Professional Documents
Culture Documents
Edited by
Roger R. Jackson
Klaus-Dieter Mathes
LEIDEN | BOSTON
Notes on Contributors vii
Introduction 1
Klaus-Dieter Mathes and Roger R. Jackson
2 The Seven Siddhi Texts (Grub pa sde bdun): Remarks on the Corpus
and Its Employment in Sa skya-Bka’ brgyud Mahāmudrā Polemical
Literature 90
Adam C. Krug
4 A Neglected Bka’ brgyud Lineage: The Rngog from Gzhung and the
Rngog pa Bka’ brgyud Transmission 142
Cécile Ducher
5 ’Jig rten gsum mgon’s Dgongs gcig on the Relation between Mahāmudrā
and the Six Yogas of Nāropa 170
Jan-Ulrich Sobisch
10 Assimilating the Great Seal: the Dge lugs pa-ization of the dge ldan bka
’brgyud Tradition of Mahāmudrā 302
Roger R. Jackson
Index 329
1 Introduction
The Grub pa sde bdun or Seven Siddhi Texts is an early corpus of Indian
Mahāmudrā works commonly grouped together with The Six Works on the
Essence (Snying po skor drug) and Advayavajra/Maitrīpa’s Twenty-five Works
on Mental Non-Engagement (Yid la mi byed pa’i chos skor nyi shu rtsa lnga).
Together, these three corpora constitute the earliest known practical canon1
of Indian Mahāmudrā texts.2 The Grub pa sde bdun is claimed by some to be
the very first collection of works on the “System of Mantra,” or mantranaya,
that were composed by the Indian mahāsiddhas, though these claims are con-
tested, and holds the far less controversial status as the textual witness to the
Mahāmudrā lineage transmission from Oḍḍiyāna.3 This essay discusses the
1 For detailed discussions of the formal/practical canon distinction, see the following sources:
Blackburn 1999: 281–310; Blackburn 2001; McDaniel 2008; Stanley 2009.
2 Jackson 2008: 151–184. Roger Jackson identifies these three corpora as an early Indian
Mahāmudrā canon. I prefer to describe these works as a Mahāmudrā practical canon, mean-
ing an institution-specific collection of works that occupy a privileged position as part of
a particular curriculum and that derive their own authority from a demonstrable connec-
tion to the scriptural traditions of the formal canon of Buddhist tantras. Their authority as
canon is a derivative of the formal canonicity attributed to the various textual traditions of
Buddhist Vajrayāna literature, but they carry a far more immediate and practical sense of
authority as part of the curriculum of a specific textual community.
3 A mgon rin po che, ed., Grub pa sde bdun dang snying po skor gsum yid la mi byed pa’i chos
skor, 7.2. The passage is translated in full below. This claim is at least partially confirmed in the
Sanskrit colophon material for Jñānasiddhi, Advayasiddhi, Vyaktabhāvānugatatattvasiddhi,
and the title for Dārikapāda’s Dpal o rgyan nas ’byung ba’i gsang ba chen po de kho na nyid
kyi man ngag, all of which locate the origin of these texts in Oḍḍiyāna. Tāranātha (1575–
1634) challenges the broader claim that the Grub pa sde bdun are the seminal corpus of the
Mantrayāna in his Rgya gar chos ’byung, pointing to a number of faults with this argument
and stating that it relies on questionable associations with these authors and the hagiograph-
ic data from Lakṣmīṅkarā’s Sahajasiddhipaddhati. Lakṣmīṅkarā’s Sahajasiddhipaddhati is a
likely source for the hagiographic material on the Grub pa sde bdun that we find in the first
various Tibetan formulations of the Grub pa sde bdun as a corpus and the role
this corpus played in the Mahāmudrā polemical literature of Sa skya and Bka’
brgyud authors from the fifteenth to seventeenth centuries. It begins with a
brief survey of Tibetan works that list the contents of the Grub pa sde bdun,
followed by a discussion of the significance this corpus holds for interpreting
and translating the contracted compound Grub snying gi skor and its various
permutations. The essay then turns to the employment of the Grub pa sde
bdun in a series of responses to Sa skya Paṇḍita’s statements in section three
of his Distinguishing the Three Vows (Sdom pa gsum gyi rab tu dbye ba) that
the Indian tantras and treatises prohibit conferral of Mahāmudrā instruction
upon a beginner who has not received the three higher consecrations associ-
ated with the doxographical category of bla na med pa’i rnal ’byor gyi rgyud or
“highest yogatantra.”
The chart in table 2.1 presents organizational schemata for the Grub pa sde bdun
in the formal canons of the Bstan ’gyur, the two primary Mahāmudrā practical
canons in which the Grub pa sde bdun serves as a foundational corpus, and a
number of works by various Tibetan authors. The chart demonstrates some de-
gree of fluidity in the enumeration of the Grub pa sde bdun in Tibetan sources,
though it is clear that a core set of texts are consistently identified as part of the
corpus. The lists of the Grub pa sde bdun in the recensions of the Bstan ’gyur are
uniform, and appear alongside texts belonging to its companion corpora,
the Snying po skor drug4 and the Yid la mi byed pa’i chos skor. The Grub pa sde
bdun features prominently in two Mahāmudrā practical canons, the Seventh
Karma pa Chos grags rgya mtsho’s (1454–1506) Phyag rgya chen po’i rgya gzhung
volume of the ’Bri gung chos mdzod. Tāranātha punctuates his critique by showing his own
bias toward the hagiographic traditions associated with the Kālacakratantra, stating, “It is
well-known among scholars that Śrī Dhāṇyakaṭaka was the place where Mantrayāna was
originally preached.” See Chimpa and Chattopadhyaya 1970: 343–345.
4 Only five of the six works in the Snying po skor drug accompany the Grub pa sde bdun
in the various Bstan ’gyur recensions. The text that is omitted in all cases is Āryadeva’s
Cittāvaraṇaviśodhana–nāma–prakaraṇa or Sems kyi sgrib pa rnam par sbyong ba zhes bya
ba’i rab tu byed pa, which is grouped in with the works of the Ārya Guhyasamāja school else-
where in both the Snar thang/Gser bris ma and Sde dge/Co ne stemma of the Bstan ’gyur.
Aside from this omission, all three corpora of the early Indian Mahāmudrā canon are listed
in order as a group in all witnesses of the Bstan ’gyur.
(henceforth Phyag chen rgya gzhung)5 and ’Bri gung Kun dga’ rin chen’s (1475–
1527) ’Bri gung bka’ brgyud chos mdzod chen mo (henceforth ’Bri gung chos
mdzod),6 where its seven texts constitute the better part of the first volume
in each collection. The Phyag chen rgya gzhung and ’Bri gung chos mdzod aug-
ment the Bstan ’gyur’s list of the Grub pa sde bdun with additional “siddhi”
texts and, in the case of the ’Bri gung chos mdzod, exegetical and historical
data on each individual work. The lists of the Grub pa sde bdun from the works
of various Tibetan authors are relatively uniform, with the exception of a list
found in the fifteenth century Sa skya polemicist Dge slong Don yod grub pa’s
(15th century) Mahāmudrā polemical work Sdom pa gsum gyi rab tu dbye ba’i
ṭīka bstan pa’i sgron me7 (henceforth Bstan pa’i sgron me) and an augmented
list found in Paṇ chen Chos kyi rgyal mtshan’s (1570–1662) Dge ldan bka’ brgyud
rin po che’i bka’ srol phyag rgya chen po’i rtsa ba rgyas par bshad pa yang gsal
sgron me (henceforth Yang gsal gron me).
5 Zhwa dmar Mi pham chos kyi blo gros, ed., Nges don phyag rgya chen po khrid mdzod. The
fourteenth Zhwa dmar rin po che compiled and published this thirteen-volume collection of
Indian Mahāmudrā works. The first three volumes are photo reproductions of a 19th-century
Dpal spungs xylograph set of the Phyag chen rgya gzhung compiled by the Seventh Karma
pa and later edited and restored by ’Jam mgon Kong sprul Blo gros mtha’ yas (1813–1899) and
Karma Bkra shis chos dpal (c. 19th century). Details on the compilation and restoration of
these volumes at Dpal spungs can be found in the dkar chag to this collection at the begin-
ning of volume three (hūṃ) in Zhwa dmar rin po che’s 1997 publication; an English summary
of this material can be found in Mathes 2011: 90–93.
6 The current Che tshang rin po che (b. 1946) has recently stated that the ’Bri gung bka’ brgyud
chos mdzod was likely compiled by the ’Bri gung patriarch Kun dga’ rin chen (1475–1527),
which would mean that the ’Bri gung practical canon took shape at roughly the same time
that the Seventh Karma pa compiled his Phyag chen rgya gzhung, or shortly thereafter. Che
tshang rin po che’s statement appears in Mathes 2014: 367.
