So just who was Śāntigarbha? Śāntigarbha, an Indian paṇḍit whom some traditional sources locate in Khri Srong lde’u btsan’s era as a prominent translator of and commentator on Yogatantra, or even as consecrator of Samye Monastery, and others associate closely with Padmasambhava, has acquired some topical historical significance in recent years specifically because he has become rather surprisingly described as relatively insignificant. To be more precise, a recent academic assertion is not so much that he is insignificant in the eyes of Western scholarship (could anything really be insignificant for Western scholarship?), nor necessarily that he was particularly insignificant to the ancient authors of the Dunhuang texts, but rather, that he is comparatively insignificant in the eyes of later rNying ma tradition, and certainly not a person of sufficient stature to venture opinions on the great Padmasambhava. This supposed relative insignificance of Śāntigarbha in the later rNying ma tradition has been juxtaposed with his portayal in a specific Dunhuang text, and the supposed contrast used, via a convoluted and to my mind elusive logic, to support the historical conclusion that for the particular authors of this specific Dunhuang text, Padmasambhava was considered merely an average (albeit respected) kind of guru, and not a uniquely mythologised being.^[1]  The precise context is the referencing of Śāntigarbha in a marginal note concerning a praise of Padmasambhava in IOL TibJ 321, the Dunhuang ms of the Phags pa Thabs kyi zhags pa padma ‘phreng gi don bsdus pa’i ‘grel pa, a tantric text on which Cathy & I have just published a 375-page tome, so we feel well-placed to address issues relating to it. In our book, we come to the opposite conclusion: While much of the Padmasambhava devotional literature had of course not yet been written, and while important sections of early proto-rNying ma literature do indeed not make Padmasambhava into a uniquely important figure, we believe this particular Dunhuang text does do precisely this. Like PT44, we believe it is crucial textual evidence for an already uniquely highly mythologised Padmasambhava, a trope found in some proto-rNying ma tantric documents but clearly absent in others. But before we begin, let me point out for those who might already feel bewildered, that these discussions of the interconnected referencing of Śāntigarbha and Padmasambhava in an old manuscript are not as trivial as they might appear. The extreme paucity of genuinely old surviving textual evidence means that our understanding of the development of the historically important Padmasambhava tradition depends quite a lot on our interpretation of just who Śāntigarbha was and what his relationship to Padmasambhava was, in this one particular text.

In this brief blog, I am therefore not primarily interested in discussing a historical Śāntigarbha for modern historical scholars—anyone interested in that is referred to Dan Martin’s monumental TibSkrit, which remains as ever the starting point of choice. Rather, I am interested in the Śāntigarbha envisaged by later rNying ma tradition, because it is his status in the later rNying ma tradition which has been used to construct, by means of a comparison, an historical argument about the early Padmasambhava tradition. Actually, I’m not entirely sure if the argument is logically coherent even if its premises were true, but that is another question—here, I will only examine its premises, to see if they are philologically supported. So, is this recent academic assertion actually true? Was (or is) Śāntigarbha really so insignificant in the later rNying ma tradition, as has recently been claimed? To my mind, the answer is ‘no’: Śāntigarbha should not in any meaningful sense be described as insignificant to the later rNying ma tradition, because I believe the Śāntigarbha of IOL Tib J 321 is intended to be the same Śāntigarbha as the ‘Great Vidyādhara of India’ who occupies such a prominent and privileged niche within the unique octadic formations that structure so much of rNying ma pa thought. I am not so far aware of Tibetan tradition specifying more than one prominent tantrist called Śāntigarbha living simultaneously—on the contrary, some traditional sources do specify that Padmasambhava’s teacher Śāntigarbha and the one who consecrated Samye etc are the same person. Nor has any modern scholar suggested such a distinction. As a result, I think the there is only one Śāntigarbha, and the elusive argument based on his mention in IOL Tib J 321, which we present in note 3 below, simply falls down.

While some seem to see rNying ma tantrism as narrowly Padmasambhava-centric, this  is clearly not the case, and never has been. There are other founding figures too, like Vimalamitra and Śāntarakṣita. Moreover, as Jeff Watt has recently reminded us in a fine series of lectures at Oxford University, later rNying ma Tantrism, in its texts, its rituals, and its iconography, is unlike that of any other Tibetan tradition because it has a symmetrical patterning running through it, consisting of densely interelated sets of eight. Although individual members of these interelated sets can vary a little from text to text, the sets themselves remain reasonably constant as structuring devices, and lend a unique shape to rNying ma tantrism. The sets include Padmasambhava and in many ways can be seen to revolve around him, but notwithstanding the hyperbole of some devotional literature, even they are not entirely or exclusively reducible to him.

Perhaps the most fundamental octad is the set of eight major yi-dams, the famous bKa’-brgyad or Eight Pronouncement Deities: Yamāntaka (‘jam dpal sku), Hayagrīva (padma gsung), Śrīheruka (yang dag thugs) Vajrāmṛta (bdud rtsi yon tan), Vajrakīla (rdo rje phur pa), Mātaraḥ (ma mo rbod gtong), *Lokastotrapūjā (’jig rten mchod bstod) and *Vajramantrabhiru (mod pa drag sngags). Thangkas often show these eight major yi-dam deities sharing a basic iconographic appearence of three heads, six arms, four legs, and two wings, differentiated only by the sets of hand symbols they hold and a few other particularities. Such a uniformity of conception across a whole set of yi-dam deities is not found in the other Tibetan traditions.  Since they are nominally at least the main meditational deities of the rNying ma school, the arrangement of their tantric textual sources also lend a structuring device for organising the sections of scriptures in the rNying ma canon (NGB), as also the treasure texts within Kongtrul’s famous collection the Rin chen gter mdzod.  Despite recent claims to the contrary, all these eight deities are moreover normally understood by the rNying ma tradition as transcendant enlightened forms of Heruka.^[2]

Connected with these eight meditational deities are the Eight Great Vidyādharas of India, often known as the Rig ’dzin gyi slob dpon brgyad, and it is here that Śāntigarbha features so prominently. The Eight Great Vidyādharas of India are often described as living in the Eight Great Cemeteries of India, and they are key sources in this world for the lineage transmissions of the Eight Pronouncement Deities (bKa’-brgyad). Lineage is extremely important in Tantric Buddhism, and the Eight Great Vidyādharas play a roughly equivalent role for the rNying ma pa to that played by Tilopa and Naropa for the bKa’ brgyud pa, or Virupa for the Sa skya pa. Thus the Eight Great Vidyādharas of India are at different moments the gurus, teachers, spiritual brethren, and disciples of Padmasambhava, his spiritual contacts several of whom remained in India but whose teachings he brought to Tibet. Each one of the eight is the teacher of one of the Eight Pronouncement Deities, and as such, these eight Vidyādharas are constantly celebrated at the level of ritual, hagiography and art. Although the lists vary a little, Śāntigarbha is normally included, often as Padmasambhava’s guru for *Vajramantrabhīru (mod pa drag sngags). Being an important Guru of Padmasambhava, illustrations of Śāntigarbha are therefore quite widespread, either as an individual, or as seated above Padmasambhava’s head in position of honour as his guru. For example, Śāntigarbha is prominent as part of the central visualisation for the Rig ‘dzin ‘dus pa, the most popular of all Padmasambhava rites that was revealed by ‘Jigs med gling pa (1729-1798), where Śāntigarbha is placed with the seven other Vidyādharas in an arc directly above and around Padmasambhava (the very name of this sādhana of course refers to the Vidyādharas, Śāntigarbha included). Likewise in a famous set of illustrations of another rNying ma set of eight, the Eight Aspects of Guru Rinpoche (gu-ru mtshan brgyad), Śāntigarbha is shown seated above Guru Rinpoche’s head as his revered guru, when Guru Rinpoche takes the wrathful form of Dorje Drollö (rdo rje gro lod). There are initiation cards showing Śāntigarbha, and needless to say, Śāntigarbha is also mentioned in numerous life stories of Guru Rinpoche, as one of his important Indian teachers.

Other rNying ma sets of eight include the Eight Great Vidyādharas of Tibet, close students of Guru Rinpoche who practiced, realised and propagated the Eight Pronouncement Deities in Tibet, as well as further sets shared with other Buddhist traditions, such as the Eight Gaurī Goddesses and the Eight Great Bodhisattvas (nye-sras brgyad). There are various ways in which the different sets of eight are made to inter-relate and correlate, but these need not concern us right now.

