Ruegg - Buddha Mind PDF
Ruegg - Buddha Mind PDF
Ruegg - Buddha Mind PDF
UNIVERSITY OF LONDON
Buddha-nature, Mind
and the Pro bletn of
Gradualistn in a
Cotnparative
Perspective
On the Transmission
and Reception oj
Buddhism in India
and Tibet
by
DAVID SEYFORT RUEGG
":-' .
19 89
Published by
SCHOOL OF ORIENTAL AND AFRICAN STUDIES
UNIVERSITY OF LONDON
CONTENTS
FOREWORD
17
56
93
138
141
147
4.
15 0
5.
6.
7.
I8~
192
ON A BHAVANAKRAMA IN BHAVYA'S
MADHY AMAKARATNAPRADlpA
206
SILENCE
209
INDEX
213
Foreword
COMPARATIVE studies in religion and philosophy have over the
years given rise to a number of questions and problems, and the
very status and validity of the comparative method in these two
fields have indeed often been the object of critical reflection and
interrogation. This has been the case in particular where either
totally different religious or historically independent philosophical traditions were the matter of comparison, even though the
phenomenological method has of course proved to be productive
in religious studies. 1
Somewhat less problematical no doubt is religious and philosophical comparison within a single culture and closely related
traditions. The specialist in Indian religion and philosophy for
example has been accustomed to compare the BrahmanicalJ
Hindu, Jaina and Indian Buddhist traditions which - whatever
their ultimate genetic relationships may be - have clearly followed distinct lines of development. Furthermore, within each of
these three traditions, the Indianist has found it meaningful to
undertake comparisons between separate currents: e.g., to name
only some of the broadest, between Vaiglavism and Saivism,
Svetambara and Digambara, or Sravakayana and Mahayana.
Reference can be made in this connexion to two previous series of
the Jordan Lectures, one by Louis Renou (Religions of Ancient
India, I953) and the other by Jan Gonda (Vi~YJuism and Sivaism,
I970). One form of comparison at least - a basically historical and
textually oriented one - has thus been well-established among
.1 For recent discussions of the notion of comparative religion reference may be made to
E. Sharpe, Comparative religion: a history (London, 1975); and F. Whaling, Contemporary
approaches to the study oj religion, i (Berlin-New York-Amsterdam, 1983), p. 165 If.
Concerning the comparison ofIndian and Western philosophy, reference can still be made
to S. Schayer, 'Indische Fhilosophie als Problem der Gegenwart' in: Jahrbuch der
Schopenhauer-GesellschaJt 15 (Heidelberg, 1928), pp. 46-69, and D. H. H. Ingalls, Journal
of Oriental Research (Madras) 22 (1954), pp. I-II; see also recently W. Halbfass, Indien und
Europa (Basel-Stuttgart, 1981). For the approach of a phenomenologist (and 'traditionalist'), see for example H. Corbin, Philosophie iranienne et philosophie comparee (paris, 1985).
FOREWORD
Chinese and Central Asian components actually known historically to have been introduced from outside), the question may
even arise as to whether the name 'Buddhism' denotes one single
entity rather than a classification embracing (more or less
polythetically) a very large number of strands held together by
family resemblances. In their work Sinologues have been wont to
focus above all on what is Chinese, and hence on discontinuities
between Chinese and Indian Buddhism; and whereas some
Tibetologists have emphasized continuities as well as differences
between Indian and Tibetan Buddhism, others have preferred to
underscore the discontinuities above all else.
In the following essays an attempt is 'made to investigate a pair
of themes in Buddhist thought by considering, in historical and
comparative outline, their treatment in some traditions ofIndian
and Tibetan Buddhism, while referring on occasion also to
parallels in non-Buddhist Indian thought (Brahmanism and
Jainism) and in Chinese Buddhism. The two themes are, schematically stated, 'nature' and 'nurture' in the twin realms of
soteriology and gnoseology, a pair of topics that call for examination in terms of the notions of 'innatism', 'spontaneism' and
'simultaneism' as contrasted with graded acquisition and reinforcement through progressive cultivation. Connected themes
are enstatic concentration (gnoseological rather than cataleptic) as
against intellectual analysis, ethical and spiritual quietism in
contrast to effort, and cataphaticism as opposed to apophaticism.
Put in these terms, these notions are of course largely 'etic' ones of
Western origin, and they require to be investigated and specified
in the light of the rich 'emic' categories belonging to the
traditions being considered. Since a full treatment of each of them
in Indian and Tibetan thought could easily fill volumes, they can
of course only be outlined in these essays. It perhaps needs to be
explicitly noted also that, whereas the co-ordinate pair of theory
(darsana = Ita ba, the8ria) and practice (carya = spyod pa, praxis)
underlies much of what is said in our Buddhist sources, praxis has
perforce to be considered here more in terms of taxonomies or
theories of practice than as spiritual experience and practice per
se. 3 As for paramartha and sa/flVrti - ultimate reality and the
3 This specification is made explicit in response to a valuable methodological observation made by Alexander Piatigorsky.
FOREWORD
FOREWORD
excludes ethical and intellectual effort or that form of understanding that focuses non-analytically on the Empty alone, in contravention of the Buddhist principle that Quiet (Samatha) and Insight
(vipasyanCi) - like means (upCiya) and discriminative understand-'
ing (prajnCi) - are co-ordinate and have to be cultivated together
either in alternation or in unison as a fully integrated syzygy
(yuganaddha). In this way, in Tibetan philosophical discourse, the
figure of the Hva San Mahayana and his teaching have come to
fulfil a practically emblematic functio:t:l, one that may in fact be
somewhat different from the position actually occupied by the
historical ho-shang Mo-ho-yen. The following study will then be
concerned as much with the impact and significance of the 'Great
Debate' ofbSam yas for the Tibetan tradition - in other words its
Wirkungsgeschichte - as with the question of what actually
happened at the discussions in which the ho-shang Mo-ho-yen was
involved toward the end of the eighth century in Tibet.
~~--------~------
in Mahayana Buddhism). In some Buddhist traditions, the complex problem of the relation (or the description of the relation)
between ultimate Awakening and means - viz. the virtues of
generosity, etc. - has been treated in terms of the dedicatory
transmutation (parilJamana = yons su bsno ba) of these virtues into
Awakening (bodhi); interestingly, however, this question of
parilJamana was scarcely thematized in the documents relating to
the 'Great Debate' of bSam yas. 5
In discussing the simultaneous/gradual polarity in Buddhist
thought, especially on the Sutra-level of the Paramitayana, it is
essential to be quite clear as to whether it is the Fruit or the Path
that is in question. Now, that the realization of the Fruit at the
very end of the practice of the Path is instantaneous (and in some
sense 'simultaneous') is generally recognized, and this was therefore hardly the issue. It is accordingly the status of the Path alongside the difficult problem of the 'homology' of Ground and
Fruit and the 'proleptic anticipation' of the latter in the former that is the problem being addressed.
The question further arises as to whether, given its positive and
cataphatic character, thetathagatagarbha theory was a syncretism, or
a symbiotic accommodation, with the atman-doctrine of Brahmanical thought - that is, in effect, a crypto-Brahmanical 'soul'-theory
in Buddhism. Or was it perhaps conceived as inclusivistic of this
atman-doctrine in the sense of Paul Hacker's 'inclusivism', i.e. as
incorporating this 'soul'-theory in a subordinate position? Or,
again, is it an authentic Buddhist treatment of a theme - and a
religious and philosophic problem - which recurs in various forms
throughout the history ofIndian thought? A related question arises
5 It is true that Sa skya PaI;t<;li ta has referred to paril1amana in proximity to some
references he made to the (dkar po) chig thub, i.e. to a spiritual factor, said to have been
assumed by the Hva
Mahayana during the 'Great Debate' of bSam yas, that is
supposed to operate as the unique and self-sufficient factor making possible the immediate
and 'simultaneous' achievement of Awakening. See for example his sDom gsum rab dbye, ff.
34a and 38b, and his sKyes bu dam pa rnams la spriti ba'i yi ge, 5b. However, in these places
paril1amani'i is not treated as either a bridge between, or as a factor permitting a leap from,
the conditioned level of activity and impurity to the unconditioned leyel. And Sa skya
PaI;tC# ta is evidently criticizing the treatment of paril1amana as a supplement to the chig thub,
that is, as supporting a factor that was supposed by its proponents to be already totally
effective and altogether self-sufficient in achieving Awakening. On paril1i'imana in
connexion with the elimination of nimitta and upalambha and with reference to the cig
char/rim gyis opposition, see also MS BN Pelliot tibetain Il6 (12.5-6) and Pelliot tibetain
823, verso.
sail
-------- ------
.-----------
FOREWORD
1986).
8 Griffiths, op. cit., p. II, describes the nirodhasamiipatti-'theory as 'even more radical in its
rejection of mental activity than are the dominant Western models for the understanding
IO
FOREWORD
II
1.
