Articles by alphabetic order
 Ā Ī Ñ Ś Ū Ö Ō
1 2 3 4 5 6 7 8 9 0

Geographical and Other Borders in the Symbolism of Padmasambhava

From Chinese Buddhist Encyclopedia
Jump to navigation Jump to search

Robert Mayer,

Ruhr-Universität, Bochum, & University of Oxford

Abstract: The mythologised founder of Tibetan Buddhism, Padmasambhava, was said to be native to Uḍḍiyāna, a geographical border region where Tibetan and Indian civilizations were in close proximity. But was he historical, or not? And either way, what might the significance have been of his origins in a border area? If the Dunhuang text PT44 is to be trusted, not only was Padmasambhava born in a border area, but the initial flourishing of his school appears to have been concentrated not predominantly in Tibet proper, but more along Tibet’s southern borderlands. Might this represent a deliberate choice to evade state controls? For the transgressive types of tantrism, with which Padmasambhava was associated, were officially restricted by the Tibetan Empire, and their texts excluded from the public translation program. Dunhuang references describe Padmasambhava enriching

his received Tantric Buddhism with indigenous Himalayan or Tibetan symbols, deities, and ritual forms, to create a new more localised Tantric Buddhism. While this implies a prior understanding of and familiarity with Himalayan cultures, it too is a development unlikely to have been approved by the State. Turning to more symbolic interpretations, Padmasambhava narratives are quite unique in Tibetan literature, for the degree to which they straddle the boundaries of history and ritual, where Padmasambhava narratives can represent narrated ritual sequences. In addition, more than any other figure within Tibetan religion, Padmasambhava is associated with the straddling of various taxanomic boundaries; yet no one inspires more widespread devotion and confidence. In this sense, Padmasambhava perhaps resembles a protective chimera, as described by the anthropologist of Tibet, Charles Ramble.

Bio: In the academic years 2017-19, Robert Mayer began contributing to a research project with several colleagues on Nyang ral Nyi ma’i ‘od zer, the great architect of Tibet’s Padmasambhava narratives, based at the Ruhr Universität, Bochum, and funded by the DFG. At the same time, he continues to be University Research Lecturer in Tibetan  and Himalayan Studies, The Oriental Institute, University of Oxford, where he has been since 2002.

Slide 1: Title page, with Padmasambhava thangka

In a deeply religious culture that has accumulated more mantras and prayers than anyone could ever enumerate, perhaps the best known prayer of all is a short verse of seven lines addressed to Padmasambhava, the legendary eighth century conversion hero and founder of Tibetan Buddhism. Its current phrasing was probably codified in the 13th century (although a single line of it has turned up within a 10th century document from Dunhuang, complete with a marginal note connecting it to Padmasambhava, contained within a longer verse of praise to Padmasambhava that became incorporated into the later rNying ma tradition, see Cantwell and Mayer 2012, 91ff). The prayer’s opening line is pertinent to my theme of borders, because it announces that Padmasambhava came to Tibet from “the north-west borders of the land of Uḍḍiyāna.”

This paper is a loosely edited version of a public talk given at the symposium “Moving Borders: Tibet in Interaction 1 with its Neighbors”, at the Asia Society, New York, May 4th-5th, 2018.

2 A great deal has been said over the last ten years about the inhabitants of borderlands. Pertinent here are Willem van Schendel’s idea of ‘Zomia’, and James C Scott’s development of it (Scott 2009) to describe various groups of Asian highlanders, as ‘state evading’ peoples, content to remain within remote and rugged highland fastnesses, an

usually foregoing literacy, if such stratagems might enable them to evade the oppressive power of the states centred on the lowlands. Scott was referring especially to South East Asia and the Eastern Himalayas, specifically excluding many Tibetan areas because he was aware that Tibetans did manage to establish effective highland governments. David Gellner, however, has suggested the extension of the idea to the territories of the modern day Pakistan and Afghanistan (Gellner 2013: 4), where the historic Uḍḍiyāna was once located.

We have almost no empirical evidence to test Scott’s hypothesis in relation to 8th century Uḍḍiyāna, nor 8th century Tibet for that matter, so we will not even try. Moreover, central Uḍḍiyāna itself, especially in its earlier history, seems to have been a centre of literate learning, and thus a far cry indeed from the Zomia of Scott’s hypothesis. Yet it is difficult not to be reminded at least of something vaguely analagous to Scott’s theory (albeit pertaining here to religious groups rather than ethnic groups), when one reflects how the growth of trangressive tantrism, both in India and Tibet, 2 seems to have been linked historically with situations or locations of weak state control. For

however sparse the available evidence for late 8th and 9th century Tibet might be, however uncertain all our speculations must necessarily remain, the few surviving clues we do have, could be interpreted as pointing to the deliberate evasion of state controls by the early school of Padmasambhava, through geographical isolation. As Matthew Kapstein has pointed out (Kapstein 2000: 155-160), in a suggestion eventually adopted by most other scholars (even if they don’t always credit him), even if the historical Padmasambhava

might have had a brief contact with the Emperor and his court, the Dunhuang text PT44 locates the main activities of Padmasambhava and his earliest school, not within Tibet proper among the Imperial court and its ministers, but along Tibet’s remote Himalayan southern fringes, where government control was quite likely less uniform. And one among several reasons for being located in such spiritually potent but politically 3 more remote regions, might have been precisely to evade the restrictions placed on religious expression by the Tibetan state. For while the Tibetan state had enthusiastically embraced various more exoteric forms of tantrism, it had also quite explicitly restricted

their own signature style of transgressive non-dual tantrism, as we know from an authentic Imperial era document, the sGra sbyor bam po gnyis pa, and again reasserted in another somewhat later text, the Dba’ bzhed. 4

In similar theme, we should also recall that it was only after the mid ninth century collapse of the Tibetan state, that transgressive tantric Buddhism began to flourish much more widely across the entire plateau, eventually dominating almost all Tibetan areas, as it continues to do today. Likewise, the emergence of the highly transgressive yoginītantras during the 9th century Pāla decline in North

The distinction between non-transgressive and trangressive forms of tantrism is a crucial one, but has not always been 2 properly articulated in some recent literature. While non-transgressive tantrism was often highly amenable to state purposes, for example the Mahāvairocanatantra in Imperial Tibet, other kinds of tantrism, such as mahāyoga, were restricted or forbidden, and regarded as subversive of morality and the state. The southern regions associated with the early school of Padmsambhava in PT 44, such as Lho brag

(in modern 3 southern Tibet), Bum thang (in modern Bhutan), and also Yang le shod (in modern Nepal), have long been considered spiritually powerful, some of them likened to a spiritual omphalos or navel, and the home of important regional deities such as the Tshe ring mched lnga. These southern regions are associated with numerous seminal spiritual figures, including not only Padmasambhava and his early school, but also Mar pa, Milarepa, Nyang ral, Guru Chos dbang, the g.Yung drung Bon master Khu tsha zla ‘od, Klong chen pa, and many more. For a more detailed discussion of these issues, see Mayer 2015.4

3 East India, was again associated with a collapse of state controls over religion (Szántó 2016: 397-8). 5 If such reflections on if, and why, the early Padmasambhava school might have been concentrated on the remote southern borders of Tibet must remain largely speculative, unfortunately, the current state of knowledge on some other issues relating to Padmsambhava are barely any less speculative.