7 Don yod grub pa’s substitution of Virwapa’s ’Chi med grub pa for either Dārikapāda’s De kho
na nyid man ngag or Yoginī Cintā’s Vyaktabhāvānugatatattvasiddhi is indicative of the ma-
nipulation of the corpus to reflect a particular sectarian identity, in this case including the
siddha author who is the ultimate source of the Sa skya lam ’bras in the list of the Grub pa sde
bdun.
table 2.1 Canonical and extra canonical lists of the Grub pa sde bdun
A B C D E F G H I J K L
Guhyasiddhi X X X X X X X X X X X X
(Padmavajra/Saroruhavajra/
Mahāsukhanātha)
Jñānasiddhi (Indrabhūti) X X X X X X X X X X X X
Prajñopāyaviniścayasiddhi X X X X X X X X X X X X
(Anaṅgavajra)
Advayasiddhi X X X X X X X X X X X X
(Lakṣmīṅkarā)
Gsang ba chen po de kho na nyid kyi X X X X X X X X X X X
man ngag (Dārikapāda)
Sahajasiddhi (Ḍombiheruka) X X X X X X X X X X X X
Vyaktabhāvānugatatattvasiddhi X X X X X X X X X X Xa
(Yoginī Cintā/Cinto/Vilāsavajra)
Lhan cig skyes pa’i grub pa X X X X
(Indrabhūti)
Lhan cig skyes pa’i grub pa gzhung X X X X
’grel (Lakṣmīṅkarā)
Dpal de kho na nyid grub pa X X
(Keralipa)
Thabs dang shes rab rnam par gtan la X X
dbab pa’i grub pa bsdus pa
(Ācārya Camari)
Dpal ’khor lo sdom pa’i snying po de X
kho na nyid grub pa
(Ācārya Jalandra)
’Chi med grub pa (Virwapa) X
Bdag byin gyis brlabs pa grub pa X
(Saraha)
D. ’Bri gung Kun dga’ rin chen, ed., ’Bri gung chos mdzod
E. ’Gos lo tsā ba Gzhon nu dpal (1392–1481), Deb ther sngon po
F. Go rams pa Bsod nams seng ge (1429–1489), Sdom pa gsum gyi rab tu dbye ba’i rnam bshad
rgyal ba’i gsung rab kyi dgongs pa bsal b
G. Dge slong Don yod grub pa (15th century), Sdom pa gsum gyi rab tu dbye ba’i ṭI ka bstan pa’i
sgron me las so thar sdom pa’i rnam bshad
H. ’Bri gung Kun dga’ rin chen, Gsan yig byin rlabs rgya mtsho’i dpal ’bar
I. Padma dkar po (1527–1592), Phyag rgya chen po’i man ngag gi bshad sbyar rgyal ba’i gan
mdzod
J. Paṇ chen Blo bzang chos kyi rgyal mtshan (1570–1662), Dge ldan bka’ brgyud rin po che’i bka’
srol phyag rgya chen po’i rtsa ba rgyas par bshad pa yang gsal sgron me
K. Dkon mchog bstan pa’i gron me (1762–1823), Phyag chen khrid kyi zin bris zhal lung bdud rtsi’i
tshigs phreng
L. Karma Bkra shis chos dpal (19th century), Phyag chen rgya gzhung glegs bam gsum pa’i dkar
chag
a Karma Bkra shis chos dpal’s dkar chag to the Dpal spungs edition of the Phyag chen rgya
gzhung mentions the text by Vilāsavajra and Yoginī Cito as separate works when this is not
in fact the case. The text attributed to Vilāsavajra in the Dpal spungs edition is identical, with
a few minor grammatical variants, to the canonical text attributed to Yoginī Cintā/Cito. This
has led me to believe that Vilāsavajra might be taken as a feminine Vilāsavajrā, which the
Tibetan transliteration of the Sanskrit name does not capture.
Tibetan authors often refer to the Grub pa sde bdun and its companion works
with the compounds Grub snying, Grub snying skor, and Grub snying gi skor.
These compounds are almost universally mistranslated in academic and non-
academic writing as something along the lines of “the essence of attainment,”
and are frequently interpreted as shorthand for Saraha’s dohā trilogy. There
are a number of reasons for this misunderstanding. First, Roerich’s translation
of ’Gos lo tsā ba Gzhon nu dpal’s (1392–1481) Deb ther sngon po consistently
takes the various formulations of the compound Grub snying gi skor as an al-
ternate name for Saraha’s dohā. This identification is not entirely inaccurate,
but it fails to recognize that the various formulations of the compound Grub
snying gi skor signify a much broader set of texts. The confusion may have
resulted from the fact that Saraha’s King, Queen, and People dohā are often
also considered a short Indian Mahāmudrā corpus in their own right referred
to as the Snying po skor gsum, or Threefold Corpus on the Essence. To com-
plicate things further, the first work in the Snying po skor drug is frequently
listed as Saraha’s Dohākoṣagīti, making it even easier to take the term snying
in the compounds Grub snying, Grub snying skor, or Grub snying kyi skor as
a stand-in for Saraha’s dohā trilogy. Modern translators who may have fol-
lowed Roerich and Gendun Chöpel’s identification of the Grub snying skor
with Saraha’s dohā are not alone in this misunderstanding. The ’Bri gung
chos mdzod also falsely associates the compound Sgrub snying [sic] with the
Grub pa sde bdun and Snying po skor gsum. The discrepancy appears in this
case between the initial title page to volume one, which records the subject
matter of the volume as the Sgrub snying [sic], and the second, hand-written
title page, which expands this compound into the full title Grub pa sde bdun
dang snying po skor gsum yid la mi byed pa’i chos skor. The actual content of
the first volume, however, is not Saraha’s dohā trilogy (i.e., the Snying po skor
gsum) but the Grub pa sde bdun, Snying po skor drug, and the Yid la mi byed pa’i
chos skor, along with a number of additional texts. There also appears to be
some precedent for interpreting the second member of the compound Grub
snying skor as an abbreviation for the Snying po skor gsum in the enumera-
tion of these corpora that is found in Paṇ chen Chos kyi rgyal mtshan’s Yang
gsal sgron me.8
Tibetan sources exhibit a fair degree of variation regarding the actual texts
that the Snying po skor drug contains, and the parameters of this corpus seem
to have been less defined than those of the Grub pa sde bdun.9 However, despite
8 My thanks to Roger Jackson for pointing this out. Following his reference to the grub sny-
ing skor as containing “the essence of the class of unsurpassed tantra,” Chos kyi rgyal mt-
shan appears to interpret the compound as a signifier for the Grub pa sde bdun and Snying
po skor gsum. Blo bzang chos kyi rgyal mtshan, Dge ldan bka’ brgyud rin po che’i bka’ srol
phyag rgya chen po’i rtsa ba rgyas par bshad pa yang gsal sgron me, 255.6–256.2: dpal mgon
po bde ba chen po’am | mtsho skyes rdo rje dang grub chen sa ra ha dang | klu sgrub zhabs
dang | ri khrod dbang phyug dang | te lo nā ro mai tri pa sogs ’phags yul gyi grub chen rnams
dang | der ma zad mar mi sgam po pa phag gru sogs bka’ brgyud gong ma rnams kyi bzhed
pa mthar thug pa’i phyag chen kyang rlung dbu mar zhugs gnas thim gsum byas pa las byung
ba’i bde chen gyi ’od gsal yin la | de ni grub pa sde bdun dang | snying po skor gsum gyi brjod
bya’i gtso bo rnal ’byor bla med kyi rgyud kyi rgyud sde rgya mtsho lta bu’i yang snying po’i
snying po yin no |. (“The Mahāmudrā that is the final accepted position of the Mahāsiddhas
of Āryavarta such as Śrī Mahāsukhanātha or Saroruhavajra, Saraha, Nāgārjunapāda,
Śabarīśvara, Telo, Nāro, Maitrīpa, etc., and additionally the Bka’ brgyud patriarchs Mar pa,
Mi la, Sgam po pa, Phag mo gru pa, etc., is the clear light of great bliss that arises from caus-
ing the vital wind to enter, come to rest, and dissolve in the central channel, and that is
the essence of the essence that accords with the ocean of the collection of the unsurpassed
yoga tantra, the principal topic of The Seven Siddhi Texts and The Threefold Corpus on the
Essence.”)
9 Listings of texts that are contained in the Snying po skor drug tend to exhibit a greater degree
of variation, which is probably symptomatic of the broader confusion as to the actual contents
of this corpus. The arrangement preserved in the various Bstan ’gyur recensions includes only
five of the six works – (1) Saraha’s Dohākoṣagīti, (2) Nāgārjunagarbha’s Caturmudrānvaya, (3)
Divākaracandra’s Prajñājñānaprakāśa, (4) Sahajavajra’s Sthitisamāsa, and (5) Kuddālapāda’s
Acintyakramopadeśa – omitting Āryadeva’s Cittaviśodhananāmaprakaraṇa. The Phyag
chen rgya gzhung preserves the following list (following the transliterated Sanskrit titles
for these works as they appear in the 1997 Dpal spungs xylograph reprint): (1) Saraha’s
Dohakoṣagīti, (2) Nāgārjuna’s Caturmudrānvaya, (3) Āryadeva’s Cittāvaraṇaviśodhananām
the fact that there are varying interpretations as to the identity of the second
member of the compound Grub snying gi skor, it can be still noted with rela-
tive certainty that the term grub as it appears in this compound signifies the
Grub pa sde bdun. It can also be said with some certainty that the compound
Grub snying gi skor and its various formulations are not meant to exclusively
signify Saraha’s dohā. This corrective allows for a richer understanding of the
transmission of the Grub pa sde bdun and the early Indian Mahāmudrā works
in Tibetan historical sources, particularly in the accounts of its transmission in
the Deb ther sngon po.
