Journal of Tibetan and Himalayan Studies
1
INDO-TIBETAN TANTRIC BUDDHIST
SCHOLASTICISM: Bhavyakīrti and His Summary of
Sāṃkhya Philosophy* (part I)
Leonard W.J. van der Kuijp
Harvard University
Center for Tibetan Studies, Sichuan University
Voor Albrecht Wezler uit dankbaarheid.
…wherever I turned my view, there was perplexity to be disentangled, and
confusion to be regulated; … adulterations were to be detected, without a settled
test of purity…
Samuel Johnson
Preface to the Dictionary of the English Language
PRELIMINARIES
Though many of their salient aspects have yet to be worked out systematically,
it is well-known that a good number of synopses of and commentaries on the
highly esoteric [and not so esoteric] tantric literature written by Indian Buddhists
from circa the late seventh or early eighth century to the twelfth century reflect the
kind of theoretical concerns and approaches to their subject-matter that we
normally expect to encounter in exegeses of Buddhist śāstra-treatises on
phenomenology (abhidharma) and analytic philosophy, and on epistemology and
logic (madhyamaka, yogācāra, pramāṇa). Indeed, aside from the points of focus of
these writings, there is really nothing else to distinguish them from the latter in
terms of method and the tacit and occasionally more overt applications of what are
2
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
recognizably rational criteria. We may thus tentatively call these concerns and
intellectual practices "scholastic" and consider them to be an vital component of
what M.T. Kapstein and R.M. Davidson respectively designated "vajrayāna
scholasticism" and "institutional esotericism."1 Inextricably conjoined with this new
phenomenon is, as a matter of course, a host of factors that made such a union
intellectually possible, socially desirable and, in the final analysis, perhaps even
hermeneutically necessary. Much was of course ultimately connected to the
discovery of, the trust in, and the reliance upon a set of rational and methodological
considerations that, beginning in the first few centuries B.C.E., certain educated
circles of the subcontinent had come to consider fruitful and meaningful avenues
for a better understanding of both oral and written texts.2 Indeed, it has been long
recognized that this was by no means an isolated phenomenon. Very similar
concerns were prevalent in pre-Han China as well as in ancient Greece. We can
safely say that these considerations marked a major turning point within what K.
Jaspers had famously called the Achsenzeit, the Axial Age, for this was the point in
time, from circa 800-200 B.C.E., that we begin to witness a global rise of the
commentary as a literary genre and much else.3
The aforesaid considerations were then expanded upon and made more
analytically creative and intellectually rewarding during the ensuing centuries. One
of their most enduring fruits was "scholasticism" with its profoundly affective
dimension, namely, the intellectual pleasures that accompanied its practice. A good
* A truncated version of this paper was first presented at the China Academy of Social Sciences,
Beijing, on December 25, 2012, and a longer one was given at the Hebrew University of
Jerusalem, Israel, on June 17, 2014, at a conference on esoteric Buddhism that was jointly
convened by the Israel Science Foundation and the Israel Institute for Advanced Studies. My
thanks go out to both institutions and the organizers of these wonderful occasions. That there is
some unavoidable overlap between this and Tomabechi Tōru's paper is detailed in Tomabechi
(2016: 81 under *).
1 See, respectively, Kapstein (2001: 236) and Davidson (2002: 114); see now also Campbell (2009:
39-47, 54-57). Whatever one may wish to call the beast, the particulars of the intellectual practice
of "vajrayāna scholasticism" was clearly a result of "institutionalized esotericism."
2 Much of this is of course contingent upon the practice of writing and the concomitant
emergence of literacy and its sociological implications, for which see the considerations in von
Hinüber (1989) and those in Falk (1993) and, recently, in Strauch (2012). Questions of literacy,
rationalism, and allied ideas in the subcontinent were addressed inter alia in Bronkhorst (2002).
3 For some recent studies of the Axial Age, see the articles in Bellah and Joas (2012), and for the
Indian subcontinent in particular, Kulke (1986) and Eisenstadt (1986). For the global rise of the
commentary as a genre and much else, see especially Farmer, Henderson, and Robinson (2002) as
well as the critical reaction to this essay voiced in Bronkhorst (2006) and the reservations
expressed in Nance (2016) about Bronkhorst (2006). For some general remarks on the role
played by tradition in philosophical analysis, but with specific reference to Vatsyāyana
Pakṣilasvāmin's (ca. 5th c.) Nyāyabhāṣya, see Preisendanz (2000). And, lastly, see also the pertinent
comments in Ganeri (2010) on the nature and scope of the commentary in Indian thought.
Journal of Tibetan and Himalayan Studies
3
deal has been published over the past decades on the inception of scholasticism in
Europe and the principal features that make up this kind of thinking and
intellectual attitude in general, which are now so close to those of us who have our
roots in the Western tradition, almost so genetically encoded in us, that we often
simply take them for granted. Risking platitude, let us just say here that, in an
Indian context, it, too, is foremost characterized by a certain kind of systematic
thinking accompanied by the application of a variety of different methods of
textual exegesis that, as in ancient Greece and China, have their origins in
intellectual practices – in Greece and China, these also often included dialogs and
to some degree rule-governed public debate - that emerged a few centuries before
the beginning of our era with the rise of the śāstra and its novel use of Sanskrit.4
The well nigh universal application of these methods by the Indian intelligenstia in
the first millennium C.E. and beyond was no doubt the result of the fact that some
sort of an agreement on method had been reached, though it is still unclear to what
degree it was socially constructed in a purposive and consensual fashion by the
intellectual elite. That said, the questions of the extent to which the earliest
practitioners of this kind of thinking actually thought about what they were doing,
such as , in what measure there was a self-reflexivity in and a self-consciousness of
their activities, and how they developed their methods [and vocabularies] are now
beginning to be investigated.5 The same holds for the social history of this new
movement that is encapsulated by the emergence of writing and reading as social
phenomena and the spread of literacy, not to mention the concomitant birth of a
new social configuration in which the author, scribe, and reader play mutually
supportive roles. Previously, the cultural and intellectual elite of the subcontinent
worked exclusively with oral religious texts like the Vedas, etc., and no doubt
derived a sense of solemn and quiet pleasure from their careful recitation and use in
ritual. The how and why of the rise of a new order within this class that began to
embed itself with written texts in the social fabric of the subcontinent still need to
be more fully appreciated and reflected upon. The custodians of the altogether
different kind of pleasure that was associated with oral texts formed this new order,
4 For some considerations on the śāstra as such, see Pollock (1985); Pollock (1989: 609, n. 32)
announced an article titled "The Idea of śāstra in Traditional India" as forthcoming in Beiträge zur
Indienforschung, but it was published in Pollock (1989a). But see now more recently the important
contribution to our understanding of the śāstra in Slaje (2007).
5 Although not unproblematic in other respects, Cabezón (1994) contains a number of interesting
remarks on the scholastic attitude and method in Buddhist thought; but see also Manevskaia
(2008) and Verhagen (2005, 2008), and now especially Nance (2011), a pathbreaking study of the
Vyākhyāyukti, Vasubandhu's (5thc.) classic treatment of Buddhist hermeneutics, and Eltschinger
(2018). In many ways derivative of late Indian Buddhist approaches to textuality and scholarship,
Dreyfus (2003) has tackled inter alia Tibetan notions of scholarship and exegesis.
4
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
namely, one that is specifically linked to penmanship, the writing and reading of
manuscripts, and ultimately their collection in libraries of a kind. Most if not all of
them belonged to the upper strata of Indian society, the Brahmins, where cultural
literacy had its highest prevalence. To be sure, we must ask ourselves whether we
might be able to ascertain the presence of certain tensions within the syntax of their
culture and where the most significant inflection points are located along the
interface of the oral and the written; and, similarly, the tension that may have been
present during this transition from a culture that was exclusively oral to one in
which a portion of the oral was gradually being replaced by the literary.6 I very
much doubt that it is simply reducible to the aesthetic and intellectual pleasure
afforded by reading and memorizing texts, and by the analysis of what was read and
memorized. The emotional and intellectual satisfaction that is won by being
engaged in a hermeneutic enterprise and its effective and affective promise of
gaining access to mental states other than our own through reading someone else's
work and reflecting on what was read is a powerful one. The intelligentsia combined
the oral as well as the literary that had developed side by side with the oral, but
never really replaced it. This is of special significance to the social and intellectual
historian of the subcontinent.
Regardless of whether one prefers to call this phenomenon "vajrayāna
scholasticism" or "institutional esotericism," in a Buddhist context it by and large
entailed the appropriation of the esoteric by the intellectual elite of the monastery.
That is to say, in the Indian subcontinent, the writing of scholarly treatises having
to do with mantra or guhyamantra in the widest sense of these words, thus giving rise
to the new literary genre of the *[guhya]mantraśāstra ([gsang] sngags kyi bstan bcos), was
to all appearances generally but not exclusively the province of the educated
monastic. The onomasticon of the vast majority of authors of treatises on esoterica
suggests that this was the case. However, the huge corpus of such writings that is
available for inspection also suggests that not a few laymen were equally in the
position to write such works, meaning, of course, that they must have had access to
decently stocked library facilities. The early tenth-century monk Bhavyakīrti uses
the expression *mantraśāstra once to describe a generic study of any of the five
classes of tantric literature.7 The welding together of the exoteric with the esoteric
did play fundamental roles in the composition of, for example, Śāntarakṣita's
*Tattvasiddhi or Trivikrama's (10thc.) Nayatrayapradīpa or Jñānākara's (11thc)
For these transitions, see in particular Moody (1987: 110-122) and Moody (2000); for a Tibetan
gloss on these and related questions, see van Schaik (2007).
7 SKAL, 394/2 [Ki, 21]; SKAL1, 565; the five are bya ba-krīya-, spyod pa-caryā-, rnal vbyor-yoga-, rnal
vbyor gyi bla ma-*yogottara-, and rnal vbyor ma-yogiṇī-tantra. For him and his oeuvre, see below.
6
Journal of Tibetan and Himalayan Studies
5
*Mantrāvatāra. All three works are excellent examples of what a *mantraśāstra might
look like.8
In Tibet, matters stood rather differently and this has everything to do with
the way in which Buddhism and its institutions had insinuated themselves into the
Tibetan social and intellectual landscape whereby it transformed its societal and
religious norms to highly significant degrees.9 From almost its inception, we find
the clergy and the laity in equal measure taking an interest in and becoming
virtuosos of its theoretical and practical aspects. The social history of Indian
Buddhist and non-Buddhist philosophy is bereft of detailed, large-scale studies of
the lives of its exponents, because of the virtual absence of sustained biographical
or autobiographical studies of their main players. True, we do have a few snippets
about the lives of the circa 800 Buddha[śrī]jñāna and [a] Śākyamitra, but these
remain thin and partly obscure.10 How different it is with the development of
Buddhism in Tibet that began in earnest in the late eighth century! Beginning in the
ninth and the tenth centuries, the rise of biographical and autobiographical writing
in the Tibetan area – examples are the Dbav/Sba bzhed and Gnubs Sangs rgyas ye
shes' (ca. 900) autobiography11 - shows that monks as well as laymen were involved
with this movement, and also demonstrates how surprisingly smoothly late Indian
Buddhism was able to transition to and continue in the Tibetan region. For
example, such monks as Lo tsā ba Rin chen bzang po (958-1055) and vGos Lo tsā
ba Khug pa Lhas btsas [?Bsod nams rtse mo] (11thc.), and such laymen as Dpyal
vByung gnas rgyal mtshan (11thc.), Sa chen Kun dgav snying po (1092-1158) and
his son Rje btsun Grags pa rgyal mtshan (1148-1216), to name but a few examples,
were great Tibetan tantric ritual masters of the early period and at the same time
also great commentators and virtuosi of this arcane literature and its associated
liturgcal and meditative practices and, what is more, the latter three were even
abbots of their monastic institutions!
Contrary to much of the later tantric lore, the meanings of tantric texts were,
in theory, not to be disclosed by those who had not been initiated in their practices.
This caveat may be considered prescriptive rather than what actually happened on
the ground, and whether this was really the case for every single interpreter of a
8 See, respectively, Steinkellner (2008) and SDE, vol. 32, no. 3712 [# 3707], 101/3-107/3 [Tsa, 5b26b], Wenta (2018) and Kano and Li (2019) - but see also, albeit very briefy, Ōhmi (2008).
9 For some remarks, see van der Kuijp (forthcoming a).
10 See Davidson (2002: 309 ff.) and, more recently, the corrective statements in Szántó (2015:
540-542)
11 For the Dbav/Sba bzhed, see van der Kuijp (2013: 133-136 nos. 44-45), and for Gnubs'
autobiography, see Esler (2014) and Dalton (2014).
6
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
tantric text is something that is still in need of evidence. Always perceptive and
sensitive to the issues at hand, Davidson has persuasively written about the idea of
secrecy in Indian tantric Buddhism writing at one point that12:
Tantric texts are discussed, haggled over, bought and sold, copied, traded and
challenged. They were hidden, revealed, forged and mythologized. In many of these
textualized discourses, the author is invariably willing, even eager, to reveal the
"opinion of the gurus" on such esoteric matters, so that esoteric language is always
given a lexicon, even while individual yogins are on record as doubting the
existence or utility of secret languages at all.
Of course, the literature bears witness to plenty exceptions to the willingness
of authors to reveal the "opinion of the gurus", and, indeed, there are not a few,
who in their commentaries, actually refer the reader to a or the "guru" for further
instruction and interpretation.
