Journal of the International Association of
Buddhist Studies
Volume 21 • Number 2 • 1998
PIERRE ARfcNES
Herm6neutique des tantra:
6tude de quelques usages du «sens cach6»
173
GEORGES DREYFUS
The Shuk-den Affair: History and Nature of a Quarrel
227
ROBERT MAYER
The Figure of MaheSvara/Rudra
in the rNin-ma-pa Tantric Tradition
271
JOHN NEWMAN
Islam in the Kalacakra Tantra
311
MAX NIHOM
Vajravinaya and VajraSaunda:
A 'Ghost' Goddess and her Syncretic Spouse
373
TILMANN VETTER
Explanations of dukkha
383
Index to JIABS 11-21, by Torn TOMABECHI
389
English summary of the article by P. Arenes
409
ROBERT MAYER*
The Figure of Mahesvara/Rudra in the rNin-ma-pa
Tantric Tradition
Most academic scholars of Buddhist Tantra should by now be reasonably
familiar with Alexis SANDERSON'S work on the dependence of the
Buddhist Yoginltantras on £aiva scriptural sources - all the more so now
that increasing quantities of this material is beginning to filter down
from the somewhat hermetic confines of the academic research seminar,
and into the bibliographical bases of such popular and best-selling works
as Gavin FLOOD'S widely praised Introduction to Hinduism, and
Miranda SHAW's more controversial Passionate Enlightenment.1
S A N D E R S O N ' S work I am referring to includes his published paper
"Vajrayana: Origin and Function",2 as well as a number of unpublished
(yet nevertheless quite well-known and widely-circulated) seminar
papers and public lectures given over the years at various universities
and institutes.3 For specialists in Tantric Buddhism, the most significant
* This paper is a slightly revised version of a lecture given to the Tantric Studies
Seminar at All Soul's College, Oxford, on April 29, 1996.
1. Gavin FLOOD, 1996. An Introduction to Hinduism. Cambridge, C.U.P.; and
Miranda Shaw, 1994. Passionate Enlightenment. Princeton, Princeton University
Press.
2. 1995: "Vajrayana: Origin and Function". In Buddhism Into the Year 2000. International Conference Proceedings. Bangkok and Los Angeles: Dhammakaya
Foundation, pp. 89-102.
3. (i) "Evidence of the Textual Dependence of the Buddhist Yoganuttaratantras on
the Tantric Saiva Canon", seminar delivered at the University of Hamburg, May
1990. (ii) "The Dependence of the Herukatantras on the Saiva Tantras of the
Vidyapltha", lecture series delivered at All Souls College, Oxford, May-June
1993. (iii) "Pious Plagiarism: Evidence of the Dependence of the Buddhist
Yoginltantras on Saiva Scriptural Sources", paper delivered at the Rijksuniversiteit te Leiden, April 11,1995. Items (i) and (ii) present excerpts from the Sanskrit
texts of the Buddhist Yoginltantras in parallel with their Saiva prototypes. Item
(iii) introduces mythic narratives as well. Although not a Tibetanist himself,
SANDERSON is able to draw significantly on Tibetan Buddhist materials from
myself and Gyurme DORJE, juxtaposing these with fascinating Saiva myths from
the Sanskrit.
Journal of the International Association of Buddhist Studies
Volume 21 • Number 2 • 1998
JIABS21.2
272
result of SANDERSON'S work has been to seriously call into question the
previously dominant view accepted by a majority of Buddhological
scholars, who had suggested that any such observable parallels between
the specifically kapalika or 'cemetery' strands within the Buddhist
Vajrayana and a number of very similar Saiva systems, were primarily
the result of both traditions arising from a common Indie cultural
substrate. 4 While Martin KALFF since the 1970's and David
SNELLGROVE since the mid-1980"s had already begun to question the
validity of this unsatisfactorily vague position on the grounds of
common sense and more generalised observation,5 it was only with the
presentation of SANDERSON'S minutely detailed and substantially documented philological analysis that we have finally been able to conclude
with a reasonable degree of certainty that such similarities are much
better explained as a result of direct Buddhist borrowings from the £aiva
sources.
In his papers, lectures and seminars, SANDERSON has analysed and
discussed the phenomena of such Vajrayana dependence on 3aivism
from a number of different perspectives and has used a number of
different types of primary sources. Firstly, he has presented detailed
philological evidence showing the movement of substantial passages of
text from specific Saiva scriptural sources into specific Buddhist Herukatantras. Secondly, he has shown how many of the general categories
of the Buddhist Vajrayana appear to be caiques on Saiva prototypes.
Thirdly, SANDERSON has cited passages of Saiva mythology, which
have provided an additional and valuable source of information for the
relation of these two Tantric traditions. In particular, SANDERSON has
shown how the Saiva myths agree in most instances with the Buddhist
ones already analysed by Tibetanists that the predominant direction of
the borrowings were from £aivism into Buddhism, even though the two
4. For an influential pioneering exposition of this view, see David SEYFORT
RUEGG, "Sur les rapports entre le Bouddhisme et le 'substrat religieux' indien et
tib&ain", Journal Asiatique 252 (1964): 7-95; see also SEYFORT RUEGG's much
later Review of David SNELLGROVE, Indo-Tibetan Buddhism: Indian Buddhists
and their Tibetan Successors in Journal of the Royal Asiatic Society 1989 (1)
p. 173.
5. SNELLGROVE, David L. 1987: Indo-Tibetan Buddhism: Indian Buddhists and
their Tibetan Successors. London: Serindia. Pp. 152ff, 462-463; and 1988:
"Categories of Buddhist Tantras". In: G. Gnoli and L Lanciotti, eds. Orientalia
Iosephi Tucci Memoriae Dicata. Serie Orientale Roma LVI. 3. Rome, IsMEO.
P.1361.
MAYER
273
traditions might offer quite different interpretations of the religious
significance of this fact.
In my recent book, A Scripture of the Ancient Tantra Collection: the
Phur-pa bcu-gnyis, I took up some themes from this third group of
sources, the mythology.6 Focusing mainly on the Buddhist sources, I
discussed the well-known Buddhist myths of the Taming of Mahes\ara/Rudra\ This myth has been dealt with at length in a number of
published secondary sources, and I expect the outlines of it are familiar
to most specialists in the field.7 In my book, I primarily understood the
6. Robert MAYER, 1996: A Scripture of the Ancient Tantra Collection, The Phur-pa
bcu-gnyis. Oxford and Gartmore: Kiscadale Publications.
7. In a nutshell, a basic core narrative more or less shared by the many variant
versions of these myths could be described as follows: typically, it might begin
by relating how the all-powerful malignance of Siva and his entourage, specifically in their more radically transgressive or kapalika tantric forms, seduced many
beings into a demonic and evil religious practice, and also came to pose a serious
threat to the survival of the Buddhist religion and even to the material welfare of
the whole world. The severity of this predicament eventually induced the assembled Buddhas of the three times and ten directions to conclude that in the exceptional case of such intractable and acutely dangerous enemies as these Saiva
deities, there remained no realistic alternative other than to tame them and convert
them to Buddhism by brute force. However, this could only be achieved if the
Buddhas themselves manifested in terryfying forms that exactly matched and
resembled the ferocious Saiva deities (viz., the Tantric Buddhist Heruka and
entourage); accordingly, such Buddhist kapalika deities or Herukas were emanated by the combined forces of all the Tathagatas, and were sent to engage the
Saiva deities in combat. Having succeeded in killing the Saiva deities, the
Buddhist Herukas are then typically described as resuscitating them and enslaving them as servants of Buddhism, giving them new Buddhist names, while likewise co-opting the entire panoply of Saiva transgressive Tantric practice as a
vehicle for disseminating Buddhist truths. It was under such circumstances, say
the Buddhist texts, that the esoteric, specifically kapdlika traditions of Vajrayana
first appeared within history, and become accessible to the human beings of our
world. Such myths appear in a great many extant Tantric texts in Sanskrit,
Chinese, and Tibetan. The main secondary sources I have consulted are as
follows: (i) R.A. STEIN, Annuaires du College de France, Anndes 71-80, and
more recently "La soumission de Rudra et autres contes tantriques", Journal
Asiatique 283.1 (1995): 121-160. (ii) Martin Michael KALFF 1979: Selected
Chapters from the Abhidhanottara-tantra. The union of female and male deities.
Unpublished PhD dissertation, Columbia University, (iii) Nobumi IYANAGA
1985: "Rdcits de la soumission de Mahes'vara par Trailokyavijaya, d'apres les
sources chinoises et japonaises", in Michel Strickmann, ed. Tantric and Taoist
Studies in Honour of RA. Stein, vol. 3, Melanges Chinois et Bouddhiques
vol. XXII. Institut Beige des Hautes fetudes Chinoises, Bruxelles. (iv) David
JIABS 21.2
274
'Taming of MaheSvara/Rudra' narrative as the charter myth (more or
less in a Malinowskian sense) of the Vajrayana in its specifically kdpdlika forms, in other words as the myth by which the Buddhists explained
and justified to themselves and to the world their co-option of so much
of the religion of their Saiva rivals. I also discussed this myth as representing Siva as a 'demon devotee' of the Buddha, in other words of repeating the widespread pattern found in Indian religions in which deities
transform hostile demons into their loyal devotees (often giving them
the specific function of guardian or protector) through the medium of
first slaying them, and then bringing them back to life.8 I also showed
how this myth can be seen as giving a Buddhist commentary (in the
Geertzian sense) upon Saiva-Buddhist relations, and how it can be seen
to document the process described by LEVI-STRAUSS as 'bricolage', in
which persisting cultural materials are re-worked to create new cultural
reconstructions.91 also looked at the myth in terms of the category of
'shamanic mediation' more recently developed by the anthropologist
Geoffrey SAMUEL, in this specific case referring to the use of altered
states of consciousness by Tantric Buddhist ritual specialists to achieve a
reconciliation or accommodation between their own tradition and conflicting extraneous cultural forces.10 Finally, I located the notion of the
taming of MaheS'vara/Rudra within its broader emic frameworks, as a
specific instance of the important Mahay ana and Tantric Buddhist category of 'dul ba or 'taming',11 which in general terms refers to a
8.
9.
10.
11.
SNELLGROVE, 1987: Indo-Tibetan Buddhism: Indian Buddhists and their
Tibetan Successors. London: Serindia. pp. 136-141. (v) Alexander W.
MACDONALD 1990: "Hindu-isation, Buddha-isation then Lama-isation or: What
happened at La-phyi?", in Indo-Tibetan Studies. Buddhica Britannica Series II.
Ed. T. Skorupski. Tring. (vi) Ronald DAVIDSON 1991: "Reflections on the
Mahes'vara Subjugation Myth: Indie Materials, Sa-skya-pa Apologetics, and the
Birth of Heruka", JIABS 14.2. etc.
cf. Alf HiLTEBEITEL 1989 (ed): Criminal Gods and Demon Devotees. Essays on
the Guardians of Popular Hinduism. Albany, SUNY Press. HILTEBEITEL's
introduction to the volume gives a clear outline.
Claude LEVI-STRAUSS 1976: The Savage Mind. London: Wiedenfield and
Nicholson. See, for example, pp. 16 ff.
Geoffrey SAMUEL 1990: Mind, Body and Culture. Cambridge: C.U.P.; and
1993: Civilized Shamans. Buddhism in Tibetan Societies. Washington: Smithsonian Institution Press.
Perhaps this concept has its roots in early Buddhism? cf. the 'Ten Epithets of the
Tathagata', as cited, for example, in TAKASAKI Jikido, trans. Rolf GlEBEL 1987:
MAYER
275
conveyor-belt process in which all non-Buddhist beings throughout the
universe are eventually to be converted to the Dharma and set
irreversibly on the path to enlightenment; more specifically, Tantric
Buddhism claims to possess unique skilful methods, by which even those
hard cases particularly resistant to taming (such as MahesVara/Rudra),
can also be converted.