Rather, I want to pose the question, if Śāntigarbha has such a central and abiding presence in the underlying mythological architecture of rNying ma ritual life, can we really describe him as insignificant in the eyes of later rNying ma tradition? Was he really a person of insufficient stature to venture opinions on the great Padmasambhava?  I think not, for reasons shown above, and especially because he is universally accepted as one of Padmasambhava’s gurus.

Even more interesting, and very importantly for our understanding of the early rNying ma, is the fact that a special association between Śāntigarbha and Padmasambhava is already indicated in a Dunhuang text, to be precise, in the marginal note attached to IOL TibJ 321 that we have already mentioned. Here we find in the main text of the commentary a verse of praise for Padmasambhava in his Padma rGyal po form, a verse taken up verbatim in later centuries by Nyang ral nyi ma’i ‘od zer into his Zangs gling ma (see blog entry below, Padmasambhava in early Tibetan myth and ritual, Part 2: IOL Tib J 321). In a marginal note at this point, Śāntigarbha is described as praising Padmasambhava, which (pace Dalton) would appear entirely unremarkable to a later rNying ma readership,  who would understand it as being exactly what one would expect from a close spiritual friend or Guru. [For a presentation of Dalton's argument which we seek to refute, see footnote 3].

But putting traditional perspectives aside, could it actually be that some such tantric connection between Padmasambhava and Śāntigarbha was already being indicated in this early text from Dunhuang? From the cumulative circumstantial evidence that such sources as IOL Tib J 321, Bu ston, the ‘Phang thang ma and mKhas pa lde’u give us on Śāntigarbha, Padmasambhava, and even their respective teachings of Yogatantra and Mahāyogatantra—as we show in our book, the Thabs zhags bridges both tantric genres and mentions both personal names—it is not at all impossible and not even particularly unlikely. Note also that in the Dunhuang text PT44, a similar close tantric relationship is albeit rather briefly indicated between Padmasambhava and someone who might transpire to be Prabhāhasti, another of the Eight Great Vidyādharas of India. We might investigate that in a future posting.

In her seminal work on the Padmsambhava hagiographies, Anne-Marie Blondeau (1980) has famously described how the traditional narratives of Padmasambhava exist in parallel ‘womb birth’ (mngal skyes) and ‘miraculous birth’ (rdzus skyes) versions.[1] Deservedly, her work has been inspirational for later generations of scholars, and she has been cited numerous times. But after a further thirty two years of scholarship, might her pioneering work now at long last deserve some slight further clarification, or a slightly different nuancing? We think so, and here is why.

In brief, Blondeau was the first to to make it widely known to Western scholarship that a ‘womb birth’ account of Padma’s birth existed in parallel to the more widespread and familiar ‘miraculous birth’ accounts. In addition, she mentioned Kong sprul’s association of the ‘womb birth’ accounts with the bKa’ ma rather than the gTer ma, and in particular, with the Phur pa transmission accounts (phur pa’i lo rgyus). Blondeau added that the account given in the Testament of Ba is more commensurable with the traditional bKa’ ma transmission of the ‘womb birth’ version of Padmasambhava’s life. As she further points out, we do not know exactly when the formalizing of the distinction between ‘womb-birth’ and ‘miraculous birth’ Padmasambhava biographies began, but we do know that the categories of ‘womb-birth’ and ‘miraculous birth’ derive from the abhidharma, and we also know that both types of Padmasambhava narratives share a very long parallel history in Tibet.

What needs revising is the occasional interpretation of Blondeau as implying that the ‘womb birth’ accounts are somehow less miraculous and more realistic than the ‘miraculous birth’ accounts, in the sense that while the one describes a natural process, the other describes a miraculous process (in fact, we don’t believe Blondeau really comes to any such simplistic conclusion, but her findings might sometimes be interpreted in this way). Be that as it may, as Cathy has been pointing out for many years now, closer analysis shows that this is not really the case at all: the ‘womb birth’ narrative is not in any way seeking to describing an ordinary event, and the’womb birth’ and ‘miraculous birth’ accounts are in fact both equally supramundane, equally miraculous, above all, both equally derived from the symbolic world of tantric ritual and its visualisations of pure perception.

If one reads the actual sources, it becomes abundantly clear that the ‘womb birth’ narratives found in the Phur pa literature that Blondeau mentions are completely integrated with the ‘miraculous birth’ accounts, occuring together within the very same sources. Both are embedded side by side within the same cycles of tantric deity teachings and practices, in which their presentation by the guru on any specific occasion is designed to generate guru devotion and a pure vision (dag snang) of all phenomena as the tantric maṇḍala. Thus, exactly like the ‘miraculous birth’ accounts, the ‘womb birth’ stories are highly symbolic and connected with tantric imagery, and so do not necessarily represent a more ‘rationalist’ strand of thinking at all. For example, Sog zlog pa’s Phur pa lo rgyus[2] forms part of the cycle of texts for the Rong zom Phur pa tradition. Its focus on a ‘womb birth’ can best be seen as expressing a Mahāyoga visionary perspective equating the physical body with the tantric deity. Hence in this account, Guru Padma is born in a physical body which is none other than the Phur pa deity and his maṇḍala: his waist is a knot like the middle section of a phur bu ritual implement, his lower body triangular in shape, again like the phur bu, while his hair is reddish brown like that of the Phur pa deity, and his eyes and mouth are semi-circular, thus resembling the three semi-circular shapes outlined by a circle around the central triangle, which is the standard graphic depiction of many Phur pa maṇḍalas. Here is Sog zlog pa’s full description:

“Called, Śāntarakṣita, (he) had a complexion of white with (a tinge) of red, the sign of the Lotus family, and his head perfected every wondrous ability. His waist was a knot, his upper body shaped to go inwards, while his lower body was triangular. His mouth and eyes were semi-circles, and his hair was reddish brown. (He was thus) born as one disfigured, (but) endowed with the phurpa’s characteristics.” (śānta rakṣi ta bya ba kha dog dkar la dmar ba’i mdangs dang ldan pas padma’i rigs kyi mtshan dang ldan zhing / sgyu rtsal thams cad rdzogs pa mgo dang sked pa rgya mdud / ro stod bcum gzhogs / ro smad zur gsum/ kha dan mig zla gam / skra kham pa ste / mi sdug pa phur pa’i mtshan nyid can zhig skyes so /, p.12)

In the ‘womb birth’ account given in the reputedly very old Bum pa nag po, a major source for all the Phur pa bKa’ ma transmissions, the accounts of the two types of birth are given together (bDud ’joms bKa’ ma version, Volume Tha: 221-225; Boord 2002: 113-115).[3] First, the ‘womb birth’ is presented, with a slightly different version of the features of the Guru’s body from Sog zlog pa’s, equally replete with potent tantric symbolism, and then there is a variant of the same story of his early years which is given in the following ‘miraculous birth’ story. The two accounts merge for the Guru’s later deeds.

A myes zhabs’ Phur pa lo rgyus, given within his commentary on the Sa skya Phur pa practice,[4] also discusses the two types of birth together. He draws a rather Levi-Straussian symbolic opposition between the two. In this case, the womb birth is said to have taken place in the eastern region of the country of Zahor, while the miraculous birth took place in the western region of the country of Urgyan, so that the residents of the two both held the Guru to be the son of their King. He stresses that there is no contradition, since both types of birth are examples of an inconceivable array of enlightened emanations which accord with the beings to be tamed.[5]

Our conclusion: rather than making a primary analytic or ‘etic’ distinction between the ‘womb birth’ and ‘lotus birth’ accounts, both of which after all are in essence closely related lo rgyus narratives often pertaining to the same tantric rituals, a better primary analytic distinction might be between such ritual narratives on the one hand, and the more historical narratives of a text like the Testament of Ba on the other hand.  The dBa’ bzhed for example refers to itself as a bKa’ mchid (royal discourse), while the sBa bzhed refers to itself as a bKa’ gtsigs (royal edict). Both titles thereby indicate that their proper context is the sphere of state, not the sphere of religious devotion or ritual. By contrast, the traditional lo rgyus accounts of ‘womb birth’ and ‘lotus birth’ alike are part of the transmission of religion. A note of caution, however: it remains a bit unclear how exactly to assess the Padma sections of the Testament of Ba, since we do not yet know who wrote them, when, or why. But that is not for now: more on that in a future blog!