What is more, in a good number of their writings on esoterica, these Indian
Buddhist doxographers bring to bear large corpora of quite heterogeneous
Buddhist literature on the texts on which they are commenting. These
hermeneutical methods gave rise to surprising doxological maneuvers and thereby
often provide us with unexpected, and sometimes quite counterintuitive, if not
misleading, insights into Indian intellectual rather than just religious history. With
the rise of tantric thought in the subcontinent, it was not long before we witness
the greatest minds of Indian Buddhism combining a profound knowledge of the
theory and practice of tantra with a keen intellectual interest in philosophical
problems addressed in the more normative texts.13 It is in this latter context that
the present paper needs to be placed. This would also appear to lie at the base of
Dol po pa Shes rab rgyal mtshan's (1292-1361) statement in the autocommentary
of his 1350s Bkav bsdu bzhi pavi don, The Intent of the Fourth Council, his last major
work; there he writes that:14
12 Davidson (2006: 62); see also the finely tuned comments in Arènes (1998) and Davidson (2002:
257-269), where the question of secrecy in tantric Buddhism is also addressed with nuance and
insight.
13 The recent astonishing work of Eltschinger (2013) and its dialogue with Davidson (2002) and
Sanderson (2009) is quite indispensable for understanding the intellectual landscape of the
subcontinent and its various transitions and transformations that took place from the 5th to the
7th century.
14 See his Bkav bsdus bzhi pavi don bstan rtsis chen povi vgrel pa, Gsung vbum, vol. 7/13, ed. Dpal brtsegs
bod yig dpe rnying zhib vjug khang, Mes povi shul bzhag, vol. 202 (Beijing: Krung govi bod rig
dpe skrun kang, 2011), 77; it is quoted [from a different manuscript of the text] in Stearns (2010:
268); for the problems associated with its dating, see van der Kuijp (2016: 137 ff.).
Journal of Tibetan and Himalayan Studies
7
rgyud ni rgyud gzhan gyis shes bya //
mdo ni mdo gzhan gyis shes bya //
mdo yang rgyud kyis shes par bya //
rgyud kyang mdo yis shes par bya //
nyis gas nyi ga shes par bya //
Tantras should be understood by means of other tantras.
Sutras should be understood by means of other sutras.
Sutras, too, should be understood by means of the tantras.
Tantras, too, should be understood by means of the sutras.
Both should be understood by means of both.
To be sure, the ahistorical tenor of this and similar statements borne from an
evident reluctance to historicise what had gestated into a full-blown corpus of
canonical texts15 that, as a result of their inclusion in a canon [for the Tibetan and
not the Indian tradition!!], no longer quite competed with one another in the market
place of ideas and ritual practice as they might have done in the Indian
subcontinent, has enormous implications for the interpretive horizon of any
Tibetan Buddhist treatise.
Flourishing most probably in the tenth century, a monk named Skal ldan grags
pa [Bhavyakīrti] or Skal ldan [*Bhavya] - he used both names to refer to himself
and I will refer to him as Bhavyakīrti - was a member of a growing group of those
late Indian Buddhist authors who combined a reasonably solid scholarly knowledge
of exoteric literature that we usually, but no doubt not always correctly, associate
with the potentially academic life of a monastic with what at least appears to be a
serious commitment to the theory, if not the practice, of tantra. We will see below
that he has been considered to be the author of several scholastic works on the
latter. While it is true that virtually every detail surrounding the genesis of the
onomasticon of Indian Buddhism has yet to be looked into, it should hardly come
15 In a late Indian context, it is probably better to call these texts as having been cherished and
given a special authenticity by the tradition, since there is no evidence of a Mahayana Buddhist
"canon", the likes of which we have in Tibet and East Asia.
8
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
as a surprise that some circles designated someone altogether different by the same
"Bhavyakīrti." Questions such as why a Buddhist was called "Nāgārjuna" or
"Dignāga," were these proper names or nicknames, or were they even monks, when
bona fide names in religion like "Dharmakīrti" or "Candrakīrti" and "Dharmamitra"
or "Śākyamitra" became part of the Indian Buddhist onomasticon, monastic or
otherwise, and do the suffixes -kīrti and -mitra indicate monastic kinship, still beg to
be answered. Presumably, then, Bhavya was not an altogether unusual name. The
editor or scribe of a manuscript of Pa tshab Lo tsā ba Nyi ma grags' (11th-12th c.)
study of the difficult points in Candrakīrti I's (7thc.) Prasannapadā, or perhaps very
possibly even Pa tshab Lo tsā ba himself, attributes the well known circa sixthcentury Prajñāpradīpa commentary on Nāgārjuna I's (2nd/3rd c.)16 Mūlamadhyamakakārikā to [a] "Bha phya kir ti" and "vBa phya kir ti," that is, therefore quite clearly,
to Bhavyakīrti.17 We do not know to whom we can assign the responsibility for this.
The extant witnesses of the translation of the Prajñāpradīpa have Legs ldan, that is,
*Bhavya as its author's name for which, in its Tibetan translation, there exists
written testimony from the eighth century onward. Of course, his name is now
generally considered to have been Bhāviveka.18 It is interesting to note here that a
manuscript of the commentary on the Prasannapadā by Pa tshab's student Thang sag
pa Ye shes vbyung gnas has both "Legs ldan byed" and "Legs ldan", that is,
"*Bhavyakara" and "*Bhavya."19 To all appearances, then, "Bhayvakīrti" as it occurs
in the manuscript of Pa tshab's text seems very fortuitous and is a rather rare
occurrence.
A good portion of what follows is devoted to Bhavyakīrti's exposition and
rejection of what he, a member of the Buddhist monastic intelligentia interested in
See the thoughtful essays of Mabbett (1998) and Walser (2002).
*Tshig gsal bavi dkav ba bshad pa, Bkav gdams gsung vbum phyogs bsgrigs, vol. 11, ed. Dpal brtsegs bod
yig dpe rnying zhib vjug khang (Chengdu: Si khron dpe skrun tshogs pa / Si khron mi rigs dpe
skrun khang, 2006), 34. Pa tshab states that he wrote this work on the basis of the instructions he
had received from his teacher, who was a scholar-trader (bla ma tshong dpon pan di ta). He may of
course most likely have to be identified as Kanakavarman, with whom he corrected his earlier
translations of Candrakīrti I's Prasannapadā, Madhyamakāvatāra, and Madhyamakāvatārabhāṣya in
Lhasa's Ra mo che temple-cum-monastery. See also the judicious remarks in Yoshimizu (2014:
183-5). For Candrakīrti I versus the later Candrakīrti II, see the lengthy discussion in Vose (2007:
27-36) and also below.
18 For the various names that we encounter for him in the literature, see Ejima (1990) and the
references in Krasser (2012: 535, n. 1). Not mentioned by these authors is that the name
"Bhāviveka" ("Snang bral") occurs in Kamalaśīla's (8thc.) pañjikā-commentary on his master
Śāntarakṣita's Tattvasaṃgraha; Jha (1986: 153). There is also "Bhāvivikta" ("Legs ldan") as in Jha
(1986: 155).
19 See, for example, Dbu ma tshig gsal gyi ti ka, 6b, 7a; for this work, see Yoshimizu (2006),
Yoshimizu-Hiroshi (2013), and now also (2018).
16
17
Journal of Tibetan and Himalayan Studies
9
the theory of tantra, considered to be the basic features of the philosophical
standpoint of the non-Buddhist Sāṃkhya tradition. Discussed and edited below,
the relevant, fairly long passage is found in his *[Pradīpoddyotanavyākhyāṭīkā]
sandhyāprakāśikā, a large and, as we shall see, a somewhat neglected subcommentary
of Candrakīrti II's (ca. 900) highly influential Pradīpoddyotana exegesis of the circa
seventh-century Guhyasamājatantra in its seventeen-chapter recension.20 The printed
Kanjur-canons of this tantra contain two different Tibetan translations. On one
hand, the Yongle, Li thang, Beijing, Sde dge, Co ne, and Ku re xylographs contain
Ravendra's and Chag Lo tsā ba Chos rje dpal's (1197-1264) revision of the earlier
translation by Śraddhākaravarman and Lo tsā ba Rin chen bzang po,21 whereas, on
the other, the Snar thang and Lhasa Zhol xylographs contain vGos Lo tsā ba Khug
pa Lhas btsas' revision of Śraddhākaravarman's and Lo tsā ba Rin chen bzang po's
translation based on explanations he had received from a certain Śrījñānakara.22 An
anonymous gloss at the end stresses the importance of vGos Lo tsā ba's revised
20 The best edition of the Sanskrit text of the Guhyasamājatantra is of course found in Matsunaga
(1978). For the first published translation of this tantra in a Western language, see Gäng (1988);
the edition and translation in Fremantle (1971) remains unpublished. A Tibetan text of the tantra
is found in SDE, vol. 25, no. 1787 [#1785], 266/2-323/3 [Ga, 1b-201/b], and an edited version
based on eight different xylograph editions is available in Bkav vgyur [dpe bsdur ma], ed. Krung govi
bod rig pa zhib vjug lte gnas kyi bkav bstan dpe sdur khang, vol. 81 (Beijing: Krung govi bod rig
pa dpe skrun khang, 2008), nos. 0464-5, 289-583; no. 0466, 584-611, contains its Rgyud phyi ma,
that is, the Uttaratantra, which is also known as the tantra's eighteenth chapter, in the translation
of Śraddhākaravarman and Lo tsā ba Rin chen bzang po (958-1055). As for the latter, the Li
thang xylograph states that: "The edition by Thang chen pa and Rgyal mtshan reng [read: ring]
mo, who compared it with the text of the "collected tantras" of Stag lung monastery, is done"
(stag lung rgyud vbum la bsdun [read: bstun] pavi thang chen pa dang rgyal mtshang reng [read: ring] mos zhu
dag grub). I am unable to identify these two men. Matsunaga's edition covers eighteen and
Fremantle's seventeen chapters. A single Sanskrit manuscript of Candrakīrti II's commentary was
edited by C. Chakravarti, who points out, in Chakravarti (1984: 1), that its readings of the basic
text of the tantra are "not close to the version" in the pioneering edition of Bhattacharyya (1931).
It is unfortunate that Chakravarti was unaware of Matsunaga's and Fremantle's earlier work, both
of which built on Bhattacharyya's edition. And it is equally regrettable that his "edition" has no
critical apparatus. An edition of the Tibetan translations, yes plural, of this work is of course still
outstanding. A preliminary but so far the most extensive listing of the various Tibetan "editions"
is found in Bentor (2010: 99-100, n. 8). Lastly, Tanemura (2015) provides a very useful survey of
this work, its contents, and its related scriptures.
21 See Bkav vgyur [dpe bsdur ma], ed. Krung govi bod rig pa zhib vjug lte gnas kyi bkav bstan dpe
sdur khang, vol. 81, no. 0464, 289-425, text-critical notes 426-441. The Co ne and Li thang
mention that a Thang chen pa and a Rgyal mtshan ring mo edited the text (zhu chen bgyi).
22 Bkav vgyur [dpe bsdur ma], ed. Krung govi bod rig pa zhib vjug lte gnas kyi bkav bstan dpe sdur
khang, vol. 81, no. 0465, 442-578, text-critical notes 579-583.
10
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
translation when reading the text together with his commentary, and that this was
the reason for its inclusion in these editions23:
vdus pavi rtsa rgyud la lo chen gyis bsgyur ba mang du snang navang / vgos kyis vgyur bcos
mdzad pa nyung bar snang la / vgrel pavi bshad pa vgos vgyur gyi steng nas byed pa la rgyud kyi
vgos vgyur gal che bar mthong nas bris pavo //
In terms of its impact on later generations in the subcontinent and Tibet, it is
undeniably true that Candrakīrti II's study exerted an influence that was far greater
in magnitude than the other exegeses of this tantra such as, for instance, the one by
an Ānandagarbha (9th c.) or a Nāgārjuna (ca. 900-?) as well as the most probably
earlier one by Rdo rje bzhad pa [*Vajrahāsa ?= Bzhad pavi rdo rje/*Hāsavajra],
among other studies.24 In connection with *Vajrahāsa's study, there is a widely
accepted Tibetan tradition maintaining that the translations of his work and that of
commentary by Viśvamitra were more or less done contemporaneously. This
widely accepted tradition, then, holds that Viśvamitra's cognate work was translated
most probably late into the second half of the eighth century during the reign of
Khri srong lde btsan (ca. 742-ca. 800). Consider, for example, the passage we
encounter in Bu ston Rin chen grub's (1290-1364) 1348 study of the Guhyasamāja
literature25:
Bkav vgyur [dpe bsdur ma], ed. Krung govi bod rig pa zhib vjug lte gnas kyi bkav bstan dpe sdur
khang, vol. 81, no. 0465, 578. I do not know when this work was first included in a manuscript
Kanjur.
24 For the first, see SDE, vol. 26, no. 1846 [#1844], 264/1-294/7 [Ji, 53b-161b] and for the second,
see SDE, vol. 26, no. 1846 [#1844], 264/1-294/7 [Ji, 53b-161b], and my note in van der Kuijp
(2014). For the third, see SDE, vol. 27, no. 1911 [#1909], 223/5-238/2 [Phi, 38a-89a]. There
*Vajrahāsa cites Nāgārjuna, the [Tattva]saṃgraha[sūtra/tantra], and twice a lung (*āgama), that is, the
Guhyagarbhatantra, on, respectively, 227/6, 228/5, and 228/5, 7 [Phi, 52b, 55b, 56b]. The
quotation from an unidentified work by this Nāgārjuna reads dngos po stong zhing gzhan ma yin // de
ma rtogs pavi dngos kyang med //, and is thus quite similar to the canonical translation of Acintyastava,
43: dngos po rnams las stong gzhan min // de med par yang dngos po med // [bhāvebhyaḥ śūnyatā nānyā na ca
bhāvo vsti tāṃ vinā /], for which see Lindtner (1987: 154).