In this paper, however, I want to focus on a somewhat different aspect
of this material. I want to address the question of what weight, or what
significance, the figure of MaheSvara/Rudra and the Saiva categories
associated with him have been accorded by the various traditions of
Tibetan Buddhism themselves. From the Tibetans' own point of view, is
MaheSvara/Rudra merely a marginal figure who appears occasionally in
a few obscure myths that might be of more interest to foreign philologists than to the tradition itself? Or is the figure of MaheSvara/Rudra
of great and central importance to the tradition as well? One part of the
answer to this question is, of course, that MaheSvara/Rudra has been
accorded a more prominent role in some areas of the diverse Tibetan
Tantric traditions than in others. In particular, I wish to single out two
areas where he is particularly significant. The first one is in the Yoginltantra traditions of Cakrasamvara.12 The second one is in the Mahayoga
traditions of the rNin-ma-pa. The hypothesis I am putting forward here
is that in the case of the Cakrasamvara tradition, the Saiva deity and the
borrowed Saiva categories are, according to most traditional voices,
acknowledged and allowed to play only a somewhat limited albeit distinctive role: their true historical and structural importance to the Cakrasamvara system is consistently understated and downplayed, even if
usually admitted in somewhat oblique and nebulous terms within certain
narrow and carefully delimited contexts. By contrast, in the Mahayoga
traditions of the rNih-ma-pa, the figure of Mahegvara/Rudra is accorded
a much more substantial part with very little restriction, and here he is
much more freely permitted to fulfil a broad and surprisingly pivotal
symbolic role of crucial importance. Once again, however, what modern
An Introduction to Buddhism, Tokyo: T5h5 Gakkai, p. 50: epithet 8 is purusadamya-sarathi, "one who controls men to be tamed".
12. The orthography 'Cakrasamvara' is deliberate: I am following SANDERSON'S
extensive but as yet unpublished research on the orthographical problems of
iamvaralsamvara, based on Saiva and Buddhist sources in both Sanskrit and
Tibetan. According to SANDERSON'Sfindings,while 'Cakrasamvara' is probably
more correct, most other instances should be spelt >-s'amvara\
JIABS 21.2
276
scholars might see as the actual historical nature of the relation of
Saivism to Tantric Buddhism is by and large tacitly ignored in favour of
highly complex metaphysical re-interpretations of that history.
I. Mahefvara in the YoginTtantras
The renowned yi dam Cakrasamvara and his consort VajravarahT are
among the most important deities of the YoginTtantras, which are more
usually designated by Tibetan doxography as the 'Mother-Tantra' {ma
rgyud) section of the Anuttarayogatantras. This is one of the areas of the
tradition very closely studied by SANDERSON, and since many readers
are probably already familiar with his work or can consult it directly, I
shall deal with this topic only very briefly.
Within Tibetan Buddhism, the Cakrasamvara Tantras are pre-eminently the speciality of the various bKa'-brgyud-pa schools, for whom these
cycles are closely connected with their famous traditions of yoga and
Mahamudra meditation. However, Cakrasamvara is also of great importance to the gSar-ma-pa traditions in general, including the dGe-lugs-pa,
the Sa-skya-pa, and so on. In the Cakrasamvara tradition, the Saiva
deities take the form of Bhairava and his consort Kalaratri. As SANDERSON has so clearly described, these non-Buddhist deities played a
specific role in the general Cakrasamvara iconography and commentarial
literature in Sanskrit, primarily as the seats of the Buddhist Herukas.13
The Tibetan tradition preserves this feature, but in addition, as one
might expect, Bhairava and Kalaratri have also come to be represented
in a number of additional indigenous Tibetan cultural developments of
the Cakrasamvara tradition. For example, there are important and elaborate pilgrimage and sacred site traditions connected with this cycle, in
which a number of holy mountains, but most notably Kailas, La-phyi
and Tsa-ri, are seen as concrete manifestations of the Cakrasamvara
mandala of deities. These mountains must surely rank among the most
famous of all the natural holy sites within the rich and so far largely
uncatalogued inventory of the Tibetan religious landscape. As concrete
manifestations of the Cakrasamvara mandala and being further
identified with some of the ancient Indian pit has described in the
YoginTtantras, in terms of the iconography and the taming narratives of
this tradition they should necessarily also include representations of
13. Origins & Functions, op.cit.; and "Pious Plagiarism", op.cit., pp.3-5.
MAYER
277
Bhairava and Kalaratri. I believe these are in fact in evidence, although
as yet I know only a very few details of their appearance.14
My readings within the Cakrasamvara literature are by no means
exhaustive; on balance, however, my impression is that consciousness of
its substantially Saiva origins and of the predominantly Saiva forms of
the deities in its pantheon are not and never have been of absolutely
overwhelming importance to the Cakrasamvara tradition as represented
in Tibet, nor to any other area of the gSar-ma-pa traditions as currently
practised, for that matter. Representations of the Saiva forms might exist
in iconography and mythology or possibly also in natural forms at
sacred sites such as Tsa-ri, but these do not seem to have been allowed to
penetrate to the inner core of the tradition, or to spread too deeply into
its sadhana and commentarial systems. Rather, the taming myths persist
as something attached to the tradition as an origin myth while remaining
slightly extraneous to the ritual practice itself and the associated higher
metaphysics, while the iconography of the Saiva deities as the Herukas'
seats is seen as little more than a symbolic detail, open to a variety of
standard Buddhist interpretations of the subjugation of evil. Even if
there might be a widespread awareness and acknowledgement of the
formal external similarities of the Cakrasamvara traditions to Saivism, as
explained by the taming narrative and commonly assented to by lamas of
this tradition, this is never allowed to become a doctrinally pivotal point:
it remains a mere detail, a mere contingency. The very names of
Bhairava and Kalaratri appear within only a few contexts in the literature, notably when the origin myth is being recounted, or when the
14. Traditional texts, such as an oral text known as the gNas chen tsdri tra'i gsol
'debs, apparently claim that there are naturally-occurring stone lirigas and yonis
found at the hamlet of rDo-mtshan (lit. 'Sexed Stones') beside Tsa-ri. According
to this tradition, these 'sexed stones' are apparently conceptualised according to
the famous Indie cosmic dismemberment narratives connected with the Saiva/
Sakta conceptions of the pltha system; in other words, as far as I understand it,
following the taming myths, they are seen as the stone symbols of MaheSvara
and his goddesses installed at Tsari (= the pltha CaritrS) and at each of the other
24 pithas, before their downfall to Heruka. Apparently these svayambhu stone
phalluses and vulvas are not obvious or easy to perceive, except by advanced
yogins. While important for Tantric meditators, they are more often seen by lay
persons as bestowing special powers of procreation and fertility. These stones
and other features of the sacred topography of Tsa-ri are described in Toni
HUBER 1993: What is a Mountain? An Ethnohistory of Representation and
Ritual at Pure Crystal Mountain in Tibet. PhD Dissertation, University of
Canterbury, Christchurch (New Zealand), pp. 82ff.
JIABS21.2
278
deities' seats are being described; otherwise, they seem to have no role
to play, and there is no need to name them. One has the impression that
were all its internal references to Saivism to become forgotten altogether, the Cakrasamvara traditions of Tibetan Buddhism could still continue more or less unchanged.
However, as I have said above, my knowledge of the Cakrasamvara
traditions is by no means entirely exhaustive, and so my perception of its
representations of Siva might yet prove to be mistaken. But if, as I
expect, my current perceptions do indeed transpire to be accurate, this
could be considered a little surprising, because there surely can be no
area of the Buddhist tradition more intimately bound up with Saivism
than the Cakrasamvara Tantras. As we know from SANDERSON'S work,
large sections of their major scriptures, such as the Laghusamvara and
the Abhidhanottara, are borrowed virtually word for word from Saiva
prototypes such as the Picumata and the Tantrasadbhdva.15 In addition,
15. For those readers unfamiliar with SANDERSON'S work, although I cannot review
all the issues here, perhaps it might be useful to present one brief quotation from
my recent book in which I review some of SANDERSON'S findings (taken from
MAYER 1996 op.cit., pp.59-60):
"A contemporary Indologist, Alexis SANDERSON, has already identified
(through textual criticism) a good quarter of all the verses in the long and important Laghusamvara as having been adapted or borrowed virtually unchanged
word-by-word from earlier Saiva texts such as the Picumata/ Brahmayamala,
the Siddhayogefvarlmata, and the Yoginisamcdra[prakarana] (which latter
appears in the third satka, or section of 6000 verses, of the composite Jayadrathayamala). This is remarkable for several reasons. Firstly, the Laghusamvara is
often considered the single most important text of the Cakrasamvara cycle.
Secondly, the quarter of the text so far demonstrably incorporated from Saiva
sources might not reveal the full extent of the dependency, since not all the corpus
of relevant Saiva texts survive; for example, the *YoginIjdlafamvara and the
*Sarvavirasamdyoga are two lost Saiva texts that were influential in the eighth
century, a period when a matrix of Buddhist Yoginltantras were produced whose
very names may have been caiques on the Saiva texts (Ddkinifdlafamvara and
Sarvabuddhasamdyoga). Thirdly, a good part of the Laghuiamvara consists of
mantroddharas and the like that are written very much in the manner of a Saiva
text, but which obviously could not be lifted in directly from Saiva sources, given
the important function of mantras as a text's unique signatures. Since the important Cakrasamvara vydkhydtantra (explanatory tantra), the Abhidhdnottara, seems
to draw on similar materials to the [proto-] Lagmdamvara, a quantity of the same
Saiva materials is found there as well, probably in an earlier form than the
Laghuiamvara as we have it now (SANDERSON op. cit. 1990; 1993; 1995).
Among other shared materials are the all-important samayah (tantric vows of
conduct). (Thanks to Alexis SANDERSON for these references)."
MAYER
279
the Cakrasamvara iconography has probably remained closer to its Saiva
prototypes than any other Buddhist iconography. On top of that, the
colourful and popular Cakrasamvara rendition of the taming myth, even
if doctrinally comparatively restricted in scope, is nevertheless extremely
widely attested in Tibet, and numerous versions of it are preserved and
frequently repeated at both learned and popular levels; at first glance,
one might possibly have expected this fact alone to have had more
impact in terms of an historical awareness of the relation of the two
traditions. Yet the acknowledgement of anything Saiva does not seem to
have been allowed to pervade very widely or deeply through the tradition. As far as I am aware, the Cakrasamvara tradition in Tibet contrives
to deny the surface meaning of its own origin myth, and presents itself
as having an exclusively Buddhist nature, often even claiming to have
been originally taught by the historical Buddha during his lifetime in a
transcendent form at the Dhanyakataka Stupa near AmaravatI in Andhra
country, simultaneous to his preaching of the Mahay ana Sutras in a more
anthropomorphic form at the Vulture's Peak in Bihar (NEWMAN 1985:
53)16. Thus, within Tibet at least, the Cakrasamvara cycle is generally
conceptualised as a product of a uniquely Buddhist environment with
scant acknowledgement of any debts to or historical contacts with
Saivism, or any realistic acknowledgement of the Saiva iconography of
its mandalas. It seems to me, then, that Buddhism has here decided to
turn a blind eye to significant aspects of its own nature and origins: there
seems to be a degree of deliberate avoidance or even denial of the full
extent of the Cakrasamvara cycle's debts to Saivism. Even the apparently obvious message of the taming narrative becomes ignored, isolated
and negated by the denying attitudes of the broader tradition as a whole.
This traditional policy of denial has inevitably had its effect on
modern Buddhology too. As recently as 1991, one of the finest of contemporary Buddhologists working in Vajrayana, Ronald DAVIDSON,
wrote of the list of 24 pithas as mentioned in the Cakrasamvara-cycle's
Siva-taming narratives, that "it cannot be immediately assumed that
[this] was a popular Hindu system subsumed into the Buddhist fold..."
DAVIDSON continues, "The Buddhist mythic contention that these [24]
places were initially £aiva cannot be accepted as fact. Instead, the list is
16. John NEWMAN 1985: "A Brief History of the KSlacakra". In The Wheel of Time,
ed. Geshe Lhundub Sopa, Roger Jackson & John Newman, Madison: Deer Park
Books, pp. 51-84.
JIABS21.2
280
developed out of such geographical lists of places noted in esoteric
Buddhist literature as early as the MahdmayurT-vidya-rajni-dhdranrJ7
In other words, DAVIDSON admits that the taming narratives of the
Cakrasamvara tradition openly assert that the Buddhists took the system
of pithas from the Saivas; yet he argues that as critical modern scholars,
we should not believe this unlikely story. To my mind, this comparatively recent quotation from a first class scholar such as DAVIDSON only
goes to show what a major long-term impact SANDERSON'S extensive,
painstakingly detailed and solidly founded philological work is likely to
have on the Vajrayana branch of Buddhology: thanks to SANDERSON,
we can now move on from educated speculation, to a terra firma of
sorts; for as SANDERSON has conclusively shown, the Buddhists did
indeed adopt the system of pithas directly from the §aiva canon, exactly
as the Buddhist taming narrative maintains!18
I do not feel this is the moment to embark on a discussion of how
things might have been in Indian Buddhism; however, as far as I can
see, there seem to be three factors that can help account for this traditional understatement of the surface message of the taming myths and
other Saiva elements within the Cakrasamvara systems of Tibet:
[i] Firstly, as far as I know, the taming myths occurs only in commentarial texts of the Yoginitantras, and not in the actual canonical YoginI
scriptures themselves. Hence they are by definition somewhat marginal
to the tradition as a whole, and from a traditional point of view there is
no compulsion for their message to make an appearance at deeper or
broader levels of sadhana and metaphysics. According to Ronald
DAVIDSON, it is not even quite clear to what extent the Cakrasamvaracycle Siva taming myths as we currently have them derive from Indie
sources, and to what extent they originate in Tibet. While R.A. STEIN
has cited the following sources which he believes to be Indie, as far as I
know he has not yet got round to publishing any findings from them: [1]
A commentary by Indrabhuti, Peking bsTan-'gyur 2129; [2] Two commentaries by Vajra, Peking bsTan-'gyur 2128 and 2140. [3] A commentary by Naropa, Peking bsTan-'gyur 4628 [4] a text he identifies only as
17. Ronald M. DAVIDSON: "Reflections on the Mahesvara Subjugation Myth: Indie
Materials, Sa-skya-pa Apologetics, and the Birth of Heruka", JIABS 14.2
(1991): 224.