As the object of so much devotion for so long, Padmasambhava inevitably became hugely mythologised. Even the most basic narratives of his lifestory soon, perhaps even in his own time, came to diverge from knowable history and enter the realm of religious myth. Hence some scholars have doubted his historicity. Most however now agree with Matthew Kapstein, who suggested that in the light of the surviving data, our least complicated option is to accept that behind the luxuriant devotional myths was a real historical figure, active predominantly on the southern Tibetan borderlands, with only occasional visits to more central locations. For things become even more complicated for the historian, if we try to shoehorn the surviving data into the theory that Padmasambhava was an entirely mythic figure, and Kapstein’s initial hypothesis of 20 years ago (Kapstein 2000: 155-160) has since been supported by data more recently come to light.

This is because, from a range of early sources, we can see that for many Tibetans especially before the 12th century, Padmasambhava seems to have been just one Indian tantric master among many others, not noticeably pre-eminent. For example, the bSam gtan mig sgron of gNubs chen Sangs rgyas ye shes mentions Padmasambhava, but cannot be considered a Padmasambhava-centric text, likewise the autobiography of the Kaḥ thog-educated Karma Pakshi. This makes it quite difficult to 6 argue that Padmasambhava was entirely mythic, since there is no suggestion that the other tantric masters among whom he is mentioned in suchlike works, were simply mythical.

Describing the North East of India in the mid-9th century, Szántó writes (p.398): “In other words, the situation is 5 similar to the one in Tibet. We have a collapsed centralized empire, lack of large-scale patronage for Buddhism, and no institutional control over religious matters. But developments in religion did not come to a halt under such circumstances. On the contrary, the Dunhuang documents testify to very vigorous activity in post-imperial times, whereas in India this is the time for the rise of the yoginītantras, which will become very popular during the phyi dar.” For example, gNubs Sangs rgyas ye shes, in his bSam gtan mig sgron, portrays Padmasambhava as a great teacher and 6 even mythologises him, but not in any obvious way more so than his various peers such as Vimalamitra, so here Padmasambhava is apparently not unique, but simply one amazing Indian siddha among others. See especially bSam gtan mig sgron’s Chapter Six on Mahāyoga, where nearly all its references to Padmasambhava or his works occur. According to Dylan Esler, gNubs refers to Padmasambhava’s Man ngag lta ba’i phreng ba on pages 192.4, 196.7, and 207; he also cites the Thabs zhags on pages 264, 271, and 289;

and there are other references to Padmasambhava in Chapter Six on pages 193, 194, 210, 222, and 277. A further reference to Padmasambhava occurs in the bSam gtan mig sgron’s Chapter 1, on page 11; yet none of these references indicate any particular Padmasambhava-centric orientation, and Padmsambhava is often mentioned as an equal alongside other authorities, such as Buddhagupta, Madhusādhu, or Vimalamitra. The copious rDzong ‘phrang srog literature, especially the versions preserved in the Bka’ ma shin tu rgyas pa (Vol Ha 2 and Vol Ye 114, 415-622), is also connected to gNubs Sangs rgyas ye shes (under the name of Yang dbang gter), and quite probably preserves some genuinely old

traditions. Here too, in the context of discussions of the bKa’ brgyad, we see no clear evidence of any avowedly Padma-centric orientation, and Padmasambhava is mentioned as one siddha among others (personal communication, Cathy Cantwell and Lopon Ogyan Tanzin). Likewise, as Charles Manson informs me (personal communication, 12th January, 2019), Karma Pakshi, who was educated at Kaḥ thog in East Tibet as late as the 1220’s, and wrote copiously on rNying ma tantras and on Mahāyoga, only rarely mentions Padmasambhava in

his autobiography: the ocassions are on pages 3, 9, 20 (twice), 35, 60, and 107. Yet, as Manson points out, there is no indication at all that Karma Pakshi is in any way Padmasambhava-centric, and for example, he cites Indrabhuti more times than he cites Padmasambhava. As far as I understand Matthew Kapstein, the same is true of a mid 12th century text of the Zur clan from gTsang in Central Tibet, that he is currently working on: they too are not in any obvious way Padmasambhava-centric. It would appear that there were once some circumstances where lamas could be rNying ma oriented in tantric practice, yet without being Padmasambhava-centric, and where Padmasambhava was understood more as one historical siddha among others, and not as a unique second Buddha.

4 Conversely, survivals from Dunhuang show that during the same period, for another group of particular Padmasambhava devotees that we think was probably mainly centred along Tibet’s southern border, he was a special guru of extraordinary accomplishment, singled out from his peers. Thus our central historical question is to figure out how the views of the particular Padmasambhava devotees, possibly centred along Tibet’s southern borders, eventually came to dominate the whole of Tibet, and it is here that the name of Nyang ral Nyi ma’i ‘od zer comes very much to the fore. But whether they describe Padmasambhava as one guru among many, or as a completely unique realised being, all agree he came from Uḍḍiyāna.

Slide 2: map of Swat and Tibet

Since Tucci, most academic scholars agree ancient Uḍḍiyāna was centred on the modern-day Swat valley of Pakistan. Swat is a long river valley, at the Western end of the Himalayas, in the foothills 7 of the Hindukush-Karakorum ranges. Its average altitude is 3,000 ft, but it is encircled by mountains up to 18,000 feet. More recently, scholars have envisaged a broader definition of a greater Uḍḍiyāna, including Gilgit to the north east. In that case, the old prayer might imply Gilgit as Padmasambhava’s birthplace. As the crow flies, Swat is only a few hundred kilometres from the Tibetophone world, while Gilgit’s dramatic mountain scenery directy adjoins Baltistan, where a dialect of Tibetan is still spoken. So Padmasambhava was either a native of Swat, where Tibetan and Indian Buddhist civilisations had already met and traded for hundreds of years; or of Gilgit, an immediate neighbour of Tibet.

But historically, Uḍḍiyāna had been much more: it was also a geographical and cultural crossroads where, over many centuries, Central Asians, Persians, and Greeks, had also left deep imprints. Its rich heritage of Buddhist artifacts have captivated generations of archaeologists and art historians, most famously the Greek-influenced Buddha statues that began to be produced in the region after the 1st century of the Common Era.

Uḍḍiyāna was a truly major centre for Buddhism, so renowned, that Buddhist pilgrimage from China and Tibet continued long after its actual Buddhist institutions had collapsed. In earlier centuries, Uḍḍiyāna, and the Gandhara region as a whole, had been key centres for the development and propogation of Mahāyāna Buddhism, where it was richly patronised by the Kuṣāṇa dynasty. They had originated in Central Asia, but extended their dominions across the North Western parts of the Indian sub-continent as well. It was from their territories that Buddhism spread through Central Asia to China. Mahāyāna Buddhists everywhere still remember the Kuṣāṇa Emperor Kaniṣka as a mighty patron, who organised a historic religious council, and built a fabulous stupa. All this is still recorded in traditional Tibetan learning.