12
FOREWORD
a 'peculiarly sinitic mode of approaching the enlightening experience'.11 The materials gathered in these essays nevertheless document the fact that the gradual/simultaneous polarity - krama/yugapad in Sanskrit and rim gyis/cig char in Tibetan - is neither uniquely
and exclusively nor pre-eminently Chinese, and that it is in fact
very well attested as a polar contrast or tension, and sometimes also
as a conflict, in the Indian Buddhist traditions too. That the cultural
and intellectual matrices and networks of concepts in which
this polarity has found expression differ appreciably between
India ,and China is of course no less clear, so that it is no doubt
legitimate to speak of distinct Indian and Chinese developments
(and also of Indic and Sinitic models in the Buddhism of Tibet).
Meaningful comparison can perhaps be most fruitfully pursued
in terms of typologies, structures and lattices of family resemblances.
As for the historical relation between Ch'an/Zen and the
teachings of the Indian Siddhas, R A. Stein and L. Gomez have
noted that it is very unlikely that Ch'an could have derived from
or been directly influenced by Indian Vajrayana or Siddha
schools; 12 and to assume that the former originates from the latter
would no doubt involve an historically unwarranted 'soft
methodology' (to borrow an expression used in another connexion by Gomez 13 ). Nevertheless, mutatis mutandis, the typological
parallels and family resemblances do seem clear enough for the
comparativist to have to address them very seriously.14
The extent to which mahCimudrCi 15 is to be seen as 'gradualist' or
'simultaneist' was moreover an important subject of reflection and
discussion in Tibet. And Sa sky a PaI,lr.1i ta for one considered what
11 Tu Wei-ming, 'Afterword: Thinking of "enlightenment" religiously', in: P. Gregory (ed.), op. cit., p. 448.
12 R. A. Stein, 'Illumination subite ou saisie simultanee: Note sur la terminologie
chinoise et tibetaine', Revue de ['histoire des religions 179 (1971), pp. 5-6, who concludes: 'Si
on se decidait it retenir les analogies [entre Ie Tch'an et Ie tantrisme indien ou la
Mahamudra], on devrait sans doute songer it des developpements paralleles'; and
L. G6mez, 'Purifying gold', in: P. Gregory (ed.), op. cit., p. 70.
13 L. G6mez, op. cit., p. 139, n. 14.
14 And the possibility of the existence of links between the Vajrayana and some trends
in at least later Ch'an (if only in Tibet) cannot be totally excluded a priori either. See
below, Chapter iii,pp. 122, 1JI-2, 137.
15 Globally described as 'gradualist' by L. G6mez, op. cit., p. 143, n. 41. But this
description would not fit well the current of the Tibetan Mahamudra tradition
represented by Zan Tshal pa for example (see Chapter iii).
13
*
The essays in this book are based on the Jordan Lectures in
Comparative Religion delivered at the School of Oriental and
16 See Sa skya Pal).4i ta, sDom gsum rab dbye (sDe dge ed.), f. 26a. On texts concealed by
the Hva san before his banishment from Tibet, and on the motif of the boot he left behind,
see the 'Alternative Tradition' of the sBa bied (ed. mGon po rgyal mtshan, Beijing, 1982,
p. 7S), apparently used by Sa skya Pal).4i ta also in his Thub pa'i dgo/is gsal (sDe dge ed.), f.
soa. The boot the Hva san is said to haye left behind in Tibet has sometimes been
interpreted as a token of the future revival there' of his teaching. Cf. below, Chapters ii and
11l.
FOREWORD
15
19
the Path, and hence the motivating 'cause' (hetu: dhatu) for
attainment of the Fruit (phala) of buddhahood. Even when the
texts do not use the term tathagatagarbha to designate this factor
making it possible for all living beings ultimately to attain
liberation and buddhahood, the importance of the theme of the
tathagatagarbha is therefore basic to the soteriology and gnoseology of the Mahayana.
To designate this same factor, certain texts use in addition the
terms (tathagata)dhatu, prakrtistha-gotra and prakrtiprabhasvara-citta,
words that have a longer history in the development of Buddhist
thought. The at least partial systematic equivalence of these terms
from the points of view of soteriology and gnoseology is set out
in several of the scriptural sources for the tathagatagarbha doctrine.
THE SOTERIOLOGICAL AND METAPHYSICAL STATUS OF
THE TATHAGATAGARBHA AS A PROBLEM IN EXEGESIS AND
HERMENEUTICS
If the fundamental role played by the notion of the 'EmbryoEssence' (garbha) , or Germ, and by the spiritual 'Element' (dhatu)
of the tathagata is accordingly clear, the metaphysical and soteriological status of the tathagatagarbha, tathagatadhatu and gotra is
somewhat less so. While the prakrtistha-gotra as the Support
(adhara) for practice of the Path is evidently situated on the
'causal' level - i.e. that of the sentient being in Sarp.sara - and
while the tathagatagarbha is said to exist in all sentient beings
without exception, the tathagatadhatu on the other hand is present
not only on this level of ordinary sentient beings but also,
evidently, on the level of buddhahood itself. This difference
makes it impossible to regard the tathagatagarbha and tathagatadhatu as simply identical in all doctrinal contexts.
The tathagatagarbha is characterized as permanent (nitya) , immutable (dhruva) , blissful (sukha) , and eternal (5a5vata) , and
sometimes we are even told that it is atman. These are epithets that
one would expect to relate to the Absolute - indeed, prima jacie,
to a substantial Absolute.
The parallelism between the tathagatagarbha (or its equivalents)
and the Vedantic atman is quite striking and it might even be
thought at first sight that a crypto-V edantic tendency has here
21
9 Ibid. Bu ston, mDzes rgyan, f. 27a, considered that this statement is, however,
intentional (dgons pa can = abhiprayika), the motive (dgos pa = prayojana) being to
introduce Tirthikas to the Buddha's teaching (by avatara1'}abhisa'1ldhi) and the intentional
foundation (dgons gzi) being the Emptiness of dharmas having discursive development
(prapaFica) in tathata, which is free from prapaFica relating to the dichotomously conceptualized binary pair Empty/not Empty (ston mi ston gi spros pal (the negative here being a case
of paryudasa-type negation). On the terms dgons pa can, dgons gzi and dgos pa, see below.
23
always said that the Buddha-nature (salis rgyas kyi rali bzin)
exists in all living beings, is this very Buddha-nature thelJ.
not self? Thus I do not teach a nihilistic theory. If, because
one does not see the Buddha-nature of all sentient beings,
one asserts the not permanent, the not self, the not blissful,
and the not very pure, it is said that one teaches nihilism.'
Then, after the ascetics had heard the explanation that this
Buddha-nature is self, they all produced the thought (citta)
directed toward supreme and perfect Awakening (anuttarasamyaksambodhi). And having at that moment entered religious life (pravraj-) , they exerted themselves on the path of
Awakening (bodhimarga).'10
But the Sutra nevertheless continues:
This Buddha-nature is not in reality atman, and it is for the
sake of sentient beings that a self is spoken of. Whereas in
virtue of the existence of causes and conditions the Tathagata has spoken of not-self (bdag med pa) as self, in reality
there is no self. Though he has spoken thus, this was no
untruth either. It is because of the existence of causes and
conditions that it is said that the self is not-self. Whereas self
exists in reality, it is with a view to the world of living
beings (loka) that it has been said that there is no self. But
that was no untruth. The Buddha-nature is not-self (bdag
med de); and if the Tathagata has spoken of 'self', this is
because a designation has been employed (btags pa yin pa'i
phyir).1 1
Elsewhere the same Sutra explains:
If what is called 'self' were an eternally permanent (kutasthanitya) dharma, there would be no freedom from suffering
(duhkha). And if what is called 'self' did not exist, pure
10 Mahiiparinirvii/!asatra, Tibetan version translated from the Chinese, vol. kha,
f. 137b-138a (quoted in Eu ston's mDzes rgyan, f. 22a-23a, and translated in Traite,
pp. 113-14).
11 Ibid. In this case Eu ston, mDzes rgyan, f. 23a, has explained that designating the
tathiigatagarbha, which is not self, as self is a case of pratipak,iibhisalfJdhi, i.e. that this
teaching was intended as a counteragent against the contempt the TIrthikas may feel for
the Dharma (because they take it to be nihilistic).
13
pp. II7-18).
14
130a-b (quoted in Bu ston's mDzes rgyan, 26a, and translated in Traite, p. 121).