25 See Dpal gsang ba vdus pavi rgyud vgrel gyi bshad thabs kyi yan lag gsang bavi sgo vbyed, The Collected
Works of Bu ston [and Sgra tshad pa] [Lhasa xylograph], part 9 (New Delhi: International Academy
of Indian Culture, 1967), 76-77; see also the undated study that is attributed to Bo dong Paṇ chen
Phyogs las rnam rgyal (1375-1451), the Gsang ba vdus pavi lung rigs man ngag ston par byed pavi bla ma
tshad mavi lo rgyus, Encyclopedia Tibetica. The Collected Works of Bo dong Paṇ chen Phyogs las rnam rgyal,
vol. 64 (New Delhi: The Tibet House, 1972), 425, with in part the very same phraseology; the
section on its development in Tibet in A mes zhabs Ngag dbang kun dgav bsod namsv (15971659) 1634 chronicle, his Dpal gsang ba vdus pavi dam pavi chos vbyung bavi tshul legs par bshad pa gsang
23
Journal of Tibetan and Himalayan Studies
11
...khri srong lde btsan gyi dus su / dpal gsang ba vdus pa rtsa bavi rgyud / rgyud phyi ma
dang bcas pa / slob dpon rdo rje bzhad pas mdzad pavi vgrel pa / bi shva mi tras mdzad pavi
vgrel pa bcas bsgyur ba las / phyis pho brang von cang rdovi dus su gsang sngags la bsnyad btags
nas spyod pa vchol ba mthong bas / gsang sngags bsgyur ba bkag go //
...aside from the fact that the basic Guhyasamājatantra together with its
*Uttaratantra, the commentary written by Master Rdo rje bzhad pa [*Vajrahāsa],
together with the commentary written by Viśvamitra, were translated at the time of
Khri srong lde btsan, later, the translation of secret spells [= tantric literature] was
prohibited, because, at the time of Pho brang von cang rdo,26 immodest behavior,
that was attributed to the secret spells, was observed.
The last portion of Bu ston's text that notes a prohibition of sorts presents us
with a more radical restriction on the translation of esoteric tantric literature during
the era of Khri gtsug lde btsan than the one we encounter in a passage of the
introductory portion in the available recensions of the Sgra sbyor bam po gnyis pa,
Grammatical Formulations in Two Bam po Fascicles.27 The earliest of these, dating from
783/795, and the later one dating from 812, each has a passage in which it was
decreed that one must first receive permission from the court if one wishes to
translate tantric texts. The manuscript found in Ta pho/bo monastery, which
appears to represent the oldest of these recensions, is at this point unfortunately
fragmentary and closes with the phrase "further, ever...?not permitted" (yang gtan la
vdus chos kun gsal bavi nyin byed, Collected Works, ed. Si khron bod yig dpe rnying myur skyob vtshol
sgrig khang, vol. 18 (Lhasa: Bod ljongs dpe rnying dpe skrun khang, 2012), 92 ff., has nothing of
the kind.
26 This appears to be another designation for Khri gtsug lde btsan (r. ca. 815-841), alias Ral pa
can, who is especially associated with this palace as he was born there, and who built a temple
there with this same name. The term pho brang with its various meanings was analyzed in
Denwood (1990).
27 See Scherrer-Schaub (2002: 287, 322-323), who juxtaposes the readings of the canonical text,
the manuscript fragment from Dunhuang, and the Ta pho/bo manuscript. See also the relevant
passage of the handwritten manuscript of the Sgra sbyor bam po gnyis pa in Dkar chag vphang thang ma
[and the] Sgra sbyor bam po gnyis pa, comp. Bod ljongs rten rdzas bshams mdzod khang (Beijing: Mi
rigs dpe skrun khang, 2003), 73. It would appear that Bu ston had this passage at hand or in mind
when he wrote his 1339 studies of tantric literature, for which see the Rgyud sde spyivi rnam par
gzhag pa rgyud sde rin po chevi mdzes rgyan, The Collected Works of Bu ston [and Sgra tshad pa] [Lhasa
xylograph], part 15 (New Delhi: International Academy of Indian Culture, 1969), 127, and the
Rgyud sde spyivi rnam par gzhag pa rgyud thams cad kyi gsang ba gsal bar byed pa, The Collected Works of Bu
ston [and Sgra tshad pa] [Lhasa xylograph], part 15 (New Delhi: International Academy of Indian
Culture, 1969), 733-734.
12
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
[...] [m]y[i] gnang / ngo /). But the text of the Dunhuang manuscript and the other
witnesses have here28:
phyin chad gzungs sngags dang rgyud bla [variant: blad] nas29 bkav stsal te / sgyur du
bcug pa ma gtogs pa / sngags kyi rgyud dang / sngags kyi tshig thu zhing bsgyur du mi gnang
ngo //
...henceforth the order was given in connection with spells and tantras:
Excepting what was prevailed upon to be translated, one is not permitted to
translate the tantric spell-texts and magic formulae of spells!
C. Scherrer-Schaub drew our attention to a Duhuang manuscript, IO 752, in
which a prohibition against "violent [ritual] activities" (drag povi las) is registered.30
In this connection, there is also an interesting passage at the beginning of Rje
btsun's survey of the Sarvadurgatipariśodhanatantra that is worth quoting; there we
read the following31:
rgyal po khri srong ldevu btsan gyi sku ring la / dbas manydzu shrī la sogs pasad mi mi
bdun gyis bsgyur zhing / sras khri lde dang / dbon po ral pa can gyi sku ring la / rma ā tsarya
rin chen mchog la sogs pas skad gsar bcad kyis bcos te gtan la phab nas bsgyur ba las / phyis lo
tsā ba gzhan gyis bsgyur ba ni med do // vdi la vkhor los sgyur ba dang me ltar vbar bavi skabs
kyi / drag povi sbyin sreg med pavang / sngon gyi rgyal blon rnams kyis sngags pa rnams kyis
mngon spyod byed du dogs nas / ma sgyur cig byas nas ma bsgyur ba yin no zhes zer ba yang /
vdi la phyis kyi lo tsā ba rnams kyis gzhan tsho vgyur dag kyang / chad pa dag vdzud par rigs
pa las / ma bcug pavi phyir rgya dpe nyid la med pa yin nam snyam du yang sems so // kha cig
For this passage, see also Simonsson (1957: 260-261) and Scherrer-Schaub (2002: 322-323).
The first text of Bu ston quoted above in n. 27 has blad nas, while the second has the
homophonous slad nas. Rnam rgyal tshe ring, Bod yig brda rnying tshig mdzod (Beijing: Krung govi
bod rig pa dpe skrun khang, 2001), 583, suggests that slad nas is an archaism for rjes nas and ched
mngag nas. Indeed, the expression blad nas does not occur in any of the dictionaries that are
available to me.
30 Scherrer-Schaub (2001: 702, 704). These rituals that induce violence are especially but not
solely associated with the abhicāra (mngon spyod) burnt offering ritual, which is one of the four
homa/sbyin sreg offering rituals.
31 See his Ngan song sbyong rgyud kyi spyi don, Sa skya pavi bkav vbum, comp. Bsod nams rgya mtsho,
vol. 4 (Tokyo: The Toyo Bunko, 1968), no. 96, 105/2-3.
28
29
Journal of Tibetan and Himalayan Studies
13
ni sngon gyi dus su li yul na chos grwa dar / livi dpe la drag povi sbyin sreg [3] yod do zhes
kyang zer ro //
[The Sarvadurgatipariśodhanatantra] was translated in the era of king Khri srong
ldevu btsan by the seven selected ones such as Dbas Mañjuśrī, etc.32 and Rma
ācārya Rin chen mchog, etc. corrected and edited the translation by means of the
new standardized terminology in the era of his son Khri lde and his grandson Ral
pa can33; apart from this translation, a translation by another, later translator does
not exist. It is alleged in this connection that the violent burnt-offering ritual (drag
povi sbyin sreg) of the sections on the universal emperor and what is blazing like fire34
were not translated due to the king and ministers, being afraid of thaumaturges
doing fierce magic (mngon spyod, *abhicāra), having said: "Do not translate these!"
However, while, in this connection, later translators had corrected other portions,
apart from the fact that it is reasonable to insert elided passages, because these had
not been inserted, I wonder whether these were absent in the Sanskrit/Indic
manuscript itself. Some have also alleged that in times of yore, religious college[s]
(chos grwa) had spread in Khotan (li yul). The passage on the violent homa-ritual was
present in the Khotanese manuscript.
Tibetan connections with Khotan are of course not far fetched and it is well
known that they were rather extensive. What is not so well known is that at least
one family in which two prominent scholars were born appears to have traced the
origins of its patriline to this Inner Asian oasis state. The capsule biographies of
Mkhas grub Dge legs dpal bzang po (1385-1438) and his younger brother Ba so
Chos kyi rgyal mtshan (1402-1473) observe that their father's family originally
hailed from Li yul Lcang ra smug po, that is, so it would appear, from a place called
Lcang ra smug po in Khotan.35
The earliest catalogs of translated scripture that are currently available, albeit
not in their original state of the early ninth century during which time they were
compiled, but in much later iterations, that is, the Lhan [or: Ldan] dkar ma and
vPhang thang ma catalogues, do not list a translation of the Guhyasamājatantra or any
For various listings of these "seven," itself an artificial and arbitrary number, see van der Kuijp
(2013: 148 ff.).
33 This passage suggests that Rje btsun was unaware that the decision to use a new terminology
(skad gsar bcad) for translations from Sanskrit [?and Khotanese] had perhaps already been made at
the end of Khri srong lde btsan's reign.
34 For these passages, see Skorupski (1983: 81 ff., 71-72).
35 On the toponym li yul lcang ra smug po, see briefly van Schaik (2016: 45-46).
32
14
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
associated commentary.36 On the other hand, the list of the titles of translated
literature by Dar ma rgyal mtshan (1227-1305), alias Bcom ldan rig[s] pavi ral gri,
does register a Tibetan translation of the Guhyasamājatantra, albeit without giving
the names of the translators, for the period of the "early propagation of Buddhism"
(snga dar) that, in his opinion, extended from roughly the eras of Khri srong lde
btsan to that of Khri gtsug lde btsan.37 Every single printed Kanjur text of the tantra
- this collection belongs to those translations of texts that were sanctioned by the
New Tradition (gsar ma) which had its inception in the first half of the eleventh
century - contained the translations that were mentioned earlier,38 and the same
holds for the manuscripts of the Shel dkar and Stog Palace Kanjurs. On the other
hand, the translations of the Guhyasamājatantra and the Guhyasamājottaratantra that
are referred to or that are actually found in the various editions of the Rnying ma[vi]
rgyud vbum, Collected Tantras of the Old Tradition, suggest that they were in fact first
translated around the year 800, even if they do not appear in the earliest catalogues!
What is more, a translation of the text was also part of the cache of Tibetan texts
that was discovered in the vicinity of Dunhuang, which predates the early eleventh
century. 39 Needless to say, a workable "diachronic" edition of the Tibetan
recensions of the Guhyasamājatantra in which all the newly available resources are
marshalled is truly a desideratum.
Bu ston's mention of Viśvamitra's commentary on the Guhyasamājottara-tantra
or the eighteenth chapter of Guhyasamājatantra - the exact position of this tract vizà-viz the Guhyasamājatantra as such, it essentially consists of a series of questionsand-answers in connection with the earlier seventeen chapters, still requires further
investigation – gives rise, on inspection, to a slight conundrum. When we look a
little more closely into this work,40 we begin to notice certain indications that it may
be of a somewhat later date than Bu ston and no doubt others before him had
supposed. In addition, one gets the suspicion that it either contains obvious
interpolations or that an Indian commentator did not write it at all. Let us first
examine his sources. The author twice mentions unnamed sutras, and he refers
many times to tantric texts that he also leaves unidentified. However, what is of
considerable importance is that he does expressly refer to a number of tantric
literary sources by title. And these include the Gdan bzhi [Catuṣpīthatantra] - P.-D.
Szántó was the first to study this work in consummate detail and tentatively
See, respectively, Lalou (1953), Dkar chag vphang thang ma [and the] Sgra sbyor bam po gnyis pa, 1-67.
Schaeffer and van der Kuijp (2009: 181).
38 See above nos. 20-22.
39 See Martin (1987: 183-184) and Mayer (2004: 130-131, n. 4).
40 SDE, vol. 26, no. 1846 [#1844], 264/1-294/ [Ji, 53b-161b].
36
37
Journal of Tibetan and Himalayan Studies
15
identified it as if not the earliest then at least one of the earliest yogiṇītantras41 -, the
Sampuṭodbhavatantra, 42 a vyākhyā-explanatory tantra of the Herukābhidhāna or
Laghuśaṃvaratantra, and the Hevajratantra, both of which are yogiṇītantras, and the
Guhyagarbhatantra, a so-called great-yogatantra (mahāyogatantra). The Guhyagarbhatantra stems from the eighth century and its dating is not in the least controversial.