Alexis Sanderson (2007) has revisted the issue of the location of Uḍḍiyāna. He takes note of the various far-flung 7 locations that have been identified with Uḍḍiyāna at different times and by diffferent sources (eastern India, the far south of India, etc), but comes to the conclusion, drawn from a variety of old textual citations, that it is located near Kashmir. See the section entitled ‘Uḍḍiyāna and Kashmir’, contained in pages 265-269 of his article ‘The Śaiva Exegesis of Kashmir’, in Mélanges tantriques à la mémoire d’Hélène Brunner. Tantric Studies in Memory of Hélène Brunner, Collection Indologie 106, EFEO, Institut français de Pondichéry (IFP), ed. Dominic Goodall and André Padoux, 2007.

But by Padmasambhava’s time of the late 8th century, Uḍḍiyāna’s and the wider region’s Mahāyāna tradition had long declined. In its place emerged tantric Buddhism, shortly to be followed by a remarkable and closely interconnected local school of philosophically non-dual Śaivism. So just as Uḍḍiyāna had been a great Mahāyāna centre from the first through fifth centuries, following the fifth and sixth centuries it had developed into an equally great symbolic centre for tantric religion, both Buddhist and Śaiva. Contemporary scholars like Deborah Klimburg-Salter, Anna Filigenzi, and Luca Maria Olivieri, are now hopeful that they are making more substantial discoveries about this later tantric phase of the wider region’s art and archaeology.

Padmasambhava’s charismatic appeal was clearly enhanced by the singular mystique of his homeland. So much so that his great 13th century devotee, Guru Chowang, can occasionally just refer to Padmasambhava as U rgyan, which is the Tibetan pronounciation of Uḍḍiyāna; and other 9 Tibetans have done the same. While U rgyan could be taken as short for U rgyan Rinpoche, meaning the Precious Guru from Uḍḍiyāna, or U rgyan Pema, meaning Padmasambhava from Uḍḍiyāna, nevertheless it’s a little bit as though Western devotees of the Dalai Lama were, in a spirit of great respect, simply to call him ‘Tibet’. U rgyan remains a popular Tibetan personal name.

Slide 3: [so-called “Siddha” 8th century stele, Swat. After A Filigenzi 2015, p.162 & Padmasambhava, 13th century painting, Tibet. HAR 160] And we also might see the Uḍḍiyāna association represented iconographically: Anna Filigenzi suggests that Padmasambhava’s distinctive visual representation in Tibetan art is

modelled on an 8th century Uḍḍiyāna iconography of what she calls the siddha, or realised tantric practitioner. While I am deeply sceptical that the image actually represents Padmsambhava (as some Buddhists seem to claim), if the detailed measurements and analysis Filigenzi made are correct and it is really holding a skull and not a begging bowl, I think it might represent a tantric figure of some type, although not necessarily Buddhist. What for me is much more significant is that Tibetan artists might well have interpreted similar iconography from Uḍḍiyāna as their prototype for representing Padmasambhava.

Padmasambhava’s popularity with Tibetans can only have been helped by the association, as a native of Uḍḍiyāna, with highland cultures, and proximity to Tibetan civilisation. Filigenzi (2015, p. 18) describes a vital, polycentric, cultural world stretching across the Hindu Kush-KarakoramHimalaya region, of which, in the eighth century, both Uḍḍiyāna and much of Tibet would have been part. At that time, Uḍḍiyāna lay in the shadow of the powerful Tibetan empire’s expansions into Central Asia. As Deborah Klimburg-Salter (2016, p. 392) points out, “From the 7th century parts of Northern Pakistan constituted the western borderland of the Tibetan Empire.” There can be

Including those connected with Deborah Klimburg-Salter’s fine Giuseppe Tucci exhibition being opened at the Asia 8 Society on the night I originally gave this lecture, and all who have worked so closely with the Italian Archaeological Mission founded by Giuseppe Tucci such as Anna Filigenzi and Luca Maria Olivieri adding to the considerable work they already achieved in their studies of the earlier Greek-influenced Gandharan Buddhist art. In his gTer ‘byung chen mo, Guru Chos dbang refers to Padmasambhava in a number of different ways, sometimes just 9 using the short form of U rgyan (see for example page 135 line 5; page 146 line 1; and page 150 line 2). Elsewhere, he calls him by a variety

of other names, for example, U rgyan Padma (page 93 line 7), Slob dpon Padma saṃ bha (page 109 line 2), and Guru Padma (132 line 3). I am not so far aware of any instances or versions of the form Padmākara in the gTer ‘byung chen mo, although I have not made any systematic search. Yet the form Padmo 'byung gnas does occur in gNubs sangs rgyas ye she’s earlier work, the bSam gtan mig sgron, on page 277.4.

6 little doubt that an educated interested native of 8th century Uḍḍiyāna, whether from Gilgit or Swat, would be acutely aware of Tibet, its Empire, and its culture. And ancient documents do seem to indicate a high degree of cultural understanding contributing to Padmasambhava’s remarkable impact upon the Tibetan imagination. Four of the earliest extant manuscripts to describe him come from Dunhuang. Two might have been inscribed in the tenth and one in the early eleventh centuries, and the fourth I am not clear about, but we suspect all are copies of older texts. Although written at different times and in different hands, three of them portray Padmasambhava incorporating

indigenous Himalayan deities, symbols, and ritual structures, into tantric Buddhism (the fourth text, IOL Tib J 644, although full of valuable data, has little iconographic content, and although it has some ritual content, it is not of a relevant kind, and moreover has not yet been exahustively analysed, so at the moment it offers us little for the present discussion). The three examples we can currently use from among these remarkable old 10 manuscripts seem to bear witness to a deliberate policy of Padmasambhava’s

school (and who knows, perhaps of Padmasambhava himself), to assimilate Indian Buddhism to Tibetan culture. They offer us direct, contemporaneous windows onto quite early stages of a long process that over the next three centuries was to issue into a distinctively Tibetan form of Vajrayāna, focused on Padmasambhava as its founding guru, and exhibiting a significant hybridity in iconography, narrative, and ritual grammar. Nowadays we call this the rNying ma pa, or the Ancient School of Tibetan Buddhism. Let’s look at all three Dunhuang sources we are consulting here in turn.

[Slide 4 PT 307. Bibliothèque nationale de France]

The first text is known by its Bibilothèque Nationale de France catalogue number, Pelliot tibétain 307, or PT307 for short. It is a narrative text for use in ritual, describing Padmasambhava taming a group of indigenous goddesses and converting them into protectors of Buddhism, in this way inducting them into the tantric Buddhist pantheon. All but one of the goddessses’s names in PT307 are clearly recognisable as those used for a set nowadays known as the brTan ma bcu gnyis, or the Twelve Established Goddesses.