The Grub pa sde bdun plays an integral role in a volley of polemical works
composed by a handful of Sa skya and Bka’ brgyud authors from the fifteenth
through seventeenth centuries. The first two polemical works from the Sa
skya side of this debate are roughly contemporary to the publications of the
Phyag chen rgya gzhung and the ’Bri gung chos mdzod, while the rebuttals from
the Bka’ brgyud side post-date the publication of both of these Mahāmudrā
practical canons. The conversation among these authors thus provides some
evidence for the measurable effect that these two publication projects had on
Bka’ brgyud Mahāmudrā polemical literature as reference works that provided
an easily accessible practical canon of authoritative Indian sources to both jus-
tify and defend the Bka’ brgyud Mahāmudrā traditions from their detractors
from among the Sa skya and elsewhere. On the whole, the degree of detail with
which our Bka’ brgyud authors discuss the Grub pa sde bdun indicates that
they were more engaged with these works than the Sa skya authors to whom
they are responding. Judging from these sources it is possible to say, with some
degree of caution, that during this period the Grub pa sde bdun held greater
influence over the Bka’ brgyud Mahāmudrā curriculum than it did among the
Sa skya, and that the publication of the Phyag chen rgya gzhung and the ’Bri
gung chos mdzod Mahāmudrā practical canons likely played a part in making
this so.
The majority of passages that draw upon the Grub pa sde bdun in the set
of Mahāmudrā polemical works analyzed here revolve around the following
statements from Sa skya Paṇḍita’s Sdom pa gsum gyu rab tu dbye ba (hence-
forth Sdom gsum rab dbye) section 3.176–179:
The King of tantra texts and major commentarial treatises also prohibit
The Great Seal to one who is unconnected with initiation.10
The primary function the Grub pa sde bdun plays in the polemical thread stem-
ming from Sa skya Paṇḍita’s Sdom gsum rab dbye revolves around the issue
of whether or not a necessary and exclusive relationship obtains between the
realization of Mahāmudrā and a disciple’s progression through the two-stage
yoga and system of four tantric consecrations associated with the textual genre
of “unsurpassed yogatantra.”11 Following a discussion of the context in which
10 I have retained Rhoton’s translation choices here to acknowledge his important work
in bringing an English translation of this text to publication. See Sakya pandita 2002:
119. The passage is located in Sa skya Paṇḍita Kun dga’ rgyal mtshan, Sdom pa gsum gyi
rab tu dbye ba, 53.5–54.1. Elsewhere in this essay, when this verse is quoted in another
work, I have relied upon my own translation, which the reader may note differs from
Rhoton’s.
11 The Grub pa sde bdun are also evoked at times to address some of Sa skya Paṇḍita’s other
accusations agains the Bka’ brgyud system of Mahāmudrā. For example, Dwags po Bkra
shis rnam rgyal’s (1512/13–1587) Phyag chen zla ba’i ’od zer invokes the corpus to push back
the corpus is most often invoked, this section analyzes passages from a handful
of authors on both sides of this Mahāmudrā debate who draw upon the Grub
pa sde bdun to support their respective positions on the relationship among
the standardized system of four tantric consecrations, realizing Mahāmudrā,
and the conferral of the guru’s blessing.
The progression of works addressed here begins in the fifteenth century
with the Sa skya pa authors Dge slong Don yod grub pa and Go rams pa Bsod
nams seng ge. It then moves to ’Brug chen Padma dkar po (1527–1592), who
is in turn challenged by the Sa skya author Mang thos Klu sgrub rgya mtsho
(1523–1596). The progression through these works culminates in a response to
Mang thos Klu sgrub rgya mtsho from Padma dkar po’s disciple Sangs rgyas rdo
rje (1569–1645). Two of these works, Go rams pa’s Sdom pa gsum gyi rab tu dbye
ba’i rnam bshad rgyal ba’i gsung rab kyi dgongs gsal ba (henceforth Dgongs gsal
ba) and Padma dkar po’s Phyag rgya chen po’i man ngag gi bshad sbyar rgyal
ba’i gan mdzod (henceforth Gan mdzod) remain integral to Sa skya and Bka’
brgyud curricula, respectively, to this day.
Dge slong Don yod grub pa’s three-volume set of commentaries to Sa skya
Paṇḍita’s Sdom gsum rab dbye mentions the Grub pa sde bdun on two sepa-
rate occasions. The first appears in his commentary to Sdom gsum rab dbye
1.244–245, in which Sa skya Paṇḍita criticizes those who say it is not necessary
to study scriptures and treatises.12 Both Dge slong Don yod grub pa and Go
rams pa identify Zhang tshal pa (1121/23–1193) as the intended target of this
verse, and both authors’ comments on the verse echo Sa skya Paṇḍita’s Thub
pa’i dgongs gsal, which includes the Grub pa sde bdun among a short list of
treatises that are integral to studying the system of Mantra.13 Don yod grub pa
against accusations from Sa skya Paṇḍita and others that Bka’ brgyud Mahāmudrā is a
Chinese doctrine in Indian garb by mentioning the Dohā skor gsum, along with the Grub
snying and Maitrīpa’s “Amanasi ” texts, as corpora of authentic Indian origin that validate
and promote subitist aproaches to non-conceptual meditation. Dakpo Tashi Namgyal
2006: 104. Like the authors examined here, Dwags po Bkra shis rnam rgyal also refers to
the Grub pa sde bdun when he argues against Sa skya Paṇḍita’s rejection of a Mahāmudrā
that is taught outside of the tantras. Although the topic of subitism is not entirely absent
from the Grub pa sde bdun, references to the corpus from the Sa skya and Bka’ brgyud au-
thors examined here tend to be employed in the context of determining the relationship
between Mahāmudrā instruction and the process of consecration.
12 Sa skya Paṇḍita, Sdom gsum rab dbye, 27.5–27.6. Tibetan: kla la rdzogs pa’i sangs rgyas kyi
|| gsung rab tshig don zab po dang || grub thob rnams dang mkhas rnams kyi || shin tu legs
par bshad pa’i chos || ’tshig gi na ya yin pas na || dgos pa med pas dor zhes zer |.
13 Sa skya Paṇḍita Kun dga’ rgyal mtshan, Thub pa’i dgongs pa rab tu gsal ba, 9.5: gsang sn-
gags yin na yang byang chub sems dpa’i ’grel pa gsum mam | grub thob rnams kyis mdzad
pa’i grub pa sde bdun nam | slob dpon rnal ’byor gyi dbang phyug bir+wa pa dang | rgyal
po in+d+ra b+hū ti dang | rdo rje dril bu pa la dogs pas mdzad pa’i bstan bcos khungs nas
and Go rams pa also both employ their own brand of ad hominem polemic in
these passages, with Don yod grub pa accusing Zhang Tshal pa’s statements of
being “nothing but nonsense” (cang la ha la la)14 and Go rams pa informing his
byung ba mnyan dgos te | mdor na sangs rgyas kyi gsungs | sdud pa pos bsdus | grub thob
kyis bsgoms | paṇ+ḍi tas bshad | lo tsā bas bsgyur | mkhas pa rnams la grags pa cig sangs
rgyas kyis bstan pa yin pas de la nyan bshad sgom sgrub byed dgos so | | de rnams las bzlog
pa’i chos gcig byung na zab zab ’dra yang sangs rgyas kyi bstan pa ma yin pas nyan bshad
bsgom bsgrub byar mi nyan no | | legs legs ’dra ba mu stegs dang chos log gzhan la’ang
bdug ste sangs rgyas kyi bstan pa ma yin pas dor la bzhag go |. (“The Secret Mantra, how-
ever, requires that one study The Three Commentaries of the Bodhisattva, The Seven Siddhi
Texts composed by the mahāsiddhas, and the treatises that were composed by the Ācārya
and Lord of Yoga Virwapa, King Indrabhūti, and Vajraghaṇṭāpa, etc., all of which are of
authentic origin. In brief, the Buddha taught, the compilers compiled, the siddhas medi-
tated, the paṇḍitas explain, the lo tsā bas translate, [and all of them] must be called the
wise ones. One must study, explain, meditate, and attain siddhi by means of what was
taught by the Buddha. If there is a single dharma that is different from these, yet is like an
imitation in tidy clothes, because it is not the Buddha’s teaching, it is not fit to be studied,
explained, meditated upon, and accomplished. There are also skilful imitations among
the tīrthikas and others [who practice a] false dharma. Because these are not the Buddha’s
teaching, one should throw them away.”)
