Dwa-rog brag and Dmu rgyal-ba blo-gros who mediated at Dwa-rog lcag-phug.* It is likely that one of these places refers to Lha-khang dmar-chags, as this is clearly where non-Buddhist religious activities in the Vestigial period were centered.

On these sites and masters, see Bellezza 1999, p. 78 (n. 21); Reynolds 2005, pp. 130, 131, 136–139.

There are three chapels (lha-khang) at Lha-khang dmar-chags: east, central and south. These structures are built against the base of formation creating the top of a natural amphitheater. The east chapel has a beehive-shaped façade decorated with red and white stripes. The central chapel contains a small window opening above the entranceway and its masonry façade is covered in clay plaster. The bulk of inscriptions and rock art at Lha-khang dmar-chags are found on the interconnected exterior wall of the central and south chapels. The south chapel also enshrines inscriptions ornamented in a manner not seen anywhere else in Upper Tibet. Most notably, a stucco fabrication affixed to the walls of the rock formation on the east end of the site boasts the earliest known painting of the Rdzogs-chen sage Ta-pi hri-tsa as well as accompanying mantras. There are also inscriptions in several outlying places at the site.

Fig. 293. The outer wall of the south and central chapels with various non-Buddhist religious inscriptions and pictographs

On the lower right side of the wall (south chapel) there is an area with a fragmentary inscription(s) and immediately above it a large letter A (situated below an endless knot inside a circle). Below the hole in the wall in the middle of the photograph there are two inscriptions reading: Oṁ and A ma. Due to a lack of adequate photographic documentation, these inscriptions are not treated separately below. For other details of the exterior wall of the south and central chapels, see figs. 294–301.

Fig. 294. Extensive fragmentary inscription on the exterior wall of the south chapel

…khyad {pa’i} ’jig sram
…mi tso mchi cho sa la
…dang rang phen bon chud
…gi gsung pa ’di
…# # # #
…ma byas
…’gyod
…rmas
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or later
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: It is not certain that the left and right halves of the various lines belong to the same inscription. Although a few individual words can be recognized, a coherent reading is elusive. Situated near the bottom of the wall.

Fig. 295. Non-Buddhist mantric syllables and pictographs exterior wall of the central chapel

A (upper)

Ā A # (lower)
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma+initial nyis-shad (lower)
Technique: white pigment and red ochre writing
Estimated date: Vestigial period or later
Proximate rock art: red ochre swastika and conch (upper), bichrome endless knot (middle), bichrome endless knot (middle right), bichrome counterclockwise swastika (lower right),
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: Situated on the upper left side of wall. The lower inscription may be an incomplete Sa-le-’od mantra.

Fig. 296. Non-Buddhist mantric syllable on the exterior wall of the south temple

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or later
Proximate rock art: bichrome counterclockwise swastika in a square
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: Situated near the top of the middle section of the wall

Fig. 297. Fragmentary Ma-tri mantra and dedication on wall of the central temple

Ōϻ {ma tri} mu ye sa le ’du
rab sgang dbon po bdag gis gsung mchog ma tri mu ye
bkod pa’i thugs rje phan yon byin brlabs (= rlabs) thus (= mthu) stobs bsod nams gyi (= kyis)
pha ma gtsor byas khams {gsum sems can} rnams kyis sangs rgyas kyi
sa la bkod par shog
Script: dbu-can and dbu-med
Grammatical marks: brda-rnying yig-mgo mdun-ma+initial nyis-shad, nyis-shad, intersyllabic tsheg
Technique: red ochre writing
Translation: “The excellent speech of ma-tri mu-ye by the nephew of Rab-sgang (Excellence Hill). Written in order that through the great blessings of the benefits of compassion, great power and religious merit, {sentient beings of the three} worlds, especially my father and mother, may be set upon the stages of the path to enlightenment.”
Estimated date: Vestigial period or later
Proximate rock art: endless knot
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: “Khams-gsum sems-can” is written in abridged form (bsdus-yig). The endless knot and inscription seem to form an integrated unit. Like the Ma-tri mantra, the endless knot appears to have been drawn as part of the dedication. The mixing of cursive and non-cursive scripts and the ordering of the lines strongly suggests that the inscription was made at two different times, the addition (in dbu-med) to supplement and augment the meaning of the earlier part (in dbu-can) Situated on the middle of the wall to the left of the entranceway to the chapel.

Fig. 298. Letter and incomplete Sa-le-’od mantra on wall of the central chapel

cha (upper)

A A dkar sa le ’od A yang (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or later
Proximate rock art: counterclockwise swastika (partially cut, upper right corner); parasol (?), triple gems, counterclockwise swastika, endless knot conch, victory banner (middle left); vase (?) (middle right), lotus (?), conch (?), Dharma wheel, fish, conch (?), and cruciform design (lower)
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: Situated between the entranceway and hole in the wall to the right.

Fig. 299. Mantric syllables on the exterior wall of central chapel

Oṁ ma
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Proximate rock art: triple gems
Estimated date: Vestigial period or later
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: Situated below fig. 295.

Fig. 300. Admonition written on the exterior wall of the central chapel

’di na su yang ma ngo (= ’ong)
dgrag (= brag)
Ang ched po tsan (= btsan) po
su yang ma
# # # # #
Script: dbu-can
Grammatical marks: none
Translation: “Whomsoever, come not here to this rock formation of greatness and mightiness. “Whomsoever, do not…”
Technique: red ochre writing
Estimated date: Vestigial period or later
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: The window opening splits the second line of the upper inscription into two parts. Situated above the entranceway to the central chapel.

Fig. 301. Non-Buddhist mantras on the exterior wall of the central chapel

A Oṁ {huṁ} (upper)

Ā {Oṁ} raṁ dza (middle)

Ā A dkar {sa le} ’od
A yang Oṁ ’du A
Ōϻ hūṁ raṁ dza (lower)
Script: dbu-can
Grammatical marks: intersyllabic tsheg, gter-tsheg (lower)
Technique: red ochre writing
Estimated date: Vestigial period or later
Proximate rock art: two counterclockwise swastikas in red ochre, counterclockwise swastika in a white pigment
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: The first and second lines of the lower inscription were superimposed on other inscriptions. The largest and clearest of which reads: A (left side). Other letters were scratched into the surface of the clay plaster in close proximity. Situated to the right of the entranceway.

Fig. 302. Mantric syllables and two other inscriptions on the exterior wall of the central chapel

A  A (upper)

trang (= drang) po rgyal mtshan (middle)
mo rgyar ngas zas nor sog (= sogs)
cog ma byas (= dbye) {pas}
# me da lung # #
# (middle)

yar # {pad} ’dur (lower)
…shar (= shes) rab
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma, intersyllabic tsheg (middle); intersyllabic tsheg (lower)
Technique: red ochre writing
Translation: Middle: “Not separating me, Drang-po rgyal-mtshan mo-rgyar, from all manner of food and wealth…” Lower: “Upwards…funerary rites…wisdom.”
Estimated date: Vestigial period or later
Proximate rock art: bichrome geometric symbol, red ochre lotus (?) and counterclockwise swastika
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: The uppermost line is comprised of two separate inscriptions. Situated below inscriptions in fig. 298.

Fig. 303. The interior of the south chapel, Lha-khang dmar-chags

A number of mantric inscriptions were written on the interior walls of this cave temple, including some in a unique style in which letters are studded with the shells of crustaceans obtained presumably from nearby Da-rog mtsho. See figs. 304–306. The masonry and adobe platform and adjoining constructions as well as the two niches must have been used for enshrining sacred images and religious ceremonies.

Fig. 304. The Sale-’od, Du-tri-su and other mantras on cave wall

Ā A dkar sa le ’od yang Ōϻ ’du
A dkar A rmad du tri su nag po zhi zhi mal mal swa hā (upper)

pum yer cig pum yer cig (lower)
Script: dbu-can
Grammatical marks: yig-mgo mdun-ma+sgab-ma+initial nyis-shad (upper)
Technique: white pigment writing on blue ground
Estimated date: Vestigial period
Proximate rock art: lower border designs consisting of interconnected swastikas (g.yung-drung lag-’brel) and the so-called Greek border design (rgya-nag lcags-ri)
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: The meaning of the lower inscription written in the imperative is unclear. It may possibly command non-Buddhists to stay away from the site. Immediately to the right of the lower inscription is a yig-mgo mdun-ma+initial shad written by a different hand. In the same white-washed band as the lower inscription are extensive but fragmentary inscriptions. Unfortunately, the available photographic documentation is inadequate to assess them.

Fig. 305. Non-Buddhist mantras and other inscriptions on prepared surface

A Ōϻ hūṁ raṁ dza Ā dkar sa le
’od A yang Ōϻ ’du A dkar A rmad du tri
su nag po zhi zhi mal mal
Script: dbu-can
Grammatical marks: yig-mgo mdun-ma+initial nyis-shad
Technique: engraved and red ochre ornamented writing
Estimated date: Vestigial period
Proximate rock art: circle and other subjects carved and outlined in red ochre and studded with shells (upper row); carved and outlined in red ochre and studded with shells circle (mandala), conjoined sun and moon, circle, conjoined sun and moon, circle, counterclockwise swastika, circle, conjoined sun and moon, conjoined sun and moon, circle, counterclockwise swastika, circle, conch (?), parasol (?), conch, circle, circle, and conch (?) (second row from top)
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: A red ochre a-chung was added to the second A in the first line. The inscription was made by etching the letters into the stucco surface affixed to the cave wall. Each letter is also starred with delicate white lake shells.

Fig. 306. Fragmentary mantras for a tutelary deity on side of a masonry platform

hūṁ khrol ta ’ga ru na… (left)
A A Ā na nag shig shig ha…mar la khri khro # ru
haṁ phad nag shig shig…
na khri khro ha ra # thun thun…

# ’du # # # khro da ya khro (upper right)

bad de rga da hla sra dab # du (lower right)
Script: dbu-can
Grammatical marks: yig-mgo mdun-ma+sgab-ma+ initial nyi-zla+nyis-shad, intersyllabic tsheg, shad (left)
Technique: red ochre writing
Estimated date: Vestigial period or later
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: These lines appear to be comprised primarily of the Yungdrung Bon mantra for Khyung-dmar, a tutelary god in the form of a red horned eagle. There are also highly fragmentary mantras, albeit less extensive, on the other open face of the platform.

Fig. 307. The façade of the east chapel

Fig. 308. Non-Buddhist mantras on stucco panel above the entranceway to the east chapel

A Ōm huṁ maṁ dza (upper)

A Ōϻ hūṁ (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing (upper), molded and red ochre writing (lower)
Estimated date: Vestigial period or later (upper), Vestigial period (lower)
Proximate rock art: counterclockwise swastika
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: The upper inscription was added later. The lower inscription was written in relief as an integral part of the red ochre tinted stucco panel.

Fig. 309. Fragmentary Sale-’od mantras on stone in the façade on right side of entranceway to the east temple

A A dkar sa le ’od A yang Oṁ ’du
Script: dbu-can
Grammatical marks: nyis-shad
Technique: red ochre writing
Estimated date: Vestigial period or later
Proximate rock art: clockwise swastika
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: There are three or four Sale-’od mantras on the stone. The clockwise swastika was probably added as a demonstration of Buddhist occupation.