Jacob Dalton first encountered this text while engaged in cataloguing the Dunhuang materials for the British Library 10 and published an article on it at that time (Dalton 2004). As he explains in his lecture at the Rubin Musem on October 14th, 2018, when working on the manuscript he missed a crucial reference to Padmasambhava’s name, which

occurred on folio 2 verso, line 7 (slobs pon pad ma sam ba…). Consequently, he concluded in his published article that IOL Tib J 644 did not directly refer to Padmasambhava by name at all and confined himself to noting links between a narrative earlier in the text and later depictions of Padmasambhava. In his article of 2004, he also tried to argue that Padmasambhava is in these Dunhuang texts portrayed as a comparatively inconsequential person, who was not yet much venerated or mythologised. Cantwell and Mayer 2016 (first published in 2013) was intended to challenge that analysis and demonstrate how Padmasambhava is in fact uniquely venerated and mythologised in the Dunhuang texts where

he is mentioned (PT 44, IOL Tib J 321, PT 307, and now also IOL Tib J 644). We are delighted that Dalton is now following other Padmasambhava scholars in accepting this more recent understanding (Dalton, this volume). However, I explain below why I think Dalton might still remain mistaken in his assessment of IOL Tib J 321 (see note 22 below). I should add, Cantwell and Mayer 2016 (2013) did not reject the testimony of IOL Tib J 644 outright, but suggested that IOL Tib J 644 be re-examined, since we suspected Dalton

2004’s analysis was not yet conclusive, and that he should return to it. Hence, we wrote “For present purposes [of Padmasambhava historical analysis], we are best advised to leave it [[[discussion]] of IOL Tib J 644] aside, pending further investigation”, and “..until a better analysis [of IOL Tib J 644] is achieved, [it] cannot be reliably taken as evidence [for Padmasambhava]..”. Until the extremely important historical testimony of IOL Tib J 644 has been more systematically interrogated, including its reference to Padmasambhava by name, that caveat must remain in place.

7 [Slide 5 Twelve Established Goddesses. HAR 99122 ]

These twelve remain important, notably as protectors of Tibet’s borders. rNying ma pas propitiate them in daily protector rites, and also give them a special niche within their tantric feast offerings. They are propitiated by the other schools too, and the failure of the Tibetan government adequately to perform their rites, is sometimes said to have beeen a factor in permitting the Chinese Communists to invade Tibet.

PT307 describes Padmasambhava and his disciple, rLang dPal gyi Seng ge, taming these Tibetan goddesses through binding them by oath, whilst gratifying them with remainder offerings. To this 11 day, the very same narrative appears in rituals. A particular offering to the Twelve Established Goddesses must be made near the end of every rNying ma t

antric feast offering, immediately following the general offering of remainders to the protectors. But before their particular offering can be made, a reminder must be narrated of the original covenant they made with Padmasambhava and rLang dPal gyi Seng ge back in the 8th century, which was the original precedent for the rite. Only after this rehearsal can the Twelve Established Goddesses receive their offerings. I find it fascinating how rLang dPal gyi Seng ge still continues to figure as Padmasambhava’s assistant in these ritual narratives, along with other details from PT307. 12

This illustrates another interesting feature of PT307: it is not structured in an Indian Buddhist manner. Rather, it follows a ubiquitous pre-Buddhist ritual narrative template called a rabs or smrang. Previous Tibetologists have written a great deal about the rabs (see for example Karmay 1998, pages 245ff and 288ff), so I won’t go into great detail here, except to say it is understood as one of the most distinctive signatures of pre-Buddhist Tibetan ritual. It entails a very particular way of integrating myth with ritual,

by narrating (often quite briefly) an original precedent for a rite, naming the legendary officiants who first did the rite, and the exact reasons and circumstances in which they did it. In this way, subsequent performers can know how and why the rite will be appropriate and effective for them. The indigenous rabs format has been accepted into rNying ma Buddhism. It often occurs embedded within Indian-style tantric rituals: For example, it is intrinsic to regular feast and protector offerings as I have just described. Longer extended forms of rabs can also be recited to introduce tantric initiations, or in pedagogical contexts, such as the opening teachings for a long tantric retreat. 13

[Slide 6: PT 44. Bibliothèque nationale de France ]

The second document, PT44, is also a rabs, and like PT307, continues to be widely used in rNying ma narrative and ritual.

It describes Padmasambhava at the Asura cave in Yang le shod in Nepal, subduing a set of four dangerous goddesses native to the Himalayan foothills. Once they are subdued, Padmasambhava procures a complete set of the Vajrakīlaya tantras from Nālandā monastery in India. He dam tsig kha chu la dga’/ mchod pa lhag la dgyes so; in Dalton’s translation (Dalton 2004): ‘They rejoice in the 11 saliva from vows. They are pleased by the remainder offerings.’ In addition, rLang dPal gyi Seng ge continues to have other ritual and narrative roles, for example, as the visionary 12 inspiration for ‘Jigs med gling pa’s famous rGyud lugs phur pa cycle. For example a narrative of Padmasambhava’s first opening of the Phur pa mandala and taming unruly deities at the 13 Asura Cave in Yang le shod,

similiar to that of PT44, will be narrated before many Phur pa retreats. In this context, the rabs is often called a Phur pa lo rgyus; see discussion immediately below. 8 comprehensively redacts these tantras, integrating the four goddesses he has just tamed into their maṇḍalas as protectors. He then teaches Vajrakīlaya to a few highly accomplished disciples at various locations along Tibet’s southern borders. And as though to confirm this narrative, an old Vajrakīlaya tantra called the Phur pa bcu gnyis, (which title is itself cited within the Dunhuang manuscript IOL TibJ321, although we cannot know if it is referring to the same text), produced, perhaps, within the early school of Padmasambhava in Tibet, includes a section on the four goddesses tamed at Yangleshod, alongside other unashamedly indigenous ritual categories. 14

[Slide 7, close up, Padmasambhava’s feathered hat, 15th cent. painting, Tibet HAR 30820 ]

A striking syncretic element in PT 44 is the manner in which Padmasambhava handles the dangerous goddesses: he confines them within his hat, where they become docile. Of course, a yogi can do anything, but this is not what Buddhists typically do, because in their symbolism, it is more customary for the guru or enlightened beings to be be placed on the head, while unruly or subservient elements like these goddeses, should more usually be placed at the feet, or on the limbs. Maybe that’s why this detail does not occur in all later versions of the narrative, and survives only within some of them, such as O rgyan gling pa’s Lha ‘dre bka’ thang. At other times, the hat is 15 switched for some other kind of container. 16

Yet we know from the triangulations of several different historical and ethnological sources, that it was the quintessential practice of pre-Buddhist shamans across much of the Himalayas and Tibet, to place protective deities symbolised by feathers on their headdresses; so that whatever might have been originally intended by the authors of PT44, its

symbolic significance to the wider Himalayan populations is unarguable. As Toni Huber and others have pointed out, wearing a magical 17 feathered headgear, to represent their protective deities, remains ubiquitous amongst the scattered surviving communities of Himalayan and Tibetan shamans across a vast geographical range, arguably constituting their most distinctive insignia. It might well be that despite its modern association with the Indic Mahākāla, the cult of the magical protective Bhutanese Raven Crown, with its depiction of the protective deity Las mgon bya rog gdong can in raven form, reflects a