The context  is the Zas gtad kyi lo rgyus, a famous appendage at the end of the dBa’ bzhed that narrates a lively debate between various groups of Tibetans about how to conduct the funeral ceremonies of the recently deceased Emperor, the great Khri Srong lde btsan himself. Some, described in the text as Bonpos,  argued for the time honoured bang so funeral, the ritual interment of the deceased Emperor within a vast tumulus filled with grave goods which would become the focal point of ritual in future years, a magnificent and elaborate system of burial that had by then already defined a great part Tibetan ceremonial life for some hundreds of years. But since such a tumulus burial entailed animal sacrifice, others, described in the text as Buddhists, strenuously disagreed: they preferred funerary rites based on such Buddhist systems as the Sarvadurgatipariśodhana and Uṣṇīṣa.

The passage that suggested lice to Wangdu and Diemberger occurs near the beginning of the story. Arriving at the debate in good time, the proponents of Bon had taken up all the rows of seats (gral), leaving no place for the late-coming Buddhist champions, headed by Vairocana. The Buddhists were facing a serious humiliation: in this important debate, they did not even have a suitably dignified seat at the table, so to speak (it seems the hierarchical arrangements of the gral were as significant indicators of status then as now).  It is at this point that Vairocana intervenes. He stood himself behind the opposing minister seated at the head of the row to the right, thus inducing the hapless minister to glance behind him;  and this is what the minister saw [the Tibetan continues as follows]:

bai ro tsa na’i smra ra’i gseb na khro chung nyungs dkar tsam shig shig snang  ba dang/ shin tu skrag nas kog gis langs pa’i shul du bai ro tsa nas bzhugs pas ban de la g.yas gral shor/ [26v7-27r1]

Wangdu and Diemberger attempted the passage as follows:

“[he] saw khro chung nyung dkar crawling around in the beard of Vairocana, and got up with a start. He was immediately replaced with Vairocana, and lost the whole right row.” (p. 96) A footnote is added, suggesting khro chung nyung dkar might indicate lice, as follows (note 381, page 96): “khro chung nyung dkar seem to indicate some peculiar seeds. Nyung dkar is in Tibet a not very common vegetable species from the seeds of which oil is obtained. Khro may stand for khra in which case it would indicate a multicoloured gathering. Perhaps the minister was scared by these seeds looking like parasites. Possibly this expression is a euphemism for nits (in which case khro = sro) and nyung dkar could be a polite form for shig, i.e. louse (in Mustang the polite form for louse is lug).”

As one can see, this is a very highly convoluted argument, based on the opinion that the words khro chung nyung dkar do not exist in the lexicon, and so must be interpreted. It also requires one to accept a number of other quite complex admitted speculations.

Is this memorable rendering, of lice in Vairocana’s beard startling a Bon po minister, accurate? It seems, alas for those who have come to love it, not. As Cathy has been reminding us for some time now, the words khro chung nyungs dkar tsam refer to something altogether less amusing, less repulsive, and more awe-inspiring, something as familiar to specialists in certain kinds of tantrism as the term ‘front-side bus’ might be to computer manufacturers, but equally obscure and misleading to anyone else. Nevertheless it nowadays only takes a quick look at the THDL or Rangjung Yeshe Dictionaries under khro chung to get a clue: khro chung nyung dkar is in one of these dictionaries usefully if tersely signposted as: “small wrathful ones; yungs dkar – nyung dkar – Mustard seeds.”  In fact, the ‘wrathful ones as small as mustard seeds’ of the Zas gtad kyi lo rgyus are a well known category notably in Mahāyoga wrathful deity systems, and indicate the emanation of vast hosts of tiny spark-like wrathful forms from the hair pores or the skin to create an awesome display. Deities or great yogis are described as doing this. In rituals, the emanation of such tiny deities is often indicated by the casting of mustard seeds (mustard seeds are commonplace in Tibetan ritual), and hence they are sometimes termed khro chung and identified with nyung dkar,  which conveys both their tiny size (as small as mustard seeds), and their ritual analogue (the throwing of mustard seeds during a ritual). These tiny wrathful forms are attested in Dunhuang texts such as IOLTibJ401, in numerous later medieval traditions such as the great Sakya Phurpa cycles, and in numerous contemporary texts such as Dudjom Rinpoche’s treasures. They have been a continuous well established feature of Mahāyoga from Dunhuang times until the present. Hence in a book we published in 2008, Early Tibetan Documents on Phur pa from Dunhuang,  we re-translated this passage from the Zas gtad kyi lo rgyus as follows:

“Amongst Vairocana’s whiskers, miniature wrathful [deities] the size of mustard seeds were appearing and amassing and, they so much terrified [his opponent that he] abruptly arose, leaving Vairocana behind, who sat down, so the right row was lost to the Buddhist monks” (p. 50).

Finally, apparent parallels to this ancient story seems to occur in other literature, including the ‘Dra ‘bag chen mo, the most famous traditional biography of Vairocana. Ani Jinba Palmo’s translation (The Great Image, page 219) mentions Vairocana arriving late at an Imperial funeral, and emitting fierce sparks, to the awe of the other attendees (no lice!). I have not yet located the original Tibetan wording.  The late Khetsun Zangpo Rinpoche’s Biographical Dictionary of Tibet also has an account of Vairocana’s life, which became a basis of A. W. Hanson-Barber’s PhD on Vairocana, and there too we find Vairocana arriving rather abruptly at the Emperor’s funeral with his student gYu sgra snying po, emitting an awesome display of sparks—clearly a version of the same story, and although Khetsun Zangpo naturally makes clear references to the ‘Dra ‘bag throughout his text, some variations in detail indicates he might in fact have drawn this section from a different source, or a different version of the ‘Dra ‘bag (A. W. Hanson-Barber, Madison-Wisconsin 1984, Life and Teachings of Vairochana, page 89).

Is this the end of the line for Vairocana’s mythical lice? I’m reasonably confident it is in the minds of Wangdu and Diemberger, who were never much convinced by the lice idea in the first place, and who, as first-class scholars, are entirely likely to be delighted by Vairocana’s belated liberation from an eternity of itching [in fact, Hilde recently emailed us, with the words "Thankyou for getting a more plausible solution to this obscure passage.."]. But I doubt its the end of the story elsewhere: the idea of Vairocana’s lice is simply too funny, too quaint, and too downright memorable, to die quietly.  Once unleashed on the global imagination, I estimate it has many decades, maybe even several centuries, to run. If there’s one thing any English language reader remembers about the dBa’ bzhed, its likely to be Vairocana’s lice.

***

Finally, for any hard-core nerds who might be mad enough to read our foolish blog, here are some examples of khro chung from the ritual literature:

[1] This first passage is from Dudjom Rinpoche’s gNam lcags spu gri bsnyen yig, found in the section commenting on the Command (bka’ bsgo ba), which is at the beginning of the ritual.

གལ་ཏེ་མི་ཉན་པའི་རིགས་གདུག་པ་ཅན་ཡོད་སྲིད་པ་ལ་རང་ཉིད་ཧེ་རུ་ཀའི་སྐུ་གསུང་ཐུགས་ལས་སྤྲུལ་པའི་ཁྲོ་ཚོགས་མཚོན་ཆ་སྣ་ཚོགས་ཐོགས་པ་གྲངས་མེད་པར་འཕྲོས་རྣམས་ཀྱི་ཞལ་ནས་དྲག་སྔགས་རྡོ་རྗེའི་གད་མོ་འབྲུག་ལྟར་སྒྲོགས་ཤིང་མཚོན་ཆས་གཟིར།

But, in case there are some evil beings who do not obey, and who remain, from the body, speech and mind of yourself, the heruka, innumerable wrathful deities emanate, carrying various kinds of weapons.  They roar out the destructive mantra [oṃ sumbhani etc.] with vajra laughter like thunder, and they stab with their weapons.

རོལ་པའི་སྟངས་སྟབས་ས་གསུམ་གཡོ་བའི་ཉམས་དང་བཅས་ཏེ་བགེགས་རྣམས་འཇིག་རྟེན་གྱི་ཁམས་མང་པོ་བརྒལ་བའི་ཕ་རོལ་ལྕགས་རི་ཆེན་པོའི་རྒྱབ་ཏུ་བསྐྲད།

They manifest (various wrathful) postures with the thought of shaking up the three worlds, and the obstacles cross over the many worldly realms, and are expelled outside the great iron mountains.

སྤྲུལ་པའི་ཁྲོ་ཚོགས་ཚུར་འདུས་ནས་སྟེང་འོག་ཕྱོགས་མཚམས་རྣམས་དང་བདག་གི་བ་སྤུའི་བུ་ག་ཀུན་ནས་ཞལ་ཕྱིར་ཕྱོགས་པས་གཏམས་ཤིང་དྲག་པོའི་ཧཱུྂ་སྒྲ་དང་བཅས་ཏེ་ཚམ་རྔམས་ཆེན་པོས་རོལ་པར་བསམ།

The host of wrathful emanations are reabsorbed back (into yourself as the Great Glorious One), and then meditate that above, below, and in all directions, and from every hair pore of your skin, they again emanate, facing outwards, and pervading everywhere.  With a really fearsome cacophony, they roar forth the destructive sound of hūṃ.