Fig. 310. Fragmentary Dpa’-bo ’bru-lnga and Sale-’od mantras on the left side of the façade of the east temple

{A} Ōϻ hūṁ raṁ dza
A dkar sa le ’od A
yang Ōϻ ’du
Script: dbu-can
Grammatical marks: intersyllabic tsheg, gter tsheg
Technique: molded and red ochre writing
Estimated date: Vestigial period
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: These inscriptions were made in relief upon a red ochre tinted stucco surface. On the middle left side of the panel Oṁ ma tri mu ye sa le ’du was written in two lines in red ochre and ye sa le ’du in a beige pigment (fragmentary Ma-tri mantra), as well as a few traces of other red ochre letters.

Fig. 311. Fragmentary Sale-’od, Du-tri-su and another mantra on the face of the rock formation

Oϻ # # # # ma…
A’ A dkar sa le ’od A yang {Ōϻ ’du}
A dkar {A} rmad du tri su {nag po} zhi zhi mal mal.
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing on a prepared ground
Estimated date: Vestigial period
Proximate rock art: superimposed on the first line of the inscription is a red ochre swastika with arms out of sync, two circles and flaming jewels; image of Ta-pi hri-tsa above the inscriptions
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: The mantras were written on a stucco surface of clay and mud affixed to the wall of the rock formation. This surface was polished and then whitewashed and lined with red ochre. The highly fragmentary top line of the inscription is where the Ma-tri mantra would be in the Yungdrung Bon tradition of arranging the three main sacred ejaculations. However, it is unclear what is written there.

Fig. 312. A close-up view of the saint Ta-pi hri-tsa

This image of Ta-pi hri-tsa, the 24th master of the Zhang-zhung snyan-rgyud lineage, appears to have had a dedicatory function, as well as legitimizing and protecting religious successors who occupied the site. It is the only known surviving non-Buddhist fresco in western Tibet. Unfortunately, it has been heavily damaged, precluding a detailed iconographic analysis of its esthetic elements. In Bon texts, Ta-pi hri-tsa is described as a translucent white figure devoid of clothing. While he is white-colored in the Da-rog mtsho rock painting, it is not clear that he is naked. Rather, he may possibly be cast as a bodhisattva wearing garments and ornaments. In the fresco, as well as in more conventional Bon renditions, Ta-pi hri-tsa’s hands meet on his lap in a gesture of meditational equipoise. The trilobate aureole, stepped throne, prominent folded legs, and palette are stylistic traits suggesting that it was painted ca. 1000–1250 CE.

Fig. 313. Sale-’od mantra

A A
dkar sale ’od
A yang Oṁ du
Script: dbu-can
Grammatical marks: yig-mgo mdun-ma+sgab-ma+initial nyis-shad
Estimated date: Vestigial period or later
Technique: red ochre writing
Location: Lha-khang dmar-chags, Da-rog mtsho

Fig. 314. Mantric syllables on the rock formation

A (upper)

A (lower)
Script: dbu-can (upper), precursory dbu-chen (lower)
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: five counterclockwise swastikas added at different times
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: These are the only inscriptions at the site that potentially predate the Vestigial period. Located some meters west and below the stucco panel in fig. 308.

Fig. 315. Inscription located at the base of the Lha-khang dmar-chags formation

A
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma
Technique: red ochre writing on white pigment ground
Estimated date: Vestigial period or later
Proximate rock art: an array of bichrome geometric symbols
Location: Lha-khang dmar-chags, Da-rog mtsho
Note: The geometric symbols above the inscription remain unidentified. These have the appearance of a sigillary script but they do not constitute conventional Tibetan letters. These symbols may possibly be a kind of visionary writing or ciphers of the mkha’-’gro gsang-yig or gter-yig genres. The symbols and inscription may possibly have been made by the same individual.

Rdzong pi-phi, Bar-yangs

Rdzong pi-phi is the name of a well-built five-sided fortress or other elite structure in the Bar-yangs region of ’Brong-pa county, formerly ’Brong-pa tsho-dgu. The inscriptions and pictographs are located at a lower site on the same hill, which consists of an ancient cave residence. On this site, see Bellezza 2014a, pp. 32–35.

Fig. 316. Non-Buddhist mantric syllables

A A
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Early Historic period
Proximate rock art: crescent moon, sun, tree, crescent moon, curvilinear subject
Location: Rdzong phi pi, ’Brong-pa tsho-dgu (’Brong-ba)
Note: The A on the right is partially cut in the photograph. The inscription and pictographs appear to form an integral composition. This composition can be assigned to the Early Historic period. Two other red ochre inscriptions are found nearby. One of them reads: Oṁ A huṁ (in vertical row) and is accompanied by a counterclockwise swastika. Above the façade of the cave there is a red ochre clockwise swastika as well as another dbu-can inscription that seems to read: jan gso nam pa ’chi. This inscription appears to convey the depreciation of the site. On this site, see Bellezza 2014a, pp. 32–35.

Ri-rgyal, Sger-rtse

Ri-rgyal is a tall, flat-topped mountain with an extensive boulder field with petroglyphs and funerary structures. The inscriptions are situated on the summit of the mountain. For more on Ri-rgyal, see Bellezza 2002a, pp. 131–135.

Fig. 317. Inscription

rga
{dung na}
Script: rudimentary dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Early Historic period
Location: Ri-rgyal, Sger-rtse
Note: The correct reading of the second line is unclear.

Fig. 318. Incomplete Ma-ṇi and other inscriptions

rgyang
ga
kar
Oṁ
Oṁ ma ni
Script: rudimentary dbu-can
Grammatical marks: initial nyis-shad, nyis-shad
Technique: carved writing
Estimated date: Early Historic period
Proximate rock art: yak, probably wild
Location: Ri-rgyal, Sger-rtse
Note: Although the inscriptions and rock art appear to form an integral composition, the syntactic connection between the lines in unclear. The uppermost line seems to signal a ritual operation, this word being related to rgyang-bu, an element of mdos rituals in Classical Tibetan texts. Kar appears to be the Old Tibetan word for white (dkar), as in a white yak. Also see Flight of the Khyung, 2010, December, fig. 12.

Far Western Tibet

For the purposes of of this study, this region includes Gu-ge, Sgar and Ru-thog. The ancient rock inscriptions of Far Western Tibet are carved, with just two exceptions that were painted.* The paleography of the region includes a few inscriptions comparable to those in Ladakh, Northern Pakistan and Wakhan. As discussed in Part 3, there is a higher proportion of Buddhist rock inscriptions in Far Western Tibet than in other regions of Upper Tibet.

The letter A is also written in red ochre on the roof of a cave called Gyam-brag phug in southwestern Tibet, see Bellezza 2014b, p. 165.

Rdu-ru-can, Gu-ge

Rdu-ru-can is spread over a steep rocky slope, which plunges off the esplanade bounding the A-yi-la range to the north. Situated in Rtsa-mda’ county, some of the inscriptions are found on boulders making up the foundations of all-stone corbelled residential structures. Converted to corrals, very little of the buildings remains. On these early domiciliary remains, see April 2016 Flight of the Khyung.

Fig. 319. Ma-ṇi mantra

Oṁ ma ṇi pad me hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: three stags, horseman coursing wild ungulate, carnivore pursing wild ungulate, stepped shrine (partially visible on far left)
Location: Rdu-ru-can, Gu-ge (Rtsa-mda’)
Note: The mantra was superimposed on the rock art. It may have been written to symbolically subjugate or purge the older rock art with its violent theme.

Fig. 320. Mantra

Ōṁ sa
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rdu-ru-can, Gu-ge (Rtsa-mda’)
Note: The precise significance of this mantra is unclear.

Fig. 321. Mantric syllable

Ōṁ
Script: rudimentary dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rdu-ru-can, Gu-ge (Rtsa-mda’)

Fig. 322. Inscription

Ōḿ ma ṇǐ pad me hūṁ
Script: rudimentary dbu-can
Grammatical marks: initial nyis-shad, intersyllabic tsheg, nyis-shad
Technique: carved writing
Estimated date: possibly Early Historic period
Location: Rdu-ru-can, Gu-ge (Rtsa-mda’)

Fig. 323. Ma-ṇi and another Buddhist mantra

Oṁ ma ṇi pad me hūṁ
E waṁ
Script: dbu-can
Grammatical marks:
Technique: carved writing
Estimated date: Possibly Early Historic period
Location: Rdu-ru-can, Gu-ge (Rtsa-mda’)
Note: E-wam (evam) is a Sanskrit word of affirmation.

Fig. 324. Mantric syllables

Oṁ Oṁ (upper)

Oṁ
A (lower)
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rdu-ru-can, Gu-ge (Rtsa-mda’)
Note: The first Oṁ in the upper inscription appears to have two sna-ldan marks flanking the vowel sign. The second Oṁ in the upper inscription has no sna-ldan mark. There appears to be an incomplete letter at the bottom of the image.

Fig. 325. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rdu-ru-can, Gu-ge (Rtsa-mda’)
Note: There may possibly be an O vowel mark above the letter.

Fig. 326. Ma-ṇi mantras

{Oṁ} ma ṇi pad me hūṁ
Oṁ ma ṇi pad me hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rdu-ru-can, Gu-ge (Rtsa-mda’)

Fig. 327. Mantric syllable

Oṁ
Script: rudimentary dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rdu-ru-can, Gu-ge (Rtsa-mda’)
Note: There appears to be another letter (a-chung?) below the syllable.

Gser-sgam, Gu-ge

Gser-sgam is a dark-colored calcareous formation hemming in the south side of the Glang-chen gsang-po valley in Rtsa-mda’ county. Once a significant agricultural settlement, Gser-sgam had its own Buddhist monastery and other monumental remains during the time of the Gu-ge kingdom.

Fig. 328. Mantric syllable

Ōṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: minor carving below syllable
Location: Gser-sgam, Gu-ge (Rtsa-mda’)
Note: This inscription was carved on limestone, a rock that does not re-patinate (unless covered by a harder mineral veneer).

Brag-gyam, Sgar

Brag-gyam is the name of a blue limestone formation near the line of control with India (Ladakh). The site boasts a collection of at least 68 mchod-rten engraved at the foot of an escarpment. These mchod-rten and accompanying inscriptions belong a tradition of rock art and epigraphy that is much more highly developed in the adjoining territory of Ladakh. On this site, see Bellezza 2017-2018.

Fig. 329. Mchod-rten dedications

co gru zhang btsun thar byang gĭs
bzhengs (upper)

snang za mo rgyan (lower)
bsod
namsu
zhengs (= bzhengs)
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Translation: Lower: “Established for merit by Snang-za mo-rgyan”.
Estimated date: Early Historic period
Proximate rock art: mchod-rten, minor carving below syllable
Location: Brag-gyam, Sgar
Note: Both inscriptions refer to the carving of proximate mchod-rten. Given its syntax, the translation of the upper inscription is ambiguous. It is not clear whether the third and fourth word elements (zhang-btsun, thar-byang) constitute part of the proper name or a description of the carver of a mchod-rten. Zhang may either be part of the proper name or designate a female member of the Co-gru clan. Co-gru (Cog-gru [sic]) was a prominent clan in the affairs of the Western Tibetan Plateau in the Early Historic Period). Btsun may be part of the proper name or denote a ‘monk’. Similarly, thar-byang (liberated and purified) may be a description of the function or aspiration behind the rock art or part of the proper name of the individual who made it. Also see Bellezza 2008, p. 187 (n. 193).