For example, Chapter 9 of the Phur pa bcu gnyis mentions the ‘go ba’i lha lnga, a set of five personal deities that 14 Tibetans believe emerge with every human being at birth, and then continue to hover protectively around their body until the person dies. I am grateful to Stefan Mang and Peter Woods, graduate students at the Rangjung Yeshe Institute in Boudha, Nepal, 15 who have corrected me on this point. Initially I had not encountered in any of the several later texts I consulted the continuing use of the guru’s hat as

receptacle for the goddesses, but Stefan Marg has now pointed out to me that such a reference does in fact occur in O rgyan gling pa’s Lha ‘dre bka’ thang. See here: O rgyan gling pa, Bka’ thang sde lnga (Pe cin: mi rigs dpe skrun khang), 23. See:!rid=W17319 . In addition, Stefan Marg and Peter Woods have informed me that the narrative continues within some oral traditions in contemporary Nepal. For example, in some early texts it's a casket, while bDud ‘jom Rin po che’s gNams lcags spu gri Lo rgyus has a 16 small vessel. See Mayer forthcoming, where I discuss feathered headgear in relation to the Mokotoff ms, Daniel Berounsky’s work 17 on the gnyan ‘bum, various Dunhuang texts, the Chinese T’ang annals, Toni Huber’s research among contemporary Himalayan shamans, and more.

9 similar theme. Feathered or avian ritual headgear is also a tradition of great antiquity, found in 18 T'ang descriptions of Tibetan royal priests (wu) wearing bird hats, in ancient Tibetan legends such 19 as the Zhang Zhung kings with feathered crowns (see Vitali 2008, but note also Huber’s critique of Martin and Vitali’s interpretations, in Huber 2013, page 279), in non-Buddhist Dunhuang texts describing bird or vulture feathers placed ritually on the head, and also profusely witnessed in 20 surviving early Tibetan

documents, such as the Mokotoff ms. And, as we can see, perhaps uniquely in Tibetan Buddhist iconography, Padmasambhava (and thus by extension, a handful of his later representatives such as Rig 'dzin rgod ldem and rDo rje gling pa), has vulture feathers as his headgear or in his hat. As the Tibetan art historian Jeff Watts writes: “The vulture feather decoration at the peak of a lotus hat appears to have originated with the iconography of Padmasambhava sometime during or after the lifetime of Nyangral Nyima Ozer. The symbolic use of the feather [in the hat] seems to be almost exclusively used in the Nyingma Tradition of Tibetan Buddhism.” 21 Huber has suggested, entirely plausibly, that such feathered headgear in the iconography of Padmasambhava, is a reference to indigenous religious iconography.

IOLTibJ321 Wings

[Slide 8. Dunhuang IOL TibJ 321, Noble Noose of Methods Tantra. British Library]

The third Dunhuang Padmasambhava manuscript is an excellently well preserved tantric commentary on the Noble Noose of Methods Tantra, complete in 85 folios, with the scriptural root text embedded as lemmata. It includes a verse of praise to Padmasambhava as a transcendent being who is not of this world, and informs us at three different points that Padmasambhava produced this text, although emphasising he does so out of his spiritual accomplishment as a siddha, not through ordinary egotistical fabrication (rang bzo). It clarifies further, that whatever is spoken out of a mind of pure awareness can be called ‘tantra’, and that in this sense, an accomplished siddha’s tongue can

Tibetan protector rites frequently incorporate indigenous elements, and Cathy Cantwell has pointed out to me, that in 18 contemporary rNying ma ritual in Bhutan, we find the contemporary usage of the Vajra Border Guard (rdo rje mtshams pa) wearing the crow's head hat – as an accomplished tantric practitioner, he needs to see and keep out the bad spirits, so he sits in the doorway during the sgrub chen. Cantwell surmises he is supposed to see the spirits via the hat, which has further eyes on it. And it is definitely

meant to be his protective deities sitting on the head to notice the spirits for him. However, she is not clear if this crow's head hat is a general thing for the Vajra Border Guard, or if it's a localised Bhutanese custom. It is also not clear to us if the protective spirits on the Vajra Border Guard’s crow’s head hat, as with the Bhutanese royal crown and Las mgon bya rog gdong can, are now understood as wisdom protectors linked to Mahākāla, who even in Indian sources is often accompanied by crows or ravens, and who

even has crow-headed forms; or whether they are seen as unruly worldly spirits as in the PT44 narrative. Xin Tang Shu, 130ff, as translated in Pelliot 1961, cited in Walter 2009: 33.19 The term bya ru, literally 'bird horns', occurs several times in ancient non-Buddhist Dunhuang texts, to indicate bird 20 or vulture feathers placed ritually on the head. See Huber 2013: 278-9; and Bellezza 2013: 69 note 90; 230)

10 produce utterances directly expressing the turning of the wheel of the Vajrayāna dharma in the Buddhafield of Akaniṣṭha. 22 [Slide 9. Heruka without wings HAR 3314473, and Heruka with wings HAR 303 ]

Although this text at first glance appears to be composed out of familiar Indian tantric elements, there is one notable exception: it places huge wings on the central deity, which otherwise resembles a standard Indian wrathful male form known as Heruka.

I searched far and wide for a winged Indian counterpart, without success (of course there are a variety of winged deities in the long history of Indian religions, but I limited my search to tantric sources, both Śaiva and Buddhist). I consulted many leading experts, but none could locate an Indian winged tantric Buddhist heruka, nor even one from Uḍḍiyāna. The closest I got was a Śaiva form called Ākāśabhairava, but it turned out to be from Himalayan Nepal, not India. 23

By contrast, scholars of indigenous Himalayan religion, especially those dealing with ancient Tibet, found the wings more familiar. Toni Huber, Daniel Berounsky, and others, describe a pre-Buddhist Tibetan religion often obsessed with avian imagery. A very important implement of pre-Buddhist priests and shamans, accompanying their feathered headresses, was a wand made from a vulture’s wing. Such wands are well attested in old sources, and described in detail in a non-Buddhist Dunhuang text. They are still widely used by contemporary shamans (for a detailed discussion, see Mayer, forthcoming (2019).