Sanderson has suggested that it is likely that the Herukābhidhāna, which he holds to
be the earliest yogiṇītantra, dates from the ninth century.43 Since the Hevajratantra is
certainly later than the Herukābhidhāna, the Sampuṭodbhavatantra must therefore be
even later. Thus, we now find ourselves perhaps in the tenth century or even sightly
later. If Viśvamitra flourished in the eighth century, then these Indic sources must
have existed in one form or another during the middle of the eighth century, and
Sanderson's suggestion needs to be rephrased if not discarded. But we know
nothing about Viśvamitra and the question of this work's authorship is of course by
no means established by the mere fact that it says in the single line at the very end
of the Tibetan translation that he was its author. However, for the sake of
convenience, I continue to call the author of this text Viśvamitra. Such an excellent
connoisseur of tantric literature as Bu ston certainly did not see any reason to raise
an eyebrow about its authorship or relative chronology, neither in his 1335
catalogue of the Zhwa lu Tanjur manuscript, nor in his later assessments of the
canonical tantric literature as a whole. While the sources available to me do not
signal any problems with its provenance, I nonetheless think that the integrity of
this work as a translation of a purely Indic text is not entirely beyond question. This,
I submit, is borne out by the fact that it contains references to "some
Indian/Sanskrit manuscripts" (rgya gar gyi yi ge la la) and the "Indian language" (rgya
gar gyi sgra), that is, Sanskrit. The first occurs in conjunction with his interpretation
of Guhyasamājottaratantra, 141a, 44 and the second in connection with
Guhyasamājottaratantra, 190d.45 Surely, these references to Indian manuscripts and
Sanskrit rather than simply to other manuscripts in what is ostensibly a Sanskrit
text must strike one as strange. And this makes little sense, unless of course,
Viśvamitra or the author of these two passages addressed an audience other than
an Indian one, namely, one that was not conversant with Sanskrit. One might
therefore reasonably argue that this work is at a minimum what Davidson has so
aptly called a "gray text," where gray texts are texts that "...do not accord with the
Szántó (2008: 3); see also Szántó (2012).
vol. 26, no. 1846 [#1844], 266/6, 277/6 [Ji, 63a, 101b]. The quotation from the
Sampuṭa[udbhavatantra] reads: kha yi thad dkar gdud par bya //.
43 Sanderson (2009: 164).
44 See Matsunaga (1978: 123) and SDE, vol. 26, no. 1846 [#1844], 286/7 [Ji, 133b].
45 See Matsunaga (1978: 127) and SDE, vol. 26, no. 1846 [#1844], 291/4 [Ji, 149b].
41
42 SDE,
16
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
prototypes of works wholly Indian or wholly Tibetan."46 This idea is a very good
beginning, although what still needs to be fully worked out are precisely those
specific features, aside from obvious ones such as internal references, etc., on
which basis we can more or less decisively judge a work to be prototypically Indic
or Tibetan.47
There is of course much food for thought here, but we ought to return to
the topic at hand, that is, Bhavyakīrti and his *Sandhyāprakāśikā, before we stray too
far afield! For starters, then, we have to be frank and acknowledge that the Tibetan
tradition as a whole, albeit with some exceptions, 48 has paid scant attention to
Bhavyakīrti and almost the same holds for the recent secondary literature on
Guhyasamāja practices. Why this should be so is not altogether clear to me, but it
does make me feel a tad uneasy, especially because I have invested off and on a
considerable amount of time in him and his oeuvre! Among the first [in English] to
have used the Tibetan translation of his work at some length, A. Wayman
memorably called it "a highly opinionated commentary" – I am not exactly sure
what he may have meant with this statement -, but I believe it was M. Broido who
actually made extensive use of his work and gave us the first indication of its
potential value.49 And sometime in 2004, I began to read him for no particular
reason. The Ālokamālā by Prajñāmitra (ca. 9th c.), alias Kambala, had already peaked
my attention ever since I had read Lindtner's well-known edition and translation,
and I soon noticed to my delight that Bhavyakīrti had quoted, with and without
attribution, a very large portion of this work in his commentary. I thus set out to
collect the numerous citations and hoped to discuss these in the foreseeable future.
Then, in August of 2005, Tomabechi Tōru told me in conversation in Vienna that
he had also worked extensively on Bhavyakīrti's work while he was a student at
Kyoto University. He never published any of his findings, but he did speak on him
Davidson (2002a: 212).
This problematic was also not lost on the better Tibetan scholars. For example, in particular
but not exclusively in connection with the Guhyasamāja literature, Bu ston sometimes judges a
work to be "Indic" (rgya gar ma) or "Tibetan" (bod ma), albeit without theorizing how he had
arrived at these judgements; see here also the valuable remarks in Wedemeyer (2009) and (2014).
Such adjudications began in the eleventh century with the literature on "false tantric [works]"
(sngags log), for which see Martin (2001: 111 ff.).
48 For example, whereas Tsong kha pa Blo bzang grags pa (1357-1419) cites him numerous times
in his large commentary on the Paṇcakrama of a Nāgārjuna - see Kilty (2013: 631, index) –, Kun
dgav snying po (1575-1634), alias Tāranātha, does not once cite him once in his equally large
commentary on the same, for which see his Rim lngavi vgrel chen rdo rje vchang chen povi dgongs pa,
Gsung vbum dpe bsdur ma, ed. Dpal brtsegs bod yig dpe rnying zhib vjug khang, vol. 10/45 (Beijing:
Krung govi bod rig pa dpe skrun khang, 2008).
49 See, respectively, Wayman (1977: 103) and Broido (1988: 95-97).
46
47
Journal of Tibetan and Himalayan Studies
17
and his work at the 2014 conference of the International Association of Buddhist
Studies in Vienna, which I did not attend. Happily, he recently published some of
his significant findings in an important article, therewith giving Bhavyakīrti a welldeserved place on the map of Indo-Tibetan studies.50
Interestingly, Matsunaga's and Khang dkar's works on Buddhist tantra in India
barely register him, while Davidson does not even once mention his name, let alone
discuss his oeuvre, in his large and rewarding study of the development of Indian
Buddhist tantra.51 Virtually the same holds for Sanderson's more recent magisterial
study of the genesis of the tantric traditions in India among Saivite and Buddhist
communities, although he does briefly note him in connection with his study of the
shorter Herukābhidhānatantra, the proto Laghuśaṃvaratantra/ Cakraśaṃvaratantra.52
But matters are quite different with his student Szántó, who has written a rather
extensive note on him or, rather them, for he distinguishes between two different
Bhavyakīrti-s.53 In doing so, he put forth the idea that the first was the author of
the *[Śrī]Vīramanoramā, a commentary on the Herukābhidhānatantra,54 and that the
other was the author of the *Sandhyāprakāśikā, a commentary on the difficult points
(pañjikā, dkav gnas) of the well known Pradīpoddyotana, Candrakīrti II's exegesis of the
Guhyasamājatantra in seventeen chapters. But more about this below.
That notwithstanding, authors of histories of the Guhyasamāja cycles such as
Bu ston, the one attributed to Bo dong Paṇ chen, and A mes zhabs barely mention
his name55; the same holds for the [Cakra]śaṃvara histories of Bu ston, Go rams pa
Bsod nams seng ge (1429-1489), and A mes zhabs. 56 The only reasonable
50 Tomabechi (2016). In the meantime, Matsuda Kazunobo published an edition of
Ratnākaraśānti's (ca. 1000) Prajñāpāramitābhāvanākrama in which a number of the Ālokamālā's
verses are cited; see Matsuda (2018: 29-31).
51 See, respectively, Matsunaga (1980: 233-315), Khang dkar (1994: 287-330), and Davidson
(2002).
52 Sanderson (2009: 158-9, n. 363-4).
53 Szántó (2012: 42-8); his remarks were in part excerpted in Del Toso (2014: 515-6).
54 Szántó (2010) discusses the arguments Bhavyakīrti's presents in defense of "a host of
antinomian pratices" in his *Vīramanoramā.
55 The Guhyasamāja histories in question that I consulted are the Gsang ba vdus pavi lung rigs man
ngag ston par byed pavi bla ma tshad mavi lo rgyus, 1-106, the text of the Dpal gsang ba vdus pavi rgyud vgrel
gyi bshad thabs kyi yan lag gsang bavi sgo vbyed, 383-585, which is incomplete and ends with a
biography of Bu ston, and the Dpal gsang ba vdus pavi dam pavi chos vbyung bavi tshul legs par bshad pa
gsang vdus chos kun gsal bavi nyin byed, 75-76.
56 For the [Cakra]śaṃvara histories, see Bu ston's Bde mchog spyi rnam don gsal, The Collected Works of
Bu ston [and Sgra tshad pa] [Lhasa xylograph], part 6 (New Delhi: International Academy of
Indian Culture, 1966), 1-117, of 1354, Go rams pa's undated Bde mchog chos vbyung brgyud pavi rnam
thar dang bcas pa, Sa skya pavi bkav vbum, comp. Bsod nams rgya mtsho, vol. 15 (Tokyo: The Toyo
Bunko, 1969), no. 82, 52/1-66/1, and A mes zhabs' 1624 Dpal vkhor lo bde mchog gi dam pavi chos
vbyung bavi tshul legs par bshad pa yid bzhin rin po chevi phreng ba dgos vdod kun vbyung, Collected Works, ed.
18
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
conclusion that we can draw from this state of affairs is that the Indo-Tibetan
tradition knew virtually nothing about his works and days and, in particular, did not
hold his *Sandhyāprakāśikā in very high regard. Indeed, history has relegated him to
her margins and shadows, a place that I think does little justice to him. But then I
use different criteria for judging the significance of his work. In this paper, then, I
seek to join Tomabechi in placing him more firmly on the map of late Indian
Buddhist intellectual history, a place that I believe he fully deserves. And if it turns
out that he really does not, that the tradition was correct in ignoring him and that
my confidence in this work was as utterly misplaced as would be this long article,
well, I guess, so be it!
I. BHAVYAKĪRTI AND HIS INTELLECTUAL ENVIRONMENT
At some yet to be more or less accurately ascertained point in time, but most
likely in the late eighth or early ninth century, a tradition parallel to, though serially
and chronologically certainly not identical with, the teacher-disciple lines of the
transmission of exoteric Madhyamaka thought emerged in the Indian subcontinent.
It embedded itself in the various lines along which the Guhyasamājatantra and its
associated smaller texts and spiritual practices were handed down. The inception of
this line began with one called Nāgārjuna (ca. 900) and in this essay, I designate him
Nāgārjuna II. This tradition became known as the vPhags pa (*Ārya) school of this
textual and practice-oriented corpus; the other mainline of transmission of the
Guhyasamāja cycle is the older Ye shes (*Jñāna) school, which is so-called after its
putative "founder" Buddha[śrī]jñāna (ca. 800) - the two designations, vPhags pa
and Ye shes, *Ārya and *Jñāna, are of course purely Tibetan in origin and have no
Indic precedent. 57 Although both the Madhyamaka and the *Ārya-school counted
a[n Ārya] Nāgārjuna, an Āryadeva, a Bhavya[kīrti, a Bhavyakara or a Bhāviveka],
and a Candrakīrti among their exponents, they differ significantly in the
chronological succession of these four. For example, unless we credulously
attribute a rather long life-span to Nāgārjuna the Madhyamaka philosopher, as does
the late Indian tradition and, arguably, under the influence of the *Ārya-school and
the alchemical doctrines that may have made this possible, he could not have been
the teacher of Candrakīrti the Mādhyamika or Candrakīrti, the exponent of the
Guhyasamājatantra's *Ārya-school. And Bhavya (6th c.) [or Bhāviveka] the
Si khron bod yig dpe rnying myur skyob vtshol sgrig khang, vol. 22 (Lhasa: Bod ljongs dpe rnying
dpe skrun khang, 2012), 1-255.
57 For the *Ārya school, see Wedemeyer (2007: 7-63) and Kilty (2013: 43-87), and for the *Jñāna
school, see Szántó [2013/2014] 2015).
Journal of Tibetan and Himalayan Studies
19
Mādhyamika certainly flourished before Candrakīrti I the Mādhyamika, while in the
*Ārya-school, he must have lived posterior to [its] Candrakīrti [II]. Similarly, and in
flat contradistinction to some recent contributions on the subject, it is in my
opinion extremely doubtful, to put it as diplomatically as possible, that Candrakīrti
the Mādhyamika is to be identified with the Ārya tradition's Candrakīrti.58 Aside
from Bhavyakīrti, the *Ārya-tradition knows of a Legs ldan byed [*Bhavyakara] [or:
Legs ldan vbyed (*Bhāviveka)], who, too, authored a subcommentary on
Candrakīrti II’s Pradīpoddyotana.59 Notwithstanding the similarity of their names, I
am disinclined to hold that Bhavyakīrti and the latter are the same person, if only
because I can see no compelling reason or evidence to do so. Indeed, the
differences in the contents of their work on the Pradīpoddyotana are so compelling
that it would be quite difficult, if not impossible, to argue for their identity.
Whatever else can be said of this thorough going conflation of members of
the Madhyamaka and the *Ārya-tradition, is that it was initially, clearly purposive.