14 Dge slong Don yod grub pa, Sdom pa gsum gyi rab tu dbye ba’i ṭīka bstan pa’i sgron me las
so thar sdom pa’i rnam bshad, 421.5–422.2: gnyis pa | sde snod la thos bsam mi dgos par ’dod
pa dgag pa ni | zhang tshal pa’am hwa shang gi ston pa cang la ha la la | rdzogs pa’i sangs
rgyas gi gsung rab tshig don zab mo mdo rgyud rnams dang grub thob rnal ’byor gyi dbang
phyug rnams kyis mdzad pa’i grub pa sde bdun zhes grags te | slob dpon pa d+ma ba dz+ra
gyi gsang ba grub pa | in+d+ra b+h+ū ti’i ye shes grub pa |bir+wa pa’i ’chi med grub pa | ḍoṃ
bi he ru ka’i lhan skyes grub pa | sa ra ha’i bdag byin gyis brlabs pa grub pa | yan lag med pa’i
rdo rje’i thabs dang shes rab gtan la dbab pa grub pa | lak+ṣ mis mdzad pa’i gnyis med grub
pa | rgyan drug la sogs pa’i mkhas pa rnams kyis sde snod kyi don rigs pas shin tu legs par
dpyad cing gtan la phab pa’i chos dbu ma rigs tshogs dang | tshad ma sde bdun sogs tshig gi
na ya yin pas na yang dag pa’i don bsgom pa la de dag dgos pa med pas dar bya yin no zhes
zer ro |. (“Second, ‘Refuting that one does not need to study and contemplate the collec-
tions of scriptures and treatises’: The teaching of Zhang tshal pa or Hwa shang is nothing
but a bunch of blabbering. The system of verses is the systems of the sūtras and tantras of
the profound meaning of the scriptural word of the perfect buddhas, the so-called Seven
Siddhi Texts that were composed by the lords of yoga, the siddhas – Ācārya Padmavajra’s
Guhyasiddhi, Indrabhūti’s Jñānasiddhi, Virwapa’s *Amṛtasiddhi, Ḍoṃbiheruka’s
Sahajasiddhi, Saraha’s Svādhiṣṭhānasiddhi, Anaṅgavajra’s Prajñopāyaviniścayasiddhi,
the Advayasiddhi composed by Lakṣmī – [and] [Nāgārjuna’s] Collection of Works on the
Logic of the Middle Way and [Dharmakīrti’s] Seven Epistemological Works, etc., which
is the teaching of the wise ones such as the six ornaments, etc., who thoroughly ana-
lyzed and correctly determined the meaning of the collection scriptures using logic. But
[Zhang tshal pa and Hwa shang] say that one should throw these out because they are not
necessary!”)
reader that the passage refers to “Zhang Tshal pa and some rag-wearing Bka’
[brgyud pas]” (zhang tshal pa dang | bka’ phyag pa la la |).15
The Grub pa sde bdun is invoked again in volume three of Don yod grub pa’s
commentary to the mantra vow section of the Sdom gsum rab dbye. His addi-
tions to the root text of Sdom gsum rab dbye 3.179 are highlighted in bold in the
following excerpt:
Don yod grub pa follows this passage with the following quotes from the
Hevajratantra and Indrabhūti’s Jñānasiddhi (1.32):
15 Go rams pa Bsod nams seng ge, Sdom pa gsum gyi rab tu dbye pa’i rnam bshad rgyal ba’i
gsung rab kyi dgongs pa gsal ba, 133.5: la la rdzogs pa’i zhes sogs tshigs bcad gsum ste |
zhang mtshal pa dang | bka’ phyag pa la la | rdzogs pa’i sangs rgyas kyi gsung rab sde snod
gsum dang | rgyud sde bzhis bsdus pa’i tshig don zab mo rnams dang | de dag gi dgongs ’grel
grub thob rnams kyis legs par bshad pa’i grub pa sde bdun dang | snying po skor drug la
sogs pa rnams dang | mkhas pa rgyan drug la sogs pa rnams kyis shin tu legs par bshad pa’i
chos sa sde dang | rigs tshogs la sogs pa rnams ni tshig gi na ya sogs so |. (“The three verses
[beginning with] ‘Some [say] the Perfectly,’ etc., [refer to] Zhang Tshal pa and some rag-
wearing Bka’ [brgyud pas]. The three baskets of the perfect Buddha’s teachings and the
profound meanings of the verses contained in the four classes of tantra, The Seven Siddhi
Texts and The Sixfold Corpus on the Essence, etc., which are the accurate explanations by
the siddhas who commented on the meaning of those works, and the dharma that was
exceedingly well explained by the wise ones who are the six ornaments etc., works such
as The Collection on the Stages and The Collection of Logical Arguments, etc., is the textual
system.”)
16 Dge slong Don yod grub pa, Sdom pa gsum gyi rab tu dbye ba’i ṭI ka bstan pa’i sgron me las
sngags sdom pa’i rnam bshad, 571.1–572.2: rgyud kyi rgyal po gzhan kye rdo rje sogs dang |
bstan bcos chen po gzhan grub pa sde bdun sogs las kyang | dbang bskur dag dang ma ’brel
ba de la phyag rgya chen po rtogs pa bkag ste |. My translation differs from that of Rhoton.
The divergence is based on Don yod grub pa’s gloss of the verse de la phyag rgya chen po
bkag as de la phyag rgya chen po rtogs bkag ste, which I believe pushes the la don particle
toward the possessive sense, and in turn the verb bkag to its more common connotation
of “to refute.” Rhoton’s translation might be amended here to “Other King of tantra texts
and great treatises also refute that one who is not endowed with the consecration pos-
sesses Mahāmudrā.” I believe this actually captures the point of the verse more accurately.
And in response,
The innate is not expressed some other way.
It is not attained somewhere [else].
It shall be understood based on one’s merit
And the teaching of method during the guru [offering]. (HT 1.8.36)17
As it says in Jñānasiddhi,
By attaining true supreme gnosis
That is devoid of all conceptual thought,
One who receives the vajra gnosis consecration
Shall attain the supreme siddhi. (JS 1.32)18
Both passages are cited here to support the Sa skya view that in order to at-
tain the mahāmudrāsiddhi, the disciple’s own cultivation of non-dual gnosis
through practicing the two-stage yoga must then be joined with the guru’s
consecration and blessing. The point is made in contrast to one Bka’ brgyud
approach to Mahāmudrā that identifies the guru’s blessing as the primary
determinant of any disciple’s realization of Mahāmudrā. This implies, as the
Bka’ brgyud would like to argue, that as long as the disciple receives the proper
blessing from the guru, the sequence of four consecrations and their atten-
dant moments and levels of joy as systematized in the Hevajratantra might
be abridged or done away with entirely. Thus the Bka’ brgyud argument leaves
room for the potential conferral of Mahāmudrā upon someone “who does not
have the consecrations,” while the Sa skya pa approach draws a more system-
atized and necessary relationship between the disciple progressing through
the “proper” consecration sequence in tandem with their generation of gnosis
through the stages of the tantric yogas and their eventual realization of the
mahāmudrāsiddhi through combining this meditative insight with the guru’s
blessing. The Sa skya position that Don yod grub pa presents here, following
Sa skya Paṇḍita, thus limits its understanding of an effective method for the
realization of Mahāmudrā to those systems that are contained within the class
17 H T 1.8.36 is also the verse that ’Bri gung patriarch ’Jig rten mgon po (1143–1217) uses to
identify guru devotion as the single means for realization, a point with which Sa skya
Paṇḍita takes issue in the section of the Sdom gsum rab dbye that immediately follows the
current passage. See Sobisch 2011: 225.
18 Don yod grub pa, Sngags sdom pa’i rnam bshad, 571.2–571.4: brtag gnyis las | de nas rnal
’byor ma zhus pa | phyag rgya chen po ji lta bu | zhes pa’i lan du | gzhan gyis brjod min lhan
cig skyes || gang du yang ni mi rnyed de || bla ma’i dus thabs bstan pa dang || bdag gi bsod
nams las shes bya || zhes gsungs so || ye shes grub pa las | rtog pa thams cad rnam spangs
pa’i || ye shes mchog bzang thob pa yi || rdo rje’i ye shes dbang bskur bas || dngos grub mchog
ni sgrub par bya || zhes gsungs so |.
19 On Sa skya Paṇḍita’s view of Mahāmudrā see Stenzel 2014: 199–228.
20 This is my own translation from the Sanskrit, which the Tibetan translation matches quite
well. For the text see David Snellgrove, ed., The Hevajra Tantra, 88–91.