25
the perfection of discriminative understanding (prajniipiiramitii), as the reverse (viparyaya) of attachment to the postulation of the non-existent self of the heterodox, who see a self
in the five Groups as objects of appropriation.1s For all the
heterodox have imagined as self a thing consisting of matter
(rupa) and the other [skandhas], [but] which does not have
this [self] as its nature; and this thing as grasped by them is
always not-self (aniitman) by reason of the [very] fact that it
does not conform to the characteristic of self (iitmalak~alJa).16
But the Tathagata has attained the supreme limit of the nonsubstantiality of all things (sarvadharmanairiitmya) by means
of exact gnosis; and this non-substantiality as seen by him is
always considered as self because it conforms to the proper
characteristic of not being self (aniitmalak~alJa). For [here]
non-substantiality (nairiitmya) is held to be self in the manner
of 'that which is fixed in the mode of non-fixation'. 17
From the viewpoint of the Buddhist, then, the situation is that
the Tirthikas' conception of iitman does not, and cannot, really
correspond to their own definition of the iitman; and for this
reason it is unacceptable. 18 Hence it is first said in the passage
quoted above that the self - i.e. the self of the speculative
iitmaviida (to which the iitman of the MahiiparinirviilJasutra and the
other comparable Buddhist scriptures only appears to correspond) - does not exist. 19 But this does not imply that' the reality
15 Ratnagotravibhaga-Commentary i.36, p. 31.IO-I2: paiicasilpadanaskandhqv atmadariinam anyattrthyanam asadatmagrahabhirativiparyayeIJa prajiiaparamitabhavanayal; paramatmaparamitadhigamal; phalaltl dra,{avyam. On the 'showing' (paridipana) of lokottaradharmas by
the counteragent (pratipak,a) of laukikadharmas, see also Ratnagatravibhaga-Commentary i.
154-5 (p. 76.19).
16 i.e. to the postulated definition of an atman.
17 The Ratnagatravibhaga-Cammentary i. 36 reads here (p. 31.13-16): tathagata!; punar
yathabhutajiianena sarvadharmanairatmyaparaparamipraptal; / tac casya nairatmyam anatmalak,aIJena yathadarsanam avisaltlvaditvat sarvakalam atmabhipreta nairatmyam evatmeti krtva /
yathOktaltl sthita 'sthaiiayogeneti /- Cf. Mahayanasutriilaltlkarabha,ya ix. 23 on paramatman.
Cf. A,{asahasrika Prajiiapiiramitii i, p. 8, for the application of the antiphrastic statement
susthita 'sthiinayogena.
18 See also Mahiiyiinasutriilaltlkiirabhii,ya vi. 2. Cf. Aryadeva, Catu(!sataka x. 3ab (quoted
in Candrakirti's Prasannapadii ix. 12, avatarat]ikii): yas taviltma mamaniitma teniitmilniyaman
na sal; / 'Your self is for me not self (bdag min), so that it is not selffor lack of certainty (ma
ties phyir).'
19
20
27
28
view as its 'intentional ground' (dgoris gzi) the theory of nairatmya, sunyata and bhtttakoti 27 as generally understood in the
Mahayana. As will be seen below, the ;mthor of the Tarkajvala
evidently also had a similar view of the purport of the tathagatagarbha doctrine.
This concept of intentionality is far from being a merely ad hoc
hermeneutical device of some commentators, and it has behind it
a long history in Indian semantics and semiotics. Already in
Pataiijali's Mahabha~ya, vivak~a 'intention to express' appears as a
factor that can determine the use of a word-form. This vivak~a
may be either ordinary and 'mundane' (laukikl vivak~a), which
means that it corresponds with ordinary linguistic usage as
covered by the usual rules of grammar; or it can on the other
hand depend on the intention that a speaker has in a given set of
circumstances and context (prayoktrf vivak~a). 28 When used
independently, an expression is determined by the speaker's
intention to express; but if an expression is conditioned by
external circumstances also it is said to be dependent. 29 The later
Sanskrit grammarians then extended the scope of this principle of
intention. Jinendrabuddhi, the commentator on the Kasika, states
that the determining factor in word-formation may be not only
the existence of a corresponding object referred to but also the
speaker's intention to express. 30 Saral).adeva (twelfth century)
also considers word-formation to be dependent on vivak~a. 31 We
furthermore read in a paribha~asutra of the CandravyakaraIJa
(no. 68): 'The determination of the desired [word form] results
from pervasion by an intention to express (vivak~avyapter i~tavasa
yah).' Hence, according to this developed doctrine, it is the
27 See LarikiivatiirasiJtra ii, p. 78.6-7, and vi, p. 223.3-4. On the concepts of dgons gzi
and dran don see D. Seyfort Ruegg, Journal of Indian Philosophy, I3 (1985), pp. 309-25, and
15 (I987), pp. 1-4
28 Pataiijali, Mahiibhii~ya V. i. 16, which explains: laukikT vivak,ii yatra priiyasya
sampratyayaiJ / priiya iti loko vyapadisyate. Here loka = priiya is the general, as opposed to
the individual in priiyoktrT vivak,ii.
29 Cf. Mahiibhii!iYa I. ii. 59: tad yadii sviitantrye/1a vivak~ii tadii bahuvacana/tl blzavi,yati,
yadii piiratantrye/1a tadaikavacanadvivacane bhavi,yataiJ /1 (comment on Pal).ini L ii. 59 asmado
dvayos cal. Cf. Mahiibhii,ya III. i. 87 (p. 67.12-13).
30 Nyiisa on KiiSikii I. i. 16: na hlha sabdasiistre vastunaiJ sattaiva sabdasa/tlskiirasya
pradhiinam kiira/1am / ki/tl tarhi / vivak,ii ca f.
31 Durghatavrtti on II. ii. 8: vivak,iidhTnii sabdavyutpattiiJ. Cf. on I. iii. 36.
29
32 Cf. L. Renou, Terminologie grammaticale du sanskrit, s.v. vivak,a. See also Hetiraja on
Vakyapad.ya III. iii (sambandhaO). r concerning the prayoktur abhipraya~ 'intention of the
user [of speech]', which is one of the three things conveyed by the use of words, together
with the atm.yal1;l riipam of the word (Bhartrhari's svariipa) and the artha~ phalasadhanah
(Bharq;hari's bah yo 'rthah). Bhart,hari himself (III. iii. ra) speaks of the jiianal1;l prayoktu~.
Helaraja observes that the relation between word and the speaker's intention is one of
effect and cause (karyakaralJabhava, rather than the vacyavacakabhava which obtains
between artha and sabda).
33 Prama~avarttika i. 327ab: vivak,a niyame hetuh sal1;lketas tatprakasanah. The autocommentary explains: vivak,aya hi sabdo 'rthe niyamyate, na svabhavata~, tasya kvacid
apratibandhena sarvatra tulyatvat Iyatrapi pratibandhas tadabhidhananiyamabhavat I sarvasabdail! kara~anal1;l abhidhanaprasatigat I tasmad vivak,aprakasanayabhiprayanivedanalak,a~a~
sal1;lketa~ kriyate I (cf. ii. r6). In i. 65 Dharmaklrti contrasts the idea that words (vaca~) are
dependent on vivak,a with the idea that they are dependent on a thing (vastuvaSa).
3I
many authorities to constitute a third semantic function additional to, and distinct from, both primary denotation (abhidha)
and secondary indication (lak.)a1Ja). Indeed, in his famous commentary on the Dhvanyaloka, Abhinavagupta has stated that this
suggestive function of words belongs to the domain of prayojana. 38
.
It is of the greatest interest to observe that practically the same
semiological concept and criteria have been applied in Buddhist
hermeneutics to identify and define a scriptural statement that is
intentional (abhiprayika) and in need of interpretative elicitation
in a sense other than the obvious surface one (neyartha) since it is
non-definitive within the frame of a given philosophical system.
Thus, before a scriptural passage can be characterized as being
abhiprayika and neyartha, its meaning must be shown to be
cancelled by another teaching the meaning of which is, in a
particular doctrinal system, recognized as final and definitive
(nrtCirtha) and non-abhiprayika. In addition, it must refer indirectly, by a kind of philosophical-systematic 'implicature', 39 to a
certain 'deep meaning' which has not been directly conveyed to
the addressed disciple by the statement in question but which is
final and definitive. And, thirdly, the use of a non-definitive
statement requiring further interpretative elicitation in another
sense (neyartha) must be conditioned by a definable and legitimate
motive (prayojana) of the speaker - in the case of a Sutra the
Buddha himself who, in virtue of his expertness in means
(upayakausalya) , employs an abhiprayika and neyartha statement as
a device (upaya) in order to benefit his listener. 40
38 Abhinavagupta, Locana i. IT dhvananaftl prayojimavi~ayam (p. 148 0. Cf. i. 4
(p. 60-61) and iii. 33 (p. 441-2).
39 For this use of the terni 'implicature', see the present writer's article in Journal oj
Indian Philosophy, 13 (1985), pp. 313 and 316 f
40 The hermeneutical status of the Buddha's teaching would thus differ significantly
from that of the Vedic sruti, inasmuch as the latter is considered by the Mlmarp.sakas to be
authorless (apauru~eya). In his Svavrtti on the PramiitJaviirttika i (Svarthanumanapariccheda)
325, Dharmaklrti - followed by KarI)akagomin and Manorathanandin - observes that,
according to this assumption of Jaimini's Mlmarp.sa school, there could be no vivak~ii
(linking sabda with artha) , no saftlketa and no abhipriiya - in other words no authorial
intention - for the Veda.