(note:) With the destructive mantra recitation, music and bdellium incense, the vajra master casts mustard seeds (“thun rdzas”, power substances) at the obstacles, which symbolise the miniature wrathful deities.  

[2] The Sa skya Phur chen example is mentioned in our book. The Sa skya Phur chen (15v.4) speaks of the twenty-one thousand (body) hairs of oneself as the deity, filled with miniature Vajrakumāras.  In this example, it is in the main deity visualisation.  It is also included, word-for-word, in the short Sa skya Phur pa practice, the Nges don thig le of the Sa skya Throne-holder, Ngag dbang Kun dga’ blo gros (1729-1783):

རང་གི་ཕུང་པོ་ཁམས་དང་སྐྱེ་མཆེད་སྐྲ་ཉི་ཁྲི་ཆིག་སྟོང་བ་སྤུ་བྱེ་བ་ཕྲག་ཕྱེད་དང་བཞི་ལ་སོགས་པ་ཐམས་ཅད་ཀྱང་རྡོ་རྗེ་གཞོན་ནུ་རང་དང་འདྲ་བས་གོ་མཚམས་མེད་པར་རབ་ཏུ་གཏམས་པར་གྱུར།

All my own skandhas, dhātus, āyatanas, twenty-one thousand hairs, and thirty-five million pores of the skin are entirely* filled with Vajrakumāras like myself, leaving no gaps in between.

*The Kun dga’ blo gros Vol. Ga edition omits, rab tu, entirely.  (Note that the Sa skya Phur chen, f.15v.5, also omits it, but there is of course no need for the text to follow the Sa skya Phur chen exactly, and a copyist’ familiarity with the Sa skya Phur chen might have been responsible for such a scribal omission.)

According to Kongtrul (1813-1899), he lived during the second rab byung, which ran from  1087 to 1146, although some sources put his birth date earlier, in 1024. He was certainly one of the most brilliant and prolific treasure revealers of his time, and his influence on Tibetan religion remains impressive to this day. With the help of Ven Tenpa Yungdrung, Jean-Luc Achard, and others, we have been translating one of his most celebrated treasures, the Ka ba nag po or Black Pillar, which lies at the core of the contemporary Bon Phur pa tradition.

Yet Khu tsha zla ‘od did not only reveal Bon treasures: his treasure contributions to Buddhism, medicine and astrology were significant enough for him to find a place in the great collection of Buddhist treasures, the Rin chen gter mdzod, to feature in some Buddhist lineage prayers, and even to become considered a major Buddhist incarnation. Kongtrul for example saw him as the reincarnation of Vairocana, while Jamyang Khyenste Wangpo (1820-1892) saw him as one of the thirteen prophesied reincarnations of the royal figure rGyal sras lha rje, and hence a primary past incarnation of the great treasure revealers Orgyan lingpa (14^th century) and Khyentse Wangpo himself, amongst others (this information from Matthew Akester’s outstanding but so far unpublished translation of Khyentse Wangpo’s biography).  Thus it was Jamyang Khyentse Wangpo who rediscovered as a yang gter Khu tsha zla ‘od’s otherwise lost treasure, the rTsa gsum rigs bsdus spyi spungs chen mo. It was also Jamyang Khyentse Wangpo who believed that one of his own treasures, the rTsa gsum gtso bsdus spyi ‘dus snying thig, along with Jatson Nyingpo’s famous dKon mchog spyi’i ‘dus, were further iterations of Khu tsha zla ‘od’s rTsa gsum rigs bsdus spyi spungs. Nevertheless, the only arguably Buddhist treasure from Khu tsha zla ‘od that survives uninterruptedly, unmediated by a 19^th century yang gter, is a very short bcud len text. In addition however, Khu tsha zla ‘od was a famous doctor, and hence sometimes known as Ku sa sman pa. He was also a philosophical or doctrinal thinker. His commentary on Dzogchen, the mKhas pa mi bzhi’i ‘grel pa, has recently been studied by Matthew Kapstein, who shows him to have been a highly creative early adopter within Tibet of the logical thought of rNgog Blo ldan shes rab (1059–1109) (as Kapstein points out, it is the chronology implied by this feature of his thinking which serves to support the second rab byung dates as given by Kongtrul, rather than the first rab byung birth date of 1024 given in some other sources).

The Ka ba nag po comes in 125 pages, but it is only a small fraction of Khu tsha zla ‘od’s Phur pa revelation:  in addition to this root text, there are also eight other explanatory tantras, as well as commentarial and practice texts, together covering the entire gamut of Phur pa learning and performance.  Contemporary Bon Phur pa practice, including the rituals done at year-end (dgu gtor) in the Bon monatery of Triten Norbutse, are still based on Khu tsha zla ‘od’s all-encompassing Phur pa revelations, achieving a degree of integration between root text, practice text, and interpretation, that might well have delighted the late Steve Jobs. Some 600 photos and several videos of the year-end rituals at Triten Norbutse, specially made for us by Kemi Tsewang, will soon be visible on further pages of this blog.

There is a lot that could and should be said about Khu tsha zla ‘od. One could talk about the interesting stories of his childhood and family background; his prolific treasure recoveries, especially those found at sPa gro in Bhutan; the great fame of his medical practice; his mention in the biography of Guru Chos dbang; what evidence there is for him serving both Buddhist and Bon communities; the evolution of later beliefs linking his treasures with those of the 17^th century Buddhist terton Jatson Nyingpo (1585-1656); his innovatory uptake and adaptation of pramāṇa; the way his Phur pa treasures relate to the Buddhist Phur pa tradition; why, when so many other details differ, the gTing ‘dzin gsum or Three Samādhis in his Phur pa treasures are exactly the same as in Buddhist Mahāyoga; and so on.  But in this short blog article, we are going to mention only one particular aspect of his ouevre: the way in which he envisages and presents a great many minor Tibetan deities within his Ka ba nag po root tantra. Our method of analysis depends substantially upon paying fine and close attention to subtle nuances and differences in the sphere of ritual, yet it seems to us that it might yield useful historical conclusions. Once again, we are reminded of the point we have made elsewhere, that ritual texts and liturgies can offer important data to the historian, if one knows what one is looking for.

The Ka ba nag po is remarkable for its presentation of unusually long and detailed lists of minor deities, including, for example, numerous species of female sman. In Buddhist Phur pa rituals, such deities would most likely come with a taming narrative attached, either implicit or explicit. We would understand how Padmasambhava had tamed them in days gone by and appointed them as guardians of the dharma, and how we should in our own time, as the authentic dharma heirs of Padmasambhava, remind them of those vows, and reassign them to their appointed tasks with an offering.  Yet in the Ka ba nag po, the emphasis is subtly but tellingly different.  In Khu tsha zla ‘od’s treasure, such deities do not need taming, since they have been an integral part of the nirvāṇic maṇḍala of the main Phur pa deity since beginningless time. In short, it is made quite explicit that they are nothing less than the emanations of the main Phur pa deity, and have never been anything else, an interpretation which is still upheld by the contemporary commentarial tradition.

Of course, Buddhist Phur pa can and must also adopt such a perspective on occasion: there are many ritual moments when pure vision is emphasised and all phenomena whatsoever are perceived as primordially pure according to their ultimate nature. But this is balanced by a relative viewpoint, in which such worldly spirits and demons do indeed exist on a conventional level as potentially problematic forces in need of careful management.  It is in the interpretation of the relative viewpoint that the Ka ba nag po differs: here the sman (and all other suchlike categories, such as bdud) are clearly described as direct emanations of the central enlightened Phur pa deity even from a relative point of view, and thus in no need of taming at any level.

What might  Khu tsha zla ’od be intending with this?  To approach this question, let us first turn to Sam van Shaik’s recent blog entry, in which he revisits F.W. Thomas’s work on one of the wooden sticks from Miran (IOL Tib N 255). These inscribed sticks have been shown to be older than most of the Dunhuang texts, and so the analyis of IOL Tib N 255 can offer us, among other things, chronologically reliable evidence that the sman were amongst the categories of deity recognised by Tibetans as early as the 9^th century. In IOL Tib N 255, the sman are in fact mentioned alongside such terms as yul lha and g.yang, within the context of funerary rites in which Bon and gShen priests worked together in partnership. In other words, this and other such evidence suggest a strong probability that the sman were a component part of a ritual belief system that pre-dated the accelerated uptake of Buddhism subsequent to the visit of Śāntarakṣita and the promotion of Dharma by Khri Srong lde’u btsan.