Of course, there is the possibility that Don yod grub pa had a copy of the
Hevajratantra on hand that substituted HT 1.8.36 for the description that we
find in the current canonical edition. But barring this possibility, it seems
strange that an author would leave himself vulnerable to criticism by manip-
ulating such a well-known canonical source to suit his own purposes. After
all, all one would have to do to challenge his argument is to point to this ob-
vious misquote from the Hevajratantra, a potentially embarrassing observa-
tion for a scholar from a tradition in which the Hevajratantra plays such an
important role. It is equally intriguing that none of the Bka’ brgyud authors
who respond to this passage as it is preserved here and in Go rams pa’s com-
mentary to the Sdon gsum rab dbye seem to notice that the passage is blatantly
misrepresented.21
Without Don yod grub pa’s manipulation of the text, the Hevajratantra’s
description of Mahāmudrā in these passages seemingly has nothing at all to
do with a soteriological absolute that is realized through the combination of
the two-stage yoga and fourfold sequence of tantric initiations. Instead, the
actual sequence of verses in the Hevajratantra presents a list of characteris-
tics becoming of an “ideal” or “superior” (i.e., mahā) mudrā or “consort.” For
a tradition that has come to see Mahāmudrā as bearing the single, mono-
lithic meaning of the highest realization, the often-messy reality of the way
in which the term is used across textual traditions would represent a notable
inconvenience. Don yod grub pa effectively sidesteps this inconvenience by
manipulating his source text and substituting a verse that supports the Sa skya
view of Mahāmudrā. This might have implications regarding the anticipated
behavior of the textual community toward whom he directs his three-volume
exegesis of the Sdom gsum rab dbye, providing some indication of the frequen-
cy with which his readers were expected to actually double-check such cita-
tions from canonical works against their original sources. The fact that Don
yod grub pa and others are able to repackage and manipulate their source texts
so easily also tells us something about the priority that Sa skya textual commu-
nities granted to material in their own practical canon over the sources for that
material in the broader formal canon of the Bka’ ’gyur. The perpetuation of this
particular reading of the Hevajratantra’s presentation of Mahāmudrā among
the Sa skya thus functions as a case in point for the formulation of sect- and
institution-specific textual communities in Tibet. It also provides a glimpse
of how the polemical applications of practical canon formation in Tibet can
21 ’Brug chen Padma dkar po points out a similar problem in Sa skya Paṇḍita’s misrepresen-
tation of the verse quoted above from the Caturmudrānvaya. For an extensive treatment
of this topic see Mathes 2016: 309–340. Padma dkar po does not, however, seem to have
noticed this problem in Go rams pa’s commentary to the Sdom gsum rab dbye.
As for the second [topic],22 the ten verses that begin with “Our,” etc., the
first three verses illustrate the cause [of Mahāmudrā], verse four illustrates
the intrinsic essence [of Mahāmudrā], then two verses illustrate the time
that it is attained, then two verses refute the concept [of Mahāmudrā as it
is understood] among others. After that, two verses illustrate the type of
scripture in which one who is intent upon attainment of Mahāmudrā en-
gages. If one wishes to understand the meaning of these verses in detail,
one can understand [this] through The Seven Siddhi Texts that were com-
posed by the Ācāryas who attained the siddhi that is the ultimate realiza-
tion of the entire class of Mahāyoga tantras.23
In line with this reference to the Grub pa sde bdun, Go rams pa’s expansion
of Don yod grub pa’s commentary on Sdom gsum rab dbye 3.179 follows thir-
teen folio sides later, in his section on how the Sa skya Mahāmudrā “is in
accord with other tantras and śāstras” (rgyud dang bstan bcos gzhan dang
mthun pa). Jñānasiddhi 1.32 makes another appearance in Go rams pa’s work,
and he provides a more expansive commentary incorporating quoted mate-
rial from the Saṃpuṭatantra (Saṃ bu ṭi [sic]), the Guhyakośasūtra (Gsang
ba mdzod gyi mdo), and an unnamed work by Āryadeva. He then refers to
22 Being the sub-topic “Our Own Tradition’s Definition of Mahāmudrā” (rang lugs kyi phyag
chen ngos bzung ba).
23 Go rams pa Bsod nams seng ge, Sdom gsum rnam bshad, 209.5–210.1: gnyis pa ni nged kyi
zhes sogs bcu | tshig rkang dang po gsum gyis rgyu bstan | bzhi pas rang gi ngo bo bstan | de
nas gnyis kyis ’grub pa’i dus bstan no || de nas gnyis kyis gzhan du rtog pa dgag | de nas gnyis
kyis phyag chen bsgrub par ’dod pas gang la ’jug pa’i lung bstan no || ’di dag gi don zhib tu
rtogs par ’dod na rnal ’byor chen po’i rgyud sde rnams kyi dgongs pa mthar thug ‘grub pa
thob pa’i slob dpon rnams kyis mdzad pa’i grub pa sde bdun las shes bar bya’o |. For the root
text, see Sa skya Paṇḍita, Sdom gsum rab dbye, 52.3. For a translation of the root text see
Sakya Pandita, A Clear Differentiation, 117.
another work from the Grub pa sde bdun, citing chapter three of Anaṅgavajra’s
Prajñopāyaviniścayasiddhi, on the bodhicittābhiṣeka (byang chub sems kyi
dbang bskur).24 The verse reads:
“Since attaining the supreme siddhi of Mahāmudrā accords with the ve-
hicle of the perfections, since abandoning the obscurations abandoned
24 Anaṅgavajra, Prajñopāyaviniścayasiddhiḥ (Thabs dang shes rab rnam par gtan la dbab pa
grub pa), in Guhyādi–Aṣṭasiddhi–saṅgraha, Skt. 75–77; Tib. 122–128.
25 Go rams pa, Sdom gsum rnam bshad, 223.2–223.4: zhes pa dang | thabs dang shes rab
rnam par gtan la dbab pa grub pa las || bder gshegs gnas kyi dkyil ’khor du || rgyud kyi
lam gyi rjes ’brang nas || mkhas pa gang tshe dbang bskur na || sangs rgyas thams cad
mngon sum yin || dpag med ’jig rten khams dbang phyug || bdag byin brlabs pa’i rim
thob pa || zhes gsungs so ||. Go rams pa’s partial quote of PUVS 3.3 gives the impression
that the subject of the verse is the “lord of infinite world systems” (dpag med ’jig rten
khams dbang phyug) when, in fact, this is the object of the agent, rendered in the in-
strumental case, that appears in the omitted pādas PUVS 3.3cd. The complete set of
Sanskrit verses reads: mantramārgānusāreṇa abhiṣikto yadā budhaḥ | pratyakṣaṃ
sarvabuddhānāṃ maṇḍale sugatālaye (3.2) anantalokadhātvīśo grāhas tathāpi dhīmatā |
svādhiṣṭhānakramaṃ prāpya samayakṣatibhīruṇā (3.3). (“According to the custom of the
mantra path, when the wise one was consecrated, [he was] in the presence of all of the
buddhas in the maṇḍala, the abode of the Sugatas. (3.2) Likewise, after an intelligent one
who fears losing samaya attains the self-consecration stage, they too will perceive the
lord of Infinite world systems.”) See Anaṅgavajra, Prajñopāyaviniścayasiddhiḥ, Skt. 75–77;
Tib. 122–128.
[on the path of] seeing accords with the secret mantra [vehicle], it un-
ties the knot of the central channel.” Such talk is senseless babbling.
The critical point of the texts referenced above [is expressed in Sa skya
Paṇḍita’s verse that reads] “Here it is refuted that someone not endowed
with the consecrations has Mahāmudrā.” This verse explains that there
is no Mahāmudrā in the vehicle of the perfections because such a siddhi
contradicts the exegetical tradition.26
Go rams pa’s final statement on the Grub pa sde bdun as an authoritative cor-
pus argues that they provide irrefutable evidence that Mahāmudrā cannot be
properly taught or fully realized without the disciple’s progression through the
tantric yogas and the series of consecrations. The Bka’ brgyud side of this argu-
ment, however, employs the very same references from the Grub pa sde bdun
to argue precisely the opposite.
The references to the Grub pa sde bdun from Sa skya Paṇḍita, Don yod grub
pa, and Go rams pa exhibit a trend toward greater exegetical engagement with
the texts this corpus contains, yet this engagement remains relatively limited.
For these authors, it would seem that the fact that the Grub pa sde bdun sup-
port the Sa skya position on Mahāmudrā is largely self-evident. A few verses
are cited, but the reader is for the most part instructed to read these works
on their own, and as the example of Don yod grub pa and Go rams pa’s treat-
ment of HT 2.8.1 indicates, it is quite possible that their readers did not in fact
explore the Grub pa sde bdun any further than the verses provided by this com-
mentarial tradition – nor were they expected to.
In contrast, Padma dkar po’s Gan mdzod begins with detailed descriptions of
each work contained in the Grub pa sde bdun. Writing nearly a generation after
the publication of the Seventh Karma pa and Kun dga’ rin chen’s Mahāmudrā
practical canon projects, a period that marks the height of the Grub pa sde
bdun’s popularity as a corpus of authentic Indian Mahāmudrā teachings,
Padma dkar po’s Gan mdzod devotes thirty folio sides in the beginning of its
first section to “A Detailed Analysis of Mahāmudrā Texts” (gzhung phyag rgya
chen po’i rab dbye) that focuses on the three core Indian Mahāmudrā corpora.
He organizes his analysis according to the ’Brug pa Bka’ brgyud patriarch Chos
kyi Gtsang pa rgya ras pa’s (1161–1211) three categories of supplemental works
26 Go rams pa, Sdom gsum rnam bshad, 223.4–223.5: kha cig ’dir | phyag chen mchog gi dngos
grub thob pa ni | phar phyin theg pa bltar na | mthong spang gi sgrib pa spangs ba dang |
gsang sngags pa ltar na rtsa dbu ma’i mdud pa grol ba la zer zhes smra ba ni bab chol te |
gong du drangs pa’i lung rnams dang | ’dir dbang bskur dag dang ma ’brel ba || de la phyag
rgya chen po bkag | zhes phar phyin theg pa la phyag rgya chen po med par bshad pa dang
dngos grub ’gal ba’i phyir ro |.