One of its perhaps unwitting exponents was Atiśa as we can glean from his little
text on the two levels of reality in which he somewhat ambiguously states that [a]
Candrakīrti was a disciple of [a] Nāgārjuna and from his study of Madhyamaka in
which his list of the writings of [a] Nāgārjuna includes the Mūlamadhyamakakārikā,
the Pañcakrama, and the Bodhicittavivaraṇa!60 And whatever else one might wish to
say about it, it is safe to say that this conflation resulted in the production of very
skewed and, it must be emphasized, ultimately wrong relative chronologies for such
figures as Rāhulabhadra – see below - and Saraha, not to mention the relative
chronologies that are asserted for bona fide Mādhyamika authors. We cannot go into
this rather weighty conundrum any further, but we ought to be fully cognizant of
the fact that this mapping of the Madhyamaka tradition onto a transmission of the
Guhyasamājatantra had far-reaching consequences for the way in which the history
of a corpus of Buddhist ideas was pursued in late Buddhist India and Tibet. Taken
solely on its own terms, this wholesale adoption resulted in very interesting and
unexpected hermeneutics of the history of Indian Buddhism in Tibet. However, on
another, more profound level, the conflation of these two series had very
problematic and, it may perhaps even turn out, catastrophic consequences for
See the remarks in van der Kuijp (2014).
vol. 25, no. 1794 [#1792], 390/3-4/3 [A, 201b-12a]; this brief study was translated into
Tibetan by the Indian Mahopādhyāya Rgyal ba mchog and the Tibetan Lo tsā ba Shākya brtson
vgrus.
60 See, respectively, Lindtner (1981: 191, 194), and Miyazaki (2007: 59-60) and Apple (2010: 173),
and, of course, a good portion of the earlier Guhyasamājatantra literature belonging to the *Ārya
tradition in which this relationship is reiterated. For the Bodhicittavivaraṇa, see van der Kuijp
(2013a).
58
59 SDE,
20
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
subsequent Madhyamaka and Guhyasamājatantra scholarship in Tibet, in general.61 Y.
Bentor discussed an important case in point in connection with the Sanskrit text[s]
and the Tibetan translations (and interpretations) of Guhyasamājatantra, II: 3, via a
Tibetan version of the Pradīpoddyotana, which led Tsong kha pa and his disciple
Mkhas grub to argue that it promotes a Prāsaṅgika-Madhyamaka point of view.
The relevant wording in that Tibetan version is quite absent from the comments
Bhavyakīrti made in the *Sandhyāprakāśikā as well as in *Karuṇāśrī's evidently later
study, the *Pradipoddyotanapañjikā, in which it does not figure at all.62
If we are to be engaged in the study of Indian Buddhist intellectual history at
all, then it is crucial for us to be as clear as possible on who authored what and
when, relatively speaking. If no such clarities can be obtained , then it is just as
important for us to be aware of the blind spots that are present in the historical
narrative, regardless of whether we or the tradition are its authors, when the
available sources and our intellectual resources are unable to create such
transparencies, even after the strictest available criteria for ascertaining them have
been formulated and applied. A history of this kind has almost everything to do
with the text[s] we attribute to one and the same person. For these create the
coordinates of his location and floruit in the realm of the history of ideas. In an
Indian context, and with but very few exceptions, what we [can] know about a
thinker is invariably and exclusively co-extensive with his oeuvre, the way it
embodies and discloses its own sense of setting and context, and the subsequent
reactions it engendered over time. As an almost universal rule, virtually nothing is
transmitted, and thus known, about his life as such. And the little that may on rare
occasions have been handed down is so much suffused with legend and mystery
that, though of signal interest and importance on its own terms, these snippets of
how the tradition viewed an aspect of his life, or what it assumed his life to have
been, does not much for an appreciation of his life as such. Bhavyakīrti waxes
somewhat autobiographical in the concluding remarks of his large
*Sandhyāprakāśikā.63 To summarize: In the first place, he was a monk and not a
layman. Having traveled east, he first studied tantra. He then went to Urgyan, that
is, Uḍḍiyāna, located in what is now Swāt in Pakistan, for further studies. There he
Among one of many examples, we may cite the profoundly problematic juxtaposition in Dol
po pavs Ri chos nges don rgya mtsho of Candrakīrti I and Candrakīrti II in terms of their positions on
the tathāgata-matrix; see his Gsung vbum, vol. 2/13, ed. Dpal brtsegs bod yig dpe rnying zhib vjug
khang, Mes povi shul bzhag, vol. 197 (Beijing: Krung govi bod rig dpe skrun kang, 2011), 84. See
also Hopkins (2006: 106-107).
62 SKAL, 436/1 f. [Ki, 148a f.]; SKAL1, 915f. For Karuṇāśrī's work, see below.
63 SKAL, 521/7-522/1 [Khi, 154b-155a]; SKAL1, 367.
61
Journal of Tibetan and Himalayan Studies
21
studied with yogis and was visited in a dream by Vajrasattva, the tantric
embodiment of buddha, who bestowed on him various initiations. He then singles
out a certain Sangs rgyas zhi ba [*Buddhaśānti] as his master. It may come as a
surprise and it may prove to be of some importance that this man is not mentioned
in connection with the Guhyasamājatantra lineages of Bhavyakīrti that are outlined by
Bu ston, Ngor chen, and A mes zhabs.
Thus, in a profound way, Bhavyakīrti can very well be reduced to his oeuvre;
one might almost speak of an ontological identity between him and his work.
2. BHAVYAKĪRTI AND HIS OEUVRE
All the available catalogues and indeed every available collection of the
Tibetan Tanjur-canon contain three works that are explicitly signed by or asigned
to him. It should go without saying that it is at this point not possible to decide
unequivocally that the individual who seems to have signed his name to these three,
or to whom these three are ascribed, is one and the same Bhavyakīrti, but my
working hypothesis is that he is, even if there is some evidence to suggest that the
author of the *Vīramanoramā and the *Sandhyāprakāśikā may have been two
different individuals. The three works in question are:
1. *Vīramanoramā or *Śrīcakraśaṃvarasyapañjikā]
2. *Pañcakramapañjikā
3. *Sandhyāprakāśikā or *Pradīpoddyotanavyākhyāṭīkā
I will first briefly discuss the first two of these three, and then go into
somewhat greater detail with the third. As the main title suggests, the first tract is
an examination of difficult points (pañjikā) of the Herukābhidhāna and was translated
by the team of Dharmaśrībhadra and Lo tsā ba Rin chen bzang po at an
unspecified time and place. 64 The second is a succinct commentary on some
difficult points of Nāgārjuna II's Pañcakrama.65 The Indian grand-abbot Rgyal ba
mchog (*Jinottara/ Jayottara) and the Tibetan Lo tsā ba Shākya brtson vgrus
translated it sometime in the eleventh century and it contains nothing particularly
64 SDE, vol. 23, no. 1408 [#1405], 274/1-285/4 [Ma, 1a-41a] – the Tibetan translation titles it
Laghucakrasaṃvaratantrapañjikā.
65 SDE, vol. 26, no. 1840 [#1838], 170/2-171/7 [Chi, 1b-7b].
22
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
noteworthy or useful for my present purpose. However, the first is of some interest
as it sheds light on the author and his intellectual milieu. In the "author's colophon"
(mdzad byang) and concluding line, the author refers to himself by "I Skal ldan
[*Bhavya]" (bdag skal ldan)" and by "Skal ldan grags pa" [*Bhavyakīrti]." The name
"*Bhavya" also occurs in the introductory verses and on other occasions in the
text. 66 In an obviously self-confident moment where he unabashedly situates
himself in the history of Indian Buddhism, he mentions in one breath in the
colophon [a] Nāgārjuna, "who analyzed phenomena" - this would appear to be a
reference to the second-century Madhyamaka master -, [b] the Buddhist poet Ārya
Śūra (?4thc.), [c] the grammarian Candragomi[n], and [d] "me *Bhavya," an expert in
"the way of mantra" (sngags tshul, *mantranaya) that is, tantra.67 Whatever else this
juxtaposition may tell us about him, it is clear that, aside from having certain
immodesty about him, he flourished after Candragomin concerning whom the
scholarly consensus is that he lived in the seventh century. Of course, worth noting
is that he nowhere mentions Candrakīrti I or Candrakīrti II, but, then, neither man
had any connections with the Herukābhidhāna. In terms of its format and structure,
the exegesis is unusual and thus rather interesting. Chapter by chapter, the text
often proceeds simply by way of a third-person juxtaposition of Bhavyakīrti's views
with those of other commentators whom it often names, and this makes one
wonder whether the author of this work was not one of his disciples, while it
simultaneously raises the question whether this work should then be understood as
a record of his lectures on the tantra as taken down by one or more of his disciples.
In any event, the other commentators mentioned there include the following:
1.Koṅkaṇa
6.Ḍombi
2. Indrabhūti
7.Spyod pa pa/Cāryapā[da]
3. Padma
8.Kumārī
4. Dpal U rgyan pa
9.Dri sbyor
5. Kambala
10. Āryadeva68
See, for example, SDE, vol. 23, no. 1408 [#1405], 274/3, 6, 275/1, 6 [Ma, 2a, 3b, 4b, 7a].
Sde srid Sangs rgyas rgya mtsho (1653-1701) cites the verse in his 1687 Bai ḍūrya g.ya sel [based
on the Lhasa Zhol xylograph], vol. II (New Delhi: T. Tsepal Taikhang, 1971), 601. For an
exploration of the term *mantranaya, see Wallace (2011).
68 He appears to have been overlooked in the listing provided in Szántó (2012: 47). The
*Vīramanoramā cites two passages in verse that are attributed to Āryadeva; see SDE, vol. 23, no.
1408 [#1405], 282/4 [Ma, 30b] and Bstan vgyur [dpe bsdur ma], ed. Krung govi bod yi shes rig zhib
66
67
Journal of Tibetan and Himalayan Studies
23
Koṅkaṇa is by far the most quoted individual in this study, but there are some
problems with this name. In the first place, it is of course foremost a toponym.
D.C. Sirkar registers "Koṅkaṇa" as the name of a "strip of land" in southwest India
on the Arabian Sea, and D.G. White identified it in the context of the Hindu tantric
deity "Mother of Koṅkaṇa" as "an apparent reference to the coastal strip of
modern-day Maharashtra, Goa, and Karnataka."69
Bhavyakīrti writes at the outset of his *Sandhyāprakāśikā that70:
kong ka na yi grong khyer du //
dpal gyi ri yi rtse mor ni //
skye bos dben pavi gnas nyid du //
dang por de la gnas byas na //
mi dang lha yi bdag por mtshungs //
mchog tu gyur pavi chos gsungs pa
sgra gcan vdzin gyi zhal snga nas //
sa steng vdi la rgyal gyur cig /
mnyes pa de las de nyid ni //
theg chen rnal vbyor pas thob pa //
sa thob gzhan gyi don la brtson //
dpal ldan klu sgrub rgyal gyur cig //
vjug lte gnas kyi bkav bstan dpe sdur khang, vol. 9 (Beijing: Krung govi bod kyi shes rig dpe
skrun khang, 1997), no. 0304, 72. The first stanza also survives as a quotation Jñānakīrti's circa
1000 *Tattvāvatāra… prakaraṇa; see Bstan vgyur [dpe bsdur ma], ed. Krung govi bod kyi shes rig zhib
vjug lte gnas kyi bkav bstan dpe sdur khang, vol. 41, no. 2407, 141. Neither is found in any of the
works by or attributed to any one of the Āryadevas. We may add that Szántó (2012: 21, n. 16)
distinguishes between as many as "three Tantric Āryadevas."
69 See, respectively, Sircar (1960: 101) and White (2003: 113). Mallinson (2019: 7 ff.) gives a
detailed overview of the religious diversity of the region.
70 SKAL, 394/2 [Ki, 21]; SKAL1, 564. This passage is cited by many Tibetan scholars, including
Tsong kha pa and Mkhas grub Dge legs dpal bzang po, see, respectively, Kilty (2013: 67-68) and
Cabezón (1992: 85, 437, n. 250) [= Khang dkar and Fujinaka (2001: 110)].
24
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
de las de nyid rin chen ni //
thob pa vjig rten gsum du grags //
vdus pavi rgya mtshovi pha rol phyin //
zla grags gsal byed rgyal gyur cig //
de yis byas pavi vgrel pa ni //
mthav drug bshad pas zab pa nyid //
de ni deng bdag ma lus par //
skal ldan grags pas gsal bshad bya //
When he first made his residence there,
In a place devoid of people,
In a town (grong khyer)71 of Koṅkana,
On the peak of Śrīparvata,
Like the lord of man and the gods [= the Buddha],
[He] pronounced the supreme doctrine;
May the words of Rāhula[bhadra],72
Be victorious on this earth!
From him, having attained happiness,73
By means of the Great Vehicle yoga on reality,
Exerting himself for the benefit of others to attain a Bodhisattva level;
Of course, Tibetan grong khyer regularly reflects Sanskrit nagara, so that one has to wonder
about the intent of "devoid of people" (skye bos dben, *vijana)!
72 For him, see, for example, Lamotte (1970: 1372-1375), and the references cited there.
73 As Kilty's translation suggests, this might be an allusion to him having attained the first,
pramuditā, Bodhisattva level, as we find in the Mañjuśrīmūlakalpa, for which, however, the usual
Tibetan equivalent is rab [tu] dgav [ba]; see also the remarks in He-van der Kuijp (2014: 327-329).
71
Journal of Tibetan and Himalayan Studies
25
May glorious Nāgārjuna be victorious!
From him, he attained the jewel state,74
Renowned in the three worlds,
He crossed the ocean of the Guhyasamāja;
May Candrakīrti the illuminator75 be victorious!
Because the commentary written by him
Explained the six alternatives,76 that profundity,
Will now be clearly explained by,
Me, *Bhavyakīrti, in toto.