In Buddhist hermeneutics as traditionally practised, there can be no question of
radically relativizing the intended purport of a canonical utterance or text (so-called semantic
autonomy) and banishing the idea of authorial intention (so-called authorial irrelevance)
33
43
34
(sa1flvrti) reference to a personal entity (pudgala) , such a provisional teaching being in accord with the inclinations of these still
immature addressees. 44 And in his Madhyamakakarikas Nagarjuna
has observed that the Buddha sometimes made use of the
designation (prajiiapita) 'atman' and sometimes taught (deSita)
anatman - in the manner so to say of the first two positions of a
tetralemma (catu.$koti) - whilst on other occasions he taught that
there is neither atman nor anatman (xviii.6). According to Candrakirti's Prasannapada these three forms of teaching are meant to
accord with the mental inclinations (asaya) of distinct arid
progressively more advanced types of disciples (hinamadhyotkr-!tavineyajana). The form of teaching based on a tetralemma is
regarded by Nagarjuna as an anusasana (xviii.8), that is, according
to Candraklrti, as a progressive and graded teaching (anupurvya
sasanam) adapted in each stage to different degrees of addressees to
be trained (vineyajana) by the Buddha. 45
Although only further detailed investigation can perhaps
determine whether the literary or the philosophical application of
these semiological and hermeneutical principles is older, the fact
remains that virtually the same notions have found employment
in both areas. This parallelism underscores once again the
fundamental importance for the history of Indian thought of its
exegetical and hermeneutical methodology based on analysis of
language and meaning, as distinct from particular doctrines
which of course vary considerably from school to school. These
methods employed by the Indian. thinkers appear, then, as a
unifying thread behind the very great diversity of philosophical
schools and even religions.
44 Asanga, Mahiiyiinasaytlgraha 2.31 with Vasubandhu's Bhii,ya (D, f. 154b), and
*Asvabhava's Upanibandhana (D, f. 233b) which interprets the allusion to a pudgala in
terms of an upapiiduka sattva.
The references to avatiiralJiibhisaytldhi as a means of reducing the addressed disciple's
terror (uttriisa) in the Mahiiyiinasiitriilaytlkiira (xii. 16) and Bhii,ya - as well as in Sthiramati's
V,ttibhii,ya (D, f. 24ob) and *Asvabhava's Tikii (D, f. I07a) - concern exclusively the
Sravaka's introduction to the Mahayana by the provisional teaching bearing on the
existence ofriipa, etc., and not the introduction of the 'outsider' by provisional allusions to
a pudgala (or iitman). On abhisaytldhi and related terms, see D. Seyfort Ruegg in: C. Caillat
et al. (ed.), Formes dialectales dans les litteratures indo-aryennes (Paris, 1989), p. 299 ff.
45 Cf. Kiisyapaparivarta 57; Nagarjuna, Ratniivalf ii. 3-4; and Aryadeva, Catul;sataka
xiv. 21 with CandrakIrti's comment which mentions a tetralemma based on the iitman
notion. See the analysis in D. Seyfort Ruegg,Journal of Indian Philosophy 5 (1977), pp. 7-9
35
It is clear, however, that so powerful a hermeneutical instrument as the idea of an intentionally motivated 'surface' teaching
of provisional or non-definitive meaning requiring interpretation
in a sense other than the obvious surface one, and opposed to a
'deep' teaching of final and definitive meaning, had to be handled
with care and restraint - and no doubt also as sparingly as possible
- in order not to be tainted with arbitrariness and disregard for a
canonical corpus. Moreover, it is plain that the mere existence of
a motive behind a teaching cannot alone suffice to justify the
conclusion that it is 'intentional' (abhiprayika) in the technical
sense in question, in other words that it is of provisional or nondefinitive value; for any teaching at all is motivated to the extent
that its author has in view an intended meaning when he
communicates it. This is indeed the reason the hermeneuticians
have insisted that the other two above-mentioned conditions
should also be satisfied.
II
36
47 It is to be noted that this latter form of the tathagatagarbha teaching, which is on the
face of it comparable with the atmavada, seems to be less frequently met with. See for
example Mahabheri"'tra, f. 18Ib, quoted in Bu ston's mDzes rgyan, fo1. 6a (translated in
Traite, p. 79). But contrast e.g. Mahaparinirval]a",tra, kha, f. 134b, quoted in mDzes rgyan,
f. 18a (translated in Traite, p. 103). In this connexion, the doctrines of the TibetanJo nan
pa school, and the related doctrines of other schools, pose a problem that will have to be
treated separately.
48 Ratnagotravibhaga-Commentary i. 154-5, quoting the Srimaladevisil1}hanadasutra, fo!'
445a.
49 See also Lalikavatarasutra ii, p. 79.8-9, and vi, p. 223.3-4, already cited. The gzan
stoli of the Jo nan pa and similar schools, as opposed to the rali stori, needs to be considered
in this cannexion.
50 See Ratnagotravibhiiga i. 160 along with several of the Tibetan commentaries. On the
hermeneutical tradition that assimilates the tathagatagarbha and sunyata theories, and which
is represented for example by the Tibetan dGe lugs pas, see Theorie du tathagatagarbha et du
gotra, p. 402 f.
37
393, 4 0 3.
39
least partial parallelism between the Buddhist's positive description of absolute reality both by means of attributes that are the
reverse of those properly applicable to the relative level and by
means of characterization through inseparable constitutive qualities on the one side, and on the other Sarp.kara's treatment of satya
'truth', jncma 'knowledge' and ananta 'infinite' as constitutive
qualities of brahman. According to his Bha~ya on the Taittiriyopani.)ad, satya,jnCina and ananta are in fact distinctive characteristics
proper only to brahman, which they thus differentiate from all
else. Hence these essential characteristics are not mere attributes
inhering in brahman as one member of a class; for brahman - the
real- is unique. These epithets therefore define its uniqueness by
delimiting (: niyantrtva) it from all that it is not, i.e. the unreal. 54
Despite very important differences in the philosophical background and problematics between Sarp.kara and the Buddhist
philosophers in question, the similarity in the procedures and
methods used to define ultimate reality is thus remarkable.
The suggestion that certain Buddhist and Brahmanical notions
of the absolute have some points of contact is therefore not
unworthy of consideration. Indeed, we have already encountered
some passages where the MahCiparinirvCiIJasutra represents the
Tirthikas as readily assenting to that particular formulation of the
Buddha's teaching in which the existence of the Buddha-nature
or tathCigatagarbha is emphasized. And elsewhere the same Sutra
goes so far as to say that BrahmaIJ.a Tirthikas have borrowed
some of the Buddha's teachings and incorporated them in their
own scriptures. 55 Ifhe may be supposed to have been influenced,
however indirectly, by earlier Buddhist thinking on the problem
of the paramCirtha, Sarp.kara could be a case in point.
In a remarkable passage the LatikCivatCirasiitra moreover points
to a certain equivalence between various names and notions of the
54 It has been suggested by M. Biardeau that this theory of the Taittiriyopani~adbha~ya
represents a final stage in the development of Sarp.kara's thought, following on his
apophaticism; see Indo-IranianJournal3 (1959), p. 100 (cf.Journal asiatique 1957, p. 371 f.).
But P. Hacker has placed this Bha~ya in the middle period of Sarp.kara's philosophical
development; see Wiener Zeitschrift far die Kunde Sad und Ostasiens 12-13 (1968),
pp. 129-30, 135, 147. Cf. also G. Maximilien, Wiener Zeitschriftfar die Kunde Siidasiens 19
(1975), p. II7 f.
55 See Bu ston, mDzes rgyan, f. 22a-b (translated in Traite p. II3-14) and f. 27a
(translated in Traite, p. 123-4).
57
41
Madhyamakah~dayakarikas
and TarkaJvala dealing with the Vedanta that this is not so. For nairatmya is precisely the absence of
the self-nature characteristic of the Vedantic atman. Nor could
that which is anatman be at the same time atman because of the
incompatibility (virodha) of the two. And no comparison may be
drawn with, for example, horse-nature that is at the same time
not cow-nature (ba Ian gi no bo ma yin pa), that is, with a case of
implicative and presuppositional (paryudasa) negation rather than
of the non-implicative andpresuppositionless prasaJya-negation
involved in nairatmya. Accordingly, this bhava-nihsvabhavata
could never be an atman that is either a creator-agent (byed pa
po = kart" as in the theistic systems) or an enjoyer-agent (za ba
po = bhokt~, as in the Sarp.khya), these entities being as unreal as
the proverbial son of a barren woman. In so far as a cognitive
object without self-nature is made the object of thought, it might
perhaps be conceived of e.g. as single and unique; but this will be
a mere mental construction, which is the source of imputation
(samaropa).62
It is clear, then, that the authors of the basic Mahayana-Sastras
were no more ready. to admit any form of substantialism on the
level of the unconditioned than they were on that of the
conditioned: the Mahayanist pudgalanairatmya and dharmanairatmy a make this quite impossible.
Moreover, for the canonical texts teaching the tathagatagarbha
and Buddha-nature, the Middle Way eschewing both eternalism
and nihilism remained valid. And the affirmation of an absolute
or atman opposed to anatman or nairatmya in a dichotomously
conceptualized binary pair (vikalpa) based on discursive proliferation (prapanca) would, therefore, be no more acceptable than the
purely nihilistic position of a dogmatic denial of the absolute.