Fig. 330. Mchod-rten dedication

bshen (= gshen) rgyal
{bkris}
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Proximate rock art: mchod-rten (partially visible)
Estimated date: Early Historic period
Location: Brag-gyam, Sgar
Note: Gshen-rgyal is either a proper name or denotes the profession of a bon high priest. Bkris is usually a conjunction of bkra-shis (good luck), but in this Early Historic period context it is likely that bris (to write) was intended. Also see Bellezza 2008, p. 187 (n. 193).

Gong-ra, Ra-bang

Gong-ra is the name of a rocky hill overlooking the south side of Ra-bang mtsho in Ru-thog. On this site, see Bellezza 2002a, p. 244.

Fig. 331. Label

g.yag
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: wild yak
Location: Gong-ra, Ra-bang
Note: The inscription identifies the proximate animal which was carved at a much earlier date. The inscriber did not designate the yak as a wild variant (’brong / g.yag-rgod). Also see March 2012 Flight of the Khyung, fig. 3.

Fig. 332. Clan inscription

khyung pho ni ra la chib
Script: rudimentary dbu-can
Grammatical marks: intersyllabic tsheg
Technique: carved writing
Translation: “The Khyung-pho (C.T. = po) [clan] is that which rides the goat.”
Estimated date: Early Historic period
Proximate rock art: minor carving below syllable
Location: Gong-ra, Ra-bang
Note: The inscription may possibly refer to a clan god (ru-lha) of the Khyung-po clan that is mounted on a goat. Also see January 2013 Flight of the Khyung, fig. 1.

Fig. 333. Clan inscription

khyung pho za dag
{spyar} khyed {ra} bzhin
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Imperial period
Proximate rock art: minor carving below syllable
Location: Gong-ra, Ra-bang
Note: Pertaining to a group of female (?) khyung-pho (C.T. = po) clan members apparently written in the second person.

Skabs-ren spungs-ri, Ra-bang

Skabs-ren spungs-ri, a rocky hill, is situated in the Skabs-ren valley, one of the seven tributary valleys of Ra-bang and Khul-pa (in Ru-thog). On this site, see Bellezza 2002a, pp. 135–140.

Fig. 334. Ma-ṇi mantra

Ōṁ ma ṇi padme hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: minor carving below syllable
Location: Skabs-ren spungs-ri, Ra-bang

Chu-mkhar gyam sgrub-phug, Rdo-dmar

Chu-mkhar gyam sgrub-phug is a large cave suspended in a large limestone formation in the northern part of Ru-thog. On top of the formation are the remains of a citadel known as Chu-mkhar gyam. On these sites, see Bellezza 2002a, pp. 34, 35, 145, 146.

Fig. 335. Ma-ṇi mantra

Oṁ ma ṇi
padme hūṁ {hri}
Script: dbu-can
Grammatical marks: none
Technique: yellow ochre or orpiment writing
Estimated date: Vestigial period
Proximate rock art: minor carving below syllable
Location: Chu-mkhar gyam sgrub-phug, Rdo-dmar
Note: The first and the last two syllables were written in a non-standard fashion.

Mchod-rten sbug sna-kha, Ra-bang

Mchod-rten sbug sna-kha is situated at the confluence of the Lha-skyes valley and a small tributary valley in Ra-bang (Ru-thog). The site consists of vertical and inclined rock panels rising a maximum of 3 m above the valley floor.

Fig. 336. Mantric syllable

A
Script: rudimentary dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Early Historic period
Proximate rock art: older wild yak
Location: Mchod-rten sbug sna-kha, Ra-bang
Note: What appears to be a letter A was carved as an integral feature of a mchod-rten, which exhibits archaic design features. For this mchod-rten, see Bellezza 2017-2018, fig. 18q.

Brag-gdong, Ra-bang

Brag-gdong is located on the north shore of Glog-phug mtsho. This site extends for nearly 3 km, encompassing various lakeside escarpments. On this site, see Bellezza 2002a, pp. 142–145.

Fig. 337. Inscription

Khyung pho
Script: rudimentary dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Early Historic period
Location: Brag-gdong East, Ra-bang
Note: A clan name.

Fig. 338. Inscription

Ōṁ # # # # #

{go} # shig
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Location: Brag-gdong East, Ra-bang
Note: Due to the light conditions in which it was taken, the photograph is inadequate for assessment of the inscription.

Fig. 339. Mantric syllables

A

A
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Proximate rock art: tiered shrine (upper middle), tiered shrine or geometric design (lower left) and minor carvings
Location: Brag-gdong West, Ru-thog

Fig. 340. Inscription

{’pags (C.T. = phags) pa}
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Translation: {“Exalted.”}
Estimated date: Vestigial period or earlier
Location: Brag-gdong West, Ru-thog
Note: This appears to be a Buddhist epithet. There seems to be another a-chung to the left of the inscription.

Glog-phug mkhar, Ra-bang

The rock art site of Glog-phug mkhar is situated at the foot of the slopes on which stands the archaic fortress of Glog-phug mkhar (Ru-thog). On this site, see Bellezza 2001, pp. 105, 106.

Fig. 341. Mantric syllables and Ma-ṇi mantra

Oṁ (upper left)

Ōṁ (upper right)

Oṁ ma ni pad me huṁ (middle)

Oṁ A hūṁ da (lower left)

Oṁ (lower right)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (middle)
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: two tiered shrines, quadruped, other minor carvings
Location: Glog-phug mkhar, Ra-bang
Note: The rock art may have been made along with one or more of the inscriptions. Also see Bellezza 2001, pp. 224, 372 (fig. 10.106).

Rwa-’brog ’phrang, Ra-bang

Rwa-’brog ’phrang extends for over 1 km along the south shore of Rwa-’brog mtsho in Ru-thog. The series of cliffs at this location host an extensive array of rock art. On this site, see Bellezza 2001, pp. 211–224.

Fig. 342. Inscription

na {yul}
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or earlier
Proximate rock art: horse to right (partially visible in photograph)
Location: Rwa-’brog ’phrang, Ra-bang

Fig. 343. Ma-ṇi mantra

Oḿ ma ni pad me hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or earlier
Proximate rock art: body of wild ungulate above
Location: Rwa-’brog ’phrang, Ru-thog

Fig. 344. Ma-ṇi mantras

Oṁ ma ni pad me hūṁ
Oṁ ma ni pad me hūṁ
# # hri
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or earlier
Proximate rock art: anthropomorphic figure
Location: Rwa-’brog ’phrang, Ru-thog
Note: The last two syllables of the upper line wrap around the middle line. The lower line was partially superimposed on the rock art.

Fig. 345. Syllable

rgwa
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Location: Rwa-’brog ’phrang, Ru-thog
Note: This may be the name of a clan.

Fig. 346. Inscription

A
{tshogs dbang ngo
man ma brgyos}
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Translation: {“The Empowerment Tshogs.”}
Estimated date: Vestigial period or somewhat earlier
Location: Rwa-’brog ’phrang, Ru-thog
Note: Tshogs is an assembly of offerings. The meaning of the third line is unclear. It may possibly have to do with sexual intercourse. However, some of the letters appear to have been retouched.

Fig. 347. Ma-ṇi mantras, complete and incomplete

Oḿ ma ṇi pad me hūḿ (upper)

A ma ṇa pad ma (middle)

A ma (lower)
Script: dbu-can
Grammatical marks: initial shad, shad (upper)
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rwa-’brog ’phrang, Ru-thog

Fig. 348. Mantric syllables

Oḿ

Oḿ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Proximate rock art: standing deity with what appears to be a sword, ewer
Estimated date: Vestigial period
Location: Rwa-’brog ’phrang, Ru-thog
Note: The Buddhist mantra (one-hundred-syllable mantra for the god Rdo-rje sems-dpa’, etc.) was carved over the two older syllables, suggesting a cleansing or purging action was intended. The older syllables must have belonged to a non-Buddhist sect. The ewer appears to have been carved by the maker of one or both older mantric syllables. Also see Bellezza 2001, pp. 224, 372 (fig. 10.107).

Fig. 349. Non-Buddhist dedication

khyung pho mdo
sob sa gyis
{lha} khrǐs (= bris)
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Translation: “Khyung-pho mdo’s facsimile made as the {divine} written/drawn.”
Estimated date: Early Historic period
Proximate rock art: two counterclockwise swastikas
Location: Rwa-’brog ’phrang, Ru-thog
Note: It appears that the upper swastika represents a ‘facsimile of a ritual object’ (sob-sa) made by the inscriber. This is supported by the carving technique and wear of the inscription and swastikas. The two swastikas establish the non-Buddhist identity of the inscriber cum artist.

Fig. 350. Ma-ṇi mantra

Ōḿ ma ṇi pad me hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: conjoined sun and moon and tiered shrine
Location: Rwa-’brog ’phrang, Ru-thog
Note: The conjoined sun and moon and tiered shrine (probably a simplified mchod-rten) appear to have been made by the inscriber as an integral composition. Also see Bellezza 2001, pp. 224, 371 (fig. 10.105); 2017-2018, fig. 17l.

Fig. 351. Clan label

khyung
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Early Historic period
Location: Rwa-’brog ’phrang, Ru-thog

Fig. 352. Ma-ṇi mantra

Ōɱ ma ṇi pad me hūḿ
Script: dbu-can
Grammatical marks: yig-mgo mdun-ma+sgab-ma+hooked style nyi-zla+initial shad, intersyllabic tsheg, shad
Technique: carved writing
Estimated date: Vestigial period
Location: Rwa-’brog ’phrang, Ru-thog

Fig. 353. Ma-ṇi mantra

Ōϻ ma ṇi
pad me hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rwa-’brog ’phrang, Ru-thog

Fig. 354. Ma-ṇi mantra

Oḿ ma ṇi pad me
{huṁ}
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rwa-’brog ’phrang, Ra-bang
Note: There appears to be an unfinished letter above the inscription. Also see Bellezza 2001, pp. 224, 373 (fig. 10.108).

Sna-kha sogs, Khul-pa

Sna-kha sogs is a rocky prow that juts into the Khul-pa valley, funneling traffic in the valley close by it. This site is situated in Ru-thog. As with other conspicuous rock art and epigraphic sites, Sna-kha-sogs was meant to be seen by travelers to the area.