In now valorising the sman as fully enlightened emanations of the nirvāṇic maṇḍala of the central Phur pa yidam deity, and by refusing to accept that they needed to undergo the humiliation of taming before being worshipped, Khu tsha zla ‘od in the 12^th century seems to be making a powerful statement. He is asserting the value and dignity of those deities and beliefs that were present and flourishing in Tibet before the great flood of Indian Buddhism became so dominant a few centuries earlier, and rejecting their complete relegation to the folkloric margins. This strategy in the Ka ba nag po is quite possibly of a piece with his strategy in the mKhas pa mi bzhi’i ‘grel pa, as reported by Kapstein. In that more doctrinal text, Khu tsha zla ‘od inserts as an equal partner the indigenous intellectual category of bla into his discussion of the prestigious Indian philosophical category cluster of kun gzhi’i rnam shes (Sanskrit: ālayavijñāna), yid (Sanskrit: manas) and sems (Sanskrit: citta), thereby asserting the equal profundity and value of the indigenous category to the Indian ones.

It seems that Khu tsha zla ‘od’s followers were intended to have the best of both worlds. On the one hand, they could fully enjoy excellently made calques upon Indian-style Mahāyoga, and Buddhist logic-inflected commentaries on Dzogchen, that promised to offer, in ritual and contemplative terms, everything the Buddhist equivalents could offer.  On the other hand, they could enjoy these without the need to repudiate their ancestral culture, deprecate their native gods, or disparage their indigenous intellectual categories. We can only surmise that Khu tsha zla ‘od was adressing an audience for whom such considerations were important­—who knows, perhaps even an audience including some proud and loyal descendents of the once great Bon and gShen priests of old.

Although remote enough, Himalayan, and to the south, I can see why the Orgyan Ling temple in the Mon Tawang region in modern day Arunachal Pradesh might not have any such legends attached to it. It was not nearly obscure enough, and it ended up being sacked by a hostile army, with its many Nyingmapa treasures destroyed or carried away. Yet despite this rather serious setback, I would like to suggest that albeit in a rather tortuous way and perhaps inadvertently, it did nevertheless achieve one of the prime purposes ascribed to a ‘Hidden Land’. My reasons for proposing this have nothing to do with mystical revelations from Guru Rinpoche nor signs from the dakinis, but derive instead from the prosaic practice of philology. Orgyan Ling temple succeeded in preserving intact for posterity an important and famous Nyingma tantra in a considerably purer and more complete form than anywhere else in Tibet. Not only that, but it also preserved a number of other very rare Nyingma tantras not so far found anywhere else in the Tibetan cultural world.

Orgyan Ling was founded in the late 15th century by Orgyan Zangpo, the younger brother of the famous Bhutanese terton Pema Lingpa (1450-1521). In 1683, Orgyan Ling subsequently became the birthplace of Orgyan Zangpo’s famous descendent, the Sixth Dalai Lama Tsangyang Gyamtso (1683-1706).  Thus it was that in 1699, the Fifth Dalai Lama’s regent, the great Desi Sangye Gyatso, greatly extended Orgyan Ling’s buildings and commissioned new volumes for its library.  So despite its remote situation at the very edge of the Tibetan cultural sphere, its galactic network of family and personal connections enabled Orgyan Ling’s library to acquire some very fine books indeed, including a copy of the Ancient Tantra Collection (rNying ma’i rgyud ‘bum), and two Kanjur manuscripts.

The older of the Kanjurs was made in silver and gold, and both Kanjurs contained a quota of sixty Nyingma tantras freely intermingled with their Sarma tantras. This number greatly exceeds the merely twenty-four squeezed into the carefully segregated doxographical ghetto that had been reserved for them by Situ Gewai Lodro (1309-1364) in his Tshalpa Kanjur catalogue. Yet the Thempangma Kanjur traditions don’t have any Nyingma Tantra section at all. Orgyan Ling was evidently such a Nyingma stronghold that even its Kanjurs unashamedly sported unusually large numbers of Nyingma tantras, unsegregated from the Sarma tantras. We can only dream about what its long lost Ancient Tantra Collection might once have contained.

But fame is a two-edged sword which can prove fatal to any Himalayan valley’s chances of successfully qualifying as a ‘Hidden Land’. Not content with merely deposing him in Lhasa, after his death Tsangyang Gyamtso’s Mongolian enemies extended their sectarian vendetta to his birthplace, sacking the monastery of Orgyan Ling around 1714 in the course of a campaign against Bhutan.

Fortunately however, a quantity of the monastery’s surviving valuable books and artifacts were transferred to the Geluk establishment of Ganden Namgyal Lhatse in Tawang, even though the gold and silver Kanjur was partially destroyed, and the Ancient Tantra Collection disappeared.

Among the volumes transferred to Ganden Namgyal Lhatse was the Kanjur manuscript commissioned by the Desi Sangye Gyatso in 1699, beautifully calligraphed by master scribes especially hired from the E region of Central Tibet. Fortunately, this Kanjur still survives, together with its full complement of sixty Nyingma tantras. But it is not its master calligraphy that concerns us (although that is certainly impressive), nor its fine thick paper: rather, it is the presumably lost exemplars from which the master scribe must have made his copy, and the textual traditions they represented.

Over the last few years, Cathy and I have been engaged in making a critical edition of an important Nyingma tantra called the  ‘Phags pa thabs kyi zhags pa padma ‘phreng gi don bsdus pa, which one might translate as A Noble Noose of Methods, the Lotus Garland Synopsis (hereafter abbreviated as Thabs zhags). In addition, we have simultaneously edited its commentary.

The Thabs zhags is a magnificent example of Mahāyoga literature, often juxtaposed with the Guhyagarbha in classic doxographies, and its commentary is one of the few remaining genuinely early word-for-word commentaries on a Mahāyoga tantra. While we don’t adhere to the unrealistically perfectionist dogma that every text in existence must be exhaustively critically edited as a matter of principle, as academic scholars we do share the perception of numerous learned Tibetan lamas of the past and the present, that a certain proportion of texts in the popular editions of the Tibetan canon can end up in rather bad shape, and are ideally in need of critical editing. It is specifically these damaged texts that we seek to restore.

The Thabs zhags and its commentary are among such damaged texts. The commentary survives in three different Tenjur editions, yet in all three instances, around a third of its text has become completely lost: for example, everything between Chapter Six and the end of Chapter 10 has disappeared, and there are other significant losses of text too.

In most of its more popular canonical versions, the root tantra has only slightly less serious problems. These more popular versions derive either from the Tshalpa branches of the Kanjur (which here must includes the Derge Ancient Tantra Collection, since it borrowed the relevant woodblocks from the Derge Kanjur), or from the Bhutanese Ancient Tantra Collection manuscripts. Yet they reveal a pattern of confusion: the respective renderings of the final section of the very long Chapter 11, for example, are so radically different in the Tshalpa Kanjur and Bhutanese recensions, that to the untrained eye it seems hard to credit that they are reproducing the same text at all.

To remedy this situation, we assembled twenty one different witnesses from every conceivable source, and began to try to piece together the evidence.

The commentary was reasonably easy to fix. Whilst the transmitted tradition of the Tenjur was irrevocably damaged, ‘new’ evidence in the form of a partially complete 10th century Dunhuang manuscript was discovered in the early 20th century and brought to London by Sir Aurel Stein. This enabled us to restore the text to its complete form, almost certainly for the first time in several centuries, since, by great good fortune, those parts missing in the Tenjur were preserved in the Dunhuang manuscript, while those parts missing in the Dunhuang manuscript were preserved in the Tenjur. It became a simple matter of splicing.

The root text proved more complex, but nevertheless exhaustive collation and stemmatic analysis permitted us a clear result. It became evident that the currently three most popular versions – those of the Tshalpa Kanjurs, of the Derge Ancient Tantra Collection, and of the Mtshams brag Ancient Tantra Collection from Bhutan – represented two distinct attempts to reconstruct the root text from out of the commentary. At some stage in its long history, the root text must have been lost in the main centres of Nyingma learning, perhaps during one of those periodic persecutions of the Nyingmapas that made them want to retire to the ‘Hidden Lands’ in the first place.  Two different attempts were then made to recover the root text by extracting its verses out of the word-by-word commentary, one attempt witnessed in the Tshalpa Kanjurs and the Derge Ancient Tantra Collection, and the other attempt witnessed in the Bhutanese Ancient  Tantra Collection manuscripts of Mtshams brag, Sgra med rtse, Sgang steng-A and Sgang steng-B.