18. See in particular Origin and Function, op.cit.; and Pious Plagiarism, p. 15, with
Sanskrit texts pp. 3-4.
MAYER
281
P. No.2624 [sic]. Following STEIN's lead, Ronald DAVIDSON reports
that the two commentaries by Indrabhuti and Suravajra do indeed make
the subjugation of MaheSvara "part of the lore surrounding the advent of
the Cakrasamvara Tantras."19 Having read these two texts, however,
DAVIDSON'S conclusion is that they are too brief to account for the
fully comprehensive earliest known Tibetan version of the Cakrasamvara
taming myth written by the early Sa-skya-pa patriarch Grags-pa rgyalmtshan, who lived from 1167-1216. One feature of Grags-pa rgyalmtshan's analysis mentioned by DAVIDSON was that the Buddhists'
arch-fiend figure of Mahe£vara/Rudra himself becomes re-interpreted as
an aspect of the Buddha of primordial enlightenment, an interpretation
highly typical of the rNin-ma-pa exegesis of the taming narratives.
DAVIDSON concludes, "so far as I am able to determine, fully developed forms [of the Cakrasamvara taming myth] occur only in indigenous Tibetan language materials, and the text of a Tibetan author of the
twelfth-thirteenth century appears to be the earliest version".
(DAVIDSON 1991 op.cit., p. 204).
[ii] A second reason for the understatement of the debts to Saivism is
that Tibetan Buddhism did not have to co-exist with Saivism in the same
way that Indian or Nepalese Buddhism has had to do. Hence Tibetan
approaches to Hinduism in any of its aspects has always tended to reduce
Hindu categories into convenient sets of mainly abstract symbols for
purely domestic intellectual or scholastic consumption. With the actual
presence of Hinduism so distant, there was no need to take it seriously as
a living historical reality. The reduction by learned clerics of its few
references to Saivism into purely abstract sets of symbols, is precisely
what seems to have happened in the case of the Cakrasamvara cycle in
Tibet.
[iii] Thirdly, the general tendency of the gSar-ma-pa traditions is to see
the Tantras as the utterances of the historical Buddha, even if uttered by
him in the transcendent form of Vajradhara. In other words, the gSarnia-pa tend to support a closed canon, rejecting the idea of ongoing
revelation and the continuous addition of new scripture to the existing
19. DAVIDSON gives exact citations: fricakrasamvaratantrarajasambarasamuccaya-vrtti. To. 1413, rgyud 'grel, vol. tsa, fol. 4ab; MMatantrahrdayasamgrahdbhidhanottaratantra-malamulavrtti, To 1414, rgyud 'grel, vol. tsa, fol.
121a7.
JIABS21.2
282
canonical collections through the 'shamanic' (in S A M U E L ' S special
technical sense) processes of colonising or 'taming' extraneous religious
systems such as Saivism. Thus the historical implications of the taming
myth, that so clearly identify the origins of the Yoginltantras as an historical Buddhist response to Saivism involving a stategy of co-option,
are possibly something of an embarrassment: they suggest an uncomfortable truth, one that best remains understated.
II. The Taming myth in the rNin-ma-pa Tantric scriptures.
The other area of Tibetan Buddhism I wish to discuss in which the
MahesVara/Rudra myths are important lies within the Mahayoga traditions of the rNin-ma-pa. Here, however, the figure of MaheSvara/
Rudra takes on a very full and central role indeed, within the very heart
of sadhana, within a great many aspects of ritual, and at the very
deepest levels of commentarial exegesis. Here, little or no effort is made
to ignore or contain the taming myths (with all their implicit
implications of dependence), nor to deny Buddhist Tantrism's debts to
the figures of Mahesvara/Rudra, even if these debts become heavily reinterpreted in Buddhism's favour. On the contrary, the very dependence
of Vajrayana Buddhism on this hostile and alien figure is itself elevated
to a valuable spiritual truth, and the Saiva deity himself is accorded a
pervasive and pivotal role in rNin-ma-pa metaphysics, soteriology, and
ritual. In the rest of this paper, I want to mainly look at this complex
figure of Rudra in the rNin-ma-pa tantric tradition.
Unlike the Cakrasamvara traditions, the taming of Mahesvara /Rudra
narrative plays a significant part in a great many of the most important
canonical Tantric scriptures of the rNih-ma-pa tradition. It is true, of
course, that much of the main rNih-ma-pa tantric canon, the rNin-ma 'i
rgyud-'bum (henceforth NGB), consists of materials composed or
compiled in Tibet, rather than direct translations from Sanskrit as in the
case of most of the Tantric texts of the gSar-ma-pa canon, the Kanjur.
However, in integrating the taming myths into their scriptures, the rNihma-pa were certainly remaining true to a much older Sanskrit Vajrayana
tradition. Arguably the oldest and most important of all versions of the
Mahesvara/Rudra taming myths is that found in the Sarvatathdgatatattvasamgraha
(henceforth STTS), the basic text of the older
Yogatantra tradition, which became the source of the very wide dispersal
of the taming myth in Far Eastern Tantric Buddhism. Likewise, taming
myths occur in scriptures of the Mahayoga traditions found in the
MAYER
283
Tibetan Kanjur as well as in the NGB, and which are generally thought
to have had an Indie origin, such as the Candraguhyatilaka-mahatantrardja,20 and the text known to the Tibetans as the *Guhyagarbha-tattvavinitcaya, which SANDERSON has possibly identified from Sanskrit
sources as originally bearing the name of Guhyakoto.
In the rNin-ma-pa Tantras, the scriptural references to the taming
myths are not merely fleeting or marginal. On the contrary, they assume
the greatest possible significance, occupying large sections within the
most important of the rNiri-ma-pa Tantras. The taming myth occupies
an entire chapter of the *Guhyagarbha-tattvavini6caya, for example, and
this text is widely considered the most important single scripture within
the rNin-ma-pa Mahayoga tradition as a whole. In the Mdo dgotis-pa
'dus-pa, a text of uncertain origin said to have been translated from the
'Bru-sha language, which is the root text for the rNin-ma-pa Anuyoga
tradition but also very influential for Mahayoga, the highly extended
taming narrative occupies no less than eleven chapters, (chs. 20-31) and,
in the words of Matthew KAPSTEIN, this taming narrative "has become
the organising metaphor of the text as a whole".21 No one has yet
attempted a comprehensive survey of the frequency of scriptural occur20. This occurrence of the taming myth is reported by DAVIDSON op.cit. p. 203,
based on the version of the Candraguhyatilaka-mahatantrardja as found in sDedge rgyud-'bum, vol. ja, fols. 281a-287a. It might be of interest to note that I
recently made a brief and cursory ad hoc comparison of the opening chapter of
the versions of the Candraguhyatilaka-mahatantraraja as contained in the sTog
Kanjur (vol. 97 CHA, 226a3-297b5) and in the mTshams-brag NGB (vol 18
folios 357 ff); I found that substantial portions at the beginning of the text seemed
to be dramatically at variance between the different editions, while other portions
were more or less the same. Note also that quite different translators from different historical periods are mentioned in the colophons of the different versions.
21. See his "Samantabhadra and Rudra: Innate Enlightenment and Radical Evil in
Tibetan Rnying-ma-pa Buddhism", in F. Reynolds and D. Tracy, eds., Discourse
and Practice, SUNY 1992, p. 66. R.A. STEIN has recenUy written a characteristically valuable and interesting paper in which he proposes that this extended
version from the Mdo dgons-pa 'dus-pa differs from Sanskrit texts such as the
STTS and the *Guhyagarbha-tattvavinis'caya in that it talks more of Rudra than
of MaheSvara; STEIN further suggests that it might be the source for the detailed
taming narratives found in a number of other Mahayoga scriptures, including
some he has closely studied that were placed in the rNying-rgyud section of the
Kanjur, such as the Me-lce 'bar-ba (P466), the bDud-rtsi chen-po (P464, and an
untitled text that accompanies them (P465) (STEIN 1995, op.cit). The Mdo dgonspa 'dus-pa version of the taming narrative has also probably been very influential
on important gTer-ma texts such as the Padma bka'i than.
JIABS21.2
284
rences of the taming narratives within the rNiri-ma-pa Tantric canons,
but I have encountered them in casual readings of several other rNirima-pa tantras, such as the dPal rdo-rje phur-pa'i b§ad-rgyud dri-med
'od, [Thimpu NGB Sa, 28]; in the Phur-pa mya-nan-las-'das-pa 'i rgyud
chen-po [Thimpu NGB Sa, 28]; and in also in a doxographically more
important text that I have looked at more carefully, the Phur-pa bcugnis, one of the main root texts for the Vajrakflaya tradition [Thimpu
NGB Dza, 19]. STEIN has also reported extended taming myths in a
number of other rNiri-ma-pa Tantras, such as the Me-lce 'bar-ba
(P466), the bDud-rtsi chen-po (P464), and an untitled text that accompanies them (P465) (STEIN 1995, op.cit.). From the above evidence, it
is not unreasonable to assume that taming narratives will eventually be
found in a significant proportion of rNiri-ma-pa tantric scriptures. For
example, they are quite likely to turn up in several more of the nineteen
*Guhyagarbhatantras preserved in the NGB, and not merely in the single
if most important specimen from among this voluminous literature
studied by Gyurme DORJE, namely the *Guhyagarbha-tattvavini&caya
mentioned above.22 On doctrinal grounds, we can also predict a good
likelihood of finding taming narratives in several more of the rNiri-mapa tantras connected with major Heruka figures such as VajrakTlaya,
HayagrTva, and so on.
The taming myth in Klon-chen-pa's Phyogs-bcu mun-sel
If the taming of Mahegvara/Rudra narratives loom so large within the
rNiri-ma-pa scriptures, the question arises as to why? What important
messages do they convey? To approach this complex question, I want to
begin by citing a few key passages from a definitive traditional commentary on the *Guhyagarbha-tattvavinUcaya called the Phyogs-bcu
mun-sel, 'Dispersing the darkness of the ten directions', written by that
most influential of all rNiri-ma-pa commentators, the fourteenth century
sage Klori-chen-pa. This extensive commentary has been translated in
full by Gyurme DORJE (DORJE 1987, op.cit.). Here, I can only summarise a few salient points from its treatment of the taming narrative.
Klori-chen-pa provides a comprehensive exegesis of the taming myth
that aims only to bring out its soteriological implications: actual histori22. DORJE, Gyurme 1987: The Guhyagarbhatantra and its Fourteenth Century
Tibetan Commentary Phyogs-bcu mun-sel. Unpublished PhD thesis, London:
SOAS.
MAYER
285
cal fact, in the shape of the relations of Buddhism with Saivism in India,
are of no interest to him. Now, it is a famous feature of this particular
commentary that Klon-chen-pa here seeks to interpret even the root text
for all Mahayoga, the *Guhyagarbha-tattvavini£caya or rGyud gsart-ba
snin-po, to some degree at least from the doctrinally higher and
philosophically even less dualistic point of view of Atiyoga or rDzogschen. Hence it is that within this profound soteriological exegesis of the
taming myth, Klori-chen-pa adopts above all the uncompromisingly and
quite radically non-dual perspectives characteristic of Atiyoga, focusing
more on themes of primal purity than on more typical Mahayoga themes
of transformation. It is this bias towards Atiyoga which probably
underpins his strong emphasis on the proposal that even the evil demon
Mahegvara, who here as elsewhere throughout the Mahayoga literature is
consistently identified as the chief and worst of all the Maras, is
ultimately an expression of primordial wisdom.