Fig. 355. Buddhist rdo-rje dedication

spyĭ ti sde myang rmang la snang
gis bris khyung po
tha chun (= chung) gyĭs brgyis (C.T. = bgyis) pa’o (right)

A gas (left)
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Translation: “Written by Myang rmang-la snang of the Spyĭ ti district. Done by the youngest Khyung po.”
Estimated date: Imperial period
Proximate rock art: two ritual thunderbolts (above inscription) and clockwise swastika (to right of first line)
Location: Sna-kha sogs, Khul-pa
Note: The object of the dedication appears to be one or both rdo-rje. This sacred symbol as well as the clockwise swastika establish the Buddhist identity of the composition. The inscription constitutes the oldest known written reference to Spiti (Spyi-ti), one of five districts (sde) or Lower (Smad) Zhang Zhung in the Imperial period. Also see Bellezza 2008, pp. 186, 187 (n. 193); 2017b, fig. 1; June 2016 Flight of the Khyung, fig. 58.

Fig. 356. Fragmentary Ma-ṇi mantra

Oṁ ma ṇi pad
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Sna-kha sogs, Khul-pa

Fig. 357. Ma-ṇi mantra

Oṁ ma ṇi pad
me hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Sna-kha sogs, Khul-pa
Note: The hūṁ syllable appears to have been added subsequently.

Fig. 358. Clan and personal name

Khyung pho lǐ la brtsan
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Early Historic period
Proximate rock art: peacock (upper left), quadruped (middle left), mchod-rten (lower middle), and anthropomorphic figure (lower right)
Location: Sna-kha sogs, Khul-pa
Note: This is the only inscription documented in Upper Tibet that contains a word element () that may possibly be derived from the Zhang-zhing language. Also see 2008, pp. 186, 187 (n. 193). Some of the rock art may have been made along with the inscription. On the mchod-rten, see Bellezza 2017-2018, fig. 16r.

Fig. 359. Rigs-gsum mgon-po mantras

# # # pad me hūṁ
Oḿ ba ki sha ri mum
Oḿ badzra pa ni hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Sna-kha sogs, Khul-pa
Note: The upper mantra has been largely obliterated. A superfluous n was added to the lower mantra.

Fig. 360. Ma-ṇi mantra

{Ōϻ} ma {ṇi} pad me hūṁ
Script: dbu-can
Grammatical marks: brda-rnying yig-mgo mdun-ma+sgab-ma+initial nyisshad, intersyllabic tsheg, shad
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Sna-kha sogs, Khul-pa
Note: Situated on same boulder as in fig. 359. The lower syllable hi and Ma-ṇi mantra are more recent additions.

Fig. 361. Inscriptions

si (upper right)

legs (lower left)
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Sna-kha sogs, Khul-pa
Note: The later relief engraving was made on top of an older inscription, of which only the letter na and what appears to be a superscribed vowel sign (gi-gu) is readable. This writing may possibly have been part of the lower left inscription.

Fig. 362. Mantric syllables

ma (upper)
Oṁ

Oṁ (lower)
ma
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Sna-kha sogs, Khul-pa

Fig. 363. A prayerful word

mkhyen
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Location: Sna-kha sogs, Khul-pa

Nag-skyom, Khul-pa

Nag-skyom is situated on the edge of the Khul-pa valley in Ru-thog. This small site consists of three panels of petroglyphs as well as one inscription.

Fig. 364. Inscription

nyi ’od gi sde
chu nad gyas {}
Script: dbu-can
Grammatical marks: none
Translation: “The district of sunlight.”
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Nag-skyom, Khul-pa
Note: A reading of the second line is problematic due to ambiguity concerning the division of syllables and the retouching of letters.

Ri-mo gdong, Sde-rog

The famous rock art site of Ri-mo gdong is situated 45 km south of Ru-thog rdzong and the modern county headquarters of Ru-thog. The escarpment here rises above the banks of the Ru-gsum/Rog-gsum river. The site straddles the main route north to Ru-thog, which at this point is a relatively narrow passageway, endowing it with a conspicuous aspect and strategic importance. On the geographical importance of this site, see October 2014 Flight of the Khyung, fig. 27.

Fig. 365. Inscription

nga’i
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Translation: “Mine”
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: mchod-rten
Location: Ri-mo gdong, Sde-rog
Note: The inscription seems to refer to the ownership of the mchod-rten. For this mchod-rten, see Bellezza 2017-2018, fig. 10j.

Fig. 366. Ma-ṇi mantras

Ōḿ ma ṇi pad me hūḿ (upper)

Oḿ ma ṇi pad me hūṁ (middle left)

Ōḿ ma ṇi pad me hūṁ (middle right)

Ōϻ ma ṇi pad me hūḿ (lower)
Script: dbu-can
Grammatical marks: initial nyis-shad, intersyllabic tsheg (upper); initial nyis-shad (middle left); yig-mgo mdun-ma+sgab-ma, intersyllabic tsheg (middle right); brda-rnying yig-mgo mdun-ma+sgab-ma+initial shad, intersyllabic tsheg (lower)
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: mchod-rten flanked on each side by a conjoined sun and moon
Location: Ri-mo gdong, Sde-rog
Note: The mchod-rten and upper left inscription appear to form an integral composition. On this mchod-rten, see Bellezza 2017-2018, fig. 17k.

Fig. 367. Ma-ṇi mantra

Oḿ ma ṇi pad me hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Ri-mo gdong, Sde-rog

Fig. 368.

Oṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Ri-mo gdong, Sde-rog

She-rang sna-kha shar-ma, Sde-rog

She-rang sna-kha shar-ma is located at the confluence of the She-rang and Ru-gsum valleys in Ru-thog. It occupies a small but prominent east-facing rock face. Except what is pictured below the inscriptions, there is no other rock art at this site.

Fig. 369. Ma-ṇi mantras

Khyung

Oṁ ma ṇi pad me hūḿ

Oḿ ma ṇi pad # hūṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: two conjoined sun and moon symbols flanking middle inscription and line of eight triangular figures
Location: She-rang sna-kha shar-ma
Note: The middle inscription and row of triangular figures appear to form an integral composition. The row of figures appears to depict offering cakes (gtor-ma). It appears that at least one of these inscriptions was made by a member of the Khyung clan.

Ru-thog rdzong

Amid the extensive ruins of the Ru-thog acropolis there is a single boulder with rock art and an inscription. It situated on the south side of the hill in what was once part of the ancient village and monastic quarter. On this site see Bellezza 2001, pp. 102–104; 2011, p. 88; 2014a, p. 153.

Fig. 370. Ma-ṇi mantras

Oṁ ma ṇi pad # #

Oṁ ma hum Oṁ ma ṇi pad {me hūṁ}
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: two quadrupeds, mchod-rten and minor carvings
Location: Ru-thog rdzong
Note: The last two syllables of the lower inscription are cut in the photograph, as is a lower inscription. On the mchod-rten, see Bellezza 2017-2018, fig. 9c.

Rno-ba g.yang-rdo, O-byang

Rno-ba g.yang-do is located a few kilometers upstream of the township headquarters of O-byang at the confluence of Rdo-dmar river and a side valley. Situated in Ru-thog, much of the site was destroyed in recent years by the construction of a dam and hydroelectric plant.

Fig. 371. Mantric syllable

Oṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Location: Rno-ba g.yang-rdo

Addendum

The following 23 images and rock inscriptions were added after the above work was compiled. With some 50,000 images from Upper Tibet to contend with, these were not immediately apparent to me. In fact, there are still roughly 10,000 images pending digitization, some of which were not accessible for inspection. Photographs that could not be consulted for this work include a small but indeterminate number of ancient rock inscriptions from the region.*

On a pargetted surface in a cave called Skyid-sgrom sgo-gru bzhi in Gzhung-smad (Shan-rtsa county), in addition to rock art consisting of a mchod-rten and two counterclockwise swastikas, there are three inscriptions reading ha, hri, and a partially effaced Ma-tri painted painted in different colored pigments by various people. The chorten fresco and the inscriptions appear to date to the Vestigial period.

Thang-lha’i rten-khang

Not far from Thang-lha’i rten-khang are the ruins of a Buddhist hermitage, Ma-ṇi wall and cremation grounds (dur-khrod).* There are also two stone pillar (rdo-ring) that are probably associated with the founding of the hermitage. This site is situated in a hanging valley at the southern foot of the Gnyan-chen thang-lha massif, in Snying-drung. The smaller rdo-ring boasts two inscriptions.

This large shrine for the mountain god Gnyan-chen thang-lha sits upon a hilltop at the base of the massif, near the ’brog-pa settlement of Bar-gling. In the oral tradition of the region, its construction is attributed to the Fifth Dalai Lama. It was rebuilt after being destroyed in the Chinese Cultural Revolution. The rten-khang is comprised of a chapel three stories in height and approximately 8 m in length on each of its four sides with a small extension at the entrance. The present shrine is confined to the middle floor and is only opened on special occasions, when high lamas like those from Stag-lung monastery come to conduct observances for Gnyan-chen thang-lha. It is said that over a period of more than 250 years until the Cultural Revolution, Thang-lha’i rten-khang acquired great wealth through the bestowal of gifts. It reportedly enshrined an assortment of armaments, precious stones and metals, ivory and even a rhinoceros horn. Local elders say that the Lhasa government sent a representative from Rnam-rgyal grwa-tshang every other year to make offerings to Gnyan-chen thang-lha. I visited the site in October 1997.

Fig. 372. Buddhist mantras

haṁ (left)
ka
ṣha
ma
la
{wa
ra
yam}

A (right)
ra
pa
{tsa
ni}
ḍhi
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Location: near Thang-lha’i rten-khang
Note: The mantra on the left side of the pillar consists of interconnecting letters in a vertical line, the only ancient inscription arranged in this fashion in this survey. The reading of the mantra on the right side of the pillar comes from my field notes, as most of it is not visible in the photograph.

Near Gu-ru sgrub-phug

For more inscriptions from this site, see figs. 64–84.

Fig. 373. Non-Buddhist mantric seed syllable

Oṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: two counterclockwise swastikas
Location: near Gu-ru sgrub-phug, Bkra-shis do-chung
Note: The inscription and rock art appear to form an integral composition.

Sri-gcod ’bum-pa, Bkra-shis do-chung

For more inscriptions from this site, see figs. 85–87.

Fig. 374. Mantric seed syllables

A
{Oṁ}
huṁ
A
Script: rudimentary dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or perhaps somewhat earlier
Proximate rock art: mchod-rten
Location: Sri-gcod ’bum-pa, Bkra-shis do-chung
Note: These syllables are arrayed inside the vase and the adjacent tier of the base of a pictographic mchod-rten. The inscription appears to be an integral part of the composition. If the second syllable is as presented, it confirms that this is a non-Buddhist version of a mchod-rten, one of many such specimens at Bkra-shis do. Also see Bellezza 2001, pp. 211, 342 (fig. 10.47); 2017-2018, fig. 16e.

Khyi-rgan gag-pa do

For more inscriptions from this site, see figs. 203, 204.

Fig. 375. Non-Buddhist inscription

sa
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: two counterclockwise swastikas
Location: Khyi-rgan gag-pa do, Gnam-mtsho
Note: The swastikas and letter form one composition.

Fig. 376. Ma-tri mantra

ye mu
sa
le
’du
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: rectilinear subject
Location: Khyi-rgan gag-pa do, Gnam-mtsho
Note: The ’du is not visible in this photo and is situated at some distance from the rest of the inscription. The letters of this mantra were scattered across the rock face perhaps to conceal them from rivals. The Oṁ and ma of the mantra were not photographed, and it is not clear whether they exist.