Śrīparvata is usually considered to be located in South India, in the vicinity of
the stupa of Dhāṇyakaṭaka, in Amarāvatī, in present-day Āndhra Pradesh. 77
Mabbett suggested that it designates a hill in the Nāgārjunakoṇda valley and, earlier,
White rightly pointed out that "śrīparvata - generally identified with a holy peak
located in the central Deccan plateau ... of South India - is a highly generic term for
a mountain or hilltop."78 Whatever the veracity of our sources, we will for now
have to keep in mind that the toponym "Koṅkaṇa" might not always refer to the
same place. And, indeed, "Koṅkaṇa" pops up in a number of places. Recently,
Davidson translated an interesting autobiographical moment that he uncovered in
the very large Kosalālaṃkāra by [a] Śākyamitra (8th c.). There we learn that he, too,
had studied in Koṅkaṇa and *Vaidyapāda (9th c.) notes in another pertinent
biographical vignette that Buddhajñānapāda had spent some time in Koṅkaṇa as
well.79 We may mention in passing that there is also at least one other treatise in
This may be an allusion to the most sublime of the five diferent types of individuals/disciples
that Candrakīrti II isolated at the outset of his work; see Chakravarti (1984: 4) {= Bstan vgyur [dpe
bsdur ma], ed. Krung govi bod yi shes rig zhib vjug lte gnas kyi bkav bstan dpe sdur khang, vol. 15,
no. 0687, 833-834}. For the "jewel-like disciple", see Kilty (2013: Index, 639 sv jewel).
75 This, I submit, is obviously an oblique reference to his Pradīpoddyotana commentary, where
Sanskrit uddyotana reflects Tibetan gsal [bar] byed [pa].
76 See Arènes (2002).
77 MacDonald (1970: 181); for Śrīparvata, see also Roçu (1969: 30 ff., 39 ff., 49 ff.).
78 Mabbett (1998: 344) and White (1996: 60).
79 Davidson (2002: 159, 312). For *Vaidyapāda [or *Vitapāda], see Szántó (2016: 540-541, n. 6).
74
26
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
which Koṅkaṇa is noted. This occurs in *Dharmamitra's Tshig rab tu gsal ba
commentary on the Abhisamayālaṃkāra - only available in Tibetan translation, the
Sde dge xylograph gives Prasannapadā, as its subtitle, though it is more commonly
[and correctly] known as the *Prasphuṭapadā - where we encounter the phrase koṅ
ka ṇa pa dag, that is, "Individuals from Koṅkaṇa," in a narrative where he points to
regionally different interpretations of the notion that those with dull faculties have
no discriminative insight.80
But "Koṅkaṇa" is not only used as a toponym in Tibetan translations of Indic
texts. In the translation of Bhavyakīrti's *Vīramanojñā, it is also used as a personal
name. The historian and virtuoso of Tantric literature and practice, Tāranātha,
wrote in his 1608 survey of Buddhism in India that a Vagiśvarakīrti (10th-11th c.)
had received the transmission of the Laghucakrasaṃvaratantra from a *Vajrahāsa and
that the latter had been a disciple of Jayabhadra.81 He adds that this Jayabhadra,
who hailed from Sri Lanka, had also written a commentary on the tantra. And
where had he done so? In Koṅkaṇa! I am not altogether sure whence Tāranātha
derived this information, because the colophon of Jayabhadra's pañjikācommentary on the Herukābhidhāna merely says that he wrote it [and other works]
in Tāra monastery ?in Mahābimba/?Mahāviśva 82 - these other tracts possibly
included a panegyric, an evocation (sgrub thabs, sādhana) and a maṇḍalavidhi.83 Be this
as it may, the certain identity of this Jayabhadra with Bhavyakīrti's "Koṅkaṇa" is
borne out when we compare the references of the latter to either what Koṅkaṇa
had said (bshad) or to his position - as a rule, the Tibetan translators marked the
latter with the honorific bzhed, "avers"; the standard vdod, "claims," is used in
connection with others and also with Bhavyakīrti himself - with what we read in
Jayabhadra's pañjikā-commentary. The very first three opinions Bhavyakīrti
attributes to Koṅkaṇa are84:
80 SDE,
81 Rgya
vol. 33, no. 3801 [# 3796], 190/2 [Nya, 43b.]
gar chos vbyung, ed. Thub bstan nyi ma (Chengdu: Si khron mi rigs dpe skrun khang, 1986),
283-284, 310-311 [= Dam pavi chos rin po che vphags pavi yul du ji ltar dar bavi tshul gsal bar ston pa dgos
vdod kun vbyung, Jo nang rje btsun tā ra nā thavi gsung vbum dpe bsdur ma, vol. 33/45, Mes povi shul
bzhag, vol. 75, ed. Dpal brtsegs bod yig dpe rnying zhib vjug khang (Beijing: Krung govi bod rig
pa dpe skrun khang, 2008), 224, 245-6]; see also Chimpa and Chattopadhyaya (1980: 296, 325)
82 On these, see Szántó (2012: 40, n. 61).
83 On Jayabhadra and his main work, see Sugiki (2001: 91-105). Szántó (2012: 43) arrived at the
same conclusion that Koṅkaṇa and Jayabhadra are one and the same person.
84 SDE, vol. 23, no. 1408 [#1405], 274/5-5/1 [Ma, 3a-5a].
Journal of Tibetan and Himalayan Studies
27
1. dkyil vkhor du vjug pa la sogs pa nas gsang bavi dbang bskur bavi bar du ni dgos pavo
// dus phyis devi vbras bu thob pa ni kongka navi zhal snga nas bzhed do //
2. de yang syaṃ na hid sra ba ne zhes bya bavi mthav ma phyis pa gang yin pavo //
zhes kongka navi zhal snga nas gyis bshad do //
3. khams gsum gyi bdag nyid can gyi vkhor lo gsum vdir bshad par bya ba yin pavi phyir
ro zhes kong ka navi zhal snga nas bshad do //
The corresponding three passages in the Tibetan translation of Jayabhadra's
pañjikā are85:
1. [bcom ldan vdas kyis gdul byavi skye bo vjug par mdzad pavi phyir] dkyil vkhor du
vjug pa la sogs pa nas gsang dbang bskur bavi bar ni dgos pa yin la / [devi vog tu] dus phyis des
vbras bu thob pa ni dgos pavi dgos pa yin no //
2. de yang sa yan du sra ba na zhes bya bavi yi gevi tha ma gzung bavi phyir te...
3. khams gsum pavi bdag nyid du vkhor lo gsum bshad par bya ba yin no //
These [and the others citations] compare so favorably to the Sanskrit text of
Jaya-bhadra's work that, though the evidence is not wholly airtight, the chances are
much better than merely good that Jayabhadra is none other than Bhavyakīrti's
"Koṅkaṇa." What is more, in the Cakrasaṃvaratantra commentary that is included
in Sa chen Kun dgav snying po's collected works, mention is made of Kong ka na
Rgyal ba bzang po as well as almost immediately thereafter of a claim made by a
person he styles Master Kong ka na!86 One wonders if these are in fact two
different individuals. Taking into consideration, as we must, that these Tibetan
texts most probably had different translators and that, again probably, neither the
original Sanskrit manuscripts on which they are based, nor, and here we can be
even more certain, the manuscript transmission of their Tibetan translations were
immaculate, these samples leave very little doubt that Jayabhadra was in all
probability Bhavyakīrti's "Koṅkaṇa."
85 SDE, vol. 23, no. 1409 [#1406], 285/5-7 [Ma, 41b-42b]; for the Sanskrit texts, see Sugiki (2001:
105, l. 13-14; 106, l. 22; 106, l. 28).
86 See the Dpal vkhor lo bde mchog gi rtsa bavi rgyud gyi ṭī ka mu tig phreng ba, Collected Works, vol. 2
(Dehra Dun: Sakya Center, 1992-1993), 552.
28
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
The foregoing has significant implications where Bhavyakīrti's dates are
concerned. If, indeed, Vagiśvarakīrti were a grand-disciple of Jayabhadra, then the
latter may have lived in circa 900 or perhaps just slightly before that. Bhavyakīrti
knew the latter's work, and this means that he must have flourished after him,
perhaps not even before circa 950, at the earliest. Further "soft" evidence for this
rather rough chronology is found when we look into the various curricular records
(gsan yig/thob yig). For example, the one by Bu ston suggests there were, respectively,
five or six positions [or generations] separating him from, respectively, Sa skya
Paṇḍita (1182-1251) and Chag Lo tsā ba.87 In the first, the penultimate position is
occupied by the Kashmirian scholar Śākyaśrībhadra (1127-1225). In both instances,
then, Bhavyakīrti's floruit can be triangulated at around the middle of the tenth
century, and in both he is said to have been the grand-student of Candrakīrti II.88
Before turning to what is arguably Bhavyakīrti's major work, namely his large
study of the Pradīpoddyotana, it is worth our while briefly to consider Szántó's
thoughtful arguments for supposing that there were two different Bhavyakīrtis,
Bhavyakịrti I, the author of the *Vīramanoramā and the later Bhavyakīrti II, the
author of the *Sandhyāprakāśikā; he writes89:
…a quotation 'from the Catuṣpīṭha' in the *Sandhyāprakāśikā (20r-20v)90 on the
very same topic, rosaries, shows a text that is very different from the one quoted in
the *Vīramanoramā. I find it very unlikely that the same author would quote two
different recensions of the same passage for the same – rather banal – topic.
I am not sure how 'unlikely' it is that "the same author would quote two
different recensions of the same passage for the same…topic" and thus how
compelling this argument might be, but it would be time to begin looking for
parallel instances. Aside from the quotation "from the Catuṣpīṭha," Bhavyakīrti [?II]
See, respectively, Bla ma dam pa rnams kyis rjes su bzung bavi tshul bkav drin rjes su dran par byed pa,
The Collected Works of Bu ston [and Sgra tshad pa] [Lhasa xylograph], part 26 (New Delhi:
International Academy of Indian Culture, 1971), 98; On p. 99/1, there is transmission lineage in
which he or his namesake figures that is obviously contaminated; it reads: Buddha - Maitreya Nāgārjuna - Bhavyakīrti - Candrakīrti - Dharmakīrti - Tilopa...; see also Ngor chen Kun dgav
bzang po (1382-1456), Thob yig rgya mtsho, Sa skya pavi bkav vbum, comp. Bsod nams rgya mtsho,
vol. 10 (Tokyo: The Toyo Bunko, 1969), no. 36, 46/4.
88 See also the note on him in Campbell (2009: 80-81).
89 Szántó (2012: 43, n. 83).
90 SKAL, 399/4 [Ki,20a]; SKAL1, 607.
87
Journal of Tibetan and Himalayan Studies
29
also cites from the Rdo rje gdan bzhi [Vajracatuṣpītha],91 and Szántó indicated that the
title Rdo rje gdan bzhi can at times reflect Catuṣpītha.92 Further, it is quite true that on
stylistic grounds the *Vīramanoramā and the *Sandhyāprakāśikā could hardly be more
different! The first is virtually written in dialog form whereby the positions of
interlocutors are tabulated in terms of what the authors had stated what they
claimed to be the case; reminder: it is not insignificant that the Tibetan translators
used vdod for Bhavyakīrti's claim [and for those of some others] and bzhed for
Jayabhadra's. The use of named interlocutors is therefore somewhat reminiscent of,
for instance, the very succinct pañjikā-commentary on the Sarvabuddhasamāyogatantra
that is attributed to [an] Indrabhūti; it is extant in the Tibetan translation of
Śraddhākaravarman and Lo tsā ba Rin chen bzang po.93 For this work consists in
large part of a conversation between Indrabhūti and Śākyamitra, and in terms of
content, there is significant overlap between this text and the *Vīramanoramā.
Further, the *Vīramanoramā has two notices in its preamble and one in the very
brief colophon, all of which clearly indicate that it was written by a Bhavyakīrti,
even if the main body of the text only cites the claims of "Bhayvakīrti" or
"Bhavya," and nothing there is said or written in the first person. This could of
course very well indicate that it was in fact written by someone else whose name
may but only may have been "Bhavyakīrti". Szántó also points out that "there are
significant doctrinal differences" between the two and gives as an example the fact
that "in the *Vīramanoramā the culminating abhiṣekha[-empowerment] is the
guhyābhiṣekha" – this is the "secret empowerment" - the *Sandhyāprakāśikā includes
the "higher" empowerment, the so-called insight-gnosis empowerment
(prajñājñānābhiṣekha). This is of course not prima facie insignificant, but it is also not
entirely convincing as part of the argument for two Bhavyakīrtis, in my opinion.
First, neither Jayabhadra nor Bhavaykīrti [I] comments on the Herukābhidhāna in toto;
rather, these are commentaries on what they considered were the text's difficult
passages. Bhavabhaṭta, alias Bhavabhadra, who in Szántó's view postdates
Bhavyakīrti [I], 94 does mention, for example, the *prajñā[jñānā]bhiṣekha in
connection with his comment on the Herukābhidhāna at III: 7 – on the other hand,
his text of the Herukābhidhāna is more "complete".95 And this is the only place in
91 SKAL, 500/1 [Khi, 78a]; SKAL1,
92 Szántó (2012: 27, n. 27).
186.
93 SDE, vol. 25, no. 1663 [#1661], 1-6/3 [La, 1-19b]. Other individuals who are mentioned in this
interesting work include *Hāsavajra and Ku ku ra [= ?*Kukurāja].