Firstly, Klon-chen-pa affirms that the Buddhist tantric wrathful deities
exist primordially, and are particularly connected with the head cakra,
just as the Buddhist peaceful deities exist primordially and are
particularly connected with the heart cakra. He explains that this is
entirely in accord with Buddhist doctrines on the trikaya and on the
subtle body. It is therefore only to illustrate the principles of the Vajrayana teachings to trainee sentient beings, that the Rudra-taming episode
is enacted at all; in other words, Klon-chen-pa is saying that although
the wrathful Buddhist kdpdlika deities exist primordially, they skilfully
manifest themselves to beings via the occasion of the taming of Rudra.
Thus Rudra himself is not really what he seems: he is in reality an expression of primordial wisdom, appearing as a demon to be subdued in
order to provide an occasion by which the principles of the secret
mantrayana can be made clear to beings. In that he constitutes the unique
condition for the revelation of the supreme Vajray§na itself, he is thus
superior even to a Buddha's emanation in the usual sutric parlance.
If all this begs some questions about Rudra's behaviour in harming
and killing so many sentient beings through illness, calamity, and false
teachings, as so graphically described in the main body of the narrative,
Klon-chen-pa compares the situation with that described in the Aiigulimald-sutra. He writes:
This is reminiscent of a common sutra which reveals that when the dwarf
Angulimaiiya had formed a garland of the fingers of 999 men he had slain, he
was tamed by the Buddha and ultimately shown to have no defect. However it
JIABS 21.2
286
was in fact an emanation of the Tathagatas who slew phantom human beings of
his own emanation so that the garland was strung... (translation from DORJE
op.cit., p. 1088)
Thus Rudra is a phantasmagorical expression of Buddhist skilful means,
who uniquely can demonstrate a negative example to sentient beings,
without actually harming anyone. On the other hand, anyone foolish
enough to attempt to emulate Rudra's outward behaviour would undoubtedly cause great harm, and this fact becomes a very important
point in rNih-ma-pa writings on ethics, doctrine and the Tantric samaya
vows, as I shall describe later on.
In typical Buddhist fashion, Rudra becomes an archetype and is
multiplied. There can be any number of Rudras in different worlds. We
can all become Rudras if we misunderstand the tantras, taking them
literally with no insight into Buddhist metaphysical truths. Thus Rudra
becomes transformed from an historical Saiva deity, into a generalised
symbol of evil with a very broad application. He has become elevated
from history into myth, from a specific Hindu god to a universal symbol
of evil equivalent to Mara.
Kloh-chen-pa summarises this first part of his exegesis:
Rudra attained Buddhahood as Samantabhadra in primordial original time, and
then, in order to subdue the Mara who appeared within his self-manifesting
cakras, he became manifest in and of himself...." (DORJE op.cit., p. 1092)
In brief, just as the lion throne appears to symbolise the presence of the four
kinds of fearlessness, Rudra appears to be subdued in order to symbolise victory
over all demons and outside aggressors, and the complete mastery which overwhelms proud spirits. At the time when enlightenment is attained, Mara must
appear to be subdued. Thus when the subjugation of Rudra, as the first and foremost of the host of proud demons or Maras, is revealed, one attains mastery over
the appearances of the self-manifesting cakras.
It is crucial for an understanding of Kloh-chen-pa to realise that Rudra is
the chief of all Maras. He remains the closest thing to a Buddhist Satan,
an inversion of all Buddhist values. It is only through the radical nondualism of Kloh-chen-pa's Atiyoga type of exegesis that even the chief
of the Maras is realised to be an expression of primordial wisdom like
Samantabhadra; for what is enlightenment other than the conquest of
Mara, and how could Mara be conquered if he did not manifest?
Therefore there can be no enlightenment without Mara, and so Mara
must be an expression of primordial wisdom. As the foremost of the
Maras, Rudra is also indispensable for enlightenment.
MAYER
287
It is worth noting that in this non-dual interpretation of Rudra as primordially enlightened, Klon-chen-pa is not diverging from the early Saskya-pa exegesis of the Cakrasamvara taming myth. As DAVIDSON has
pointed out, the Sa-skya-pa patriarch Grags-pa rgyal-mtshan took a very
similar view in his work dPal he ru ka'i byun tshul, a title which
translates as "How Heruka was born".23 Here Grags-pa rgyal-mtshan
distinguished two levels of the Cakrasamvara tradition of the taming
myth, the provisional meaning (neydrtha) and the definitive meaning
(nitdrtha). While the provisional meaning took the taming narrative at
face value as a story of Buddhism conquering the wicked deities of the
Saiva pantheon, according to the definitive meaning, the tamer and the
tamed become non-differentiated. The Saiva deities, including Bhairava
and Kalaratri, are all seen as emanations of Mahes*vara, who is in turn
seen as an emanation of the Buddha Vajradhara. Likewise, all the
Buddhist deities who effect the conversion are emanations of Sri
Heruka, who is himself an emanation of Vajradhara. Thus, from the
point of view of the definitive meaning, all the characters in the taming
myth are emanations of Mahavajradhara (DAVIDSON 1991 op.cit.,
p. 208). However, as DAVIDSON points out, Grags-pa rgyal-mtshan
cited no written sources for this interpretation of his: he simply states, it
is "culled from the speech of my guru".24 As a major early commentator
of his own hereditary family lineage of the rNin-ma Phur-pa tradition as
well as of the newer tantras such as Hevajra and Cakrasamvara, it is
theoretically not impossible that Grags-pa rgyal-mtshan in fact borrowed
this idea from the Vajrakllaya literature or other rNin-ma-pa sources.
Until we have read more of the surviving Indie Cakrasamvara commentarial texts such as those from the Tenjur mentioned above, we cannot
easily assess how much of the Cakrasamvara taming myth is indigenous
to Tibet. However, my own hunch is that it probably does have a substantially Indie basis; otherwise, it would be unlikely to have established
so secure a place for itself within the writings of such seminal gSar-mapa authorities as Grags-pa rgyal-mtshan and Bu-ston, endowed as they
were with such highly-developed critical faculties.
23. Sa-skya bka'-'bum (henceforth SKB), III.298.4.2-300.2.6 (bSod-nams rgyamtsho, ed., The Complete Works of the Great Masters of the Sa Skya Sect of the
Tibetan Buddhism (Tokyo: Toyo Bunko 1968).
24. Dpal he ru ka 'i byun tshul mam par btag pa bla ma 7 gsun las curi zad btus te.
SKB III.300.2.6; cited in DAVIDSON op.cit., p. 231, n. 8
JIABS21.2
288
Chapter Seven of the Phur-pa bcu-gnis.
I now want to consider more specific applications of the figure of Mahesvara/Rudra within the Vajrakllaya tradition, beginning by looking at
the taming narrative of Chapter 7 of the Phur-pa bcu-gnis-kyi rgyud, an
important root-text from the NGB, counted by that tradition as one of
the Eighteen [Root] Tantras of Mahayoga (tantra sde bco-brgyad).251
want to show how the single episode of the taming of Mahes'vara/Rudra
can be seen as the central, pivotal moment for the unfolding of this
whole scripture: for it is out of this single great and complex act of the
taming of MaheSvara/Rudra that the Phur-pa bcu-gnis seeks to derive
all the most important teachings of the Vajrakllaya tradition, including
the key doctrinal positions, the major iconographic features, the basic
visualisation sequences used in Vajrakllaya sddhanas, the major Vajrakllaya ritual specialities, and the key Vajrakllaya textual passages, which
latter function both as scripture and as the most important liturgical
passages. Surprising though it might seem to those unfamiliar with the
rNiri-ma-pa Mahayoga traditions, according to the anonymous authors of
the Phur-pa bcu-gnis, all such absolutely fundamental components of
this most pre-eminent among rNin-ma-pa yi dam soteriological systems
should thus be seen as being directly derived from or directly linked to
the taming narrative, and hence to the figure of MaheSvara/Rudra. In
this way the figure of Mahes'vara/Rudra seems to be invested with a
scriptural prominence and a quite crucial doctrinal and even spiritual
importance that I have not yet encountered in my readings of the Cakrasamvara tradition (although I would not be entirely surprised if something similar were eventually to turn up in some early precursors of the
developed Cakrasamvara tradition).
25. The full title is Phur pa bcu gfiis kyi rgyud ces bya ba theg pa chen po'i mdo.
There are in fact several texts in the NGB with quite similar titles to this, but the
particular one I am referring to is perhaps traditionally considered the most important, since it has been selected as one of the special Eighteen Tantras. My
analysis below is based on my consultation of this text as contained in the
following editions of the NGB: the sDe-dge xylograph NGB, vol. PA, ff.l76r25 lv; a manuscript NGB held in the National Archives, Kathmandu, vol MA,
ff.37r-129v; the reprint edition of the mTshams-brag manuscript NGB preserved
in the National Library, Thimpu, vol. DZA, 393r-507r (pages 785-1013); the
Thimpu reprint of the gTing-skyes dgon-pa-byang monastery manuscript NGB,
vpl. DZA, lr-lOOr (pages 1-199); and the Waddell Manuscript NGB, mainly
held in the India Office Library, London, vol DZA, lr-91r.
MAYER
289
Perhaps I should begin with a brief summary of the chapter. It is entitled 'Chapter Seven, How the arrogant [gods] were tamed', gdug pa can
btul ba 'i le 'u ste bdun pa 'o //, and is found within pages 82-92 of the
Thimpu reprint NGB, within pages 871-884 of the mTshams-brag
Manuscript NGB, and within folios 206r to 210v of the sDe-dge xylograph NGB. This substantial prose chapter is dedicated entirely to a
lengthy description of the taming of Mahes\ara/Rudra and his entourage, containing no other material. To my mind, the narrative clearly
appears here, as in many other rNiri-ma-pa Tantras, as a charter-myth
for Buddhist kapalikaism. More specifically, it also gives the context for
the first appearance of the main Vajrakllaya mandala itself, which up to
this point had not yet made its appearance in the text. The narrative
begins with the Buddhas noticing that the Saiva deities were causing
terrible harm to the world by their savage attacks against the Buddhist
religion. The Buddhas also perceived that the Saiva deities could never
be influenced by peaceful means; hence Vajrakllaya manifested a special
wrathful form with one thousand heads, a thousand arms, and ten billion
feet, dwelling in a cemetery palace. From this form in turn was emanated the six-armed, four-legged, three-headed basic form of Vajrakumara
so familiar from rNiri-ma-pa sddhana texts, here called the 'Excellent
Son', embracing his consort Ekajata. Easily victorious, Vajrakumara
subdued and trampled the arrogant Saiva gods underfoot. To mark his
victory, he was then invested by Vajrakllaya with the emblems of the
Saiva gods, such as the khatvdnga and the other kapalika regalia. To
proceed with the conquest of the remainder of the Saiva pantheon, the
Buddhist deities next emanated some further forms, known as the
'Kilayas of the Five Families', i.e. Buddhakilaya, Vajrakllaya, Ratnakllaya, Padmakllaya, and Karmakllaya; each of these had upper bodies
similar to Vajrakumara's, while their lower bodies were shaped like
triangular kilas. Upon this, the subsidiary Saiva deity Vighnaraja (or
Ganesa) with all his retinue of vighna deities (Tib. bgegs; as with the
Sanskrit, literally, = 'obstacles') was summoned; to the accompaniment
of some potent 'vajra verses', the assembled Saiva vighna deities were
killed, and their remainders roasted or burned and eaten.
Most significantly, in my opinion, it is only at this point within the
scripture, after these great acts of conquest had been carried out, that the
interlocutor figure becomes transformed for the first time from the
general, less esoteric tantric form of Vajradharma, to the specifically
kapalika, more esoteric Tantric form of Karmaheruka, which latter form
JIABS21.2
290
he retains for the duration of the remainder of the scripture. Thus transformed into Karmaheruka, the interlocutor now asks Lord Vajrakllaya
how the yogins of future ages should emulate this great deed of subduing the Saiva deities? Vajrakllaya replies with twenty-three lines of the
most famous root verses of the Vajrakllaya tradition, verses that are
found repeated verbatim in virtually every Vajrakllaya gter ma, sadhana, liturgy, commentary, and so on, verses which have also attracted
far more important commentarial attention than any other section of the
Vajrakllaya literature. They are known by heart in some version or another by most serious Tibetan practitioners of this cycle, and they also
occur in the rTsa-ba'i dum-bu or Vajrakilayamiilatantrakhanda™ the
only fragment of Vajrakllaya literature to have gained entry into the
Kanjur, in this case through the efforts of Grags-pa rgyal-mtshan's
famous successor, Sa-skya Pandita. These root verses not only give the
key teachings on the practice of iiberative killing' (sgrol ba, moksa) for
which the rNin-ma-pa Mahayoga in general, and the Vajrakllaya cycle in
particular are so famous, but, according to the commentarial tradition,
they also give the most crucial of all instructions on the main Mahayoga
contemplative soteriology itself.27 After this important episode, the
26. Peking Kanjur 78, 3; sTog Palace Kanjur Catalogue no. 405; sDe-dge Kanjur,
Toh. no. 439; Ulan Bator Kanjur Handlist, no. 469.