Se-mo do, Gnam-mtsho

For more inscriptions from this site, see figs. 233–236.

Fig. 377. Non-Buddhist inscription and mantra

bon phug (upper)

A ma (lower)
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or later
Proximate rock art: three counterclockwise swastikas
Location: Se-mo do, Klu-khang
Note: The swastikas appear to have been made with one or both inscriptions.

Chos-lung O-rgyan bsam-gtan-gling, Gzims-phug

For another inscription from this site, see fig. 259.

Fig. 378. Non-Buddhist mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika (partially visible)
Location: Chos-lung O-rgyan bsam-gtan-gling
Note: The inscription and swastika appear to have been made as an integral composition.

Fig. 379. Non-Buddhist mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: counterclockwise swastika (partially visible)
Location: Chos-lung O-rgyan bsam-gtan-gling
Note: The inscription and two counterclockwise swastikas arrayed below it appear to have been made as an integral composition.

Skyid-sgrom dgon-pa, Gzhung-smad

This site consists of a cave complex and associated residential ruins. The Buddhist inscriptions found here appear to have been created to bring this ancient sanctuary under the auspices of Buddhism. For more on this site, see Bellezza 2014a, pp. 395–397.

Fig. 380. Rigs-gsum mgon-po mantras

Ōϻ ba gi shwa ri muṁ Ōϻ ma ṇi padme hūṁ Ōϻ badzra pa ni hūṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing and relief carving
Estimated date: Vestigial period
Location: Skyid-sgrom dgon-pa, Gzhung-smad

Fig. 381. Buddhist seed syllables

Om A huṁ
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: white pigment writing
Estimated date: Vestigial period
Location: Skyid-sgrom dgon-pa, Gzhung-smad

Stong-shong brag-khung

The two inscriptions at this site are found in an escarpment bounding the north side of Rtsid-skud mtsho in Srin-ya township, Shan-rtsa county (part of the erstwhile Nag-tshang district). In content and technique, the non-Buddhist inscriptions of Stong-shong brag-khung are of the type commonly encountered in the Eastern Byang-thang. For more on this site and the lore of the surrounding area, see Bellezza 2001, pp. 134–138.

Fig. 382. Incomplete Sale-’od mantra

le ’od
A yang {Oṁ ’du}
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Stong-shong brag-khung
Note: The vowel sign over the syllable ’od is not visible.

Fig. 383. Mantric syllable

A
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Stong-shong brag-khung
Note: There is a third inscription in this cave that reads: A ma, but no photograph is available.

Gnas kun-bzang

The sole rock inscription recorded at this ancient troglodytic settlement is inside a small cave, one of at least 14 located in an escarpment towering above Dang-chung mtsho. This site is in Dang-chung township, Nyi-ma county (formerly part of the Nag-tshang district). On Gnas kun-bzang and its inscription, also see Bellezza 2001, pp. 112–114.

Fig. 384. Buddhist mantric syllables

{Ōϻ}
hūṁ
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period
Proximate rock art: clockwise swastika with dots between the four arms
Location: Gnas kun-bzang
Note: The inscription and swastika appear to form a unified composition.

G.yung-drung lha-rtse, Dang-ra g.yu-mtsho

This site is located on the circumambulatory trail around Dang-ra g.yu-mtsho, in what is now ’Om-bu township, Nyi-ma county (once part of the Nag-tshang district). Above the site is the famous Yungdrung Bon Rdzogs-chen hermitage of G.yung-drung lha-rtse founded in the 11th century CE. This is one of the most westerly examples of a group of early non-Buddhist scripts used in the rock epigraphy of the eastern half of the Byang-thang. The area of the inscription (east side of Dang-ra g.yu-mtsho) still constitutes Yungdrung Bon ecclesiastic territory. On this site and the inscription, see Bellezza 1997a, p. 393.

Fig. 385. Mantric syllables

A
A
sa
#
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: G.yung-drung lha-rtse
Note: The style of the A and sa are associated with non-Buddhist red ochre inscriptions of the Eastern Byang-thang.

Dar-chung, Dang-ra g.yu-mtsho

Dar-chung is another site on the east side of the circumambulatory trail around Dang-ra -g.yu-mtsho, in ’Om-bu township (formerly part of the Nag-tshang district).

Fig. 386. Non-Buddhist seed syllables

A Oṁ huṁ
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period
Location: Dar-chung
Note: The inscription is found inside the middle part of the base of an engraved mchod-rten. On this mchod-rten, see Bellezza 1997a, pp. 388, 389; 2017-2018, fig. 18w.

Do dril-bu, Da-rog mtsho

Do dril-bu is located in Da-rog mtsho, Lung-dkar township, ’Brong-pa county (formerly part of ’Brong tsho-dgu). Like Se-mo do, this is an insular site. The island is home to a group of ancient all-stone corbelled residences. The inscriptions at this site are written on the interior walls of the residential structure designated RS7. The mantric inscriptions were made by a non-Buddhist practitioner occupying one of the ancient residences. For more on this site, see Bellezza 2014a, pp. 473–478.

Fig. 387. Non-Buddhist mantras

Ā (upper left)

ma syaṁ khaṁ (lower left)

A (middle)
Ōϻ

Ōm ma tri {mu ye sa} le ’du (right)
Script: dbu-can
Grammatical marks: intersyllabic tsheg (right)
Technique: white pigment writing
Estimated date: Vestigial period or later
Location: Do dril-bu, Da-rog mtsho
Proximate rock art: two series of boxed-in dots and tree-like subject
Note: There are traces of a lower inscription on the right side of the photograph. All these inscriptions may have been made by the same individual. The syllables syaṁ and khaṁ are framed by thick lines. The series of dots accompanying them may have been made by the artist to keep track of the length of occupancy.

Tham-ka-can, Rtsa-mda’

This site occupies one side of a defile through which the road from Tho-ling to Khyung-lung passes. It is named after more than a dozen square and circular geometric subjects that resemble the imprint of seals. However, each one is unique and was painted by hand. These pictographs are located on the roof of an overhand several meters overhead. In order to paint them a ladder, scaffolding or some other kind of extension was required.

Fig. 388. Inscription

A
na
ya
{rga}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Tham-ka-can, Rtsa-mda’
Note: Nearby is a counterclockwise swastika painted in what appears to be the same ochre pigment, suggesting a non-Buddhist identity for the inscription.

Fig. 389. Inscription

{me nya}
Script: dbu-can
Grammatical marks: intersyllabic tsheg
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Location: Tham-ka-can, Rtsa-mda’
Note: Situated below inscription in fig. 388. Nearby is a red ochre counterclockwise swastika.

Fig. 390. Mantric syllables

{Oṁ}
{Oṁ}
Script: dbu-can
Grammatical marks: none
Technique: red ochre writing
Estimated date: Vestigial period or somewhat earlier
Proximate rock art: three geometric subjects
Location: Tham-ka-can, Rtsa-mda’
Note: Situated on the roof of an overhang above reach. The geometric subjects were painted more boldly in a different red ochre hue and at a later date. They partially obscure the lower inscription.

Rwa-’brog ’phrang, Ra-bang

For more inscriptions from this site, see figs. 342–354.

Fig. 391. Mantra

Oṁ ni
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rwa-’brog ’phrang, Ra-bang
Proximate rock art: tiered shrine above (partially visible)
Note: The incised syllable hri was added in a later period.

Fig. 392. Rigs-gsum mgon-po mantras

Ōḿ ’wa (C.T. = wa) gi shwa ri mung (= muṁ) Ōḿ ma ṇi padme hūṁ Ōϻ badzra dzar pa ni hūṁ
Script: dbu-can
Grammatical marks: brda-rnyig yig-mgo mdun-ma+sgab-ma+initial nyis-shad, intersyllabic tsheg
Technique: carved writing
Estimated date: Vestigial period or somewhat earlier
Location: Rwa-’brog ’phrang, Ra-bang
Proximate rock art: mchod-rten (partially visible)
Note: The inscription is situated inside the middle section of a mchod-rten and they may have been made at the same time. There appears also to be an O vowel sign above gi. The syllable dzar is unnecessary. The later Buddhist inscriptions flanking the Rigs-gsum mgon-po mantra are not relevant to this study. On the mchod-rten and inscription, also see Bellezza 2001, pp. 224, 370 (fig. 10.102); 2017-2018, fig. 20d.

Fig. 393. Two letters

na na
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Vestigial period or earlier
Location: Rwa-’brog ’phrang, Ra-bang

Brag-gyam, Sgar

For more inscriptions from this site, see figs. 329, 330.

Fig. 394. Mantric seed syllables

Oϻ huṁ A
Script: dbu-can
Grammatical marks: none
Technique: carved writing
Estimated date: Early Historic period
Location: Brag-gyam, Sgar
Proximate rock art: tiered shrine above (partially visible)
Note: The inscription is located inside the bum-pa of a carved mchod-rten. Due to the sequence of the three syllables the religious affiliation of the inscription is uncertain. On this mchod-rten, see Bellezza 2017-2018, fig. 19q.

 

Bibliography

Achard, Jean-Luc. 1998. “Bon zhig khyung nag and the Rig pa gcer mthong Tradition of rDzogs chen”, in The Tibet Journal, vol. 23 (no. 4), pp. 28–57. Dharamsala: Library of Tibetan Works and Archives.

Akester, Mathew. 2016. Jamyang Khyentsé Wangpo’s Guide to Central Tibet. Chicago: Serindia.

Bacot, J. / Thomas, F. W. / Toussaint, Ch. 1940-1946. Documents de Touen-houang relatifs à l’histoire du Tibet. Annales du Musée Guimet. Paris: Librairie Orientaliste Paul Guenther.

Beckwith, C. I. 1987. The Tibetan Empire in Central Asia: A History of the Struggle for Great Power among the Tibetans, Turks, Arabs, and Chinese during the Early Middle Ages. Princeton: Princeton University Press.

_____In press-a. “Facsimiles of the Earliest Ritual Architecture in Tibet: A Comprehensive Survey of Stepped Shrines in the Rock Art of Upper Tibet”, in asianarts.com, International Association of Tibetan Studies, Conference XIV, Bergen, June 19–25, 2016.

_____In press-b: “Discerning Bon and Zhang Zhung on the Western Tibetan Plateau: Designing an archaeological nomenclature for Upper Tibet, Ladakh and Spiti Based on a study of cognate rock art”, for International Conference of Shang Shung Cultural Studies Beijing, September 18–21, 2015.