Once we had the full commentary available to us together with the evidence from our twenty one witnesses, it became unfortunately evident that neither of these brave attempts at reconstruction had been entirely successful. Both in their different ways and at different junctures mistakenly introduced commentarial passages into the root text, whilst also excluding passages of root text through the mistaken assumption that they were commentary. Nor did they agree in their errors – hence for example their radically different interpretations of Chapter 11.

But this is where the ‘Hidden Land’ principle comes into play. While these famous mainstream texts of the Tshalpa Kanjur, the Derge Ancient Tantra Collection, and the Bhutanese Ancient Tantra Collection were uniformly confused about what was root text and what was commentary, all around that great geographical arc that comprises the southern Himalayan fringe of the Tibetan cultural zone, in sometimes obscure and ignored local monasteries, versions of the text survived that derived not from one or another attempt to reconstruct the root text from the commentary, but directly, from the original root text itself. In the forgotten ruins of an old library in Hemis, along the Nepalese borders, and even in far-off Bathang in the North-East of Tibet, ancient versions of the text, or their direct copies, survived. It was through studying these that we were able to restore the text to its original boundaries, differentiating root text from commentary with clarity and ease. Applying stemmatics to all the witnesses together, we were also able restore many other original readings.

But the last word belongs to the excellent Orgyan Ling manuscript. It can only be consulted in one place, in the library of the Central University for Tibetan Studies in Sarnath, and microfilm is not easily available. Hence we did not get to see it at all until a provisional version of our edition was already complete. To our delight and astonishment, it agreed with our stemmatic reconstruction at every significant juncture. Not only that, but its superior scribe from E had even eliminated spelling errors and many other small problems, just as our laborious edition had attempted to do. We found ourselves face to face with a beautifully calligraphed late 17th century version of the text that was virtually identical in all important respects to the one we had so painstakingly reconstructed over the previous three years. The only differences were that the Orgyan Ling manuscript preserved a few more archaic spellings than we saw fit, and its scribe had fallen for one minor slip of the pen when describing a mudra.

The moral of the story is this: we should not allow ourselves to become excessively swayed by the grand reputations and religious-political support invested in the few mainstream editions of the canons. Don’t forget the ‘Hidden Land’ principle, which tells us that just occasionally the most valuable editions can be located in the least expected places. This was something well understood by the great Tibetan critical editors of the past, men like Situ Panchen, Tsongkhapa or Kathok Rigdzin Tsewang Norbu, who became dissatisfied if they found reason to suspect that the standard canonical fare was deficient, and who left no stone unturned in seeking out rare witnesses in Sanskrit and Tibetan to complete their own critical editions.

Current palaeographical knowledge locates the Dunhuang manuscript to the latter half of the tenth century. However the text itself must be older than its Dunhuang witness:  our critical edition can demonstrate that an ancestor of the versions from the local Kanjurs of Bathang, Hemis, and Tawang, and from the South Central Tibetan rNying ma’i rgyud ‘bum editions, was older than the Dunhuang manuscript by more than one copying at the least.

The Dunhuang manuscript mentions Padmasambhava four times: once in the marginal notes at the beginning, twice in the marginal notes near the end, and once within the main text of the commentary itself, also near the end. The references are somewhat enigmatic, and we will be publishing on them at greater length elsewhere, so here I will only review our findings in brief.

Ken Eastman in the 1980′s was the first to look at these references, and had tentatively suggested they might be presenting Padmasambhava as the human author of the commentary. Jake Dalton and Sam van Schaik follow him in taking much the same line, albeit quite assertively and no longer tentatively.[2] However, despite the difficulty of the materials and the rather complicated way in which the root text, commentary and marginal notes cross-reference, neither Ken Eastman, Jake Dalton or Sam van Schaik had time to study the text in much depth or for very long, and none have written more than a few pages on it.

After a much more detailed study, it now appears quite uncertain that Padmasambhava is being represented as the author of the commentary, and it is important that the scholarly community takes note of this fact. Rather, there is a distinct emphasis on portraying him as a sublime realised being with exceptional access to the Tathāgata’s secret teachings, and quite possibly even as the source of the root tantra itself.

The references it makes to Padmasambhava are not entirely clear and unambiguous, since they assume the reader already has such information, but what is clear and unambiguous is that these are references to an exceptional, mythologised being, and not to an ordinary human teacher. At its end, the main text of the commentary lavishly praises Padmasambhava as padma rgyal po, the ‘Lotus King’, in verses which the accompanying notes explain are being addressed by Śāntigarbha to Padmasambhava. It is fascinating that these verses comprise a precise form of laudatory words picked up two centuries later by Nyang ral Nyi ma’i ‘od zer and the wider hagiographical tradition in their own praises of Padmasambhava, and Nyang ral again specifically links these words to the Lotus King, a form which remains canonical as one of the famous Eight Aspects of Guru Rinpoche.

Figure 1: Padma rGyal po, as depicted in the Ritual Dance of the Guru’s Eight Aspects (gu ru mtshan brgyad ’chams), Jangsa Monastery, Kalimpong, 2009. Photo by Cathy Cantwell.

The praises from IOLTibJ321 and from Nyang ral's Zangs gling ma

As you can see in the diagram, the verses say that Sam bha ba is “he who has attained the supreme siddhi, of great wonder, Padma rGyal po [The Lotus King] (who) is not worldly; (he who) unravels from the expanse the tathāgata’s great secret pith instructions“. Note the use of the Tibetan word ma ‘gyur here in the Dunhuang manuscript. It is not a very natural Tibetan usage, and like many phrases in these old manuscripts, it’s not entirely clear what it means. Where ‘gyur ba occurs in Tibetan literature, it is often as a translation of the Sanskrit bhūta. Literally, both ‘gyur ba and bhūta mean ‘become’, but most often, the idiomatic meaning of the Sanskrit is simply ‘who is’. Tibetans tended to translate bhūta literally rather than idiomatically, thus giving ‘gyur ba (to become) simply to render ‘who is’.[3] So there is some albeit rather tenuous suggestion here that the verse of praise might have somewhere in its background or prehistory, something translated into Tibetan from Sanskrit.

The marginal notes attached to this praise are slightly ambiguous, explaining that after examination, Śāntigarbha finds either Padmasambhava himself, or his teachings, flawless, and is praising him.^[4] Right at the start of the text, the notes already told us that while the Buddha has condensed (the meanings) of the root text (‘bu tas bsdus), it was Sambhava who produced or made (them) (sam ba bhas byas)—a similar meaning to Śāntigarbha’s praise of him here for unravelling the secret great pith instructions of the tathāgata from the expanse.

Finally, a page above this praise which ends the commentarial text, right at the end of the root tantra itself, a marginal note refers to Padmasambhava.  This annotation seems possibly to suggest that what has gone before, namely the speech of the tantra, which has emerged naturally out of sameness, was demonstrated by Padmasambhava without any personal fabrication or rang gzo.  There follows an explanation of how, given this natural emergence, when a noble being speaks with pure awareness, the resulting utterance is Tantra.[5]

Thus, Padmasambhava seems to be closely involved with the Buddha’s original teaching of the tantra, in terms that go some distance to making him sound like a treasure revealer of some kind, and so one might speculate if the name ‘Padma’ in the texts’s title might conceivably be referring to Padmasambhava.  This, however, is unlikely: the first chapter of the commentary, which discusses the title, gives no hint of the word, “padma”, having any such implication.

I should add, these findings in IOL Tib J321 have not been remarked upon by previous scholars, but we feel they might add significantly to our knowledge of 10^th century representations of Padmasambhava. We will deal with them at very much greater length in our forthcoming book.

Finally, a note on the citation from Nyang ral Nyi ma’i ‘od zer: Lewis Doney of the School of Oriental and African Studies, University of London, has worked on critically editing Nyang ral’s Guru Padma hagiography.  He argues convincingly that the earliest and historically most influential recension is that represented by two manuscripts in the National Archives in Kathmandu and two manuscripts from Bhutan, which he classifies as ZL3.  The version of ZL3 used here is Lewis Doney’s identification in the Kathmandu National Archives.  We have emended rtog in line 2 to rtogs, found in all the other witnesses of ZL3.  The Rin chen gter mdzod chen mo version (Paro: Ngodrup and Sherab Drimay, Kyichu Monastery, 1976, Volume Ka: 25), which has in modern times become the most widely used version, incorporates later material.  It gives a variant second line for this verse (rtogs ba bla med mchog tu gyur pa yis/).