27. Innumerable variants of these verses occur within the multifarious Phurpa traditions taken as a whole (see MAYER op.cit. pages 212-215). Here, I show only the
version given in Chapter 7 of the Phur-pa bcu-ghis itself, mainly following the
sDe-dge xylograph NGB, vol PA, folio 209r, but including some of the main
variant readings taken from other editions: / rdo rje khros pas ie sdah gcod II
mtshon chen srion po 'bar ba yis II nam mkha 7 dkyil nas thigs par [thig pa] tor
II srog gi go ru far ba dan II shin gi go ru bsgom par bya II shin rjes bsgral ba 7
dam tshig ni II gsad tin mnan pa ma yin te II phuh po rdo rjer gtams byas nas II
mam par fes pa rdo rjer bsgom II rdo rje gzon nu 7 rigs 'dzin rnams II srid pa
rdo rje grub mdzod cig II srid pa rdo rje phur pa 'i lha I [however, the previous
four yig-rkang are omitted in sDe-dge; I take them here from mTshams-brag ms.
vol DZA, p 880] / ye ies khro bo 'grub [or grub] par mdzod II saris rgyas kun
gyi ye s"es sku II nan hid rdo rje chos dbyihs las II 'bar ba 'i khro bo mi bzad pa II
sku yi dbyig tu [sDe-dge reads dbyings su] bdag bskyed cih II thabs kyi spyod
pas 'gro don du II byams dan shin rjes gah 'dul ba II sans rgyas 'phrin las
rdzogs bya 7 phyir II dbah dan byin rlabs 'dir stsol cig II de rjes phyi nan gsah
gsum dbah II byin rlabs bdag la stsal nas ni II de hid du ni mi snad [or snang]
'gyur II de nas sras mchog yab yum gyis II ghis med byah chub sems las ni II
'byuh ba 7 gsah shags 'di yin no I The soteriological meaning of these key verses
is analysed at considerable length by 'Jam-mgon kon-sprul blo-'gros mtha'-yas
in a famous commentary on the rTsa-ba 7 dum-bu: see his dPal rdo rje phur pa
MAYER
291
taming narrative continues with the emanation of the well-known Mahayoga versions of the Ten Wrathful Deities' (dafakrodha), along with
their female consorts and their zoomorphic attendants. By now, the
Saiva deities have been reduced to a pulp, upon which the special
Buddhist waste-disposal deity Ucchusmakrodha is emanated. The latter
consumes the mess, through which process the Saiva pantheon become
revived once more; now they take the service-names of 'Grub-pa Lanka'
(sic), promising to protect future Buddhist yogins, and offering themselves as the seats of the Buddhist deities. Next, VajrakTlaya copulates
with each one of the consorts of the Saiva deities, and from these
unions, the series of goddesses Gauri etc., Sinhama etc., and Ankusa
etc., are born. After being used in this way, the Saiva female deities
themselves are bound under oath as servants, and consigned to the outside of the mandala (as the protective Twenty-eight IsVaris). With this,
the emanation of the VajrakTlaya mandala of deities is complete.
Let us now look at how this narrative is interpreted and exploited by
the tradition. To my mind, the taming narrative clearly signals sacrificial
motifs, as I believe is also the case in Chapter 15 of the *GuhyagarbhatattvaviniScaya. The main implement used by the Buddhist deities in
their work is the kila, which, as I have shown elsewhere, shares distinct
iconographic and ritual features with the yupa, or sacrificial stake
(MAYER 1991).28 The Saiva deities are slain, consumed by the Buddhist
deities, digested by them, and then excreted, a symbolic representation
of the transformation of impurity still widely current in India and which
in itself has sacrificial overtones, here with the digestive fire of the
deities analogous to the transformative fire of the sacrifice.29 Through
rtsa ba 'i rgyud kyi dum bu 'i 'grel pa shin po bsdus pa dpal chert dgyes pa 'i ial
lun ies bya ba (henceforth DG), 85 ff.
28. Robert MAYER 1991: "Observations on the Tibetan Phur-pa and the Indian
Kila", in The Buddhist Forum vol. II, London: SOAS. A considerable quantity
of data further reinforcing my original association of the phur pa with the yupa
has come to light since this article was published. The reader might like to note
that in this article, a computer-generated hyper-correction resulted in the two quite
distinct Tibetan words phur pa and phur bu becoming conflated as a meaningless
single word, *phur ba.
29. See Jonathan PARRY, 1985. "Death and Digestion: the Symbolism of Food and
Eating in North Indian Mortuary Rites". In Man vol. 20.4, pp. 612-630. PARRY
writes: "Digestion is thought to distil the good and nourishing part of food from
the bad waste products; and it is employed in a wide range of cultural contexts. It
is argued that by ingesting and digesting the deceased, his impure sins are elimi-
JIABS21.2
292
this process, the impure 'Proud Gods' are transmuted into pure aspects
of the VajrakTlaya mandala: in other words, we can see the entire
process as a rite of passage, through which Rudrahood sheds its delusive
aspects and achieves maturity into Herukahood.30 In this sense, the
taming of Rudra is a symbol of the entire Buddhist path. But let us look
more minutely at what the Phur-pa bcu-gnis gives us out of this great
sacrifice of Manes' vara /Rudra.
To start with, the taming narrative becomes the opportunity to introduce the specifically VajrakTlaya mandala for the first time. Before this
point in the Phur-pa bcu-gnis, only general Vajrayana categories had
been discussed, and only the non-specific peaceful mandalas had been
described in any great detail. Now, the main subject matter or the central
mandala of the text is finally introduced (a moment marked by the
change in the name of the interlocutor from Vajradharma to Karmaheruka, as I have already mentioned above). The manner of this introduction of the main subject matter is also noteworthy: the detailed
description of the VajrakTlaya or Vajrakumara mandala precisely and
exactly matches the step by step stages of visualisation followed in the
sadhana traditions. Now, we must remember that VajrakTlaya is a Mahayoga cycle, and in the rNin-ma-pa tradition, Mahayoga is understood to
put a special emphasis on the visualisation processes of the development
stage, skyed rim or utpattikrama, which constitutes its main contemplative technique. Hence in giving the actual visualisations that are the main
basis of the main practice through which yogins approach their main
spiritual objective of identification with the deity Vajrakumara, the text
is indeed at this point finally offering up its central tenet. The actual
narrative runs as follows: in order to tame Mahesvara and his retinues,
the teacher of this Tantra, the Lord, the Master of Supreme Secrets
[VajrakTlaya], does the following acts (I paraphrase and summarise the
text):
nated, while his pure essence is distilled and translated by the 'digestive fire' of
the stomach to the other world - as the corpse is transmitted by the fire of
cremation, and offerings to the gods by the sacrificial fire" (summary, p. 612).
30. For analyses of sacrifice as rites of passage, see Charles MALAMOUD'S learned
discussion of Vedic ritual, Cuire le monde, Paris 1989: 248 ff; for a broader
theoretical view, see also Edmund LEACH, Culture and Communication: the
logic by which symbols are connected. Cambridge University Press 1976; especially pp. 77-93.
MAYER
293
Firstly, to build up the deity's palace, he emanates the syllables e,
yam, ram, ma, sum, kem and bhrum, along with the utterance of the
associated mantras, e akaia hum, karma raksa ham, ram vajra jvala ram,
mahdrakta jvala mandala, sum samaya hum, kem nirrti mandala, and
bhrum bhanda (sic) jnanacakra jvala mandala. From these are built up
the mandalas of the five elements and the cosmic mountain that form
the foundation of the deity's palace: the triangular pyramidal blue
mandala of the space element, identified with Samantabhadii, the green
mandala of the air element shaped (in this case) like a crossed vajra and
identified with Samayatara, the triangular red mandala of the fire element identified with PandaravasinI, the circular white or red mandala of
the water element identified with Mamakl, and the square yellow
fnandala of the earth element identified with Buddhalocana; upon these
bases arises the Mount Meru of skeletons, with the immeasurable blazing
skull palace of the deity on the very top.31 Those familiar with
Vajrakllaya sddhana texts will easily recognise that this sequence is
virtually identical to the ones commonly found in the sddhana traditions
of the Sa-skya-pa and rNin-ma-pa alike; it can be found in numerous
texts of all types, for example the long Sa-skya-pa Phur-chen, the short
and popular Sa-skya-pa Nes-don thig-le (henceforth NT),32 or the
extensive rNiri-ma-pa gNam-lcags sPu-gri (henceforth NP),33 etc. etc.
Next in the taming narrative of the Phur-pa bcu-ghis we get the emanation of Vajrakumara's important main consort of union, DIptacakra,34
through the utterance of the mantra om vajra kilikllaya mahakrodhi hum.
31. gTin-skyes dgon-pa-byah ms. reprint, vol. DZA, pages 83-85; sDe-dge xylograph, vol. PA, folios 206v-207r; mTshams-brag ms. reprint, vol DZA, pages
872-875.
32. My edition of the Phur Chen is from Rajpur, in 79 folios. NT can be found at
sGrub-thabs kun-btus, vol. PA, pp. 165-169.
33. dPal rDo-rje Phur-bu bDud- joms gNam-lcags sPu-gri, from The Collected
Works ofH.H. bDud-joms Rin-po-che, vols. 10 and 11. n.d., n.p.(for the fullest
description of this sequence, see its bsfien yig folios 92ff).
34. Kon-sprul explains her name as follows: "DIpta means blazing, and cakra means
wheel". He goes on to explain the wheel specifically as a wheel of destruction
that kills enemies (DG, folio 101, lines 1-2). The rendering of her name given by
Martin BOORD as Trptacakra', interpreted by BOORD to indicate a 'wheel of
bliss', is not attested in any of the commentaries I have so far seen.
Unfortunately, BOORD does not cite any sources for his unusual rendering of the
consort's name. See Martin J. BOORD 1993: The Cult of the Deity Vajraklla.
Tring: Institute of Buddhist Studies.