Bellezza, John V. 2017-2018. A Comprehensive Survey of Stepped Shrines in the Rock Art of Upper Tibet: With Archaeological, Cultural and Historical Comments, in November, December 2017 and January 2018 Flight of the Khyung, Tibet Archaeology. http://www.tibetarchaeology.com/november-2017/ http://www.tibetarchaeology.com/december-2017/ http://www.tibetarchaeology.com/january-2018/

2017a: “The Swastika, Stepped Shrine, Priest, Horned Eagle, and Wild Yak Rider: Prominent antecedents of Bon figurative and symbolic traditions in the rock art of Upper Tibet”, in Revue d’Etudes Tibétaines, no. 42, October 2017, pp. 5–38. Paris: CNRS.
http://himalaya.socanth.cam.ac.uk/collections/journals/ret/pdf/ret_42_01.pdf

2017b: “The Rock Art of Spiti: A General Introduction”, in Revue d’etudes tibétaines, vol. 41, pp. 56–85. The Spiti Valley: Recovering the Past and exploring the Present (eds. Y. Laurent and D. Pritzker). Paris: CNRS. http://www.himalaya.socanth.cam.ac.uk/collections/journals/ret/pdf/ret_41_03.pdf

2014a. Antiquities of Zhang Zhung: A Comprehensive Inventory of Pre-Buddhist Sites on the Tibetan Upland, Residential Monuments, vol. 1. Miscellaneous Series – 28. Sarnath: Central University of Tibetan Studies. Online version, 2011: Tibetan & Himalayan Library (THlib.org). http://www.thlib.org/bellezza

_____2014b. Antiquities of Zhang Zhung: A Comprehensive Inventory of Pre-Buddhist Sites on the Tibetan Upland, Ceremonial Monuments, vol. 2. Miscellaneous Series – 29. Sarnath: Central University of Tibetan Studies. Online version, 2011: Tibetan & Himalayan Library (THlib.org): http://www.thlib.org/bellezza

_____2014c. The Dawn of Tibet: The Ancient Civilization on the Roof of the World. Lanham: Rowman & Littlefield.

_____2014d. “Straddling the Millennial Divide: A case study of persistence and change in the Tibetan ritual tradition based on the Gnag rabs of Gathang Bumpa and Eternal Bon documents, circa 900–1100 CE”, in Revue d’etudes tibétaines, vol. 29, pp. 155–243. Paris: CNRS.
http://himalaya.socanth.cam.ac.uk/collections/journals/ret/pdf/ret_29_07.pdf

_____2013. Death and Beyond in Ancient Tibet: Archaic Concepts and Practices in a Thousand-Year-Old Illuminated Funerary Manuscript and Old Tibetan Funerary Documents of Gathang Bumpa and Dunhuang. Philosophisch-Historische Klasse Denkschriften, vol. 454. Wien: Österreichischen Akademie der Wissenschaften.

_____2011. “Territorial Characteristics of the Pre-Buddhist Zhang-zhung Paleocultural Entity: A Comparative Analysis of Archaeological Evidence and Popular Bon Literary Sources”, in Emerging Bon: The Formation of Bon Traditions in Tibet at the Turn of the First Millennium AD (ed. H. Blezer), pp. 51–116. PIATS 2006: Proceedings of the Eleventh Seminar of the International Association for Tibetan Studies, Königswinter 2006. Halle: International Institute for Tibetan and Buddhist Studies GmbH.

_____2010. “gShen-rab Myi-bo: His life and times according to Tibet’s earliest literary sources”, in Revue d’etudes tibétaines, no. 19, pp. 31–118. Paris: CNRS.
http://www.himalaya.socanth.cam.ac.uk/collections/journals/ret/pdf/ret_19_03.pdf

_____2008. Zhang Zhung: Foundations of Civilization in Tibet. A Historical and Ethnoarchaeological Study of the Monuments, Rock Art, Texts and Oral Tradition of the Ancient Tibetan Upland. Philosophisch-Historische Klasse Denkschriften, vol. 368. Wien: Österreichischen Akademie der Wissenschaften.

_____2002a. Antiquities of Upper Tibet: Pre-Buddhist Archaeological Sites on the High Plateau. Findings of the Upper Tibet Circumnavigation Expedition, 2000). Delhi: Adroit.

_____2002b. “Gods, Hunting and Society: Animals in the Ancient Cave Paintings of Celestial Lake in Northern Tibet”, in East and West, vol. 52 (no. 1–4), pp. 347–396. Rome: IsMEO.

_____2001. Antiquities of Northern Tibet: Archaeological Discoveries on the High Plateau. (Findings of the Changthang Circuit Expedition, 1999). Delhi: Adroit.

_____2000. “Bon Rock Paintings at gNam mtsho: Glimpses of the Ancient Religion of Northern Tibet”, in Rock Art Research, vol. 17 (no. 1), pp. 35–55. Melbourne: AURA and IFRAO.

_____1997a. Divine Dyads: Ancient Civilization in Tibet. Dharamsala: Library of Tibetan Works and Archives.

_____1997b. “Notes on Three Series of Unusual Symbols Discovered on the Byang thang”, in East and West, vol. 47 (no. 1–4), pp. 395–405. Rome: IsMEO.

Bertsch, Wolfgang. 2002. The Currency of Tibet. Dharamsala: Library of Tibetan Works and Archives.

Blezer, Henk / Gurung, Kalsang Norbu / Rath, Saraju. 2013. “Where to Look for the Origins of Zhang zhung-related Scripts?”, in The Journal of the International Association for Bon Research, vol. 1, pp. 99–174. New Horizons in Bon Studies 3. Montreal: The International Association for Bon Research.
http://www.digitalhimalaya.com/collections/journals/jiabr/index.php?selection=1

Bruneau, Laurianne. 2011. “Influence of the Indian Cultural area in Ladakh in the 1st millennium AD: The Rock Inscriptions Evidence”, in Purātattva – The Bulletin of the Indian Archaeological Society, no. 41, pp. 179–190. Delhi.

Bruneau, Laurianne and Bellezza, John V. 2013. “The Rock Art of Upper Tibet and Ladakh: Inner Asian cultural adaptation, regional differentiation and the ‘Western Tibetan Plateau Style’”, in Revue d’etudes tibétaines, vol. 28, pp. 5–161. Paris: CNRS.
http://himalaya.socanth.cam.ac.uk/collections/journals/ret/pdf/ret_28.pdf

Bushell, S. W. 1880. “The Early History of Tibet. From Chinese Sources”, in The Journal of the Royal Asiatic Society of Great Britain and Ireland, New Series, vol. 12 (no. 4), pp. 435–541. Cambridge: Cambridge University Press

Carter, Martha L. 1998. “Three Silver Vessels from Tibet’s Earliest Historical Era: A Preliminary Study”, in Cleveland Studies in the History of Art, vol. 3, pp. 22–47. Cleveland Museum of Art.

Chang, Garma C. C. 1999. The Hundred Thousand Songs of Milarepa. Boston: Shambhala.

Dagkar, Namgyal Nyima Dagkar. 1999. “gShen: The Ancestral Clan of Rin then bzang po”, in The Tibet Journal, vol. 24 (no. 2), pp. 45–59. Dharamsala: Library of Tibetan Works and Archives.

Dargyay, Eva K. 1989. “Srong-Btsan Sgam-Po’s “Biography” in the Maṇi Bka’ ‘Bum”, in Acta Orientalia Academiae Scientiarum Hungaricae, vol. 43 (no. 2-3), pp. 247–257. Budapest: Akadémiai Kiadó.

Denwood, Philip. 2007. “The Tibetans in the Western Himalayas and Karakoram, Seventh-Eleventh Centuries: Rock art and inscriptions”, in Journal of Inner Asian Art & Archaeology, vol. 2, pp. 49–58. London.

_____1980. “Temple and Rock Inscriptions at Alchi”, in The Cultural Heritage of Ladakh. Zanskar and the Cave Temples of Ladakh. (by D. L. Snellgrove and T. Skorupski), vol. 2, pp. 117–163. Warminster: Aris & Phillips.

Denwood, Philip and Howard, Neil F. 1990. “Inscriptions at Balukhar and Char Zampa and Archaeological Observations on the Fort of Balukhar and its Environs”, in Indo-Tibetan Studies: Papers in Honour and Appreciation of Professor David L. Snellgrove’s Contribution to Indo-Tibetan studies (ed. T. Skorupski), pp. 81–88. Tring: Institute of Buddhist Studies.

Devers, Quentin. Forthcoming. Listing of the Historical Sites in Kargil District Carried out by INTACH, 2015-2016.

Devers, Quentin / Martin, Nils, Kacho, Khan, Mumtaz / Ldawa, Tashi. 2016. “The Petroglyphs of Kharool: The New Alchi of Ladakh”, in Flight of the Khyung, October 2016, Tibet Archaeology website.
http://www.tibetarchaeology.com/october-2016/

Dobrez, Patricia. 2014. “Hand Traces: Technical Aspects of Positive and Negative Hand-Marking in Rock Art”, in Arts, vol. 3, pp. 367–393. Basel: MDPI.
http://www.mdpi.com/2076-0752/3/4/367/pdf

Dotson, Brandon. 2012. “At the Behest of the Mountain: Clans and Political Topography in Post-imperial Tibet”, in Old Tibetan Studies Dedicated to the Memory of R. E. Emmerick. Proceedings of the Tenth Seminar of the IATS, 2003 (ed. C. Scherrer-Schaub), pp. 159–204. Leiden: Brill.

_____ 2009. The Old Tibetan Annals. An Annotated Translation of Tibet’s First History. Philosophisch-Historische Klasse Denkschriften, vol. 381. Wien: Österreichischen Akademie der Wissenschaften.

_____2008. “Complementarity and opposition in early Tibetan ritual” in Journal of the American Oriental Society, vol. 128 (no. 1), pp. 41–67. Ann Arbor: American Oriental Society.

Francke, August, H. 1928. “A Tibetan Inscription on the Darkōt Pass”, in Appendix L of Stein’s Innermost Asia, pp. 1050, 1051. Oxford.

_____1915. “The Meaning of the “Om-mani-padme-hum” Formula”, in The Journal of the Royal Asiatic Society of Great Britain and Ireland, July, pp. 397–404. Cambridge: Cambridge University Press

_____1914. Antiquities of Indian Tibet, vol. 1. Personal Narrative. Archaeological survey of India, Calcutta, Superintendent Government Printing (New imperial series, vol. 38).

_____1907. “Archaeology in Western Tibet: III-Inscriptions at Basgo and Nyemo, IV-Inscriptions at Daru, V-The Rock inscriptions at Sheh”, in Indian Antiquary, vol. 36, pp. 85–98.

_____1906a. “Archaeology in Western Tibet. Khalatse: I-Inscriptions at Khalatse; II-Inscriptions at Saspola”, in Indian Antiquary, vol. 35, pp. 237–241, pp. 325–333. Bombay.

_____1906b. “Kleine Erträger einer Missionsreise nach Zanskar”, in Zeitschrift der Deustchen Morgenlandischen Gesellschaft, vol. 60, pp. 645–661. Wiesbaden.

_____1905a. “Archaeological notes on Balu-mkhar in Western Tibet”, in Indian Antiquary, vol. 34, pp. 203–210. Bombay.

_____1905b. “The rock inscriptions at Mulbe”, in Indian Antiquary, vol. 34, pp. 72–81. Bombay.

_____1903. “Some more rock-carvings from Lower Ladakh”, in Indian Antiquary, vol. 32, pp. 361–362. Bombay.

_____1902. “Notes on rock-carvings from Lower Ladakh”, in Indian Antiquary, vol. 31, pp. 398–401. Bombay.

Gibson, Todd. 1998. “Two Figures in the Early Great Perfection”, in The Tibet Journal, vol. 23 (no. 4), pp. 72–91. Dharamsala: Library of Tibetan Works and Archives.