94 Szántó (2012: 43-44).
95 The Sanskrit text in Pandey (2002: 39) has prajñābhiṣekha (*shes rab kyi dbang), whereas the
Tibetan text in Pandey (2002: 240) has the correct shes rab ye shes kyi dbang, as has SDE, vol. 25, no.
30
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
his entire commentary where he, in fact, mentions this empowerment! But when
we look at how this particular chapter fared under the scrutiny of Jayabhadra's and
Bhavyakīrti [I]'s exegetical eyes, we notice that they do not care to explain this
passage of the text at all, even though their version of the Herukābhidhāna had the
identical verse. And, indeed, their comments on the third chapter of the
Herukābhidhāna are extremely sparse!96 On the other hand, the *Sandhyāprakāśikā
does mention "four empowerments" (dbang bzhi) including the prajñājñānābhiṣekha,
even if Candrakīrti II of the Pradīpoddyotana mentions neither!97 Bhavyakīrti did so
in connection with a quotation from the Guhyasamājatantra's eighteenth chapter,
Guhyasamājottaratantra, 112c-113, whereas Candrakīrti II commented on the
Guhyasamājatantra in seventeen chapters and, as we know, did not comment on the
work known as the Guhyasamājottaratantra, its purported eighteenth chapter.
Attributed to a Nāgārjuna, the *Guhyasamājatantraṭīkā is a commentary on the
Guhyasamājatantra in which the Guhyasamājottaratantra is quoted severally, including,
again, Guhyasamājottaratantra, 112c-113, wherein the four empowerments are
mentioned, including the prajñājñānābhiṣekha, the third of these.98 As I indicated
elsewhere, the Tibetan tradition considered the authorship of this work,
purportedly by a Nāgārjuna, to be quite problematic.99 Finally, Szántó points out
that the Bhavyakīrti of the *Vīramanoramā does not once cite the Hevajratantra "even
in places where we would expect a quotation" and instead quotes a passage from
the Guhyasamājatantra, V: 1-5,100 [as well as other passages from this text]. The same
can also be said of the *Sandhyāprakāśikā in which we do not once come across an
explicit quotation from the Hevajratantra, although the author does quote the Bde
mchog [*Śaṃvara] a good number of times! However, what is probably rather
significant is that the Bhavyakīrti of the Vīramanoramā designates the
Guhyasamājatantra as a yogatantra, whereas the *Sandhyāprakāśikā explicitly
categorizes it as "a supreme yoga[tantra]" (rnal vbyor gyi bla ma).101 So, in brief, I am
not altogether convinced that these differences are, in aggregate, sufficient to
1663 [#1661]; Sanderson (2009: 209-211) compares the listings of the empowerments in
Jayabhadra and Bhavabhadra.
96 SDE, vol. 23, no. 1409 [#1406], 287/5-6 [Ma, 48b-49a] and SDE, vol. 23, no. 1408 [#1405],
277/5 [Ma, 13b].
97 SKAL, 522/1 [Khi, 155a]; SKAL1, 367-368.
98 Bstan vgyur [dpe bsdur ma], ed. Krung govi bod kyi shes rig zhib vjug lte gnas kyi bkav bstan dpe
sdur khang, vol. 15, no. 0685, 17-18. For the quotation, see Matsunaga (1978: 121) and Bkav vgyur
[dpe bsdur ma], ed. Krung govi bod rig pa zhib vjug lte gnas kyi bkav bstan dpe sdur khang, vol. 81,
no. 0466, 595.
99 van der Kuijp (2014: 122-124).
100 Szántó (2012: 45).
101 SKAL, 394/4 [Ki, 2b]; SKAL1, 565. However, he calls it a great-yogatantra in SKAL, 521/2 [Khi,
152a]; SKAL1, 360. This ambiguity is puzzling.
Journal of Tibetan and Himalayan Studies
31
warrant the conclusion that there were two Bhavyakīrtis. But we have to be aware
that there may have been two of them.
So far, Bhavyakīrti's in many ways remarkable *Sandhyāprakāśikā appears to be
only extant in a fairly readable, if certainly not always spotless, Tibetan translation
and edition by a certain Gzhon nu [= Kumāra]. It was never translated into
Chinese, but there is a chance that a portion or perhaps even the entire text is still
extant in Sanskrit. A handlist of copies of Sanskrit palm leaf manuscripts from
Tibet that are housed at the Chinese Tibetan Research Center in Beijing registers an
incomplete five-folio manuscript "of the beginning of a commentary (rgya cher bshad
pa)" on the Pradīpoddyotana and a one hundred and twenty-two folio manuscript of
what the compiler has called a Pradīpoddyotanaṭīkā.102 That is all that is said about
these and, to be sure, only their study will allow us to establish the identity of their
authors.
The Tibetan translators' colophon (vgyur byang) relates that103:
yon gyi bdag po gtso bo vbrom jo bo dkon mchog khu mtshan gyis zhus nas / rgya gar
mkhan po chen po ku mā ra nyid kyis bsgyur cing zhus te gtan la phab pavo //
The text was thus translated and edited "by Kumāra himself" (ku mā ra nyid
kyis), who had received funding for this project from "the foremost of patrons"
vBrom Jo bo Dkon mchog khu mtshan; I cannot identify the latter individual at the
moment, but suffice it to say that he must have been a well to do member of the
vBrom clan. To this, the Snar thang and Beijing xylographs add a note to the effect
that the text "ought to be given to others":
gzhan la sbyin par bya bavi chos dpal gsang ba vdus pavi rgyud kyi rgyal po chen povi vgrel
pa sgron ma gsal bar byed pavi ṭī ka dgongs pa rab gsal zhes bya bavo //
Zhongguo zangxue yanjiu zhongxin shoucangde fanwen beiye jing (Suowei jiaojuan) mulu 中国藏学研究
中心收藏的梵文贝叶经 (缩微胶卷)目录/ Krung govi bod kyi shes rig zhib vjug lte gnas su nyar bavi ta
lavi lo mavi bstan bcos (sbyin shog vdril mavi par) gyi dkar chag mdor gsal (np, nd), 59, no. 77/5, 147, no.
169; my thanks go out to V.A. Wallace for so generously giving me a copy of this valuable
catalogue.
103 SKAL, 522/1 [Khi, 155a]; SKAL1, 367-8.
102
32
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
A topical outline-cum-summary of the Pradīpoddyotana in Tibetan is attributed
to a Kumāra as well and we are told in its colophon that it was translated by Gzhon
nu bum pa can [*Kumārakalaśa] and Lo tsā ba Shākya blo gros.104 This might well
indicate that Kumāra and *Kumārakalaśa are not one and the same person. We also
learn from the colophon that this Kumāra was a disciple of Dpal Lakṣmi ka ri [*Śrī
Lakṣmikari (sic)]. Kumāra himself must be assigned to around the middle of the
eleventh century, for he refers to the extremely versatile scholar Ratnākaraśānti (ca.
1000), who himself was the author of a large study of the Guhyasamājatantra in
eighteen chapters, among many other things.105 The colophon of the translation of
Jñānavajra's Sgrub pavi spyod pa la vjug pa [?*Siddhacāryāvatāra] states that it was
rendered into Tibetan by "the great layman Btsun pa" *Kumāra and a Lha btsun pa,
where the latter was most probably the royal-monk Zhi ba vod (1016-1111) of Gu
ge with whom he, Kumāra, had co-translated the voluminous Śrīpāramādyatantra.106
I suspect that these Kumāra-s are the same person. *Kumārakalaśa co-translated
the Mañjuśrīmūlakalpa with Lo tsā ba Shākya blo gros in the monastery of Tho gling
monastery in Gu ge.107 And I therefore think it is not entirely impossible that
Gzhon nu/*Kumāra is simply a short form for Gzhon nu bum pa, that is,
*Kumārakalaśa, and that, therefore, the colophon of the outline-cum-summary of
the Pradīpoddyotana may be misleading. But this is a mere conjecture. What would
104 SDE,
vol. 25, no. 1793 [#1791], 381/4-90/3 [A, 170b-201b]. Both men also translated
Prajñāmitra's Ālokamālā and its *Hṛdānandajananī commentary by a Ngo bo nyid med pa
(Asvabhāva). In an entry for the year 1442 of his 1517 biography of vGos Lo tsā ba Gzhon nu
dpal (1392-1481), Zhwa dmar IV Chos grags ye shes (1453-1524) relates that his master had left
unfinished (vphror lus) a translation of an unidentified commentary on the Ālokamālā; see vGos lo
gzhon nu dpal gyi rnam thar, ed. Ngag dbang nor bu (Beijing: Mi rigs dpe skrun khang, 2004), 86 [=
Dpal ldan bla ma dam pa kun mkhyen pa don gyi slad du mtshan nas smos te gzhon nu dpal gyi rnam par thar
pa yon tan rin po che mchog tu dgyes pavi ljon pa, Collected Works, ed. Yangs can dgon ris med dpe
rnying myur skyob khang, vol. 1 (Beijing: Krung govi bod rig pa dpe skrun khang, 2009), 511].
This passage, as are many others, is taken verbatim from Smon lam grags pa's earlier biography of
vGos Lo tsā ba, for which see his Mkhan po rin po che thams cad mkhyen pa gzhon nuvi zhal snga nas kyi
rnam par thar pa, dbu med manuscript, bdrc.org, W26618, 120. One wonders whether this
commentary was written by Vanaratna (1384-1468) or whether it was one that is referred to by
Asvabhāva in the afterword of his commentary. The commentary to which he referred was the
one by a certain Dgav ba sbas pa (*Ānandagupta); see SDE, vol. 36, no. 3901 [# 3896], 296/6 [Ta,
108b]. Asvabhāva writes tongue-in-cheek that this work was not pleasure-engendering (dgav byed
min).
105 See, respectively, SDE, vol. 25, no. 1793 [#1791], 383/3 [A, 177a], and SDE, vol. 26, no. 1853
[#1851], 467/4-502/5, 537/1 [Ti, 202a-325a, Thi, 2a-120a].
106 SDE, vol. 26, no. 1829 [# 1827], 114/5 [Ci, 115a], and Karmay (1998: 20-21).
107 This was but one of several translations, whole or partial, of this work; see van der Kuijp
(2015: 361 ff.).
Journal of Tibetan and Himalayan Studies
33
argue against their identification are the following two examples. Written by
Prajñāmitra, the Sanskrit text of Ālokamālā, 11, reads108:
bhūtakotiś ca sā saiva tathatā saiva śūnyatā /
samatā saiva sā muktiḥ saiva vijñaptimātratā //
It is also the limit of what is real, it is thusness, it is emptiness,
It is sameness, it is liberation, it is the state of consciousness only!
*Kumārakalaśa and Lo tsā ba Shākya vod rendered this verse as follows:
yang dag mthav yang de yin te //
de nyid de bzhin nyid stong nyid //
mnyam nyid de nyid grol ba nyid //
de nyid rnam par rig tsam nyid //
It is clear that they made no effort to attempt to account for the poetics of the
yamaka (zung ldan) figure of saiva. In addition, their translation of tathā saiva śūnyatā
by de bzhin nyid stong nyid in 11b clashes with acceptable Tibetan prosodic rules.
Now consider Kumāra's rather elegant solution to the poetics of this quatrain in his
translation of the *Sandhyāprakāśikā 109:
de nyid yang dag mthav yin te //
de nyid de bzhin de stong nyid //
de nyid mnyam nyid de grol ba //
de nyid rnam rig tsam nyid de //
108 The Sanskrit, Tibetan, and English translations, with my implicit and explicit modifications,
are found in Lindtner (2003: 16-17).
109 SKAL, 433/1 [Ki, 137b]; SKAL1, 890.
34
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
We find the same glaring differences in their translation of Ālokamālā, 13-14,
which state the following110:
svasaṃvedyā tu sā saukṣmyād buddhānāṃ sūkṣmadarśinām /
mādṛśaiḥ svāśrayasthāpi sthūladhībhir na dṛśyate //
dūrasaṃjñī bhaven mokṣe na kathaṃ cana yogavit /
śūnyaḥ kalpitarūpeṇa dṛṣṭaḥ svātmani nirvṛtiḥ //
It can, however, due to its subtlety be personally experienced by the subtleseeing Buddhas. Though thusness rests in one's own body, it cannot be seen by
blockheads like me.
A yogin should in no way think that liberation is far away. When he is seen in
himself as devoid of imagined nature or duality, this cognition is extinction.
The team of *Kumārakalaśa and Lo tsā ba Shākya vod translated these verses
as follows:
rang rig de yang phra bavi phyir //
sangs rgyas rnams kyis phra bar gzigs pa //
rang la gnas kyang bdag vdra bas //
rtsing bavi phyir ni mthong ba med //
nam yang rnal vbyor rig pa yis //
thar pa ring snyam mi bsam ste //
brtags pavi ngo bos stong pa nyid//
110
Lindtner (2003: 16-17).
Journal of Tibetan and Himalayan Studies
35
rang dngos mthong bas mya ngan vdas //
But Kumāra has111:
phra bar gzigs pavi sangs rgyas kyis //
phra phyir de ni rang rig yin //
rang gnas kyang ni bdag vdra bas //
blo rags phyir ni ma mthong ngo //
rnal vbyor rig pasa nam yang ni //
thar pavi vdu shes ring mi bya //
brtags pavi ngang tshul rang bdag nyid //
stong par mthong na thar pa yin //
a
Sde dge and Co ne xylographs read rig par.