(zur ’debs) for the Bka’ brgyud Mahāmudrā tradition. Gstang pa rgya ras pa’s
first category, “The corpus of textual exegeses” (bshad pa tshig gi skor), includes
the Grub pa sde bdun, Snying po skor drug, and the Yid la mi byed pa corpo-
ra. After drawing attention to the continuity of textual exegesis on the Grub
pa sde bdun in his own lineage, Padma dkar po goes on to discuss all seven
works in the corpus, and provides short chapter-by-chapter explanations of
Guhyasiddhi, Prajñopāyaviniścayasiddhi, and Jñānasiddhi that highlight spe-
cific passages from these texts that are intended to refute the Sa skya position.
While it clearly shows a greater degree of engagement with the texts, Padma
dkar po’s discussion of the Grub pa sde bdun also manipulates its source mate-
rial in certain cases by reading a number of topics into the corpus that are not
present in the original works. His discussion of the Guhyasiddhi argues that the
text contains instructions on “the subitist path” (cig car ba’i lam bstan) as well
as “the path of passing over” (thod brgal ba’i lam) in chapters one and three,
respectively. In his discussion of “the path of passing over,” Padma dkar po pro-
vides what appears to be a doctored quote from Guhyasiddhi chapter three:
27 ’Brug chen Padma dkar po, Phyag rgya chen po’i mang ngag gi bshad sbyar rgyal ba’i gan
mdzod, 9.6–10.1: dang po’i las can sems can rnams || mtshon pas yid ches byed pa po || mkha’
dbyings rdo rje kun sbyor ba’i || reg pas ngo mtshar chen po nyid || gdams pa gang gis ’byung
gyur te || mchog tu dga ba byed pa pa’o |.
28 Padmavajra’s Guhyasiddhi only mentions three types of “joy,” ānanda, paramānanda, and
viramānanda, which would make the text a perfect example of the limitations of a rigid
interpretation of Mahāmudrā as necessarily dependent upon the realization recognized
and cultivated during the experience of the four moments of joy. Padma dkar po, how-
ever, does not make this argument at this point in the text.
29 My English translation preserves the Tibetan syntax in order to make it easier to provide
the verse correspondences for the Tibetan translation of Prajñopāyaviniścayasiddhi.
There is no third consecration in this text. So what are these Ācāryas who
are convinced that this kind of consecration ritual is unacceptable talk-
ing about?30
30 Padma dkar po, Gan mdzod, 13.2–13.5: gsum pas dbang bskur | dbang yang | de nas dpal
ldan slob dpon sogs kyis tshogs dkyil bsgrubs te gsang dbang bskur | de‘i mthar rjes gnang
sbyin pa gsungs te | byang chub sems kyi dbang bskur bas | slob ma sdig dang bral bar ’gyur |
| sangs rgyas sras mchog de la ni | | rjes su gnang ba de nas sbyin | | zhes dang | phyis mos na
tshig dbang bskur te | zab cing rgya che ba la lhag par mos na ni | | tshig gis rin chen dbang
bskur sbyin par bya zhes gsungs kyis | ’di la gsum pa‘i dbang ma byung | dbang gi cho ga‘i
’gros ’di lta bu mi ’thad na slob dpon tshad mar gyur pa des ji la gsung |.
31 The Tibetan text diverges from the Sanskrit, which leaves the term jagatprabhuḥ in the nom-
inative singular to match the term ācāryaḥ. See Anaṅgavajra, Prajñopāyaviniścayasiddhiḥ,
Skt. 75.
It is clear in these passages that the disciple is united with the consort, and
that he is consecrated while they are in union. The disciple is also, seemingly
for the first time during the rite, given a mixture of substances to ingest that
signifies his taking of the samaya. Both elements typically associated with
the guhya- and prajñājñāna-abhiṣeka are thus present here, and it is unclear
if the rite prescribes the former, the latter, or a combination of both. What
is clear is that Padma dkar po’s statement that the consecration chapter in
Prajñopāyaviniścayasiddhi does not contain a third consecration glosses over
the complexity of the passage, and it does so to his rhetorical advantage.
The second issue revolves around Padma dkar po’s statement that the chap-
ter contains a “word consecration” (tshig dbang bskur), which would function
here as a fourth consecration representing the full, simultaneous conferral of
32 The Tibetan rendering of this verse might be interpreted as the various substances men-
tioned here arising from the “pure five” aggregates. I have opted to leave out any such
interpretation because this verse actually diverges from the Sanskrit, which reads pañ-
camam vāksamudbhavam or “the fifth, which is arisen from speech.” See Anaṅgavajra,
Prajñopāyaviniścayasiddhiḥ, Skt. 76.
33 Anaṅgavajra, Prajñopāyaviniścayasiddhiḥ, Tib. 125: de nas slob dpon skal bzang gis || phyag
rgya dang ni sbyar byas te || rgyal ba’i gnas gyur byang chub sems | (20) || pad+ma’i snod du
bzhag nas ni || bkra shis glu yi tshigs bcad dang || gdugs dang rnga yab rgyal mtshan bcas ||
phyag rgyar ldan pa’i slob ma ni | (21) || ’gro ba‘i gtso bor dbang bskur bya || slob dpon dbang
phyug mchog gis ni || dbang bskur rin chen byin nas su || rang bzhin gsal zhing mngon
sbyangs pa | (22) || dam tshig nyams dga‘ sbyin par bya || rin chen chen po ga bur bcas || tsa
n+da na dmar po sbyar ba dang || rdo rje yi ni chu dang bcas | (23) || lnga po dag las yang
dag byung || ’di ni bu khyod dam tshig ste || sangs rgyas kun gyi mthun par gsungs || bzang
pos rtag tu bskyang bar byos | (24) || da ni sdom pa mnyan par gyis |.
a final Mahāmudrā instruction and realization. The term tshig, however, does
not appear as a modifier for the consecration itself, but as an adverbial form
describing the verbal expression of the consecration according to the rituals
described previously in the chapter. The Sanskrit verse reads:
One can imagine Padma dkar po’s temptation to read this as a clear exam-
ple of the guru’s imparting a “word consecration,” and by association a final
Mahāmudrā instruction that is bestowed upon the disciple in the absence of a
third consecration. The problem is, just as it is somewhat unclear whether or
not there is a third consecration in the chapter, it is also not entirely clear that
the verse in question constitutes a true “word consecration.” The two issues
34 Padma dkar po’s quote reads tshig gis, but the Beijing and Snar thang Bstan ’gyur both
read tshig gi, as does the Sarnath edition of the Tibetan text. The witnesses from Sde
dge, Co ne, and Padma dkar po, which all read an instrumental particle here, match the
vācaiva in the extant Sanskrit version of the text.
35 Anaṅgavajra, Prajñopāyaviniścayasiddhi, Skt. 77; Tib. 127. My translations of the Sanskrit
verse follows the variants for PUVS 3.38 that are noted in the Sarnath edition that read
udāragambhīranayādhimukto instead of udāragambhīranayādhimukta–.
are in fact related – both revolve around the absence of a clear and standard-
ized vocabulary for the sequence of consecrations in the text. This ambiguity
highlights another point at which an important aspect of the commentator’s
own tradition has been read into his sources with a degree of certainty that
is not borne out in the source material itself. It should also be noted that in
his subsequent comments on Prajñopāyaviniścayasiddhi, Padma dkar po ad-
mits to reading his own tradition into the text when he presents Anaṅgavajra’s
chapter on “Meditation on Ultimate Reality” (tattvabhāvanā, de kho na nyid
bsgom pa) as a teaching on the Bka’ brgyud Mahāmudrā system of four yogas
(rnal ’byor bzhi). In this case, he openly states that “chapter four does not men-
tion the names of the four yogas, but it teaches [them] according to [their]
meaning.”36 Importantly, this interpretation also rules out the possibility
that the material in chapter four is intended as the “word consecration” that
Padma dkar po identifies at the end of chapter three. This leaves two possi-
bilities for the potential inclusion of a “word consecration” in chapter three of
Prajnopāyaviniścayasiddhi – either this “word consecration” appears in liturgy
itself in the form of the “command” or rjes gnang that is imparted following the
consecration, or it is not included in the liturgy for chapter three but merely
implied in PUVS 3.38d. The former position does not make sense because the
guru’s “command” in this text is not a final instruction on the nature of reality
or Mahāmudrā. If the liturgy for this “word consecration” is merely implied
in PUVS 3.38d, then Padma dkar po’s entire argument rests on a single phrase
(tshigs gis, vācaiva) employed in a single verse for which there is no clear refer-
ent in the chapter itself. The latter option would seem to be relatively flimsy
grounds from which to make such an important claim.
Padma dkar po, likely prompted by Don yod grub pa and Go rams pa, also
cites Jñānasiddhi 1.32 in his discussion of the role of consecration, treating
Jñānasiddhi 1.32–33 and 1.37 as a brief set of summary verses corresponding to
the lengthy consecration liturgy that Indrabhūti provides later in chapter sev-
enteen. Here he refers to the consecration chapter in Jñānasiddhi as a “blessing
ritual” (byin rlabs kyi cho ga) and elaborates upon these verses with material
from Jñānasiddhi’s consecration chapter to argue that the Grub pa sde bdun
support the view that the guru’s blessing can perform the same function as
a complete set of consecrations. The ritual elements of the chapter include
the performance of a feast offering, the disciple’s offering a consort (dakṣinā)
to the guru, the return of the consort along with the guru’s blessing, and fi-
nally the guru’s command (anujñā, rjes gnang). Toward the end of Jñānasiddhi
36 Padma dkar po, Gan mdzod, 13.5–13.6. Tibetan: le’u bzhi bas rnal ’byor bzhi’i ming ma bshad
kyang don ji lta ba bshed de |.