Hatt, Robert T. 1980. “A Thirteenth Century Tibetan Reliquary. An Iconographic and Physical Analysis”, in Artibus Asiae, vol. 42 (no. 2-3), pp. 175–220. Ascona: Artibus Asiae.

Heller, Amy. 1997. “Buddhist Images and Rock Inscriptions from Eastern Tibet, Vlllth to Xth Century, Part IV”, in Proceedings of the 7th Seminar of the International Association for Tibetan Studies. Graz 1995. Tibetan Studies, vol. 1 (eds. H. Krasser et al.), pp. 385–403. Philosophisch-Historische Klasse Denkschriften, vol. 256. Vienna: Osterreichische Akademie der Wissenschaften.

Helman-Ważny, Agnieszka. 2014. The Archaeology of Tibetan Books. Tibetan Studies Library, vol. 36. Leiden: Brill.

Henss, Michael. 2014. The Cultural Monuments of Tibet. The Central Regions. The Central Tibetan Province of Ü, vol. 1. Munich: Prestel.

Hill, Nathan, W. 2010. “An Overview of Old Tibetan Synchronic Phonology”, in Transactions of the Philological Society, vol. 108 (no. 2), pp. 110–125. London.

Iwao, Kazushi / Hill, Nathan / Takeuchi, Tsuguhito. 2009. Old Tibetan Inscriptions. Tokyo: Research Institute for Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies.

Jettmar, Karl. 2008. “Petroglyphs as Evidence for Religious Configurations?” (trans. H. van Skyhawk), in Journal of Asian Civilizations, vol. 31, pp. 65–146. Islamabad: Taxila Institute of Asian Civilizations.

_____1993. “The Patolas, their Governors and their Successors”, in Antiquities of Northern Pakistan. Reports and Studies, vol. 2, pp. 77–122. Mainz: Philipp von Zabern.

_____1990. “Exploration of Baltistan”, in South Asian Archaeology 1987. Proceedings of the Ninth International Conference of the Association of South Asian Archaeologists in Western Europe, held in the Fondazione Giorgio Cini, Island, of San Giorgio Maggiore, Venice (ed. M. Taddei), pp. 801–813. South Asian Archaeology, vol. 9, part 2. Rome: Istituto Italiano per il Medio ed Estremo Oriente.

_____(ed.). 1989. Antiquities of Northern Pakistan: Rock inscriptions in the Indus Valley, vol. 1, Mainz: Philipp von Zabern.

_____1985. “Non-Buddhist traditions in the petroglyphs of the Indus valley”, in South Asian Archaeology. in South Asian Archaeology. 1983. Papers from the Seventh International Conference of the Association of South Asian Archaeologists in Western Europe, Held in the Musées Royaux d’Art et d’Histoire, Brussels (eds. J. Schotsmans and M. Taddei), pp. 751–778. Naples: Istituto Universitario Orientale Dipartimento di Studi Asiatici, Series Minor XXIII.

_____1982. Rock carvings and Inscriptions in the Northern Areas of Pakistan. Islamabad: Institute of Folk Heritage.

Jackson, David and Janice. 1988. Tibetan Thangka Painting: Methods and Materials, illustrated by Robert Beer. Ithaca: Snow Lion Publications.

Jamgӧn Kongtrul Lodrӧ. 2012. The Treasury of Knowledge, Book Six, Parts One and Two (trans. Gyurme Dorje). Indo-Tibetan Classial Learning and Buddhist Phenomenology. Boston: Snow Lion.

Jayarava. 2011. Visible Mantra: Visualizing and writing Buddhist mantras. Cambridge: Visible Mantra Books.

Jettmar, Karl, in collaboration with Ditte König and Volker Thewalt. 1989. Antiquities of Northern Pakistan, Reports and Studies. Vol. 1, Rock Inscriptions in the Indus Valley. Mainz: Philip von Zabern.

Jettmar, Karl and Sagaster, Klaus. 1993. “Ein Tibetisches Heiligtum in Punyal”, in Antiquities of Northern Pakistan. Reports and Studies, vol. 2 (ed. K. Jettmar), pp. 123–139. Mainz: Philipp von Zabern.

Kapstein, Mathew T. 2009. “The “Commentaries of the Four Clever Men”: A Doctrinal and Philosophical Corpus in the Bon po “rDzogs chen” Tradition”, in East and West, vol. 59 (no. 1-4), pp.107–130. Rome: Istituto Italiano per l’Africa e l’Oriente (IsIAO).

_____2006. The Tibetans. Oxford: Blackwell Publishing.

_____2003. The Medieval Period: c. 850–1895”, in The History of Tibet. The Development of Buddhist Paramountcy, vol. 2, pp. 523–537. London: Routledge.

_____1997. The Royal Way of Supreme Compassion”, in Religions of Tibet in Practice (D. S. Lopez), pp. 69–76. Princeton: Princeton University Press.

Karmay, Samten G. Karmay, Samten G. 2014. The Arrow and the Spindle: Studies in History, Myths, Rituals and Beliefs in Tibet, vol. 3. Kathmandu: Mandala Book Point.

_____1997. “Inscriptions Dating from the Relgn Of Btsan po Khri Lde-srong-btsan”, in Proceedings of the 7th Seminar of the International Association for Tibetan Studies. Graz 1995. Tibetan Studies, vol. 1 (eds. H. Krasser et al.), pp. 478–486. Philosophisch-Historische Klasse Denkschriften, vol. 256. Vienna: Osterreichische Akademie der Wissenschaften.

_____1972. The Treasury of Good Sayings. A Tibetan History of Bon. Reprint, Delhi: Motilal Banarsidass, 2001.

Korpon Lobsang Khenrap. 2000. “Ru thog District before the Chinese Occupation (1900-1958): A Preliminary Study of its Cultural History”, in The Tibet Journal, vol. 25 (no. 4), pp. 33–77. Dharamasala: Library of Tibetan Works and Archives.

Kyabje Yonzin Trijang Dorje Chang Losang Yeshe Tenzin Gyatso Pal Zangpo. 1982. “The Significance of the Six Syllable Mantra OM MA NI PAD ME HUM’, in The Tibet Journal, vol. 7 (no. 4), pp. 3–10. Dharamsala: Library of Tibetan Works and Archives.

Lacouperie, Terrien de. 1885. “Beginnings of Writing in and around Tibet”, in The Journal of the Royal Asiatic Society of Great Britain and Ireland, New Series, vol. 17 (no. 3), pp. 415–482. Cambridge: Cambridge University Press.

Laurent, Yannick. 2017. “On Maṇi and Epigraphy: Four Stone Inscriptions from Spiti”, in Revue d’Etudes Tibétaines, no. 39, pp. 229–265. Paris: CNRS.
http://www.himalaya.socanth.cam.ac.uk/collections/journals/ret/pdf/ret_39_05.pdf

Ligeti, Louis 1961. “Les Inscriptions Djurtchen de Tyr: La Formule Oṃ Maṇi Padme Hūṃ”, in Acta Orientalia Academiae Scientiarum Hungaricae, vol. 12 (no. 1-3), pp. 5–26. Budpaest: Akadémiai Kiadó.

Linrothe, Rob. 2003. “Rock Art of the Himalayas, Tibet and Central Asia”, in Himalayan Art Resources. Accessed May 3, 2016: https://www.himalayanart.org/search/set.cfm?setID=785

Melnick, Alison and Bell, Christopher. N.d. “The Maṇi Kambum”, in Tibetan & Himalayan Digital Library (Thlib.org)
https://collab.its.virginia.edu/wiki/renaissanceold/Ma%E1%B9%87i%20Kambum.html

Miller, Roy A. 1976. Studies in the Grammatical Tradition in Tibet. Studies in the Theory and History of Lingusitic Science, vol. 6. Amsterdam: John Benjamins.

Mini, P. V. 2010. “Preparation techniques of pigments used for traditional mural paintings of Kerala”, in Indian Journal of Traditional Knowledge, vol. 9 (4), pp. 635–639. New Dehli: NISCAIR.
http://nopr.niscair.res.in/bitstream/123456789/10308/1/IJTK%209(4)%20635-639.pdf

Mock, John. Forthcoming. “The Red Buddha Hall Road Revisited: New Information about the Tibetan and Tang Empires in Afghanistan Wakhan”.

_____2016. “Tibetans in Wakhan: New Information on Inscriptions and Rock Art”, in Revue d’Etudes Tibétaines. no. 36, pp. 121–141. Paris: CNRS.
http://himalaya.socanth.cam.ac.uk/collections/journals/ret/pdf/ret_36_05.pdf

_____2013. “Darkot Revisited: New Information on a Tibetan Inscription and mchod-rten”, in Revue d’Etudes Tibétaines, no. 27, pp. 11–19. Paris: CNRS.
http://himalaya.socanth.cam.ac.uk/collections/journals/ret/pdf/ret_27_02.pdf

Nagano, Yasuhiko. 2014a. “Introduction”, in Issues in Tibeto-Burman Historical Linguistics (ed. Y. Nagano), pp. 1–3. Senri Ethnological Studies 75. Osaka: National Museum of Ethnology.

_____2014b. “Zhangzhung and Gyarong”, in Issues in Tibeto-Burman Historical Linguistics (ed. Y. Nagano), pp. 131–149. Senri Ethnological Studies 75. Osaka: National Museum of Ethnology.

Narkyid, Ngawangthondup. 1982. “In Defence of Amdo Gendun Chomphel’s Theory of the Origin of the Tibetan Script”, in The Tibet Journal, vol. 7 (no. 3), pp. 23–34. Dharamsala: Library of Tibetan Works and Archives.

Norbu, Namkhai, C. 2009. The Light of Kailash: A History of Zhang Zhung and Tibet, vol. 1. Merigar: Shang Shung Publications.

O’Donovan, Siofra. 2004. “The Great Perfection of Non-Sectarianism: rDzogs chen in the Bon and Buddhist Traditions of Tibet”, in The Tibet Journal, vol. 29 (no. 1), pp. 56–84. Dharamsala: Library of Tibetan Works and Archives.

Orofino, Giacomella. 1990. “A note on some Tibetan petroglyphs of the Ladakh area”, in East And West, vol. 40 (no. 1-4), pp. 173–200. Rome: IsMEO.

Pa-tshab pa-sangs dbang-’dus (Pasang Wangdui) and Glang-ru nor-bu tshe-ring (eds.). 2007. Gtam shul dga’ thang ’bum pa che nas gsar du rnyed pa’i bon gyi gna’ dpe bdams bsgrigs. Lhasa: Bod-ljongs bod-yig dpe-rnying dpe-skrun khang.

Petech, Luciano. 1977. The Kingdom of Ladakh c. 950-1842 A. D. Oriental Series, vol. 51. Rome: IsMEO.