JIABS21.2
294
She too has not appeared in the text up to this point. She is closely
followed by the first manifestation of the central deity Vajrakumara
himself, the very well known form practised in sadhana and surely the
most popular yi dam or istadevata among the entire rNin-ma-pa tradition. In many respects, this is absolutely the central event of the whole
scripture. I quote from my draft translation:
Then the fearsome lord Vajrakllaya uttered 'hum hum hum vajrakTlaya sarvavighnan bam hum hum hum phatV, upon which he emanated wrathful deities
from out of his body, speech and mind. These emanated deities [killed and]
'liberated' the arrogant gods [Mahesvara etc.] in all the ten directions; and then,
regathering, they merged together in front of the Lord, and transformed themselves into the Excellent Son, Vajrakumara. He had three faces, the right one
white, the left one red, and the middle one dark blue, which were very wrathful;
and he stood with his four legs held in the posture with [two] drawn in and [two]
extended. Filling the surrounding space with his vajra wings, his dark reddish
brown hair bristled straight upwards. His head was ornamented with the blooddrinkers of the five [families], and a complete human skin was worn draped
around the upper part of his body. A fresh elephant skin was worn at his flanks,
and he was ornamented with garlands of dry and fresh heads. His four joints
were ornamented with snakes of the four varnas, and he wore a lower wrap of
tiger skin. Residing within a dark maroon mountain of cremation fire, he rested in
the embrace of his consort Ekajata. Thus [the many wrathful deities] were transformed into the single [form of Vajrakumara], who, having trampled underfoot
the vicious arrogant gods [Mahesvara etc.], stood there like a servant attending
his lord.35
The description given here is exactly that of the version of the deity as
visualised in sadhana, in this case embracing the 'liberating' aspect of
his consort, Ekajata, rather than her 'uniting' aspect Dlptacakra. So here
we learn from the taming narrative that the great yi dam Vajrakumara,
the focus of more rNin-ma-pa sadhana practice than any other single
35. From sDe-dge xylograph NGB, vol PA, folio 207v: / de nas 'jigs byed kyi bdag
po badzra kT la yas II hum hum hum badzra kl la ya I sarba bigh nan bam hunt
hum hum phat I ies brjodpas I sku gsun thugs las phyuh ba 'i khro ho de dag gis
I phyogs bcu 7 dregs pa can de dag bsgral nas I slar 'dus te I bcom Idan 'das kyi
spyan sha na sras mchog rdo rje gion nur gyur te I ial gsum pa I g-yas dkarpo
I g-yon dmar ba II dbus mthiri nag II ial tin tu yah rhams pa I tabs bii brkyan
bskum du biugs pa I rdo rje 'i gfog pas gtams pa I skra kham nag gyen du
brdzes pa I khrag 'thuh Inas dbu la brgyan citi g-yah gii ya kor du mnabs pa I
glati chen gyi ko Hon go zur mnabs pa I thod pa skam rlon gyi phreh bas brgyan
pa I sbrul rigs biis mdo biir brgyan pa I stag Ipags kyi s'am thabs can I dur
khrod kyi me ri smug nag gi nah na biugs tih I yum ral gcig ma dan 'khril pa 'i
tshul du biugs pa gcig tu gyur te I bran jo bo la bka' nod pa 'i tshul du biugs fin
gdug pa can gyi dregs pa rnams iabs 'og tu brdzis te gnas par gyur to I
MAYER
295
deity, was manifested especially and specifically to effect the killing of
the hostile Saiva deity, MaheSvara/Rudra! The point I will try to draw
out later on is that when in their daily sddhanas yogins practising Vajrakumara build up the visualisation of the elements, the palace, and the
deity in stages in exactly the same way as described here, they are quite
consciously emulating the great archetypal act of taming which the
Buddha, in his tantric aspect as the Great Lord, the Master of Supreme
Secrets, first did in a bygone age: these later yogins too are doing it for
the express purpose of killing Rudrahood, even if not the original
legendary Rudra himself, for he has already been slain. First, let us
return to the sequence of the taming narrative. The next passage reads as
follows:
Then, in order to endow [Vajrakumara] with the great [vajra] pride of the wrathful [dharmata], and to bestow empowerment [abhiseka] [upon him], the Great
Blood Drinker [Vajrakllaya] uttered the following [mantra]: 'Hum vajrakllaya
hum jhanavajra'; upon which he placed a nine-pointed vajra of wisdom [jnana]
into [Vajrakumara's] first right hand, a five-pointed vajra of the five wisdoms
into [Vajrakumara's] middle right hand, and a Mt. Meru [ktla], for piercing the
defilements [klefa], into the lower right hand. Then, placing the skull-cup of great
compassion's lust in the first left hand, empowerment was bestowed. Since
[Vajrakumara] had subjugated [all] the vicious [gods i.e. MaheSvara etc.] in [all]
the ten directions, he [also] appropriated their emblem, the khatvahga, to be
brandished as a sign of heroism in the middle left hand.36
Thus this passage describes how the hand-emblems of the deity are
generated, once again in a sequence very like that of the sddhanas, and
once again at least partly in terms of the conquest of Mahes'vara/ Rudra.
The narrative continues:
"Then, in order to [kill and] 'liberate' the [non-Buddhist] protector [deities] of the
directions [of space], the King of the Blood Drinkers Vajrakllaya once more
entered into the equanimity of a samadhi of vajra wrath; and from out of the
body, speech and mind of Vajrakllaya himself, [the following mantras] issued
forth:
36. From sDe-dge xylograph NGB, vol PA, folio 207v: / de nas khrag 'thun chen
pos khros pa 7 ha rgyal chen por byin te dbah bskur ba 7 phyir I hum badzra klla
ya hum dzha na badzra ies brjod de I ye ies kyi rdo rje rtse dgu pa ni g-yas kyi
dan por by in I ye fes lha 'i rdo rje rtse lha pa ni phyag g-yas kyi bar mar byin I
non mohs pa gzir ba'i ri rab ni g-yas kyi tha mar byin I thugs rje chags pa 'i
bhan da ni g-yon gyi dan por byin te ban bskur ro II phyogs bcu 7 gdug pa can
btul nas de'i lag cha kha twam ga phrog nas g-yon gyi bar ma na dpa' rtags su
bsnams so I
JIABS21.2
296
Om buddhakilikilaya sarvavighnan bam hum phat
Om vajrakilikilaya sarvavighnan bam hum phat
Om ratnakilikilaya sarvavighnan bam hum phat
Om padmakilikllaya sarvavighnan bam hum phat
Om karmakilikilaya sarvavighnan bam hum phat
Upon this utterance, ten activity kilayas became emanated - those of the activity
kilaya of the blood drinker's wrath, the activity kilaya of vajra wrath, the activity
kllaya of ratna wrath, the activity kilaya of padma wrath, the activity kilaya of
karma wrath, and the [five] activity kilayas [further] emanated from those of the
five families.37 [All of] these also had three faces and six arms for their upper
body, but their lower body appeared as the three sided blade of an iron kila. The
reddish brown hair upon their heads bristled up on end in a triangular [shaped]
lock, and their heads were perfected with the five [buddha] families. With seed
syllables at their hearts, they obediently took up their positions before the
Supreme Son Vajrakumara, upon thrones of [the syllable] e."38
The Vajrakflaya sadhanas often talk of three aspects to the mandala to
be visualised: the dharmakaya mandala, the sambhogakaya mandala,
and the nirmanakaya mandala. These are typically built up in visualisation in sequence, one after the other. With the figures already described
above, up to and including the figures of Buddhakllaya, Vajrakllaya,
Ratnakllaya, Padmakllaya and Karmakllaya, the dharmakaya mandala is
complete.39
37. The text here is rather obscure, in that it is unclear what the further five are. I have
been unable to resolve this problem by consulting NP, NT, DG etc.; most likely,
they are the consorts of the kilas of the five families.
38. From sDe-dge xylograph NGB, vol PA, folio 207v:-208r / de nas yah khrag
'thun gi rgyal po badzra ki la yas phyogs skyoh ba bsgral ba 7 ched du I rdo rje
khro bo 'i tin he 'dzin la shoms par zugs nas I rdo rje phur pa hid kyi sku dan
gsuh dah thugs las than to II om buddha kili ki la ya sarba bighndn bam hum
phat II om badzra ki li ki la ya sarba bighnan bam hum phat I om ratna ki li ki la
ya sarba bighnan bam hum phat I om padma kili ki la ya sarba bighndn bam
hum phat I om karma ki li kT la ya sarba bighnan bam hum phat I zes brjod pas
II khrag 'thun khro bo las kyi phur pa dan I rdo rje khro bo las kyi phur pa dan I
rin chen khro bo las kyi phur pa dan I padma khro bo las kyi phur pa dan I las
kyi khro bo las kyi phur pa dan I rigs lha las gyur pa 'i las kyi phur pa bcu 'thon
par gyur to II de yah sku stod ial gsum phyag drug pa I sku smad Icags kyi phur
bu zur gsum du snah ba I dbu skra kham pa ral pa zur gsum pa gyen du snah
ba I dbu la rigs lha rdzogs pa I thugs kha na sa bon dan Idan pa I e 'i gdan la
sras mchog rdo rje gzon nu'i spyan sha na bka' nod pa 7 tshul du biugs so I
39. To give a highly typical example of this last aspect of the dharmakaya mandala:
In NP, the kilayas of the five families immediately surround Vajrakllaya, Dlptacakra and Ekajata in the Root Mandala. They are seen as direct expressions of
Vakrakllaya in terms of the five jhanas. According to NP las byah: they all
resemble the root deity in having three heads, two wings, six arms, and in having
MAYER
297
As one might expect, the taming narrative continues to proceed in
following in the steps of the sddhana tradition, by presenting the sambhogakaya mandala next; this mainly comprises the group of Heruka
deities known as the Ten Wrathful Ones (khro-bo-bcu, dasakrodha),
along with their consorts and attendant deities. Then come the various
figures of the nirmanakaya mandala, who have a more protective function. There is neither space here nor, I feel, any need to give further descriptions of the highly complex sambhogakaya and nirmanakaya
mandalas, but I should reiterate that this absolutely fundamental threefold construction of the main central skyed rim or visualisation practice
is not described anywhere else within the twenty-four chapters of the
Phur-pa bcu-gnis other than here, in the context of the taming narrative,
just as the twenty-three lines from the famous root verses do not occur
elsewhere.
their two lower arms rolling a kila, but the lower part of their bodies are formed
into phur bus, rather like the Supreme Sons. To the right is blue Buddha-kTlaya,
supreme body, right face white, left face red, upper right hands hold meteoric iron
vajra and wheel, left ones hold khatvdtiga and fire; lower hands roll a meteoric
iron phur bu; the lower body is a conch phur bu. To the east is white Vajrakllaya, supreme mind, right face blue, left face red, upper right hands hold nine
and five spoked vajras, left ones hold iron hook and skull-cup of blood; lower
hands roll a silver phur pa; the lower body is a silver phur bu. To the south is
yellow Ratna-kllaya, supreme qualities, right face white, the left red, upper right
hands hold blazing jewel and war axe, left ones hold noose and skull-cup of
blood; lower hands roll golden phur pa; the lower body is a golden phur bu. To
the west is red Padma-kllaya, supreme speech, right face white, left blue, upper
right hands hold lotus and blazing sword, left ones hold skull-cup of blood and
iron chains; lower hands roll a jewelled copper phur pa; lower body is a copper
phur bu. To the north is green Karma-kflaya, supreme activities, right face white,
the left red, upper right hands hold crossed vajra and blazing skull staff, left ones
hold mass of flames and bell; lower hands roll a jewelled turquoise phur pa;
lower body is an iron phur bu (NP las byan: 97-98).
JIABS21.2
298
sGrol ba
I want to turn now from the standard or typical Mahayoga visualisation
practices of Vajrakumara, to a particular yogic practice known as sgrol
ba, or moksa, which is the most famous speciality or hallmark of the
Vajrakllaya tradition.40 Sgrol ba is basically a practice of ritual killing.
It takes many forms and is practised on many different occasions, but is
most popularly done as a central component of the ganacakra or tshogs
offerings connected with the extended forms of the standard sadhana
practice. In this rite, an effigy is typically made, described as a linga, to
be ritually stabbed by the yogin with a klla or phur pa. The idea is that
spiritual negativities are visualised as concentrated within the linga, and
these are then forcibly transformed into wisdom through 'slaying' the
linga with the klla. In the Vajrakllaya tradition, the linga is commonly
identified as 'Rudra', and the act of stabbing the linga with the klla can
be seen as an emulation of the original act of Vajrakumara in slaying
MaheSvara/Rudra. So here once again, a central aspect of the Vajrakllaya tradition is linked with the taming narrative, and unsurprisingly
MahesVara-Rudra can frequently become the central focus of the entire
rite. There are numerous examples of this in the sadhanas. I quote here
from Cathy CANTWELL's translation of NP:
A lifiga or effigy of Rudra is made and placed before the practitioner; the visualisation instructions are thus:41
"From the heart of oneself visualised as Vajrakumara, multitudes of wrathful
emanations are sent forth...all negative forces, lacking any independent power,
become summoned and dissolve into the linga [of Rudra]..."
While visualising as above, one recites:
... Multitudes of Rudras of breakers of samaya, hostile forces and obstacles, are
summoned.
The time for their 'killing and liberation' [sgrol] has come!'
Then, rolling the klla, one strikes at the heart of the linga, which by
now embodies all the Rudras, and recites:
40. The most comprehensive treatment of this subject so far is by Catherine M.
CANTWELL, forthcoming: "To Meditate on Consciousness as the Vajra Nature.
Ritual 'Killing and Liberation' in the rNying-ma-pa tradition", in H. Krasser et al.
eds., Tibetan Studies vol I, Vienna: Verlag der Osterreichischen Akademie der
Wissenschaften, 1997, pp. 107-118.
41. Page 130, line 6, to page 131, line 3.
MAYER
299
These Rudras of the path of confused appearances
Are 'killed and liberated' in the expanse of the unborn essence.
They are purified in the great spontaneous three kayas....
They are united in the circle of Great Bliss....
The instructions continue: after being stabbed at the heart with the kila,
the Rudra linga is to be stabbed in all its other cakras too. Then further
weapons are employed. The liriga is bombarded with magical black
mustard seeds and other power substances; it is sliced with a small ritual
sword; then each of the resulting slices is hammered with a small ritual
hammer; and finally mashed with a pestle. Each of these actions is
accompanied by the appropriate liturgy and visualisation. At the end, the
mashed remains of Rudra are offered up to Vajrakumara and his retinue
to be eaten.