Therefore, contrary to what has sometimes been suggested and despite the undeniable parallels noted above between the
problems treated and the methods used on the Buddhist and
Brahmanical sides - the tathagatagarbha theory, as well as the
theory of the paramartha with which it is connected, can not
62
43
64 See for example Itivuttaka II. ii. 6 ( 43, p. 37): atthi ... ajatarr< abhatarr< akatarr<
asatikhatarr<. no ce tarr< ... abhavissa ajatarr< ... asatikhatarr<, nayidha jatassa ... sankhatassa
nissaral')arr< paiiiiayetha . .. .
aharanettippabhavarr< naIarr< tad abhinanditurn I
tassa nissaral')arr< santarr< atakkavacararr< dhuvarr<11
ajatarr< asarnuppannarr< asokarr< virajarr< padarr<1
nirodho dukkhadharnrnanarr< sankhariipasarno sukho II
ti; and Itivuttaka III. ii. 2 ( 51, p. 46) and III. iii. 4 ( 73, p. 62): kayena arnatarr< dhaturr<
phassayitva nirupadhirr<1 ... deseti sarnrnasarnbuddho asokarr< virajarr< padarr<11 ti. Compare also
Suttanipata 204; Theragatha 521 and 947, Dharnrnapada II4; Atiguttaranikaya II 247; III 356;
Maiihimanikaya I 436-7; Sarpyuttanikaya IV 373; Patisarnbhidarnagga I 13, IS, 70; and
Saddaniti (ed. H. Smith), p. 70.
65 See for example the present writer's remarks in The study of Indian and Tibetan
thought (Leiden, 1967), pp. 9-13, 37-38.
66 Kasyapaparivarta 55-57, 64-65. See also Nagarjuna, Miilarnadhyarnakakarika xiii.
8; and Ratnagotravibhaga-Commentary i. }2-33 (p. 28.5-13).
44
T ATHAGATAGARBHA
AND
THE ABSOLUTE
45
68
69
See 'Gas 10 tsa ba gZon nu dpal, Deb ther sri on po, cha, f. lob (Roerich, p. 349).
Ratnagotravibhaga i. 9 and 153.
47
77
thereafter.
It may be recalled here that these texts too support the assimilation of the tathagatagarbha
to siinyata: tathagatagarbhajiianam eva tathagatanarrt siinyatajiianam (Srfmalasiitra, f. 445a) and
tathagatagarbhaiunyatarthanaya (Ratnagotravbhaga-Commentary i. I54-5 avatara'1ika, p.
75.I7) Compare pp. 33 and 36 above.
48
Cf Ratnagotravibhaga-Commentary i.
I2.
49
yanas in this matter rests in the precise mode of their comprehension of sunyata, which is fuller in the case of the Arya-Bodhisattva
than in the case of the Arya-Sravaka and Pratyekabuddha. 83
Following this interpretation, the references in Sutra (the Srtmalasutra, etc.) and Sastra (the Ratnagotravibhaga and its Commentary)
to the essential role of faith in understanding the paramartha have
therefore to do with the fact that only the Bodhisattva endowed
with the sharpest faculties (tlk~1Jendriya) is able to understand it in
all its aspects exclusively through his transcending discriminative
knowledge (prajna) without first having to have recourse to faith.
Consequently, faith may properly be said to be a characteristic
feature of the other two ways, that of the Sravaka and that of the
Pratyekabuddha. In sum, although the object of the Sravaka's and
the Pratyekabuddha's understanding - nairatmya and sunyata - is
the same as the Bodhisattva's, the mode of their comprehension is
not as full and all-embracing, and their understanding in this sense
is only partial. 84
and which can therefore be thought to imply that they are all
destined sooner or later to become buddhas. But the ekayana
theory, which holds that all vehicles ultimately converge in a
single course leading to buddhahood, is quite in harmony with,
and complementary to, the theC?ry of the tathagatagarbha. 85
BUDDHA-NATURE AND PARAMATMAN IN A COMPARATIVE
PERSPECTIVE
51
90
91
53
54
97
55
99
II
100 On the name Mo-ho-yen, as well as on the epithet ho-shallg (Tib. hva sari) derived
from Skt. upiidhyaya 'master', see P. Demieville, Le coneile de Lhasa (Paris, 1952), pp. 9 If.
It should be recalled that other Ho-shangs such as Me 'go/mgo and Kim (Ch. Chin,
below, n. 116) had preceded Mo-ho-yen in Tibet. See e.g. sBa bied, ed. mGon po rgyal
mtshan (Beijing, 1982) (=G), pp. 6-IO, 65, 67-68; ed. R. A. Stein (Paris, 1961) (2abs
blags ma version = S), pp. 4-6, 8-IO, 55, 57; dPa' bo gTsug lag phren ba, mKhas pa'i dga'
51011, ja, If. 1I5a5, 116a4. The Chos 'byuri by Ne'u PaZlI;li ta, the sNail gyi gtam 1\1e tog pilmi
ba, ed. T. Tsepal Taikhang, Rare Tibelan historical and literary textsfrom the library of Tsepon
W. D. Shakabpa, I (New Delhi, 1974), f. Isa (p.87), indeed mentions their presence
already at the time of King Sron btsan sgam po in the seventh century; and Bu ston (Chos
'byuti, f. 124b6) gives the name Ma ha de va tshe, while the Deb ther dmar po (f 16b = p.
35) gives the name Ma ha de ba. Cf P. Demieville, Coneile, p. I I, n. 4, on the Hva san
Mahadeva at the time of Sr011 btsan sgam po.
57
See Demieville, Candle. The source of the questions put tD Mo-ho-yen according
to Wang Hsi's Cheng-Ii chiieh is the so-called 'Brahman monk' (Condie, pp. 39-40),
presumably Kamalama.
The question and answer form of discussion was evidently used also in cases where there
appears to have been no debate, for example in T'an-k'uang's 'Dialogues' translated by
W. PachDw, A study of the twenty-two dialogues ofMahiiyiina Buddhism (Taipei, 1979), and in
Hung-jen's 'Discourse' translated by W. Pachow, Chinese Buddhism (Washington, 1980),
pp. 35-53. For T'an-k'uang and his role in connexion with the problems posed by the
'Great Debate', see below, p. 128.
102 See below, p. 66.
58
Coneile, p. 18.
The invention oj tradition is the title of a book of essays, edited by E. Hobsbawm and
T. Ranger (Cambridge, 1983), dealing with rather different situations in modern England,
Scotland, Wales, continental Europe, India and Africa; and the use of the expression here
should not be understood as implying an identification with such situations. For the bSam
yas monastic cantre, one might compare also the concept of' Lieux de memo ire' , the title of
a collection of essays edited by P. Nora (Paris, 1986).
On the constitution of tradition in the case of Ch'an Buddhism in the Sung period after a break in the clan and school traditions of the T'ang - see H. Schmidt-Glintzer, Die
ldentitat der buddhistischen Schulen und die Kompilation buddhistischer Universalgeschichten in
China (Munich, 1982), p. 27 ff.
103
104
59
60
61
109 See sBa bzed, G, pp. 65 and 68 (sNa Bye rna), and S, p. 55 (mNa Bi rna); Nan ral Ni
rna 'od zer, Chos 'byuti Me tog siiiri po, ed. R. O. Meisezahl (St. Augustin, 1985), If. 426b3
and 429b3 (Na Bi rna); dPa' bo gTsug lag phren ba, mKhas pa'i dga' stan, ja, ff. I 15a5 and
116b4 (sNags Bye rna la). Cf. Demieville, Candle, p. 41.
The rus-name here may be equivalent to gNags (as in the case of gNags Jiiana-Kumara,
the disciple of Vim ala), in which case the Bi rna/Bye rna in question would of course be a
Tibetan, rather than the Indian Vimala(mitra).
110 See sBa bzed, G. pp. 65 and 67; S, pp. 55.1 and 57.II; mKhas pa'i dga' stan, ja,
f. II5a4 and II6b4; Nan Ni rna 'od zer, Chos 'byuti Me tog siiiri po, f. 429b3 together with
f. 426b3.
On Sa mi see also G. Tucci, Minor Buddhist texts, II, pp. 9-10. A certain Myan gSa (?)
mi go cha has been mentioned in connexion with the Bodhisattva (i.e. Santarak,ita) in the
document published by F. W. Thomas, op. cit., p. 85. This document also mentions
another member of the Myan family in the same context.
111 For rNog Rin po che, compare Demieville, Cotlcile, p. 33.
112 In the mKhas pa'i dga' ston,ja, f. II8b7, Co rMa rma is described as thegzims mal pa
or chamberlain (cf. f. 12ob6) (cf. Tucci, Minor Buddhists texts, II, p. 38, and Preliminary
report on two scientific expeditions in Nepal [Rome, 1956], p. 89). Compare Chos 'byUti Me tog
siiin po, f. 429a1.