The translation of Ālokamālā, 13c, is by and large unproblematic. Lindtner's
"Though [thusness] rests in one's own body" (svāśrayasthāpi) is obviously not based
on the interpretation of *Asvabhāva's *Hṛdānandajananī commentary. For there we
read the more cogent idea that thusness "resides in one's own mind" (rang gi sems la
gnas pa). 112 Something went seriously awry with Kumāra's rendition of 14b.
*Kumārakalaśa and Lo tsā ba Shākya vod nicely rendered dūrasaṃjñī bhaven mokṣe
by thar pa ring snyam mi bsam ste,113 Kumāra's vdu shes is cute and his translation
ignores the Sanskrit syntax. Lindtner's version of 14c-d is jarring, to say the least.
The use of the lexeme rang dngos for svātmani is somewhat interesting.114 Finally,
noteworthy is that the text of the citation of Ālokamālā, 13, in Atiśa's and Nag tsho
111 SKAL, 436/6
112 SDE, vol. 36,
[Ki, 150b]; SKAL1, 921.
no. 3901 [# 3896], 284/6 [Ta, 66b].
113 For the phrase dūrasaṃjñī (ring bar mi bsam), see also Liu (2015: 19, 39) and Kano (2015: 199200).
114 For cognates, see van der Kuijp (1986: 26, 49, 'Additional Note').
36
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
Lo tsā ba Tshul khrims rgyal ba's (?1011/1012-ca. 1070) rendition of the
Madhyamakaratnapradīpa, a work that is attributed to a Bhāviveka but which, pace
Lindtner, cannot predate the ninth century,115 suggests that the text in the Sanskrit
manuscript they used for their translation was in pretty bad shape, for it reads as
follows116:
vdi ni rang rig phra ba ste //
phra ba rnams kyi spyod yul yin //
bdag cag lta bur gyur pa yi //
blo gros rtsing bas mi shes so //
The same passage from Candrakīrti II's Pradīpoddyotana is also cited in both the
*Madhyamakaratnapradīpa and the *Sandhyāprakāśikā. The case in point occurs in the
comment of Guhyasamājatantra, II: 3117:
tena bahyavastvātvākārabhāvanā nopalabhyate / na vidyate cittamayatvāt /
de bas na phyi rol gyi dngos po lta bur sgom pa dmigs su med cing yod pa ma yin te sems
kyi rang bzhin yina pavi phyir ro //
a
Sde dge and Co ne xylographs read yin.
See Ejima (1980a), He-van der Kuijp (2014: 323-329) and Del Toso (2014). There is also a
gloss in the Mngav ris Chos kyi rgyal po's (1306-1386) catalog of the Byang Ngam ring Tanjur to
the effect that Bhāviveka was not its author, "since it quotes the texts of Candrakīrti and
Dignāga"; see Bstan bcos vgyur ro vtshal gyi dkar dkar dri med vod kyi phreng ba, Jo nang dpe tshogs, vol.
23, ed. Ngag dbang kun dgav vjams dbyangs blo gros (Beijing: Mi rigs dpe skrun khang, 2010),
130-131. Aside from a serious chronological problem with dating this work, Lindtner (1984)
nonetheless gives a fine survey of its contents.
116 Lindtner (1982: 175, n. 39).
117 Chakravarti (1984: 31) {= Bstan vgyur [dpe bsdur ma], ed. Krung govi bod kyi shes rig zhib vjug
lte gnas kyi bkav bstan dpe sdur khang, vol. 15, no. 0687, 883}.
115
Journal of Tibetan and Himalayan Studies
37
Therefore, meditative cultivation on what is like an external thing has no
external reference and does not exist, because [the thing] is of the nature of the
mind.
This line reads in the *Madhyamakaratnapradīpa118:
de bas na phyi rol gyi dngos po lta bu dmigs su med cing yod pa ma yin te sems kyi rang
bzhin yin pavi phyir ro //
And in the *Sandhyāprakāśikā - the text in [….] consists of Bhavyakīrti's
comments]:119
de bas na phyi rol gyi dngos po lta bu bsgom pa dmigs su med [929] cing yod pa ma yin te
[zhes bya ba ni rnam par shes pavi rnam pa dngos po so sor snang bavi rtog pa ni med do //]
sems kyi rang bzhin yin pavi phyir ro // [zhes pa ni sems nyid so sor snang bavi phyir ro //]
Albeit under the name of Gzhon nu bum pa[vi zhabs], *Kumārakalaśa also
rendered into Tibetan the rather cerebral Srog gi vbyung ba[vi] rgyud, a work that is
contained in the Rnying ma[vi] rgyud vbum.120 He was also part of a family that
appears to have specialized in the Guhyasamājatantra and its related literature. Other
members of this family were Mantrakalaśa, Tilakakalaśa - he was also associated
with the monastery of Bya sa in Central Tibet - Tārakalaśa, and Alaṃkakalaśa - he
was associated with the monastery of Dpal Tshangs pa vbyung ba [?*Śrī
Brahmodaya in ? India or Kashmir]. 121 We learn from the colophon of the
translation of the Guhyasamājatantraṭīkā - it is attributed to a Nāgārjuna - that
Mantrakalaśa was the nephew (dbon po) of Kumārakalaśa who in turn was the son
118 SDE, vol. 34,
119 SKAL, 436/6
no. 3859 [# 3854], 506/7 [Tsha, 280b].
[Ki, 150b]; SKAL1, 921. 928-929
120 SDE, vol. 54, no. 4548, 404/7-409/6 [Nga, 53a-70a].
121 For the notes anent Tilakakalaśa and Alaṃkakalaśa, see, respectively, SDE, vol. 26, no. 1812
[#1810], 42/3 [Di, 145b] and no. 1814 [#1812], 43/2 [Di, 148b]. The co-translator of both men
was Pa tshab Nyi ma grags. In the first, we also learn that a certain Ston pa Dar ma rdo rje had
requested them to translate this little work by Nāgabodhi. Alaṃkakalaśa also worked with Steng
pa Lo tsā ba Tshul khrims vbyung gnas (1107-90). For the latter, see Yonezawa (2016).
38
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
(sras) of Tārakalaśa.122 But this conflicts what we read in the colophon of the
Tibetan translation of Lakṣmi's Pañcakrama commentary, for there it is
Mantrakalaśa who is designated as the son of Tārakalaśa!123 All men are said to be
from the subcontinent (rgya gar), which the canonical colophons often distinguish
from Kashmir (kha che). But the affix -kalaśa occurs in the names of several
Kashmirians in the eighteenth canto of the Kashmirian poet Bhaṭṭa Bilhaṇa's (11th
c.) eulogy of Vikramāditya VI (r. 1076-1126), the ruler of the Western Cālukya
Empire.124 And, finally, -kalaśa is also found among the names of a Dardic people
that still live in what is now the Chitral District of northern Pakistan.125
On the text-historical front, the literature suggests that things were not well
with its transmission. In their translation of Tāranātha’s history, Lama Chimpa and
A. Chattopadhyaya expressed the view that the first eight chapters of this text were
written by Āryadeva II and that thus only chapters nine to seventeen had come
from his pen.126 I do not know whence this was taken, but it certainly is not what
we find mentioned in this work at the end of its eighth or seventeenth chapter!
Writing in 1743, Zhu chen Tshul khrims rin chen (1697-1774) states in his
catalogue of the 1744 Sde dge xylograph of the Tanjur that some unidentified
scholars had suggested that it was the next work, the *Pradīpoddyotanaṭīkā, a
commentary on the Pradīpoddyotana's first chapter, that had been authored by
perhaps Āryadeva II; the passage reads127:
sgron ma gsal ba zhes bya bavi levu dang povi vgrel bshad slob dpon ārya de was vphags pa
lhas mdzad zer ba / vdi bod ma vdra bar vdug navang sngar gyi rnams kyis bris vdug pas
bzhugs su bcug pa yin /
It is alleged that Master Āryadeva had written a commentary on the
Pradīpoddyotana's first chapter. Even if it appears to be a Tibetan work (bod ma), I
included it, since it was written by ones of yore.
122 SDE,
123 SDE,
vol. 25, no. 1786 [# 1874], 253/6 [Sa, 283b].
vol. 26, no. 1844 [# 1842], 248/7 [Chi, 277a].
124 Banerjee and Gupta (1965: 284, 290 ff.); I owe this reference to my colleague Michael Witzel.
125 See the Wikipedia article 'Kalash People'.
126 Chimpa and Chattopadhyaya (1980: 326, n. 7).
127 Bstan vgyur dkar chag, ed. Blo bzang bstan vdzin and Don grub phun tshogs (Lhasa: Bod ljongs
mi dmangs dpe skrun khang, 1985), 662; for the entire text, see SDE, vol. 25, no. 1796 [# 1794],
522/1-536/3 [Khi, 155a-205a].
Journal of Tibetan and Himalayan Studies
39
Be this as it may, the earliest catalogs of translated scripture, namely, those by
Dar ma rgyal mtshan and his disciple Dbus pa Blo gsal Rtsod pavi seng ge (ca.
1255-?) do not register it.128 And the catalogs of Bu ston and Ngor chen do not at
all suggest that there was anything amiss with the integrity of Bhavyakīrti's work, let
alone the question of its authorship in toto.129 On the other hand, Bu ston does state
in his chronicle of the Guhyasamāja precepts that the Āryadeva in question might
have been Rngog Āryadeva (13th c.).130
A few further points: Contrary to the first two works with which he is
associated, Bhavyakīrti cites a variety of sources in his *Sandhyāprakāśikā. Among
these, are quotations from works he attributes to Śāntideva (8th c.) and [a] Kambala
[?or Prajñāmitra] that I have not yet been able to identify.131 Further, he explicitly
quotes Dharmakīrti's (7thc.) Pramāṇavārttika, Āryadeva I's (ca. 3rd c.) Catuḥśataka,
Nāgārjuna II's Bodhicittavivaraṇa and the Pañcakrama, and of course, with the greatest
frequency, the Ālokamālā.132 The first four works are the only non-tantric texts in
his entire work to which he expressly refers by title or author. But he also cites
from other non-tantric treatises, providing neither their titles nor the names of their
authors. In addition, he also cites such authors of tantric texts as a Padmavajra, Rin
chen tā/ta la, Dpal gyi sde [*Śrīsena], Dpal Rdo rje hūṃ mdzad [*Śrī
128 See, respectively, Schaeffer and van der Kuijp (2009: 209 ff.) and Bstan bcos gyi dkar chag, eightyone folio dbu med manuscript, China Nationalities Library, Cultural Palace of Nationalities, Beijing,
catalogue no. 002376, 18a.
129 See the Bstan vgyur gyi dkar chag yid bzhin nor bu dbang gi rgyal povi phreng ba, The Collected Works of
Bu ston [and Sgra tshad pa] [Lhasa xylograph], part 26 (New Delhi: International Academy of
Indian Culture, 1971), 463; see also the Bstan bcos vgyur ro vtshal gyi dkar chag yid bzhin gyi nor bu rin po
chevi za ma tog written by a triumvirate of scholars in The Collected Works of Bu ston [and Sgra tshad
pa] [Lhasa xylograph], part 28 (New Delhi: International Academy of Indian Culture, 1971), 402403, and Ngor chen's Rdo rje theg pavi bstan bcos vgyur ro vtshal gyi dkar chag, Sa skya pavi bkav vbum,
comp. Bsod nams rgya mtsho, vol. 10 (Tokyo: The Toyo Bunko, 1969), no. 157, 353/3.
130 Dpal gsang ba vdus pavi rgyud vgrel gyi bshad thabs kyi yan lag gsang bavi sgo vbyed, 72.
131 For example, the unidentified quotation from Śāntideva reads in SKAL 448/1 [Ki, 190a]; SKAL1,
1020: nyon mongs yul la med cing dbang povi tshogs lavang med la bar yang mi gnas te / de las gzhan du ma yin
gar gnas von kyang de yis vgro ba thams cad bcom par byed / vdi ni sgyu ma yin gyi de phyir sems kyi vjigs pa
vdor zhing rtsol ba brten pa gyis // shes rab kyis ni don du don med pas ni dmyal bar bdag nyid ci phyir gnod
par byed /. Bu ston reports in his study of the transmissive history of the Guhyasamājatantra and its
exegesis that some had suggested that "Śāntideva" was another name for Devākaracandra, the
author of three minor works; see his Dpal gsang ba vdus pavi rgyud vgrel gyi bshad thabs kyi yan lag gsang
bavi sgo vbyed, 76. This name is otherwise not associated with the better known Śāntideva, on
whom see de Jong (1975).
132 See, respectively, SKAL, 406/4 [Ki, 44b]: SKAL1, 666, SKAL, 406/4 [Ki, 44b]; SKAL1, 666 [=
Catuḥśataka, 105], and SKAL, 406/4 [Ki, 44b]; SKAL1, 666. In the first, the citation of
Pramāṇavārttika, I: 98, is correctly sourced - tshad ma rnam vgrel las..., whereas most of the other
citations from this work are not at all so identified.
40
INDO-TIBETAN TANTRIC BUDDHIST SCHOLASTICISM
Vajrahūṃkara].133 Future and more in-depth studies of his work as well as greater
familiarity with Indian Buddhist literature than I possess will no doubt uncover and
identify more of these overt and covert citations.
【to be continued】
133
See, respectively, SKAL, 432/7 [Ki, 137a]: SKAL1, 888; SKAL, 436/7, 464/5 [Ki, 152a, 248a]:
923, 1163; SKAL, 520/7 [Khi, 151a]: SKAL1, 358; SKAL, 521/3 [Khi, 153a]: SKAL1, 361.
SKAL1,