As well as this standard practice of sgrol ba as a means of transforming one's own negativity, there is another, much rarer form that envisages the actual killing by magic of a living human being. Needless to
say, given the Buddhist context of these rites, the tradition holds that
such literal killing should only be done in the most extreme circumstances, when the heinous demerits being earned by a potential victim
are so vast that it becomes an act of mercy to kill them, to save them
from a certain rebirth in hell. In addition, tradition holds that the performer of such rites must be highly accomplished, possessed of sufficient compassion and siddhi to unfailingly transfer the victim's consciousness to a higher realm - otherwise the performer will themselves
incur the evil karma of murder. In the rNin-ma-pa tradition, potential
victims of this rite are typically described as Buddhist yogins who have
misunderstood the real meaning of Tantric practice. Instead of taming
Rudra, they have become Rudra. Instead of a Buddhicised Mahayanacongruent kapalika practice, they have regressed into following an
uregenerated version as originally taught by Rudra. The Phur-pa bcugnis Ch. 12 describes them as follows:
"Now for the characteristics of those who have fallen away from the purpose [of
Vajrayana practice].
In performing their duties, they do various things wrong:
They handle the sacred substances and ritual utensils in public;
They practice the rites of tana and ghana [i.e. sexual union and killing], in a
[purely literal] physical [way];
They turn their backs on the view and on contemplation;
They are always ready to indulge in coarse behaviour;
They express anger, rage and pride for no reason;
They understand truthful oral instructions wrongly;
JIABS21.2
300
When offering guidance to others, their [teaching of] Dharma is false;
They are erudite in [any] lore that misleads;
Casting aside precepts and scriptural authority, yet they embark on grandiose
undertakings;
They practice assorted perversions;
Since such persons are genuine Rudras,
Even in [killing and] liberating them with the abhicara [rites],
one remains unstained by sin."42
MaheSvara/ Rudra in Mahayoga Doctrine
I want to finish this section of my paper by briefly summarising how the
figure of Mahegvara/Rudra fits into the Mahayoga doctrinal system as a
whole. In very general terms, the basic view of the rNin-ma-pa Mahayoga system is a non-dual one, not dissimilar to that of the Yoginltantras. It holds that all phenomena, whether conventionally designated
good or bad, pure or impure, should all ultimately be realised as being
from the point of view of absolute truth the inseparability of appearance
and emptiness (snan ston dbyer med). To approach this from the point of
view of relative truth, one meditates on all phenomena alike, whether
pure or impure, whether good or bad, as the primally pure mandala of
the Tantric deities. The conventional dualistic designations we impose
on phenomena as good or bad, pure or impure, are thus seen as having
no ultimate validity and are considered illusory, predicated upon false
notions of inherent existence (ran bziri) and ontological duality {ghis
'dzin). A ritual corollary of this Mahayoga doctrine is that those factors
of existence or mind designated as impurities or sins to be abandoned in
the conventional Mahayana systems, are here to be retained as potential
sources of wisdom, as the raw materials of spiritual practice; thus rather
than merely abandon such negativities as the five moral defilements or
kleia of conventional Buddhism, in Mahayoga one seeks to overcome
them through the alternative method of retaining them and exposing
their true nature as aspects of the great purity of the mandala of deities,
and, from the ultimate point of view, as appearance and emptiness
42. From sDe-dge xylograph NGB, vol PA, folio 219r: / don las Hams pa 'i mtshan
hid ni II bya ru mi ruh sna tshogs byed II rdzas dah lag cha mnon du 'dzin II tan
gan sbyor sgrol dnos por spyod II Ita ba tin 'dzin rgyab tu bor II spyod pa brlan
po dan du len II khro gtum na rgyal ran gar gtori II don gyi man nag log par go
IIpha rol 'dren la log pa 'i chos IIphyin ci log gi don la mkhas II bka' gzun bor
nas rtsom pa che II log par spyod pa mtha' dag spyod II 'di ni ru dra dnos yin te
II mhon spyod bsgral yafi sdig mi gos I
MAYER
301
inseparable. To take a specific example: in the case of the defilement of
aggression which is the main focus of the Vajrakllaya tradition, this
would imply not abandoning it, but meditating on it; meditating on
aggression should show it to have as its ultimate true nature, or to be
from the point of view of emptiness, the mirror-like wisdom (jnana) of
the Buddha Aksobhya. The method of meditating on aggression would
be to visualise it as an aspect of the primal purity of the mandala of the
Tantric deity. Thus the basic method used to achieve the transformation
of negativity in Mahayoga is to visualise all negative factors as aspects
of the wrathful Heruka deity mandala, and as components of the special
tantric offerings made to the Heruka. A central principle behind this
practice is discussed by 'Jam-mgon Kon-sprul in his commentary on the
Vajrakllaya root verses I have mentioned above. He explains that the
wrath of Herukas such as Vajrakumara is directed against the city of ego
projections built upon the duality of subject and object and discriminating discursive thoughts; such a wrath is quite unlike the hatred of personal egotism aimed at a specific enemy; it is more like the radiant sun
fiercely dispelling the gloom of discursive conceptualisation all round
(DG p.81). Similarly, the Heruka's wrathful compassion cuts through
hatred, because it can not co-exist in the mental continuum with hatred
any more than heat can co-exist with cold in a single substrate, since
compassion and hatred are mutually exclusive (DG p.82).43 The implication seems to be that simply by putting whatever negativities there
might be in direct proximity with the spiritual presence of Vajrakumara,
all such negativities will spontaneously become destroyed and their
energy will become transformed into wisdom. Thus one visualises the
Heruka's cemetery palace as replete with negativity, graphically represented by gruesome symbols drawn from kapalika iconography, because
from the relative point of view, the compassionate Heruka is understood
43. DG p.81: gan la khros na I mtshan ma 'i rnam par rtog pa gzuh 'dzin bdag rtog
gi groh khyer la khros I tshulji Itar khros na I ie sdah rati rgyud pas mig s'is
fian dgra la khros pa ha bu ma yin par ni ma 'char ba 'i gzi brjid kyis mun pa 'i
smag rum mdun na mi gnas pa Itar I DG p.82: don ni I ie sdah gcod ces pa kun
snah shih rjes ie sdah gcod pa ste I dper na tsha reg gan na yodpa na gran reg
med I gran reg gan na yodpa na tsha reg medpa de biin du I gan zag gcig gi
rgyud la ze sdah skyes tshe siiih rje med I snih rje skyes tshe ie sdah med pa
than cig mi gnas 'gal ba yin cih I Cathy Cantwell informs me that this passage is
probably not Kon-sprul's original composition: almost identical passages occur
in the NP bsnen yig, for example, suggesting an older common source (NP bsnen
yigpp.87-88).
JIABS21.2
302
to delight in his bodhisattva's acts of effortlessly and spontaneously
transmuting negativities into wisdom, and thus he constantly immerses
himself in negativity. At the same time, from the point of view of absolute truth, the Heruka mandala can be understood as the real nature of
such negativities, once their empty or ultimate nature is understood as
the inseparability of appearence and emptiness.
An additional significant point in Mahayoga doctrine is that its commentators also sometimes take a position found also in the YoginTtantras,
which argues that since beings of the present dark age are so heavily
defiled, they quite naturally have far more negativity than virtue. It
follows that pragmatically speaking, the potential good results of a spiritual approach which aims at the transformation of negativities, in
general tends to outweigh the potential good results of a path that aims
at the cultivation of positive values, by sheer force of weight. From such
a logic, it follows that environments and persons with a great deal of
negativity are the ideal and intended sphere of operations for deities such
as Vajrakumara; for if the heruka practice is being practiced successfully, the more negativities there are, the more wisdom energy will be
produced.
Now, as we have seen above, the commentarial tradition explains that
from the point of view of conventional analysis, Rudra is the Vajrayana
equivalent of Mara, the main Buddhist symbol of evil. From the absolute point of view, it also describes him as primordially pure, as an
aspect of the Buddha Samantabhadra since beginningless time, merely
appearing as evil and impure for the express purpose of allowing
enlightenment to become manifest through the process of overcoming
his apparent negativity. It would therefore make sense if in Mahayoga
ritual, the negativities visualised in sadhana with the intention that their
pure inner wisdom aspect should be revealed, are identified with Mahesvara/Rudra; and this is in fact exactly what happens. Rudra stands for
the primordial ground to be purified, or, to use the analogy of alchemy,
the base metal of defilement to be transmuted into the gold of Buddhist
enlightenment. Just as one can not alchemically make gold without first
having some base metal to transmute, so one can not manifest enlightenment without a Rudra of egohood to liberate. In particular, in the Mahayoga cycles, the specifically kapalika categories associated with Rudra in
the taming narratives are expressly identified with the negativities to be
transformed.
MAYER
303
In sddhana texts, one therefore comes across frequent references to
MaheSvara/Rudra, in nearly all cases broadly identifying him with the
negative basis of purification. In the NP las byan, currently a very popular sddhana to Vajrakumara, the cemetery palace is visualised as built up
of the dismembered fragments of Rudra's corpse, with the roof canopy,
for example, being described as made of Rudra's flayed skin (NP las
byari, Page 93, line 5). On the one hand, the palace as made of Rudra's
dismembered corpse signifies the expanse of the negativities which need
to be transmuted, within which the Heruka figure has his ideal field of
activities; yet at the same time, the Rudra cemetery palace is also seen as
an expression of pure primordial wisdom, a demonstration that confusions have primordially been of the nature of wisdom, in accordance
with the primordially pure ultimate nature of Rudra as described in
Kloh-chen-pa's commentary on the *Guhyagarbha-tattvavinis'caya, Ch.
15. Similarly, referring to the practices of a rDo rje gro lod and a Bla
sgrub ritual manual of the bDud 'joms tradition, CANTWELL notes that
the wrathful series of the so-called outer offerings (or in some cases the
inner offerings), which include items such as the flowers of the senses,
the incense of melting human fat, butter lamps of burning stomach fat,
perfume of blood or urine, food of human flesh or excrement, and
music of skull-drums and thigh-bone trumpet, are often explained as the
parts of the body of Rudra.44 By offering these to the Buddhist Heruka,
the idea is that the negativities they represent should become spontaneously revealed as wisdom, a wisdom which is already inherent within
them when they are understood from the ultimate level of emptiness.
So the specific function of the wrathful Buddhist Herukas in Mahayoga is an aggressive one: rather than attempting to directly increase
good qualities which are already inherent anyway, Herukas exist in order
to counteract the negativities that mask the primordial perfection. Their
function is primarily to destroy what needs to be destroyed, above all to
demolish dualistic thinking and intellectual clinging to the idea of inherent existence, even more than to encourage or increase any good qualities. Their wrathful, exorcistic names, epithets and representations
underline their primarily destructive orientation: Vajrakflaya is described
as cutting through hatred much more frequently than he is described as
44. Catherine M. CANTWELL 1989: An Ethnographic Account of the Religious
Practice in a Tibetan Buddhist Monastery in Northern India. Unpublished PhD
thesis, University of Kent at Canterbury. See pages 135, 180.
JIABS21.2
304
building up love, while Hayagriva is described as subduing arrogance far
more frequently than he is described as cultivating humility. Now, as we
have already seen above, (embarrassing though this might be for the
ecumenically-minded contemporary Buddhist seeking a rapprochement
with, say, Kashmiri Saivism), the Mahayoga commentarial tradition has
chosen as its main symbol of negativity the kdpalika figure of Mahesvara/Rudra, which in this context largely displaces the more traditional
Buddhist symbol of evil, Mara. It is not surprising then that one is often
reminded within the literature that a major purpose of the Buddhist
Heruka is to attack and destroy Mahes'vara/Rudra! If Mahesvara/Rudra
is the embodiment of all that needs to be destroyed, then, by an unavoidable logic, it follows that the whole spiritual purpose of Mahayoga
can be and often is symbolically expressed as the destruction of
MaheSvara/Rudra. Thus, as the NP liturgy puts it, the whole point of
Vajrakumara is that his "wrathful roar of hum\ [shall] destroy the brains
of Rudra" (NP las byan, p. 119, line 6); the whole point of his entourage is that they should "annihilate the Rudras of the teaching" (NP
las byan p. 121, line 2); as the prayer of fulfilment (bskan ba) describes
them, the Vajrakllaya deities are blessed and praised precisely because
they "[have] the mighty power to subdue the Rudras of perverse views"
(NP las byan, p. 160, line 1). Thus through achieving union with such
Heruka deities as Vajrakllaya, the Mahayoga sddhaka might aspire to
emulate the great gter ston and saint 'Ja'-tshon sitin-po, who famously
earned that highest (if not ecumenically resonant!) spiritual accolade of
the rNiri-ma-pas, bdag 'dzin ru dra 'joms pa, 'The Vanquisher of
Rudra-egohood.'
In Conclusion
In conclusion, we can see that Tibetan Buddhism seems to allow the
figure of Mahesvara/Rudra to play only a comparatively small role in
the Cakrasamvara traditions, although a considerably larger one in the
rNiri-ma-pa Mahayoga systems, where he functions as one of the most
important symbolic categories of that tradition. In both cases, however,
the Saiva deity and the traditions connected with him are predominantly
constructed in abstract metaphysical terms. The actual concrete historical
significance of the Saiva tantric traditions for Vajrayana Buddhism is
only rarely or almost never the focus of Buddhist scholastic attention.