37 As mentioned above, this terminology does not appear in the text itself.
38 Padma dkar po, Gan mdzod, 18.1–3: zhes byin rlabs kyi cho ga kho nas dbang thob pa de
rgyud thams cad kyi rdo rje slob dpon du ’os pa sogs gsungs pa’ang mthong || gzhung ’di
tshad mar yang khas len bzhin du byin gyis rlabs pas dbang bskur gyi go mi chod zer ba de
rgan ’chal kho nar zad do |.
39 The title of this text may also be taken as a veiled attack on Dwags po Bkra shis rnam
rgyal’s famous treatise on Mahāmudrā, the Zla ba’i ’od zer.
the power and ability to cause the precious teachings to wax or wane. The
great saint who has attained siddhi, who possesses the fortunate name
Padma dkar po, has composed a treatise called The Victor’s Treasury: A
Cohesive Exegesis of Mahāmudrā Instructions, in which, in order to test
the deluded scholars among the followers of the glorious Sa skya pa of
this time, he criticizes [them] with degrading words and levels numer-
ous responses and refutations. This is appropriate for a scholar, and is
the foundation of analytical logic. I have obtained permission to respond
in this work, so it is appropriate that it should be given to discerning
scholars.40
He then comments on the role of consecration in the Grub pa sde bdun and
challenges Padma dkar po’s reading of both Jñānasiddhi 1.32 and chapter 17. He
quotes Padma dkar po’s ad hominem polemic against the “childish and senile”
who argue that the guru’s blessing alone cannot perform the function of a full
sequence of consecrations, and refers to this statement as “just the senseless
babbling of someone poorly trained who was overwhelmed upon seeing the
true profundity of the tantra with the discriminating eyes of a mentally chal-
lenged fool (byis pa blo gros ma smin pa).”41 Klu grub rgya mtsho argues that
the blessing ritual in chapter seventeen is explicitly designated for a disciple
who has already been ripened through consecration and has already generated
gnosis on his or her own, in contrast to Padma dkar po, who argued that the
chapter is itself a rite for the performance of a ripening consecration.42 For Klu
grub rgya mtsho, the presence of this ripening consecration at the beginning of
Jñānasiddhi chapter seventeen, which he locates in JS 17.4–5, indicates that the
40 Mang thos Klu sgrub rgyal mtshan, Sdom gsum rab dbye’i dka’ ’grel sbas don gnad kyi sny-
ing po gsal byed las | phyag chen rtsod spong skabs kyi legs bshad nyi ma’i ‘od zer, 116.3–116.6:
dpal ldan ’brug pa sprul sku zhes snyan pa’i grags pa phyogs kyi mkhas pa rnams kyi rna
ba’i rgyan du gyur la | snying la rab dga’i bdud rtsis gsos ’debs pa | gang du sku’i snang
brnyan shar ba dang ma shar ba las | bstan pa rin po che la ’phel ’grib kyi rngo thogs par nus
pa’i mkhas shing grub pa brnyes pa’i skyes chen pad+ma dkar po zhes mtshan gyi dge legs
dang ldan pa des | phyag chen man ngag gi bshad sbyar rgyal ba’i gan mdzod ces pa’i bstan
bcos brtsams pa der | dus deng gi dpal ldan sa skya pa’i rjes ’brang dag la mkhas rmongs
kyi nyams sad pa’i phyir | nyams ldan gyi gsung gis sun ’byin dang | ’gog byed lan gyi rnam
grangs mang du gnang ba ni | mkhas pa la ’os shing | dpyad par rigs pa’i gzhir gyur la | ’di la
lan du bka’i gnang ba yang thob pas | mkhas pa dpyod ldan dag gsan par bya ba’i ’os so |.
41 Mang thos, Nyi ma‘i ’od zer, 117.3: de skad smra ba de ni | byis pa blo gros ma smin pa‘i rnam
dpyod kyi mig gi rgyud don zab mor lta ma bzod pa‘i bslab nyes kyi bab col kho nar zad do |.
42 Padma dkar po, Gan mdzod, 16.3–16.4: bcu bdun par thog mar smin pa’i dbang dgos pa
bskur ba’i tshul ’di ni mdor bstan du |. Translation: “The way that the necessary ripening
consecration is conferred is taught at the beginning of chapter seventeen.”
The Indian [master] Pha dam pa’s instructions [on these verses] say,
“The verse that reads ‘Oh compassionate one, due to [your] blessing,’ [JS
17.4a] means that the one who requests the consecration only needs to
engage the Vajrācārya. Thus, the disciple says, ‘Compassionate one, due
to [your] blessing’ [in reference to] the [Vajr]ācārya. Among Tibetans it
is said that you ‘attain the authentic supreme gnosis’ [JS 17.4b], and then
‘One gains certainty in the true nature’ with respect to that realization
of ‘the essence of self-reflexive awareness gnosis,’ and [thus the verse in
Jñānasiddhi] says,
One attains the perfect supreme gnosis and
Produces supreme certainty in the true nature
With respect to the essence of self-reflexive awareness. [JS 17.4.bcd]
Since you [i.e. the Vajrācārya] possess ‘this non-dual gnosis,’ [it] ‘does not
exist anywhere else in the world,’ [JS 17.5ab] [meaning among] us [the
43 It is possible that the Tibetan bdag ni reflects the Sanskrit variant *ātmānam api niścayam
for JS 17.4d. This variant is not reported in the Sarnath edition, nor does it appear in the
Sanskrit manuscripts for Jñānasiddhi that I currently have at my disposal (NGMPP A
134/2, A 137/4, E 1474/4, and IASWR MBB 7/4), which are all consistent with the Sarnath
edition.
Sangs rgyas rdo rje then responds to Klu grub rgya mtsho’s reading of the verse
with the following critique:
44 The literal meaning of the phrase ’bru gnyer tshul in this context might read something
like “paying attention to the details.” This translation – suggested to me by Elizabeth
Callahan in a private email correspondence – remains tentative, and further research is
needed on similar uses of the phrase to justify this reading.
45 Sangs rgyas rdo rje, Rgyal ba’i dge mtshan, 121.2–122.3: pha dam pa rgya gar ba’i man ngag
bzhin | thugs rje’i bdag nyid drin can gyis | zhes sogs ni dbang bskur ba po rdo rje’i slob dpon
kho na la sbyor dgos te slob mas | thugs rje’i bdag nyid drin can zhes slob dpon la bod nas |
khyod kyis yang dag pa’i ye shes mchog thob nas so sor rang rig pa’i ye shes kyi ngo bo rtogs
pa de la bdag nyid nges pa’i shes pa skyes so zhes ’chad pa la | yang dag ye shes mchog thob
ste || rang gi rig pa’i ngo bo la || bdag ni shin tu nges pa skyes | zhes pa ’di byung | khyod kyi
de ltar rtogs pa’i gnyis med kyi ye shes ’di nyid ni bdag cag ’gro ba gzhan la yod pa ma yin gyi
| da ni chos kyi bdud rtsi ’thung ba’i phyir mchog gi bla ma khyod la gsol ba ’debs so || chos
kyi bdud rtsi bgo bshar mdzad pa’i slad du byin rlabs dbang bskur stsal du gsol zhes ’chad pa
la | gnyis med ye shes ’di nyid ni || ’gro ba gzhan la yod ma yin | chos kyi bdud rtsi ’di ’thung
phyir | bla ma mchog la gsol ba ’debs | zhes ’bru gnyer tshul don dang mthun pa ’di ka’o || de
ltar ma yin par slob ma la sbyar na | yang dag ye shes mchog thob zin nas slar chos kyi bdud
rtsi’i phyir du gsol ba btab nas | ci zhig don du gnyer | ’dir rang gi rig pa zhes dang | bdag ni
zhes pas ’khrul gzhi byas nas go log rgyab par ’dug kyang rdzob rtags dum tsam ’brog pas los
ston | thugs bden mchis so |.
In other words, Sangs rgyas rdo rje believes that Klu grub rgya mtsho is mis-
led here by the passage’s reference to ‘self-reflexive awareness’ and the inclu-
sion of a first person pronoun in the Tibetan version of the text. This error
allowed Klu grub rgya mtsho to read this bdag ni as a subject who “has gener-
ated confidence with respect to the nature of self-reflexive awareness,” and to
read the verse in agreement with Sa skya Paṇḍita’s view of Mahāmudrā in his
Sdom gsum rab dbye as “gnosis arisen from initiation and the self-arisen gnosis
that ensues from the meditations of the two processes.”46 This, in turn, allows
Klu grub rgya mtsho to argue that the supplicant in JS 17.4–5 has already re-
ceived a “blessing consecration,” and that the opening supplication constitutes
a request for a “ripening consecration.” Without this variant in the Tibetan
translation of the text, however, it is clear that the disciple is requesting “per-
fect supreme gnosis” and “certainty as to the nature of self-reflexive aware-
ness,” which are both attained “from the blessing” (prasādāt, drin can gis). This
reading would support Padma dkar po and Sangs rgyas rdo rje’s argument that
Jñānasiddhi chapter seventeen preserves evidence from an authentic Indian
Mahāmudrā source that the Vajrācārya’s blessing, in the form of a “blessing
consecration,” can in fact confer a complete realization of Mahāmudrā.