If the Dunhuang Noble Noose of Methods seems so far to be the earliest known attestation of the winged heruka, it is by no means the last: many or most rNying ma herukas still remain depicted with wings. And if one systematically compares rNying ma meanings and symbolisms of the wings with those of their older non-Buddhist counterparts, it appears quite plausible that the Buddhist

The main text of IOL TibJ321 has the following passage (see Cantwell and Mayer 2012, page 348), which could refer 22 to a method of revealing tantric scriptures, and which, interestingly, corresponds broadly to the method of scriptural revelation described in gShen che klu dga's 10th century gter ma colophons (see Martin 2001: 50) /skyes bu gang gis rig pa de / /ngag gis ci skad brjod pa'i sgra / /thams cad ma lus tan tra zhes / 'og myin bla myed gnas mchog du / /mgon po bdag nyid chen po yis / /rdo rje 'khor lo bskor pa na / /ljags kyi dbang po bkram las gsungs/ / “ (When) this pure awareness (is produced) by any noble being whatever, whatever sound is articulated by (his) speech, all without exception is called, “tantra". In the supreme incomparable place of Akaniṣṭha, the Protector Great Being, turning the vajra wheel, speaks through disseminating the tongue's sense faculty.” To this passage it attaches the following marginal note referring to Padmasambhava: pad ma sam ba bhas rang gz[or?] byas pa ma yin bar ston, which

might be intended to mean "This demonstrates [that this text] was not created by Padmasambhava idiosyncratically". On the same folio, in the very next sentence, it praises Padmasambhava as a sublime realised being, in a verse of praise later adopted by Nyang ral into his Zangs gling ma (note that this verse of praise is certainly not a colophon, as some have thought): dngos grub mchog brnyes ya mtshan chen po'i/ 'jig rten ngam 'gyur pad ma rgyal po yis/ de bzhin gshegs pa'i man ngag gsang chen rnams/ klung [em. klong] nas bkrol mdzad de la phyag 'tshal lo/ "(I) prostate to he who has attained the supreme siddhi, of great wonder, / Padma rGyal po (who) is not worldly;/ (he who) unravels from the expanse / the tathāgata's great secret pith instructions.' For a detailed examination, see Cantwell and Mayer 2012, page 93ff, and Cantwell and Mayer 2013, pages 25ff.

Where the Dunhuang manuscript reads klung, all other extant editions of the text read klong, hence with the probable meaning of "I prostrate to you who unravel from the expanse the tathāgata's great secret pith instruction". Right at the beginning of IOL TibJ321, it gives the title: 'Phags pa thabs kyi zhags pa pad ma 'phreng gi don bsdus pa'i 'grel pa, and to this it attaches the following marginal note: rdo rje sems dpa'I dngos grub thob par bya ba dang/ bsam ba la bgegs myi 'jug cing bsam ba mthar phyin par bya ba'i

don/ 'bu tas bsdus sam ba bha byas, which might mean: (This) means that Vajrasattva's siddhis should be accomplished, no obstacles to (one's) aspirations will arise and (one's) wishes should be perfected. (These meanings have been) summarized by the Buddha, produced by Sambhava. Teun Goudriaan and Sanjukta Gupta (1981, p. 115) mention a winged Śaiva Bhairava, Ākāśabhairava, which is 23 particularly worshipped in Nepal, but Sanjukta Gupta informs me that she concluded this form is of Himalayan rather than Indian origins (personal communication, 16th November 2015).

11 heruka’s wings were an ingenious and culturally sensitive accomodation to indigenous religion (Mayer, forhcoming, 2019).

It is interesting that the main Sanskritic Guhyagarbhatantra root tantra, so popular with the rNying ma pa, does not actually mention wings anywhere, yet even its deities are often painted with wings in rNying ma art. One cannot be sure if the Noble Noose of Methods was actually produced by Padmasambhava as claimed, or by his early school, but either way, the attachment of wings to the Buddhist heruka seems a finely considered gesture towards indigenous iconography, and in keeping with what appears to be the hybridising tendencies in evidence in some other Dunhuang sources for the early Padmasambhava school.

As well as crossing geographical, cultural, and ritual, boundaries, Padmasambhava was evidently in his own time also seen to have crossed boundaries of conventions and mores. Early sources consistently associate him with highly esoteric and transgressive forms of Tantrism that were still quite new and avant-garde in his lifetime. These new methods displayed Śaiva influences, and had developed out of interactions between Śaivites and tantric Buddhists. They were feared as potentially dangerous by more conventional Buddhists, and were not initially accepted for general dissemination by the Tibetan authorities. Evidence still survives of their official restriction within the Empire in an authentic Imperial text designed to guide translators, the Sgra sbyor bam po gnyis pa, and an early historical text, the Chronicles of Ba, describes Padmsambhava being banished by the Tibetan authorities after only a short visit, because of fears concerning his potential to abuse tantric powers. More interesting still, several extremely early tantric literary sources give Padmasambhava no exceptional prominence in comparison to other tantric masters. Yet, as we have seen, a handful of Dunhuang sources do virtually apotheosise him, as a scripture-producing, deitytaming, siddha.

Interpreting the complex evidence of these conflicting accounts, Matthew Kapstein was the first to suggest that in his own time, Padmasambhava had a really powerful following only along the southern borders of Tibet, but was less widely welcome at the central Imperial institutions. However, as his school’s particular brand of tantrism gained popularity in subsequent decades, its growing prevalence eventually inspired a revision of Tibetan history, to place Padmasambhava at the centre of the Imperial conversion project. Thus he became in retrospect described as the Tibetan Emperor’s personal guru, and as Tibet’s paramount conversion hero.

The Padmasambhava narratives are also rather unique in Tibetan literature through the comprehensive way they straddle the boundaries of history and ritual: while no Tibetans (and comparatively few modern scholars) doubt his historicity, most of the narratives that describe him indiscriminately mingle mythologies connected to Tantric ritual performances, with more conventional historiography. To give just one example, hagiographies such as the Zangs gling ma, or the Padma bka' thang, give detailed accounts of Padmasambhava and his consort Mandarava doing longevity practices together, at the Maratika cave in Nepal. Yet this narrative in the hagiographies cannot be seen separately from its parallel existence in rituals: to give just one example, in the Descent of Consecrations section of the ‘Chi med srog thig sGrub chen. In similar vein, the narratives of PT 44 and PT 307 continue to be used in ritual.

It is not simply the case that the rituals are following after the hagiographies: on the contrary, the relationship is symbiotic, and in many cases, ritual is generating hagiography. For example, the traditional hagiographies describe him as a young man being sent into exile from his native

12 Uḍḍiyāna, to live in cemeteries as a penance for his accidental killing of a youth; yet this narrative is unmistakeably descriptive of an important esoteric practice known as the kāpālikavrata, which had originated in Śaivism and been adopted by Buddhism, so it is uncertain if the narrative is historic.

In that sense, Padmasambhava’s life stories can be seen as religious myth, even if they are presented as history. And since they are myths intended to illustrate ritual, they have little need to avoid contradiction. Hence many of the more puzzling episodes in his biographies can only be unlocked by understanding their relation to often quite arcane elements of tantric ritual: for example, a seminal text called the Black 100,000 (‘Bum nag) describes two completely contradictory accounts of Padmasambhava’s birth on the same folios. The first has him born conventionally from a human mother, but as a fearsomely ugly baby with the physical characteristics of the wrathful deity Vajrakīlaya. The second describes him as a radiant infant born miraculously from a lotus flower on a pure lake. Both accounts are equally mythologised, the first to conform with the Vajrakīlaya rituals, the second to conform with the devotional practices of guru yoga. For the faithful, all are true.