For Tibetan tradition, the answers are simple. Padmasambhava was a peerless guru with the vidyādhara’s control over lifespan, who became revered in Tibet when Emperor Trisongdetsen invited him, by which time he had already been a living legend in India for many centuries.

Modern academics are denied such beautiful and easy answers.  In general we are permitted to accept as valid evidence far less data than traditional Tibetan historians, and in few places is this felt more acutely than the history of Padmasambhava: for modern scholarship, the admissible historical evidence for the person or even for his representation is very slight indeed.

Following the digitisation of the Dunhuang texts over the last decade, we have happily seen a small augmentation of the available evidence for the early representations of the great Guru, even if not for the great Guru himself. Part of that augmentation has come from the discovery of a new Dunhuang source, and part from a more intensive analysis of already known sources.  However, Cathy and I are not convinced that the implications of the new source have so far been fully appreciated, nor that the bigger picture as it should now stand has been properly assessed. In this multi-part blog I want to present a more thorough interrogation of the new source of evidence, together with a fuller investigation of the already known sources, to arrive at a more complete depiction of what we can now know about the prehistory of  Padmasambhava’s early representation if we put all the available evidence together.

The most convenient summary of how the historical Padmasambhava looked to modern scholarship before the digitisation of the Dunhuang texts comes from Matthew Kapstein. Writing in 2000, the only admissible evidence then available to him was fourfold:

(i) The early historical text, the Testament of Ba, which presents Padmasambhava visiting Tibet.

(ii) The 10th century Dunhuang text PT44, which narrates Padmasambhava bringing the Vajrakīla tradition to Tibet.

(iii) An early text attributed to Padmasambhava called the Garland of Views, or man ngag lta ‘phreng, and a commentary on it by the 11^th century rNying ma sage, Rong zom

(iv) The termas of Nyang ral (1124-1192) and Guru Chowang (1212-1270), which present fully-fledged apotheoses of Padmasambhava as a fully-enlightened Buddha.

Based on this evidence, Matthew Kapstein concluded that:

(i) The Testament of Ba shows Padmasambhava quite likely did visit Tibet during Trisongdetsen’s reign.

(ii)  PT44 indicates followers of his tantric teachings were active in post-Imperial Tibet.

(iii) Rong zom’s commentary and the few Dunhuang references show that the Padmasambhava cult began its ascent during the ‘time of fragments’, between the end of Empire and the start of the gsar ma period in the late 10^th century.

(iv) Nyang ral and Chowang’s termas suggest the most massive elaboration of Padmasambhava’s cult developed from the 12th century.[1]

Since Matthew Kapstein published that in 2000, there have been two further developments. Firstly, a new Dunhuang source, PT307, was felicitously discovered by Jacob Dalton, who published an article on it (this article also deals with another Dunhung text, TibJ644, that Jake Dalton had initially hoped related to Padmasambhava, but after some last-minute discussions between us, we decided was not yet so conclusively established as initially hoped).[2]

Secondly, Cathy and I are completing a much more detailed analysis than has hitherto been attempted of the evidence from the Dunhuang text IOL Tib J 321, looking at it more carefully than Ken Eastman’s short note from the 1980’s,  Jake Dalton brief survey in 2004/6?, or Sam van Schaik’s small blog entry in 2007. It is largely these two sources of evidence that will inform my paper today, together with a reassessment of the already well-known sources PT 44 and the Testament of Ba.

So what can the new and fully admissible evidence tell us different from what Matthew Kapstein wrote in 2000? It is a tribute to his historical discipline, and the caution with which his reasoning neither exceeded nor underrated the scanty evidence, that the advances we can now report consist more of detail than of substance. Kapstein wisely put no definite dates on any particular aspect of the Padmasambhava cult, which he understood as a gradual process developing throughout the post-imperial period, coming to some sort of culmination with Nyang ral three centuries later. What is new is that we now have much stronger evidence that the mythologisation of Padmasambhava, his incorporation into ritual, and it seems even his apotheosis, began within the earlier part of the very long time frame Matthew Kapstein suggested. In other words, when portraying Padmasambhava in his famous hagiographical and historical writings, Nyang ral was developing existent themes, rather than inventing new ones. Our new evidence suggests that Padmasambhava was already the object of religious myth and ritual worship, and was probably already seen as the enlightened source of tantric scriptures, as many as two hundred years before Nyang ral, even if not yet with so much poetic elaboration. An important proviso is that we have not yet ascertained if the evidence bears witness to a widespread veneration of Padma in the tenth century, or something narrower followed only by a few. This is because the evidence currently available suggests two different views of Padma in the early sources:

• Firstly, in the context of the possibly early or mid 10th century rDzogs chen oriented bSam gtan mig sgron of gNubs sangs rgyas ye shes, he is cited as a great teacher and even mythologised, but no more so than his peers like Vimalamitra, and there is no sign of his integration into ritual (but of course, we cannot yet tell for sure if the sole available version of this text includes later interpolations or not).
• Secondly, in the possibly late 10th century Mahāyoga texts from Dunhuang, he is mythologised, incorporated into ritual, and elevated above his peers, even apotheosized. The available versions of the testament of Ba seem to broadly concur with this.

Jake Dalton has in the last five years or so emerged as one of the most frequently cited interpreters of the early rNying ma pa, and is widely renowned as a highly innovative and interesting scholar. He has recognised, as many others such as Anne-Marie Blondeau and Matthew Kapstein did before him, that there is real evidence for some kind of a Padmasambhava cult from the 10th century or earlier. However, we think his work in this particular instance has not yet exhausted the possibilities for these sources, despite his early access to them. This is because Jake has not focused enough on the ritual function of the Dunhaung texts related to Padmasambhava, nor their connections with Mahāyoga; approaching them mainly in terms of their narratives divorced from ritual context, he has several times arrived at incomplete or even slightly inaccurate conclusions. By not recognising the ritual clues, he has significantly underestimated the full import of these extraordinary sources whose evidence for the early Padmasambhava cult is in fact quite a lot richer than he realises.

I was impressed to see on a YouTube interview made on his campus that Paul Harrison of Stanford University, although not a practising Buddhist, has learned the Diamond Sūtra by heart and recites it daily, to introduce a performative understanding into his academic research on this text.  In the same way, academic scholars of tantrism might benefit sometimes by actually performing tantric rituals themselves, to get a more nuanced view of things.

In the subsequent parts of this blog, I will set out in detail the further insights we can add to Jake Dalton’s, and indeed Matthew Kapstein’s and various other scholars’ earlier findings, by approaching the Dunhuang evidence for Padmasambhava through the lens of ritual. After all, it has never been doubted that PT44 and PT307 must be connected in some way with ritual. No one has ever suggested otherwise, because both comprise stilted narratives culminating in explicitly ritual passages. The reason they have not so far been studied through the lens of ritual is probably one of methodological tradition: in the short history of their subject, unlike the anthropologists, academic tantric textual scholars have quite simply never been expected to approach their materials from a performative perspective. I wonder if it might be time for that to change.

Anyone who has read much Tibetan literature will be familiar with one of its most salient differences to our own modern conventions: the ubiquitous verbatim repetition of phrases, sections, literary structures, and even entire chapters, across many different texts. Such repetition is commonplace even where these many different texts are written by ostensibly different authors. Some modern scholars have rather condescendingly (and stupidly) characterised this as ‘plagiarism’. They have entirely missed the point. Others avert their gaze in embarrassment from such features, and instead try to emphsasise the aspects of Tibetan literary culture that chime with our own values: originality, innovation, sceptical inquiry, and so on. But they have also missed the point to some extent, because although there is absolutely no doubt that such values do show themselves in Tibetan literature, whether in individual creativity or in the cultural generation of new genres, they occur against an established backdrop of quite different traditional literary norms.

What are these traditional norms? First of all, Tibetan religious literature, including Treasure, is sometimes (not always!) de facto collectively rather than individually produced: close inspection reveals that the final product has the input of more persons than the nominal ‘author’, often extending backwards (and even forwards) over considerable stretches of time. Much is also recycled, within a literary culture that normatively envisions contributors as tradents rather than innovators: in other words, the person producing a text sees himself as passing on existing knowledge, rather than creating new knowledge from nothing (I will elaborate further on the term tradent below). Texts can be substantially modified by other hands in subsequent re-publications, even while still retaining their original authorial (or revelatory) attribution. At other times, modification can be generated silently and less deliberately via the subtly transformative medium of memorisation: we must recall that Tibetan scholars carry huge tracts of literature around in their minds, which they can access instantly without recourse to a written book, but sometimes it comes out in a form ever so slightly different from other or previous iterations. I think one can even say that the Buddhist tradition often understands authorial attributions as a conventional shorthand indicating the accepted presiding spiritual authority in a given literary instance, rather than as the sole or exclusive literary agency that created it.