The surface historical reading of the taming narrative, that so obviously
to Westerners seems to admit that the kdpalika elements of the Buddhist
MAYER
305
Vajrayana are derivative of Saivism, is not widely followed by Tibetans.
Instead, in the hands of the tradition, this underlying historical narrative
is comprehensively and systematically reconstructed into a purely
metaphysical set of symbols onto which psychological factors can be
correlated in accordance with Buddhist doctrine.
Unfortunately, the question as to why the Tibetan Vajrayana tradition
seems uninterested in its historical debts to Saivism is one I have not yet
had much opportunity to think about. Nevertheless, it might be useful at
this juncture to float a few largely speculative ideas on the subject, and I
have already mentioned some of these earlier. On the one hand, it does
seem to me at the moment that there is surely some degree of simple
denial involved, some effort to gloss over what could be construed as an
embarrassing historical fact for Buddhism: it can surely not be considered a very comfortable situation for Buddhists to constantly to have to
reflect on the substantially derivative nature of some of their most sacred
traditions. On the other hand, it is no straightforward task to ascertain to
what degree such denial is the case, for a number of reasons. Firstly, the
strictly historical understandings of events of any kind, whether shameful or glorious, have never been of much interest or relevance to
Mahayana and Vajrayana Buddhism, especially in India. The Buddhist
scholastic tradition has consistently mythologised every aspect of its
history, the shameful and the glorious to an equal degree. We can see
this from the highly mythologised traditional accounts of the Buddha's
life, the hagiographies of Buddhist saints and kings, predictions about
the decline of the £asana, and the stories about the origins of Buddhist
scriptures and Buddhist sacred sites such as stupas. The Indian Buddhist
scholastic tradition, like its many Hindu counterparts, seems to have
quite consistently sought to distil what it saw as religiously valuable
mythic narratives out of any historical events within its experience, and
not merely the embarrassing events. It is not impossible that the mythologised approaches of the figure of MaheSvara/Rudra evidenced in
Tibetan literature might derive from a typical Indian Buddhist mythological response to history, just as much as from a specific cover-up
attempt.
I am also not clear to what degree the different elements of the
Buddhist tradition in India would have found the facts of its debts to
Saivism embarrasssing. Surely the blatant borrowings of substantial unedited passages from the scriptures of the Saiva canon into the Buddhist
Yoginitantras must have been a considerable embarrassment to many
JIABS21.2
306
among the more respectable clerical elements of Indian Buddhism at the
time; yet this raw co-option of Saiva material was presumably at the
same time seen as a great coup by the 'shamanic' (in SAMUEL'S technical usage) or yogic Buddhist individuals who initiated it. Perhaps these
yogins understood their coup in terms of a continuation of the process of
the taming of Mahesvara described in already well established Yogatantra scriptures such as the STTS; this is a point that also needs to be
carefully considered. Above all, however, as LAMOTTE remarked many
years ago, we must remember the pervasive theme in pre-Tantric
Buddhist literature of the conversion of Mara and his daughters to the
Dharma. The Tantric narratives of the taming of MaheSvara /Rudra
make clear allusion to these much older stories - for example, conversion of Mara's daughters by multiple simultaneous sexual intercourse
(cf. Mahayoga) already existed in Mahayana scriptures such as the
Suramgamasamadhisutra, a text which devotes considerable attention to
the taming of Mara and his daughters. With the identification of Mara as
Mahes\ara-Rudra, an ancient Buddhist literary template (cf. $uramgamasamadhisutra, Mahasamnipata, Vimalaklrtinirdesasutra, etc.)
found a new concretely historical focus. Now there was a real, tangible
Mara out there to be converted, not just a symbolic or mythical one.
So even from a clerical point of view, the general principle of
Buddhism's co-option of Saiva kapdlika tantrism might not have been
altogether unacceptable. A basic tendency of Buddhism from its very
inception is that it seems to have normally preferred to recode and to
respond to or react to existing non-Buddhist categories, rather than to
create new ones of its own ex nihilo. Most of my readers are probably
aware of the work of Richard GOMBRICH and K.R. NORMAN, who
have shown most of the key doctrines of early Buddhism to be constructed from recycled Brahmanical categories. This includes even such
quintessential Buddhist ideas and technical terms such as karma, nirvana,
and the Middle Way, all originally articulated through the medium of
redefining existing non-Buddhist ideas. One can see another aspect of
this tendency neatly represented in texts like the Brahmajala Sutta, the
first text of the Digha Nikdya of the Pali Canon, which presents the
Buddhist point of view through the medium of a critique of sixty-two
specifically non-Buddhist theories. K.R. NORMAN writes as follows:
There have been those who thought that Buddhism was simply an offshoot of
Hinduism, while there are others who maintain that there is no trace of Hinduism
in Buddhism. The truth, as always, lies somewhere between these two extremes.
MAYER
307
What is certainly true is that Buddhism owes much, especially in terminology, to
Brahmanical Hinduism and much of the Buddha's preaching would have been
unintelligible to those who had no knowledge of Brahmanical teaching. Although
some of the technical terms of Buddhism are exclusive to that religion, much
Buddhist terminology is, in form, identical with that of brahmanism. At the same
time it must be recognized that, although the Buddha took over some of the
terminology of Brahmanical Hinduism, he gave it a new Buddhist sense.45
Although NORMAN was here discussing early Buddhism, it seems to me
his words are substantially true of later Buddhism as well. On several
significant occasions through its long history, Indian Buddhism seems to
have recreated itself anew primarily through the medium of reinterpreting or reacting to the existing categories of its opponents, in preference
to inventing new categories of its own. This process of subverting its
rivals seems to have eventually become interpreted as a central Buddhist
virtue and elevated to the level of a conscious dialectic. Perhaps the most
famous example is the way in which the Mahayana Buddhism of the
early Perfection of Wisdom literature primarily defined itself through its
critique of the Abhidharma, upon the categories of which it is by that
very token substantially dependent for its articulation. Similarly, as Paul
WILLIAMS suggested to me some years ago, it might be that Buddhist
logic invented itself in an effort to defeat the Nyaya logicians in their
own terms. Buddhist logic is therefore philosophically diametrically
opposed to Nyaya - idealist or phenomenalist as opposed to Nyaya's
naive realism - while nevertheless expressing itself almost entirely
through the medium of originally Nyaya types of discourse. The most
basic Mahayana philosophical notions often seem to encapsulate this sort
of propensity - the apohavdda claims that nothing can be defined except
in terms of what it is not, the Madhyamaka dialectic puts forward no
argument of its own but merely negates those of its opponents.
In addition, the three major philosophical strands represented in Vajrayana Buddhism, the emptiness doctrines of the Madhyamaka, the mindonly doctrines of the Yogacara, and the Buddha-nature doctrines of the
Tathagatagarbha sQtras, all share an important basic axiom. They all
hold that emptiness or the ultimate nature is already inherent within all
the profane phenomena of samsara, and can never be found or constructed outside of them. Thus the major purpose of Buddhism was seen
to be to expose the ultimately true nature already inherent in existing
45. K.R. NORMAN: "Theravada Buddhism and Brahmanical Hinduism: Brahmanical
Terms in a Buddhist Guise", Buddhist Forum II, p. 193.
JIABS21.2
308
defiled phenomena, not to try to create new structures of truth outside of
them. Truth or Dharma was seen as self-existent and all pervading; the
only purpose of Buddhism was to point it out to beings. To this kind of
thinking, if an originally non-Buddhist deva were compelled by the skilful methods of Heruka to heed the teachings of the Buddha and begin to
realise his inherent Buddha-nature, he could become just as much a
legitimate object of Buddhist devotion as any being of purely Buddhist
lineage who had achieved realisation within a conventional sangha
career. It seems to me that with such an ideological framework, many
elements within Indian Buddhism might have been reasonably unashamed and unperturbed by the fact that their religion had taken substantial borrowings from non-Buddhist sources.
The situation developing within contemporary Tibetan Buddhism is
harder to assess. The barrier of the Himalayas has meant that for around
1,000 years Tibetan Buddhism had no significant social contact with
Saivism, and as a result, 3aivism remained of a mainly abstract and
symbolic value within Tibetan thinking. It is only in very recent years
that Tibetan Buddhist refugees have had to seriously confront their
boundaries with Saivism in any concrete sense, and they have had to
begin to do so with little or no traditional template or precedent to work
from. Over recent decades, the Tibetan refugee experience of modern
Hinduism has generally been very harmonious, but it has also been politically fraught on some occasions. Several Tibetans have claimed that in
Nepal, Hindu bureaucrats have sometimes compelled Tibetan refugee
lamas to sign documents avowing their religion to be a minor subsection
of Hinduism.46 In India, Hindu fundamentalism can be seen as a threatening force by some Tibetans. In Bhutan, the Buddhist populations feel
their traditional way of life to be gravely threatened by what they see as
an engulfing tide of Hindu Nepalese settlers and colonists. As one might
expect, in such circumstances a wide variety of responses to the question
of historical relationships with Saivism seem to be forthcoming. I have
discussed the problems of Saiva/ Buddhist scriptural intertextuality with
a number of Tibetan refugee lamas living in India and Nepal, including
some major figures in the contemporary Tibetan Cakrasamvara traditions, as well as specialists in Mahayoga. My impression is that there are
46. David GELLNER likewise reports that the official Nepalese Government view has
been that Sikhism, Buddhism and Jainism are all branches of Hinduism. See his
Monk, Householder and Tantric Priest: Newar Buddhism and its Hierarchy of
Ritual, Cambridge University Press 1992, page 92.
MAYER
309
no fixed ideological positions dictated by sectarian or doctrinal affiliation; rather, responses seem to vary according to the individual. My
initial impression is that the lamas most closely connected to the yogic
side of Tibetan Buddhism, what Geoffrey SAMUEL calls its 'shamanic
current', seem to be quite happy to admit a concrete historical relationship with Saivism, while the more clerical lamas seem more reluctant to
admit to any Buddhist dependence on Saivism. Thus Khenpo A, a
widely respected scholar of the Karma Bka'-brgyud-pa and rNin-ma-pa
traditions especially favoured by some among the more yogic side of the
tradition, told me he had made a special study of the relation of the
Vajrayana to £aivism whilst studying in Benares. He was an enthusiastic
supporter of the idea that Vajrayana was derivative of Saivism, in the
specific sense that Saivism had been tamed ('dul ba) by Buddhism. He
told me that the next thing would be that Buddhism was going to tame
technology and science, which were going to become a principal
medium through which the Dharma would express itself in future centuries. B Rinpoche, a major incarnate lama with considerable background in traditional mountain retreats dedicated to both the bKa'brgyud-pa and rNift-ma-pa systems, firmly took the view that £aivism
and the Vajrayana were often virtually identical in all ritual respects, but
that Buddhism uniquely applied these rituals to the understanding of
emptiness. He described to me Saiva tannic rituals he knew of, which he
said were almost identical with those of Buddhism. However, he advised
me to dissemble when discussing this fact with Buddhists of a more
clerical or traditionalist mind set, because he said such talk would only
upset them and achieve no benefit. C Rinpoche is the lama in charge of
one of the leading yogic training centres within contemporary Tibetan
Buddhism. His centre is particularly associated with the Cakrasamvara
cycle and its associated yogas, which are practised there in the form of
long solitary retreats of many years duration. C Rinpoche clearly felt
very positively about certain aspects of Hinduism, and spoke warmly of
the virtues that can be found within the Hindu traditions. Our conversation on the subject of Saiva-Buddhist intertextuality came to a slightly
uncomfortable close when I made remarks that could be interpreted to
carry a slightly anti-Hindu nuance. D Rinpoche, who was mainly concerned with the social aspects of Buddhism and had no experience of
retreat, had entirely different attitudes. At one stage I even became anxious that he might come to blows with the Hindu pandits we met at the
Nepal-German Manuscript Preservation Project offices in Kathmandu,
JIABS21.2
310
when our conversation turned to Saiva/Buddhist parallels, and the
pandits all asserted that the Buddhists had copied their rituals from the
Hindus. We had to leave quite briskly. D's close friend, a senior monk
and a recent escapee from Tibet, had only weeks before been arrested,
beaten and, he alleged, forced to sign a humiliating document by the
Nepalese police, asserting that Buddhism was a minor offshoot of Hinduism. It will be interesting to see how Tibetan Buddhism comes to
terms over the coming years with the growing evidence of its historic
debts to Saivism.