Metaphysically, Padmasambhava came to represent a figure within history who is simultaneously beyond history. He is both a human yogin who visited Tibet, and a Buddha beyond time and space. Acting as bridge between our world and the transcendent underpins perhaps his most important role: for Padmasambhava has, for the last thousand years or so, served as the main conduit for the continuing revelation of new sacred texts in Tibet. We have already seen in the 9th or 10th century Noble Noose of Methods Tantra from Dunhuang, how he might already be portrayed as its source, and how in PT 44 he is seen as redacting the Vajrakīlaya tantras. Over time, such roles became greatly magnified. Tibetans believe that during the 8th century, Padmasambhava visited every part of Tibet, to bury tens of thousands of sacred texts, known as gter ma, or Treasure texts. His closest disciples are then reborn again and again through time, to recover the Treasure texts Padmasambhava had hidden, and teach them to Tibetans. In this way, Padmasambhava became an inexhaustible source of fresh Tantric scriptures for generation after generation of Tibetans. Most of the liturgies used by the rNying ma pa, and a great many of those used by other schools too, are Treasures of this sort. In addition, the tantric practice of guru yoga, or spiritual union, with Padmasambhava, and the fact that this is keyed into his mythology, means that devotees are constantly re-living his life-stories, and merging spiritually with the guru and his retinue. Thus Padmasambhava remains not merely a distant historical figure, but, in the Tibetan imagination, a very contemporary actor as well, constantly revealing new teachings to his devotees, and infusing their lives.

As we have seen, the figure of Padmasambhava relates to borders in many way. He comes to embody the enchanted border region of Uḍḍiyāna, where for many centuries, seminal new religious syntheses arose amidst the intermingling of different cultures. His (or his early school’s) missionary activities creatively navigated the boundaries between Buddhist tantrism and indigenous Tibetan religion, utimately to produce what we now call rNying ma Buddhism. His pioneering advocacy of esoteric Vajrayāna stood in his own day at the outermost boundaries of respectability, at the same time bridging the boundaries between Buddhism and Śaivism (even if esoteric Vajrayāna was later to become a mainstay of Tibetan Buddhism). His life stories and hagiographies straddle the boundaries of history and tantric ritual in a unique manner. His person comes to be seen as the bridge between this world and the next, and the gateway for ongoing fresh spiritual revelations. 13 Even his stay in Tibet might have been mainly limited to its southern borderlands, although this is no longer acknowledged by Tibetans.

Slide 10 Close up, Padmasambhava in ‘King Appearance’ HAR 73460

Unsurprisingly, his representations in art also crosses boundaries. Perhaps uniquely in the iconography of Tibetan gurus, Padmasambhava is depicted with the attributes usually reserved for kings, even though he holds only the non-royal status of a religious teacher. As Jeff Watt explains, “’King Appearance' in Himalayan art is a specific type of figurative form. The principal characteristics are the face often with a stern look achieved by upturned eyebrows accompanied by a mustache and goatee. The clothing is heavy and layered with multiple colours, a cloth head covering or hat sometimes with a small jeweled crown, and boots on the feet. They can hold any number of objects in their hands.” 24 In Tibetan art, this appearence is strictly limited to those who actually are kings: historical kings of India or Tibet, mythical kings such as those of Shambhala, or deity kings, such as the Guardian Kings of the Four Directions. Padmasambhava is the only guru in Tibetan art who is represented in King Appearence: no other saints, siddhas, gurus or paṇḍits of India or Tibet take this form. It is noteworthy that Padmasambhava was already described as “The Lotus King” in the Dunhuang text of the Noble Noose of Methods Tantra.

Padmasambhava’s iconography surprises in other ways too. For example, he combines clothing from India, with boots from Central Asia or Tibet. The many layers of his rich robes combine secular, monastic, and tantric elements. He displays emblems of Indian and Uḍḍiyāna tantrism, some familar within Śaivism as well as in Buddhism, alongside the vulture feathers so symbolically significant to indigenous Tibetan religion.

Slide 11. Padmasambhava, his Eight Aspects, plus two more aspects. HAR 73460

But this is only the main iconographic form of Padmasambhava, possibly developed, as we have seen from fig XXX above, around what Filigenzi calls an Uḍḍiyānasiddha” prototype. In addition to this main form, a very popular set of eight further forms, some peaceful, some wrathful, were described by his 13th century devotee Guru Chowang, to be worshipped in rituals. These eight are connected to different moments in Padmasambhava’s hagiography, where he manifests in different modes, each displaying different capabilities to meet different challenges or to benefit beings in different ways (once again, the intersection of Padmasambhava narrative with ritual). In addition to these eight, many further forms emerged over time, by now probably numbering in the thousands, most of them beleived to have been disclosed by Padmasambhava himself through the medium of Treasure texts. I am not aware of any guru, or even any deity, within the Tibetan tradition, that has as much variety of iconographic representation (and ritual worship) as Padmasambhava.

Surely there is no figure in the Tibetan imagination more complex, more varied, more consistently associated with borders and the crossing boundaries, than Padmasambhava, yet also none who inspires more devotion and confidence.

14 Slide 12. Wolf-hawk deities and other chimeras. Detail from Ka ba nag po thangka, Bon, 20th century Nepal. Courtesy of Ven Tenpa Yungdrung. Buddhist snow lion-garuḍa chimera. After Robert Beer 1999, p.74.

As a final more anthropological thought, I would like to suggest that Padmasambhava’s popularity might not be unrelated to his conceptual complexity and his crossing of boundaries. The anthropologist Charles Ramble has commented on the repeatedly manifest affinity of indigenous Tibetan culture for chimeras. In Greek mythology, the chimera was a creature with a lion's head, a goat's body, and a serpent's tail, and by extension the term now indicates any mythical beast formed from the parts of various different animals. In indigenous Tibetan ritual, chimeras figure very prominently as protective deities. For example, in the Black Pillar, a Bon text concered with the overcoming of obstacles, a majority of the hundreds of deities populating its dense maṇḍala, are chimeras of one sort or another, for example, the wolf-hawk crossbreeds (khra spyang) in this illustration.

Ramble writes as follows:

Animals that transgress culturally-sanctioned taxonomic boundaries are often the object of special beliefs. Tibetan ritual texts, especially those of …..[the indigenous religion of] Bon, sometimes feature semi-divine animals that play an important role as protectors. These creatures, though natural, are perceived as concatenations of the body-parts of numerous other natural species, and may be understood as different varieties of chimera. …… In addition to real animals the literature also features imaginary creatures that exhibit the physical or behavioural characteristics of several natural species. Each of the animals that provides a component is presented as wielding a specific type of capability, and it is the concentration of these multiple capabilities that gives the chimera, whether real or imaginary, its extraordinary power (Ramble 2014, page 13).

Perhaps, then, the unique construction of the mythologised Padmasambhava, Tibet’s predestined guru and protector, might be culturally connected to his crossing of so many taxonomic boundaries, and his consequent concentration of so many multiple capabilities within his single person.