113 The Bhadanta Lanka is not mentioned in the Chos 'bYlin Me tog siii,; po (f. 429a),
which mentions Co rna (?) only.
114 On these two see Demieville, Concile, pp. 25, 33.
115 See sBa bzed, G, p. 63-64; S, p. 54; and mKhas pa'i dga' ston,ja, f. 114b4; Chos 'bYUti
Me tog siii,; po, f. 425'al (=Ms B, f. 237a7).
The Chos 'byun Me tog siiiti po also refers to his being a disciple ofVimalamitra (f. 472a4)
and to his exile and murder (?) (f. 473a; Ms B, f. 258a) at the time of King Glan dar rna.
This source in addition mentions his connexion with dBu ru (f. 472a4). On Tin ne 'dzin
and the Myan. see also the inscription of the dBu ru Zva'i Iha khan in H. E. Richardson, A
corpus of early Tibetan inscriptions (London, 1985), pp. 43 If. Cf. Tucci, Milwr Buddhist texts,
II, p. 52 f.; below, p. 75.
62
116 This Sail si (ta) has sometimes been identified with (sBa) dPal dbyails, but the
question of his identity is far from clear. See Tucci, Minor Buddhist texts, II, pp. II-I2, 22,
24; P. Demieville, in M. Soymie (ed.), Contributions aux etudes sur Touen-houang (GenevaParis, 1979), pp. 4-7 (referring also to H. Obata).
The sBa bied (5, p. 5) mentions a rGya phrug gar mkhan Sail si. And a Sail si is
mentioned also in connexion with sBa gSal snail (G, p. 10~ 5, p. 9); he seems to have
been linked with Kim Hva sail (5, p. 10; cf. G, p. 7). Kim - Chinese Chin Ho-shang - was
the name of a Korean Ch'an master, Musang (Ch. Wu-hsiang, 694-762) who taught in
Sichuan (Chengdu). Cf. S. Yanagida, in W. Lai and L. Lancaster (eds.), Early Ch'an in
China and Tibet (Berkeley, 1983), pp. 18, 193; P. Demieville, in Jao Tsong-yi and
P. Demieville, Peintures monochromes de Dunhuang (Paris, 1978), p. 47, and in M. Soymie
(ed.), op. cit., pp. 4-5, 7, 11-12 (referring also to Z. Yamaguchi). - Wu-chu (Tib. Bu chu,
717-774), who is known to the rDzogs chen tradition (bSam gtan mig sgron) , was in
contact with Wu-hsiangjMusang in Chengdu (Demieville, Contributions, p. 5); but there
were significant differences in their views (p. 7).
117 On the question of the language(s) in which the discussions and debates were held,
their number, and of the participants in them, see P. Demieville, T'oung-Pao 56 (1970),
p. 42; Y. [maeda, Journal asiatique 1975, p. 129.
118 See sBa bied, 5, p. 62.6 (the specific reference to Nagarjuna is not in the other
version); Bu ston, Chos 'byun, f. 129b5; mKhas pa'i dga' ston, ja, f. II9a2; Nail Ni rna 'od
zer, Chos 'byun Me tog siiin po, f. 435b5.
119 The accounts of the Great Debate differ to a greater or lesser extent in the various
sourc~s, and there seems to have taken place a conllation with earlier discussion(s) in which
a certain Me mgo (?) (rather than Mo-ho-yen) was the leading Ho-shang.
63
For long the best known Tibetan historical source on the Great
Debate was no doubt Bu ston's Chos 'byun (f. I28a-I29b). In
1935 E. Obermiller - who had already published in 1932 an
English translation of the relevant section of Bu ston's history called attention also to what he supposed to be one of Bu stan's
main sources on the subject, the third BhCivanCikrama by KamalasIla. 121 This work indeed contains many passages that are perti120 See sBa bzed, S, p. 5+; Ne'u Pal).(;li ta, Chos 'byun, f. 22aI; Bu ston, Chos 'byuti,
f. I29b; Dalai Lama V, Bod kyi deb ther dpyid kyi rgyal mo'i glu dbyans, f. 39b (= p. 89);
Thu'u bkvan Bio bzan chos kyi iii rna, Grub mtha' Ie/ gyi me loti, rGya nag chapter, f. 10 if.
(with the transcriptions tun men/min and tsi'an men/tsi yan min).
For some pseudo-etymological explanations, based on Tibetan, of the originally
Chinese expressions, see sBa bzed, G, p. 64 and S, p. 54.12-13; dPa' bo gTsug lag phren ba,
mKhas pa'i dga' ston,ja, f. II5a; and Nan Ni rna 'od zer, Chos 'byun Me tog siiiti po, f. 426b.
121 E. Obermiller, Journal of the Greater India Society 2 (1935), pp. I-II. See also
Obermiller's posthumously published facsimile edition of the third Bhiivaniikrama (Moscow, 1963).
nent to the Great Debate with Mo-ho-yen; but none of Kamalasila's three Bhiivaniikramas actually mentions Mo-ho-yen. And
based as it is largely on quotations from Siitras, KamalaSlia's
treatment of the progressive stages of meditative realization
(bhiivanii = bsgom pa) involving, beside Quieting of mind (Samatha = zi gnas) , the fundamental factors of exact analytic investigation (bhutapratyavek~a = yati dag pa'i so sor rtog pa) and its
culmination in analysis of the factors of existence (dharmapravicaya = chos sin tu rnam paT 'byed pa) or discriminative knowledge
produced from meditative realization (bhiivaniimaYI pTajiiii =
bsgom pa las byuti ba'i ses Tab), and then finally in Insight
(vipasyanii lhag mthoti), refers specifically neither to any particular debate or to any individual opponent. Kamalasila's three
treatises are accordingly relevant to Buddhist philosophical theory and practice in general.
As for the historical existence of KamalaSlia and his master
Santarak~ita, it is of course very well established by their extant
works available in Sanskrit or in Tibetan translations in the bsTan
'gyur as well as by many references in original Tibetan historical
and philosophical works. The historical existence of the Hoshang Mo-ho-yenJHva sail Mahayana (or Maha yan) and of the
debate(s) in which he took part in Tibet - the two matters left
somewhat unclear in Tibetan bsTan 'gyur as well as in the
Sanskrit sources - has, despite some hesitations and obscurities in
the Tibetan traditions, been demonstrated by Wang Hsi's Tun-wu
ta-cheng cheng-Ii chiieh or 'Ratification of the true principles of the
Great Vehicle of Sudden Awakening' preserved in two Chinese
manuscripts from Dunhuang (Pelliot 4646 and Stein 2672) which
has been translated into French and commented on by Demieville
in his Le condie de Lhasa and 'Deux documents de Touen-houang
sur Ie Dhyana chinois'.122 In addition to Wang Hsi's Preface, the
Cheng-Ii chiieh comprises three 'Memorials' two of which consist
122 In: Essays on the history of Buddhism presented to Professor Zenryu Tsukamoto (Kyoto,
1961), pp. 1-27
Pelliot tibetain 823 contains a Tibetan version of a part of Wang Hsi's Cheng-Ii chueh; C
Y. Imaeda, Journal asiatique 1975, p. 127 ff. For further parallels between the Chinese
sources and Tibetan manuscripts from Tun-huang, see L. Gomez, 'The direct and gradual
approaches of the Zen master Mahayana: Fragments of the teachings of Mo-ho-yen' in:
R. Gimello and P. Gregory, Studies in Ch'an and Hua-yen (Honolulu, 1983), p. 106 if.
65
See n. 122.
M. Lalou, 'Document tibetain sur l'expansion du Dhyana chinois',journal asiatique
1939, pp. SOS-23
125 Cf. bSam gtan mig sgron or rNal 'byor mig gi bsam gtan (see below, n. 129), f. 90b3;
Tucci, Minor Buddhist texts, II, p. 21; R. Kimura, journal asiatique 1981, p. 187.
126 Cf. R. Kimura, journal asiatique 1981, p. 183 If.
127 D. Ueyama, 'The study of Tibetan Ch'an manuscripts recovered from Tun-huang:
a review of the field and its prospects' in: W. Lai and L. Lancaster, op. cit., pp. 327-49. See
also P. Demieville, 'Recents travaux sur Touen-houang', T'otmg Pao S6 (1970), pp. I-9S,
and 'L'introduction au Tibet du bouddhisme sinise d'apres les manuscrits de Touenhouang' in M. Soymie, Contributions aUx etudes sur Touen-houang, pp. I-IS, together with
G. Mala and R. Kimura, 'Additif in: M. Soymie (ed.), Nouvelles contributions aux etudes de
Touen-houang (Geneva, 198 I), pp. 321-7; G. Mala, 'Empreinte du Tch'an chez les
mystiques tibetains' in: Le Tch'an (Zen): racines et jloraisons (Hermes 4, Nouvelle serie,
Paris, I98S), pp. 387-424.
128 L. Gomez, 'Indian materials on the doctrine of Sudden Enlightenment' in: W. Lai
and L. Lancaster (eds.), op. cit., pp. 393-434; and 'The direct and gradual approaches of the
Zen master Mahayana' in: R. Gimello and P. Gregory (eds.), op. cit., pp. 69-167.