This study of the role that the Grub pa sde bdun played in the works of several
prominent Sa skya and Bka’ brgyud Mahāmudrā polemicists has brought to
light a number of points that are of broader significance for Tibetologists. It
has shown that the corpus of the Grub pa sde bdun exhibited some degree of
fluidity in the hands of various Tibetan authors, with some authors swapping
out members of the standardized list in the Bstan ’gyur for other “siddhi” texts
to bring the corpus closer in line with a particular sectarian identity and others
expanding the list of seven to include a number of additional “siddhi” works.
An important technical issue has been brought to light concerning the transla-
tion of the compounds Grub snying, Grub snying skor, and Grub snying gi skor,
which is almost universally mistaken to signify Saraha’s dohā when, in fact,
46 Sa skya Paṇḍita, Sdom gsum rab dbye, 52.3; for a translation of the root text see Sakya
Pandita 2002: 117. As mentioned above, Don yod grub pa argued that Sa skya Paṇḍita’s
position is supported in the Grub pa sde bdun without pointing to any particular text or
passage from the corpus to support his argument. Mang thos’ work thus reflects a greater
engagement with the texts contained in the Grub pa sde bdun, and it is likely that this is a
direct function of Padma dkar po’s more detailed engagement with the actual content of
these works.
it most often signifies (with some exceptions) a shortened form for the two
“earliest” corpora of Indian Mahāmudrā works, the Grub pa sde bdun and the
Snying po skor drug. The employment of the Grub pa sde bdun in Sa skya-Bka’
brgyud Mahāmudrā polemical literature from the fifteenth to the seventeenth
centuries has revealed that authors on both sides primarily drew upon the cor-
pus to clarify whether or not a necessary relationship obtains between impart-
ing and realizing the nature of Mahāmudrā and the combination of receiving
the higher tantric consecrations while progressing through the two-stage yoga
of the “unsurpassed yogatantra.” This analysis has also brought to light some
of the ways that polemical authors might twist or manipulate their sources to
support their arguments. It has also demonstrated that these authors’ engage-
ment with the Grub pa sde bdun became increasingly more sophisticated over
time. This pattern, I argue, is likely a result of the increased awareness and ac-
cessibility that the Grub pa sde bdun enjoyed due to its prominent placement
in the first volumes of two Bka’ brgyud practical canons published at the end
of the fifteenth and beginning of the sixteenth centuries, the Phyag chen rgya
gzhung and the ’Bri gung chos mdzod.
For Tibetan authors on both sides of this polemical divide as well as modern
scholars, the task of interpreting the consecration chapters from Guhyasiddhi,
Prajñopāyaviniścayasiddhi, and Jñānasiddhi is complicated by the fact that
the sequence of the consecration rituals and the terminology used to describe
them varies across all three works. This is the case despite assurance from
hagiographic sources that these three works represent a single Mahāmudrā
lineage transmission from Oḍḍiyāna. In addition, none of these works em-
ploys a consecration terminology that matches the more standardized lexicon
for the three higher consecrations – the guhya-, prajñājñāna-, and caturtha-
abhiṣekas.47 The lack of a standardized and consistent consecration-ritual se-
quence and lexicon across these three works undoubtedly made the job of Sa
skya and Bka’ brgyud Mahāmudrā polemicists that much more difficult. The
fact that the Grub pa sde bdun is widely accepted as an authoritative corpus of
Indian Mahāmudrā works meant that Sa skya and Bka’ brgyud authors were
required to find some way to read aspects of later, more standardized conse-
cration systems into the texts. In doing so, both sides grappled with a corpus
containing a series of somewhat loose internal correspondences around the
critical issue of consecration rites.
47 This fact seems to go unnoticed by both sides of the debate, which is surprising given that
the absence of this common vocabulary could only strengthen the Bka’ brgyud argument
against the more rigid conception of a proper consecration ritual among the Sa skya.
48 On the historical development of the four-stage consecration system, see Dalton 2004,
and Isaacson 2010.
49 This is particularly true with respect to the three texts that feature prominently in the Sa
skya-Bka’ brgyud debates around the relationship between the mechanics of consecra-
tion and realization of Mahāmudrā. None of these three works (Guhyasiddhi, Jñānasiddhi,
and Prajñopāyaviniścayasiddhi) is aware of any category of “unsurpassed yogatantra.” The
Guhyasiddhi is aware of the categories of kriyā and caryā tantra, but Padmavajra does not
provide us with any indication of his understanding of the class to which his primary source
text, the Guhyasamājatantra, belongs. Anaṅgavajra’s Prajñopāyaviniścayasiddhi does not
mention its primary source text, the Samputodbhāvatantra, by name, and Indrabhūti’s
Jñānasiddhi refers to its own textual sources on several occasions as yogatantra, foregoing
even the addition of the modified term mahāyoga despite the fact that it invokes several
texts that would later be categorized as part of this genre.
Guhyasiddhi, the presence of a list of only three types of joy), and the fact that
authors such as Indrabhūti refer to the textual sources for their Mahāmudrā in-
structions as yogatantras, and not mahāyogatantra or “unsurpassed yogatan-
tra,” could only play to the advantage of the Bka’ brgyud position.
Bibliography
Mang thos Klu sgrub rgyal mtshan. Sdom gsum rab dbye’i dka’ ’grel sbas don gnad kyi
snying po gsal byed las | phyag chen rtsod spong skabs kyi legs bshad nyi ma’i ’od zer.
In Mang thos klu sgrub rgya mtsho’i gsung skor 5, no. 3. Kathmandu: Sa skya rgyal
yongs gsung rab slob gnyer khang, 1999.
Sangs rgyas rdo rje. Phyag rgya chen po‘i man ngag gi bshad sbyar rgyal ba’i gan mdzod
ces bya ba‘i bstan bcos la rtsod pa spong ba‘i gtam srid gsum rnam par rgyal ba’i dge
mtshan. In ’Brug lugs chos mdzod chen mo 42, no. 3. Kathmandu: Drukpa Kagyu
Heritage Project, 200?.
Sa skya Paṇḍita Kun dga’ rgyal mtshan. Thub pa’i dgongs pa rab tu gsal ba. In Sa skya
bka‘ ’bum 10, no. 1. Kathmandu: Sachen International, 2006.
Sa skya Paṇḍita Kun dga’ rgyal mtshan. Sdom pa gsum gyi rab tu dbye ba. In Sa skya bka’
’bum 12, no. 1. Kathmandu: Sachen International, 2006.
Zhwa dmar Mi pham chos kyi blo gros, ed. Nges don phyag rgya chen po khrid mdzod.
New Delhi: Rnam par rgyal ba dpal zhwa dmar ba‘i chos sde, 1997.
Secondary Sources
Blackburn, Anne M. 1999. “Looking for the Vinaya: Monastic Discipline in the Practical
Canons of the Theravāda.” Journal of the International Association of Buddhist
Studies 22.2: 281–310.
Blackburn, Anne M. 2001. Buddhist Learning and Textual Practice in Eighteenth Century
Lankan Monastic Culture. Princeton: Princeton University Press.
Chimpa, Lama and Alaka Chattopadhyaya, trans. 1970. [Jo nang rje btsun tā ra nā tha.]
Tāranātha’s History of Buddhism in India. Edited by Debriprasad Chattopadhyaya.
Simla: Indian Institute of Advanced Study.
Dakpo Tashi Namgyal. 2006. Mahamudra: The Moonlight – Quintessence of Mind
and Meditation. Translated by Lobsang P. Lhalungpa. Somerville, MA: Wisdom
Publications.
Dalton, Jacob P. 2004. “The Development of Perfection: The Interiorization of Buddhist
Ritual in the Eighth and Ninth Centuries.” In Journal of Indian Philosophy 32:
1–30.
Isaacson, Harunaga. 2010. “Observations on the Development of the Ritual of Initiation
(abhiṣeka) in the Higher Buddhist Tantric Systems.” In Hindu and Buddhist Initiations
in India and Nepal. Edited by Astrid Zotter and Christof Zotter: 261–279. Wiesbaden:
Harrassowitz Verlag.
Jackson, Roger R. 2008. “The Indian Mahāmudrā ‘Canon(s)’: A Preliminary Sketch.”
Journal of the Indian International Association of Buddhist Studies 9: 151–184.
Mathes, Klaus-Dieter. 2011. “The Collection of ‘Indian Mahāmudrā Works’ (phyag
chen rgya gzhung) Compiled by the Seventh Karma pa Chos grags rgya mtsho.” In
Mahāmudrā and the Bka’-brgyud Tradition. Edited by Roger R. Jackson and Matthew
T. Kapstein, 89–127. (PIATS 2006: Tibetan Studies: Proceedings of the Eleventh