All this bears little resemblance to modern literary ideals, in which the author is constructed somewhat heroically as an individual creative source. Yet despite a vague general awareness of such differences, Tibetologists have never systematically addressed the issue, and we now have numerous detailed studies of works traditionally attributed to famous Tibetan sources, without further investigation into what such attribution might actually entail in each individual case. I have often felt that the time is long overdue for a new analysis of Tibetan religious authorship, and some new items of vocabulary to describe it.

After some years vaguely wondering how exactly to articulate such an approach and  vocabulary, in May 2008, I was rewarded with an answer. I invited Jonathan Silk to give a guest lecture, and aware of my interests, he obliged by delivering a wonderful paper entitled What Can Students of Indian Buddhist Literature Learn from Biblical Text Criticism? Although aimed at Sanskritists, it was obvious that what Jon was saying also applied to Tibetan studies. Suddenly, I came to the realisation that right here under my nose, in my own Wolfson College, Talmudic scholarship had already articulated much of the approach and vocabulary I needed – and yet I had not been aware of it![1]

Talmudic scholars no longer depend on the conventional modernist language of ‘authorship’ and ‘work’. Instead, they can speak of ‘tradents’, who  ’re-anthologise’ existing ‘lemmata’ and ‘microforms’, sometimes anonymously, within the context of a culture of extraordinary textual memorisation and the ubiquitous synchronous interactions of written and oral modes of text.  We have a lot to learn from them, because Tibetan religious literature is in some important respects closer to Medieval Hebraic literature than to modern literature.

What do these terms mean, and why are they useful for us? Obviously, I can only give very brief answers here. Firstly, the word ‘tradent’ indicates a producer of text who sees as his main project the passing on of existing spiritual truths, rather than the invention of new ones ex nihilo. Since he is largely engaged in passing on existing truths, he tends to seek out existing materials of proven Dharmic worth, to use as building blocks with which to construct his new text. At the most elemental level, these building blocks comprise well established fundamental Dharmic categories, such as ‘The Three Jewels’, ‘The Four Activities’, ‘The Hundred Peaceful and Wrathful Deities’, and so on. Some Hebraists call such fundamental categories ‘lemmata’. Longer passages, such as a paragraph or chapter comprising composites of such lemmata, are also legitimately reproducible either approximately or verbatim, according to Tibetan norms. Some Hebraists would call such reproducible composites that are not yet a complete work ‘microforms’. Finally, a complete work, such as the Guhyasamāja Tantra, or a commentary upon it, is called a ‘macroform’. The way such literary constructions are put together resembles an ‘anthological’ model: tradents select existing lemmata and microforms and re-anthologise them to make new wholes.

I cannot agree with those modern scholars who find such literature lacking in creativity. A good analogy is Lego: imagine if you ask two persons to make a square palace out of Lego bricks, one a great sculptor and architect, the other just an ordinary person. Clearly, the results would not be the same in quality, notwithstanding the restriction placed on the materials used and the outcome required. In the same way, some Tibetan tradents can produce works of astonishing subtlety and brilliance, while others can be unremarkable and predictable (of course, if they follow their cultural template accurately enough, both will at least produce a work of some value).

Literary production in Tibetan Buddhism, as anywhere else,  is a process, and in order to understand it, we need to track its processes minutely, step by step, bit by bit, stage by stage. Fortunately, we have been awarded funding from the AHRC[2] to do just this, in an international project based here at Oxford. Our local personnel are Vesna Wallace and Cathy and myself, while our international partners and consultants include Janet Gyatso, Sarah Jacoby, Matthew Kapstein, Jonathan Silk, Lopon P. Ogyan Tanzin, and Antonio Terrone. Part of the project is simply to minutely track all the processes, over several generations, that gave us some of the terma literature we know so well today, while another part will be to achieve critically-aware knowledge transfers from Hebrew studies and the English medievalists into Tibetology. Through this, we aspire to help catalyse a broader debate on what authorship really means in Tibetan religious writing as a whole, in other genres beyond terma, so that our analysis might contribute to the understanding of Tibetan religious writings as a whole. I hope it will be a rewarding study for all concerned.

[1] Peter Schaeffer wrote much of his most important work at Wolfson, while I was a graduate student there, yet I never encountered it at the time.  Jacob Neusner, Martin Jaffee and others working in medieval Hebraic literature have of course contributed equally to the debate.

[2] The Arts and Humanities Research Council, the main source for Humanities research in the UK.

Last year, Cathy Cantwell established that a substantial component text from one of the most important early terma collections, Nyang ral’s famous Deshek Dupa (bde gshegs ‘dus pa), corresponds exactly to the Dunhuang manuscript IOL TibJ 331.III.[1] While the Dunhuang version is anonymous, Nyang ral claimed the original author of his terma was Vimalamitra, a famous Indian scholar traditionally associated with the Dzogchen teachings and held to have come to Tibet in the late 8th century, at the time of Padmasambhava and Emperor Trisongdetsen.

IOL TibJ 331.III is a text that we studied in detail some years ago, in one of our earlier projects on Dunhuang manuscripts.[2] Our earlier study established that this very same text also appears, albeit in slightly different order, as chapters 8-11 of a historically transmitted Nyingma tantra called the ‘Perfection of Activities Tantra’ (‘phrin las phun sum tshogs pa’i rgyud). In less exact replications, it also occurs in some other Nyingma tantras as well (ibid. 76-87). Clearly, this is a text with many ‘incarnations’.

Leonard van der Kuijp has established Nyang ral’s dates as 1124-1192. We know the Dunhuang texts are a lot older, and current palaeographic thinking attributes the handwriting style of IOL TibJ 331.III to the mid-tenth century, which is just before the start of the Sarma (gsar-ma) or New Translation period, and two hundred years before Nyang ral. We also know that Nyang ral was a very major scholar of the Nyingma tantras, although we do not yet know exactly which ones were studied by him or in which redactions.

What are we to make of Nyang ral’s terma discovery? Might he have extracted his terma from within the pages of canonical Nyingma tantras, such as chapters 8-11 of the ‘Perfection of Activities Tantra’? As far as we currently know, not very likely, since none of the versions we have so far found in any of the canonical tantras are as close to the Dunhuang text as is Nyang ral’s terma. According to current evidence, the most likely hypothesis is that Nyang ral found an old stand-alone manuscript corresponding to IOL TibJ 331.III, perhaps one that had been lost for some time, and put it back into circulation, supplying in the process a substantial commentary, and an attribution to Vimalamitra.

Cathy’s discovery that the text of Nyang ral’s terma existed at Dunhuang about 200 years before Nyang ral’s time raises many important questions about early terma, and about early Nyingma tantric literature as a whole. We are planning a substantial research project on these issues, all the more so since a detailed look at Nyang ral’s phurpa writings has long been on our agenda.

It is widely held within Tibetan Studies that whilst Bon terma were often simply conceived of as lost old manuscripts that eventually surfaced again, Nyingma terma are more typically associated with and largely dependent on additional visionary procedures, in which the text revealer ‘remembers’ the transmission of the teaching to him by Padmasambhava or some other great master in a past life. Clearly, as Dan Hirshberg suggested, and as some others have more tentatively suggested before him, a reliance on visionary procedures to produce the text need not always or on all occasions have been the only method. Whilst extant redactions of Nyang ral’s Deshek Dupa can contain such items as the terma punctuation mark (gter shad) and the Ḍākinī Symbolic Script (mkha’ gro brda yig), which nowadays are suggestive of the mystic processes of remembering, we must also conclude that at least one substantial stand-alone component part of the huge Deshek Dupa cycle looks like it was primarily based on a found manuscript.

In a subsequent posting, I will mention some other correspondences between Dunhuang manuscripts and Nyang ral’s writings, concerning early Nyingma traditions of Padmasambhava.

[1] Dunhuang is a Buddhist site in North Western China where a substantial library of texts was recovered intact in the early 20th century, undisturbed since their interment in the early 11th century. It contains a many manuscripts from an earlier date still.

[2] See pages 68-135 of our Early Tibetan Documents on Phur pa from Dunhuang.Vienna, The Austrian Academy of Sciences Press, 2008.