Reynolds, John M. 2005. The Oral Tradition from Zhang-Zhung: An Introduction to the Bonpo Dzogchen Teachings of the Oral Tradition from Zhang-zhung known as the Zhang-zhung snyan-rgyud. Kathmandu: Vajra Publications.

Richardson, H. E. 1998. High Peaks, Pure Earth. Collected Writings on Tibetan History and Culture. London: Serindia Publications.

_____1985. A Corpus of Early Tibetan Inscriptions. London: Royal Asiatic Society.

Schaik, Sam van. 2016. “Tibetan Buddhism in Central Asia: Geopolitics and Group Dynamics”, in Transfer of Buddhism Across Central Asian Networks (7th to 13th Centuries) (ed. C Meinert), pp. 57–81. Dynamics in the History of Religions, vol. 8. Leiden: Brill.

_____2014. “Towards a Tibetan Paleography: A Preliminary Typology of Writing Styles in Early Tibet”, in Manuscript Cultures: Mapping the Field (eds. J.-U. Sobisch and J. B. Quenzer), pp. 299–337. Berlin: de Gruyter.

_____2013. “Dating Early Tibetan Manuscripts: A Paleographical Method”, in Scribes, Texts, and Rituals in Early Tibet and Dunhuang: Papers based on Conference Panel “Old Tibetan Studies III” at the 12th Seminar of the International Association for Tibetan Studies, Vancouver, August 2010 (eds. B. Dotson et al.), pp. 119–135. Wiesbaden: Ludwig Reichert.

_____2012. “The Origin of the Headless Script (dbu med) in Tibet” in Tibeto-Burmese Linguistics (ed. N. Hill), pp. 331–369. Leiden: Brill.

_____2011. “A New Look at the Tibetan Invention of Writing”, in New Studies of the Old Tibetan Documents: Philology, History and Religion (eds. Imaeda et al.), pp. 45–96. Old Tibetan Documents Monograph Series, vol. 3. Tokyo: Research Institute for Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies.

Scherrer-Schaub, Cristina, A. 2014. “A Perusal of Early Tibetan Inscriptions in the Light of the Buddhist World of the 7th to 9th Centuries A.D.”, in Epigraphic Evidence in the Pre-Modern Buddhist World: Proceedings of the Eponymous Conference Held in Vienna, 14-15 Oct. 2011 (ed. K. Tropper). Wiener Studien zur Tibetologie und Buddhismuskunde, vol. 85. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien.

_____2013. “Classifying, Questioning and Interpreting Tibetan Inscriptions”, in Tibetan Inscriptions: Proceedings of a Panel Held at the Twelfth Seminar of the International Association for Tibetan Studies, Vancouver 2010 (eds. K. Tropper and C. Scherrer-Schaub), pp. 139–169. Brill’s Tibetan Studies Library Series, vol. 32. Leiden: Brill.

_____2002. “Enacting Words: A Diplomatic Analysis of the Imperial Decrees (Bkas Bcad) and their Application in the Sgra Sbyor Bam Po Gñis Pa Tradition”, in Journal of the International Association of Buddhist Studies, vol. 25 (no. 1-2), pp. 263–340.

_____1999. “Towards a methodology for the study of old Tibetan manuscripts: Dunhuang and Tabo.”, in Tabo Studies II: Manuscripts, Texts, Inscriptions and the Arts (eds. C. A. Scherrer-Schaub and E. Steinkellner), pp. 3–36. Rome: Is.I.A.O.

Scherrer-Schaub, Cristina A. and Bonani, George. 2002. “Establishing a typology of the old Tibetan manuscripts: A multidisciplinary approach.”, in Dunhuang Manuscript Forgeries (ed. S. Whitfield), pp. 184–215. London: The British Library.

Schuh, Dieter. 2013. “Tibetischen Inschriften ins Maul geschaut: Beobachtungen zu Stein- und Felsinschriften sowie den Schriften des 7. bis 9. Jahrhunderts in Tibet”, in Nepalica-Tibetica Festgabe for Christoph Cüppers, vol. 2 (eds. F.-K. Ehrhard and P. Maurer), pp. 143–184. Beiträge zur Zentralasienforschung, vol. 28. Andiast: International Institute for Tibetan and Buddhist Studies GmbH.

Schwieger, Peter. 1996. “The Lineage of the Noble House of Ga-zi in East Tibet”, in Kailash, vol. 18 (no. 3,4), pp. 115–132.

Sha-bo mkha’-byams. 2015. “Ri bo hi ma la ya’i nub rgyud du gnas pa’i bod btsan po’i skabs kyi rdo brkos yig rnying khag gcig gleng ba”, in Bod ljongs zhib ’jug (Tibetan Studies), 2016.1, pp. 68–75. Lhasa.

Snellgrove, David. and Richardson, H. E. 1968. A Cultural History of Tibet. Reprint, Boulder: Prajñā Press, 1980.

Sørensen, Per K. 1994. Tibetan Buddhist Historiography. The mirror Illuminating the Royal Genealogies: An Annotated Translation of the XIVth Century Tibetan Chronicle: rGyal-rabs gsal-baʾi me-long. Asiatische Forschungen: Monographienreihe zur Geschichte, Kultur und Sprache der Völker Ost- und Zentralasiens, vol. 128. Weisbaden: Harrassowitz.

Suolang Wangdui (Bsod-nams dbang-’dus). 1994. Art of Tibetan Rock Paintings. Introduction by Li Yongxian and Huo Wei. Chengdu: Sichuan People’s Publishing House.

Takeuchi, Tsuguhito. 2013. “Formation and Transformation of Old Tibetan”, in Journal of Research Institute: Historical Development of the Tibetan Languages, vol. 49, pp. 3–17. Kobe: Kobe City University of Foreign Studies.

_____2012a. “Old Tibetan Buddhist Texts from the post-Tibetan Imperial Period (mid-9 C. to late 10 C.)”, in Old Tibetan Studies Dedicated to the Memory of R. E. Emmerick. Proceedings of the Tenth Seminar of the IATS, 2003 (ed. C. Scherrer-Schaub), pp. 205–214. Leiden: Brill.

_____2012b. “Old Tibetan Rock Inscriptions Near Alchi”, in Historical Developments in Tibetan Languages. Proceedings of the workshop B of the 17th Himalayan Languages (Kobe 6th-9th, September, 2011. Symposium) (eds. T. Takeuchi and N. Hayashi). Journal of Research Institute, vol. 49, 2012. pp. 29–71. Kobe: Research Institute of Foreign Studies, Kobe University of Foreign Studies.

Thakur, Laxman, S. 2001. Buddhism in the Western Himalaya: A Study of Tabo Monastery. Delhi: Oxford University Press.

_____1997. “Tibetan Historical Inscriptions from Kinnaur and Lahaul-Spiti: A Survey of Recent Discoveries”, in Proceedings of the 7th Seminar of the International Association for Tibetan Studies. Graz 1995. Tibetan Studies, vol. 2 (eds. H. Krasser et al.), pp. 967–979. Philosophisch-Historische Klasse Denkschriften, vol. 256. Vienna: Osterreichische Akademie der Wissenschaften.

Thomas, F. W. 1933. “Tibetan Documents concerning Chinese Turketan. VI: The Tibetan Army”, in Journal of the Royal Asiatic Society of Great Britain and Ireland, no. 3 (1933), pp. 537–568. Cambridge: Cambridge University Press.

Tshe-ring chos-rgyal and Zla-ba tshe-ring. 1994. “Gsar du brnyed pa’i spu hreng gi spyan ras gzigs kyi rdo ring las byung ba’i gtam dpyad”, in Gangs ljongs rig gnas, vol. 2, pp. 4–20. Lhasa.

Tucci, Giuseppe. 1977. “On Swāt. The Dards and Connected Problems”, in East and West, vol. 27 (no. 1-4), pp. 9–103. Rome: Istituto Italiano per l’Africa e l’Oriente (IsIAO).

Uebach, Helga. 1999. “On the thirty-seven holy places of the Bon-pos in the Tibetan Empire” in Indica et Tibetica. Studia Tibetica et Mongolica (Festschrift Manfred Taube) (eds. H. Eimer, M. Hahn, M. Schetelich, and P. Wyzlic), vol. 34, pp. 261-277. Swisttal-Odendorf: Indica et Tibetica Verlag.

Uray, Geza. 1972a. “Queen Sar-Mar-Kar’s Songs in the Old Tibetan Chronicle”, in Acta Orientalia Academiae Scientiarum Hungaricae, vol. 25, pp. 5–38. Budapest: Akadémiai Kiadó.

_____Uray, G. 1972b. “The Narrative of Legislation and Organization of the Mkhaspa’i Dga’-ston. The Origins of the Traditions Concerning Sroň-Brcan Sgam-po as First Legislator and Organizer of Tibet”, in Acta Orientalia Academiae Scientiarum Hungaricae, vol. 26 (no. 1), pp. 11–68. Budapest: Akadémiai Kiadó.

_____1955. “On the Tibetan Letters “Ba” And “Wa”: Contribution to the Origin and History of the Tibetan Alphabet”, in Acta Orientalia Academiae Scientiarum Hungaricae, vol. 5 (no. 1-2), pp. 101– 122. Budapest: Akadémiai Kiadó.

Venturi, Frederica. 2008. “An Old Tibetan Document on the Uighurs: A New Translation and Interpretation”, in Journal of Asian History, vol. 42 (no. 1), pp. 1–35. Wiesbaden: Harrassowitz.

Vitali, Roberto. 1996. The Kingdoms of Gu.ge Pu.hrang. According to mNga’.ris rgyal.rabs by Gu.ge mkhan.chen Ngag.dbang grags.pa. Dharamsala: Tho.ling gtsug.lag.khang lo.gcig.stong ’khor.ba’i rjes.dran.mdzad sgo’i go.sgrig tshogs.chung.

Walter, Michael, L. and Beckwith, Christopher, I. 2010. “The Dating and Interpretation of the Old Tibetan Inscriptions”, in Central Asiatic Journal, vol. 54 (no. 2), pp. 291–319. Weisbaden: Harrassowitz.

Wangdu, Pasang and Diemberger, Hilda. 2000. dBa’ bzhed: The Royal Narrative Concerning the Bringing of the Buddha’s Doctrine to Tibet. Philosophisch-historische Klasse, Denkschriften, vol. 291. Wien: Österreichischen Akademie der Wissenschaften.

Zeisler, Bettina. 2009-2010. “East of the Moon and West of the Sun? Approaches to a Land with Many Names, North of Ancient India and South of Khotan. in The Tibet Journal, vol. 34 (no. 3) to vol. 35 (no. 2), pp. 371–463. Dharamsala: Library of Tibetan Works and Archives.

Zla-ba tshe-ring / Suo Wenqing / Dbang-’dus / Bsod-nams dbang-ldan-a. 2000. Precious Deposits: Historical Relics of Tibet, China. Prehistoric Age and Tubo Period. Beijing: Morning Glory Publishers.

Zla-ba tshe-ring / Suo Wenqing / Dbang-’dus / Bsod-nams dbang-ldan-b. 2000. Precious Deposits: Historical Relics of Tibet, China. The Period of Sepratist Regimes. Beijing: Morning Glory Publishers.

Previous Post