123
124
66
67
68
That a true debate took place towards the end of the eighth
century at the Byan chub glin of bSam yas between Kamalasi:la
and the Tibetan Gradualists on the one side and the Hva san
Mahayana, i.e. Mo-ho-yen, supported by his Tibetan 'Simultaneist' followers on the other under the aegis of the Tibetan ruler
Khri Sronlde btsan is stated in both available texts of the sBa bzed
(G, pp. 67/5 and S, pp. 57-62). This chronicle connected with
the first Tibetan monastic centre of bSam yas (see G, p. 82.II)
contains records of the sBa family, members of which are reputed
to have participated in the Great Debate. 134 And the sBa records
rna y well be our oldest chas 'byuri source on the subject; at all
events the sBa bzed has been described as the 'matrix' (phyi ma) i.e. the textual source - of all Tibetan chronicles (rgyal rabs) and
religious histories (chas 'byuri) as well as the record (bka' gtsigs,
etc.) of bSam yas. 135
The Supplemented Version (.Zabs btags ma) of the sBa bzed
became widely available to modern scholarship only in 1961,
when R. A. Stein published a facsimile edition of it. And another,
unsupplemented version of the sBa bzed was published as recently
as 1980 in Beijing by mGon po rgyal mtshan. The title sBa bzed
appears in Tibetan sources also in the forms rBa bzed, dBa' bzed,
dPa' bzed and sPa bzed. 136 The reference in each case appears to be
134 The colophon of the sBa bied G p. 82 mentions the view that, in the title sBa bied,
sBa refers to sBa gSal snan, as well as to another view that this rus-name refers to sBa San
si. This colophon then adds that the sBa bied is in fact widely regarded as the record of
bSam yas (bsam yas bka' thams [= than?]). A version known as the Mi sPa bied (sic) is
connected with gSal snan and San si in Sum pa mkhan po Ye ses dpal 'byor's dPag bsam
!jon bzati, f. IOIas (p. ISS).
135 See the title-page of the 1978 Dharamsala edition of the Zabs btags rna version.
136 The form dPa' bied is found in Sa skya PaI:19i ta Kun dga' rgyal mtshan, sKyes bu
dam pa mams la sprin ba'i yi ge (sDe dge ed.), f. 72b4, while the Thub pa'i dgons pa rab tu gsal
ba, f. sob2, has both dBa' bied (but dPa' bied in the N-GMPP Ms from Nepal, f. 72aS) and
'Ba' bied; and the same text mentions in addition a rCyal bied (cf. above, n. 134). 'Gos
gZon nu dpal's Deb ther stion po, ka, f. 20a4, and Sum pa mkhan po's dPag bsam !jon bzali,
f. IOIas (p. ISS), have sPa bied (gtsmi rna). The spelling rBa bied is found in Bu ston's Chos
'byun (f. 93b2), and in dPa' bo gTsug lag phren ba's mKhas pa'i dga' ston, ja, f. I2oas.
As for the family name, the forms sBa, rBa, dBa', dBa's 'Ba' and Bha are all found (as
69
70
k~ita
71
sBa bzed published by Stein must for the reasons mentioned above
be considered as a whole to be much later than the eighth
century, and although the recensions of the sBa bzed now
available to us differ in wording and in many details, there would
nevertheless seem to exist no compelling reason to reject as
completely spurious and unreliable the matter on which the
recensions agree in substance. And there is reason to think that
both these recensions contain ancient records or traditions (bzed
lugs) that could go back to members of the sEa family which
played so important a r<~le at the time of the foundation of bSam
yas and the controversy between the 'Gradualists' and 'Simultaneists' in late eighth-century Tibet, and that we thus have
reflected (however indirectly) in our texts of the sBa bzed the
views of major participants in these events.
Bod kyi rgyal rabs, an old chronicle in some three folios only by
the Sa sky a hierarch Grags pa rgyal mtshan (rr47-12I6), has not
entered into doctrinal matters, and this very short work contains
no reference to the Hva san and the controversy between
Gradualists and Simultaneists. And in the Chos la 'jug pa'i sgo by
the Sa skya hierarch bSod nams rtse mo (rr42-rr82), a work that
does include briefly at the end some Chos 'byun type material
relating to Tibet, we also find no mention of these matters. It
seems, then, that in the Sa skya historical tradition Sa sky a PaI).4i
ta (rr82-I25I) was the first of the great hierarchs to direct his
attention to the controversy between Santarak~ita's and Kamalasila's school and the Ho-shangs, which he has done in several of
his writings.
In the history by Ne'ujNel pa PaI).4i ta Grags pa smon lam blo
gros, the sNon gyi gtam Me tog phren ba (dated to 1283 or 1343)
which on many points follows traditions different from those
found in the Chos 'byun of the author's contemporary Eu ston,141
141 Cf G. Roerich, The Blue annals (Calcutta, 1949); p. viii; H. Uebach, in M. Brauen
and P. Kvaeme (eds.), Tibetan studies (Zurich, 1978), pp. 219-230; and in B. Aziz and
M. Kapstein (eds.), Soundings in Tibetan civilization (New Delhi, 1985), p. 147 f
72
we find an allusion to the debate (rtsod pa) between the sTon min
and the rTse min (f. 22ar), the Tibetan transcriptions of the
Chinese names of the two opposed tendencies of'Simultaneism'
and 'Gradualism'. Ne'u Pa.Q.<;li ta has however provided no details
about the doctrinal points at issue except to say, very interestingly, that the Hva san Mahayana taught a doctrine that was in
agreement with the Mahamudra (phyag rgya chen po 'thun pa'i
grub mtha' bzuri, f. 2rb5). This point had already been made
earlier by Sa sky a Pa.Q.<;li ta. 142 Ne'u Pa.Q.<;li ta concludes his
account of the matter by stating ( 22ar) that the monk
Mahayana was defeated and reporting that it was thereupon
decreed that only Dharma '(chos 'ba' zig) - that is, evidently, the
Dharma taught by Santarak~ita and KamalaSlla following Nagarjuna - should henceforth be practised in Tibet, and that nonDharma (chos ma yin pa) should not be practised ( 22a2).
The account of the controversy and ensuing debate between
the 'Gradualists' and 'Simultaneists' provided by Bu ston Rin
chen grub (r290-r364) in his gSuri rab rin po che'i mdzod
(if. r27a-r29b) has hitherto been no doubt the best known one
since it has been drawn upon by a number of later Tibetan
historians. Bu ston recounts how, in view of disagreements
between Santarak~ita's followers such as Ye ses dbail po and the
Hva sail Mahayana, Kamalaslla was invited to Tibet to replace his
master Santarak~ita, who had recently died as the result of an
accident. A debate was then organized with the King sitting in
the centre as witness and arbiter, and with the Hva san placed to
his right and KamalaSlla accompanied by the Tsen min to his left.
Bu ston explains that the Chinese term tsen min (pa) corresponds
to rim gyis pa 'Gradualist', and that the Chinese term ton mun (pa)
corresponds to gcig car ba 'Simultaneist'. According to Bu ston's
account, the points at issue were the Hva sail's teaching of the
need for simultaneous engagement (gcig car [du] 'jug pa) while
giving up all activity and thinking and the teaching of Kamalaslla's school concerning the need for gradual engagement (rim gyis
142 See below, p. lorfE 'Brug pa Kun legs has also spoken of a phyag rgya hva san gi Ita
ba in his gSUIi. 'bum, kha, f. 14a (quoted by R. A. Stein, Revue de I'histoire des religions 179
[1971], p. 10 note). Compare also the colophon of the bSam gtan mig sgron ascribed to
gNubs Sans rgyas ye ses (Leh, 1974), f. 254a, for references to the relation between the (da
Ita'i) phyag chen and the teaching of the Hva san Mahayana.
73
143
74
else would be of help in this task (f. 21a5). Another historian from
the Karma pa school, dPa' bo gTsug lag phren ba (1504-66), has
on the contrary dealt in detail with the Great Debate and with the
role of Kamalaslla in his mKhas pa'i dga' ston Ga, if. II4a-I22b).
The treatment of the Great Debate and ofKamalaSIla's role in the
Chos 'byuri bstan pa'i padma rgyas pa'i iiin byed by the 'Brug pa
bKa' brgyud pa Padma dkar po (1527-1592) is very closely
related to Bu ston's account and thus to that of the sBa bzed. 144
Thus, in the earlier Tibetan historical literature up to the middle
of the fourteenth century as available to us until recently, apart from
the sBa bzed the exact school-links of which are not altogether dear,
the encounter and controversy in Tibet between the 'Gradualists'
and 'Simultaneists' and the Great Debate are found mentioned
above all in sources connected with the Sa skya pa and, probably, the
bKa' gdams pa schools. Later the subject was treated by bKa' brgyud
pa authors too such as dPa' bo gTsug lag phren ba, who-has based his
account on the sBa bzed (including the 'Alternative Tradition'
reproduced in the unsupplemented recension of the sBa bzed) and
has also quoted (f. II9b) Bu ston's Chos 'byuri.