’Bum nag. Rdo rje phur pa’i bshad ’bum slob dpon rnam gsum gyi dgongs pa slob dpon chen po padmas mkhar chen bza’ la gdams pa / phur ’grel ’bum nag ces gzhan las khyad par ’phags pa’i thugs kyi ti la ka. Two editions: (i) bDud ’joms bKa’ma: Rñiṅ ma Bka’ ma rgyas pa, vol. tha, pp. 215-557. (ii) Phur pa ’bum nag and Phur pa’i ’grel chen bdud rtsi dri med, Two rare Vajrakīla teachings from the miraculous Lotus-born Gu-ru Rin-po-che Padmasambhava, reproduced from rare mss. from the library of Lopon Sonam Sangpo. Gonpo Tseten, Gangtok 1976, pp.1-229 (TBRC W27485). For an English translation, see Boord, M.J., 2002. A Bolt of Lightning From The Blue: the vast commentary of Vajrakīla that clearly defines the essential points. Annotated translations, including Phur ’grel ’bum nag as transmitted to Ye-shes mtsho-rgyal. Berlin: edition khordong.
Beer, Robert, 1999: The Encyclopedia of Tibetan Symbols and Motifs, London, Serindia.

15 Bellezza, J. V. 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. Wien, Verlag der Österreichischen Akademie der Wissenschaften.
Catherine Cantwell and Robert Mayer, 2012: A Noble Noose of Methods, the Lotus Garland Synopsis: A Mahāyoga Tantra and its Commentary, Vienna, The Austrian Academy of Sciences Press.

Cathy Cantwell and Rob Mayer, 2016 “Representations of Padmasambhava in early post-Imperial Tibet”, in ZAS 45 (2016), Christoph Cüppers, Robert Mayer and Michael Walter (Editors) Tibet after Empire: Culture, Society and Religion between 850-1000 Proceedings of the Seminar held in Lumbini, Nepal, March 2011, pages 41-76
Dalton, Jacob, 2004. “The Early Development of the Padmasambhava Legend in Tibet: A Study of IOL Tib J 644 and Pelliot tibétain 307”. The Journal of the American Oriental Society 124.4: 759-772.
Eastman, K.W. 1983 "Mahāyoga texts at Tun-huang." Bulletin of Institute of Buddhist Cultural Studies 22: 42-60.
Filigenzi, Anna, 2015: Art and Landscape. Buddhist Rock Sculptures of Late Antique Swat/ Uḍḍiyāna. With contributions by L.M. Olivieri and P. Rockwell. Wien, Österreichische Akademie der Wissenschaften, Philosophisch-Historische Klasse Denkschriften, 462. Band.
Gellner, David N., 2013. ‘Northern South Asia’s Diverse Borders, from Kachchh to Mizoram’, in Borderland Lives in Northern South Asia, ed. David N. Gellner. Durham and London, Duke University Press.

gNubs chen sangs rgyas ye shes, n.d. rNal-’byor mig-gi bsam-gtan or bSam-gtan migsgron: A treatise on bhāvanā and dhyāna and the relationships between the various approaches to Buddhist contemplative practice, Reproduced from a manuscript made presumably from an Eastern Tibetan print by ’Khor-gdong gter-sprul ’Chi-med rig-’dzin, Smanrtsis shesrig spendzod, vol.74, Leh: Tashigangpa, 1974 Goudriaan, T. & Gupta, S. 1981 Hindu Tantric and Śākta Literature. (Wiesbaden, Harrasowitz).
Guru Chos dbang, n.d. gTer ‘byung chen mo, pages 75-193, within Gu ru chos dbang gi rnam dang zhal gdams, Rin chen gter mdzod chen po’i rgyab chos Vols 8-9, Ugyen Tempa’i Gyaltsen, Paro, 1979. TBRC Work Number 23802

Huber, Toni, 2013: ‘The Iconography of gShen Priests in the Ethnographic Context of the Extended Eastern Himalayas, and Reflections on the Development of Bon Religion’, pp. 263-294 in F-K Ehrhard & Petra Maurer (Hrsg). Nepalica-Tibetica Festgabe für Christoph Cüppers, Band I. Andiast, IITBS
Kapstein, Matthew, 2000: The Tibetan Assimilation of Buddhism. Conversion, Contestation and Memory. New York, Oxford University Press.

Karmay, Samten G., 1998. The Arrow and the Spindle: Studies in History, Myths, Rituals and Beliefs in Tibet. Kathmandu: Mandala Book Point.
Klimburg-Salter, Deborah, 2016: “Along the Pilgrimage Routes between Uḍḍiyāṇa and Tibet: The Gilgit MSS covers and the Tibetan decorated book cover.’ In: Tibet in Dialogue with its Neighbours: History, Culture and Art of Central and West Tibet, 8th to 15th century. Eds: E. Forte, Liang Jun Yan, D. Klimburg-Salter, Zhang Yun, and H. Tauscher. Vienna/Beijing, OAW/CTRC

Mayer, Robert, 2015: 'We Swear Our Grandparents Were There', in The illuminating Mirror, Tibetan Studies in Honour of Per Per K. Soerensen, ed. O. Czaja and G. Hazod, Reichert Verlag, Wiesbaden.

Mayer, Robert, forthcoming (2019): “How did the Heruka get his wings? And why did the Guru have a feather in his cap? Avian symbolism in rNying ma iconography.” To appear in Archiv Orientalni, ed. D. Berounsky, Prague.
Pelliot, P., 1961: Histoire ancienne du Tibet. (Paris, Librairie d'Amérique et d'Orient, Adrien-Maisonneuve).
Ramble, Charles, 2014: 'Real and imaginary Tibetan chimeras and their special powers.' In Mongolo-Tibetica Pragensia ’14, Volume 7, No. 2, Special Issue "Indigenous Elements in Tibetan Religions", Edited by Daniel Berounský. Charles University, Prague

Sanderson, Alexis, 2007: ‘The Śaiva Exegesis of Kashmir’, in Mélanges tantriques à la mémoire d’Hélène Brunner. Tantric Studies in Memory of Hélène Brunner, Collection Indologie 106, EFEO, Institut français de Pondichéry (IFP), ed. Dominic Goodall and André Padoux, 2007.
Scott, James C., 2009: The Art of Not Being Governed: An Anarchist History of Upland South-east Asia. New H
aven, CT: Yale University Press.

Szántó, Péter-Dániel, 2016: ‘Before a Critical Edition of the Sampuṭa”, in ZAS 45 (2016), Christoph Cüppers, Robert Mayer and Michael Walter (Editors) Tibet after Empire: Culture, Society and Religion between 850-1000 Proceedings of the Seminar held in Lumbini, Nepal, March 2011, pages 397-422.

Vitali, R. 2008: 'A Tentative Classification of the Bya Ru Can Kings Of Zhang Zhung', in Pommaret, Françoise and Jean-Luc Achard (eds), Tibetan Studies in Honor of Samten Karmay, Revue d'Etudes Tibétaines Vol.15, November 2008. Paris, Langues et Cultures de l'Aire Tibétaines, CNRS.

Walter, M. 2009: Buddhism and Empire. The Political and Religious Culture of Early Tibet. Leiden, Brill