II.
NAN N1 MA
'OD ZER
75
The date of birth of Nan ral has been variously given as II24
and II36, and his death has been placed in either II92 or 1204. He
is renowned as a master of the rDzogs chen, and as a Discoverer
(gter stan) of both rNm rna and Bon po texts (gter ma).146 He thus
belongs to' a tradition quite different from the author(s) of the sBa
bzed, from Bu ston and the other gSar rna pa authors. of later
Chos 'byuns who follow the sBa bzed and/or Bu ston in their
accounts of the Great Debate, and from Sa skya Pal).c;li tao
Ni rna 'od zer's family name Nan/Myan (both spellings have
the same pronunciation) might possibly suggest a link, however
distant, with the tradition of Myan/Nan Tin ne 'dzin (bzan
pO),147 the preceptor of the young Khri IDe sron btsan (Sad na
legs, reigned c. SOO-SI5?), a disciple ofVimalamitra from whom
he received the sNin thig teaching, and hence a very important
early teacher of the rDzogs chen. As already noted above, Myan
Tin ne ' dzin bzan po figures as an opponent of sBa Yeses dban
po. Another master evidently connected with the Myan/Nan
tradition, Myan/Nan Sa mi, is also known as the associate of a
Hva san sometimes known as Me 'go/mgo and of sNa Bye rna
(sBa bzed, G, p. 67-6S, and S, p. 57).
The Ch05 'byuti Me tog sniti po accordingly is one of our earliest datable Tibetan sources concerning the encounter between
'Gradualists' and 'Simultaneists' and the Great Debate. Its author
is senior by at least half a century to Sa skya Pal).c;li ta, who has
hitherto been our oldest securely datable authority on the subject.
The Chos 'byuti Me tog sniti po may then be surpassed in antiquity
only by sBa records incorporated in the sBa bzed, and perhaps by
146 On Nan ral and the question of 'ambivalent' - i.e. Buddhist (rDzogs chen) and
Bon po - gter stons, see A. M. Blondeau in 1. Ligeti (ed.), Tibetan and Buddhist studies
commemorating the 200th anniversary of the birth of Alexander Csoma de Koros (Budapest,
1984), p. 77 ff., and MeisezahI's introduction to his facsimile edition of the Chos 'byun.
147 Richardson, A corpus of early Tibetan inscriptions, p. 44, regards such a connexion as
perhaps speculative. The Myan ban Tin ne 'dzin (bzan po) founded the dBu ru Zva'i lha
khan 50 miles north-east oflHa sa on the Man ra chu. As for the Nan bdag and Nan ral Ni
rna 'od zer, though born in lHo brag, hisgdan sa is to be found in Myan stod according to
the 'Dzam glin rgyas Mad (f. 65a; Wylie, p. 71). And the Myan clan to which Tin ne 'dzin
belonged may have originated in the upper valley of the Myan chu around rGya mda'
according to Richardson (Corpus, p. 44; see also Bulletin of Tibetology 4 (1967), p. 19 n. 10
on the location of Myan). Eva Dargyay has listed Myan Tin ne 'dzin as an ancestor of Nan
ral Ni rna 'od zer in her Rise of esoteric Buddhism in Tibet (Delhi, 1977), p. 57, but without
clearly giving her source.
another source stemming from the same school as Nan ral, the
bSam gtan mig sgron ascribed to gNubs Sans rgyas ye ses the date
of which is, however, not established with certainty and which
does not in any case explicitly mention the Great Debate even
though it has much to say on the two schools of thought that then
confronted each other.
It is at all events very important for the question of the
reliability of the Tibetan accounts of the Great Debate that the
version given in the Chos 'byuri Me tog siiiri po is in many parts
close to, and sometimes indeed practically identical in wording
with, the sBa bzed. In particular, this text relates the defeat of the
Hva san Mahayana by the 'Gradualists' headed by Kamalasila and
his subsequent departure from Tibet. The only possible conclusion seems then to be that if Nan ral did not actually follow the
sBa bzed in one of its recensions, he was making use of either a
source of the extant recensions of the sBa bzed or of some other
text closely related to it.
The Chos 'byuri Me tog siiiri po thus seems effectively to dispose
of the suspicion that the account of the Great Debate and of the
Hva san's defeat by Kamalasila's school to be found in the sBa
bzed, in Eu ston's Chos 'byuri and in the many later sources that
give this version of events was nothing but a tendentious sectarian
fabrication of Atisa's bKa' gdams pas, and of other gSar ma pas
such as the Sa skya pas and dGe lugs pas, which was concocted in
order to discredit the traditions including the Dhyana ones
associated with the rNin ma pasjrDzogs chen pas. 148 It has to be
recalled also that in his Deb ther dmar po, composed in 1346 after
the Chos 'byuns of both Ne'u Pal;u;li ta and Eu ston, Kun dga' rdo
rje has made no point of mentioning the Great Debate or the H va
san's defeat, something he might be expected to have done had
this version of events been a fabrication of the gSar mas pas. It is
equally noteworthy that whereas Sa skya Pa1].c;li ta attached
importance to criticizing the Hva San's teachings, the Great
Debate and the Hva San's defeat have been mentioned by neither
of his two great predecessors as hierarchs of Sa skya - Grags pa
148 On the relationship between the sBa hied and the rNiD. rna tradition, see also Sog
bzlog pa, Nes don 'hrug sgra, f. 6a f.
77
rgyal mtshan in his Bod kyi rgyal rabs and bSod nams rtse mo in
his Chos la 'jug pa'i sgo - although these two authors were
contemporaries of the rNin rna pi Nan ral who has recounted
these things.
The account given by the rDzogs chen pa Nan ral of the Great
Debate and especially of the Hva san's defeat is all the more
significant for the evaluation of the authenticity and reliability of
the sources containing them because at least as early as the time of
Klon chen rab 'byams pa (I308-I363) rDzogs chen pa authorities
have inclined to look on the Hva san Mahayana's teachings with
favour (see below, p. I02).
79
At all events, the fact that the Chos 'byuri Me tog siiiri po does
mention a formal debate in which KamalaSi:la and his followers
prevailed over the Hva san Mahayana lends no support to the
opinion that the account found in the sEa bzed and in Bu ston's
Chos 'byun represents nothing but a deliberate and tendentious
distortion of events by opponents of the 'Simultaneists' and of the
rNin ma/rDzogs chen traditions motivated by sectarian bias. An
argument ex silentio based on the absence of an explicit reference
in the Chinese documents to a formal debate between the Hoshang and Kamalasi1a or his Tibetan followers cannot be regarded
as conclusive, especially in view of the clearly polemical character
of Wang Hsi's Cheng-Ii chueh studied by Demieville.
(2) In agreement with other sources, the Chos 'byuri Me tog siiiri
po (f. 433a-b) has placed the master dPal dbyans among the
opponents of the Hva san and the cig char 'jug pa or 'Chinese
theory' (rgya'i Ita ba) of the sTon min pas. This connexion is in
accord with the attachment of dPal dbyans to the 'Ba'lsBa family
(for example in the sEa bzed). Tucci 151 has entertained the
hypothesis that since a (certain) dPal dbyans is stated to have
belonged to the gNan family, and indeed to have been a disciple
of Vimalamitra and gNags Jnanakumara,152 the attachment of
the participant in the debate named dPal dbyans to the party of
Kamalasi1a and sBa Ye ses dban po may simply reflect a desire on
the part of the author(s) of the sEa bzed to glorify their own
family. But the fact that the Chos 'byuri Me tog siiiri po also regards
this dPal dbyans as an associate of Kamalaslla and Yeses dban po
goes against this hypothesis; and the alternative view considered
by Tucci, namely that we have here two different masters having
the same name, is the likely one. It should be noted, moreover,
that the words ascribed to dPal dbyans in this Chos 'byun, in one
version of the sEa bzed (G, p. 70) and in the rnKhas pa'i dga' stan
(ja, 117b) are ascribed in the tabs btags rna version of the sEa
bzed (S, p. 59) to a certain San si, a name (or title) borne by
another member of the 'Ba' family (S, p. 17. IS). (Strangely, the
151 G. Tucci, Minor Buddhist texts, ii, pp. 20-21, ISO. Cf. Demieville, in: M. Soymie
(ed.), Contributions aux etudes sur Touen-houang, p. II (referring to Z. Yamaguchi).
152 See mKhas pa'i dga' ston, tha, f. 25a (p. 215.7); Deb ther slion po, ga, f. 2a (p. 104). On
gNan dPal dbyans see R. Kimura, Journal asiatique, 1981, pp. 191-2.
80
152a The problem is compounded by the fact that in another historical work giving an
account of the Great Debate a~d also ascribed to Nan ral - the Mi rje Ihas mdzad Bymi
chub Se/'ns dpa' sems dpa' chm po chos rgyal mes dbon rnam gsum gyi rnam par thar pa, Rin po
che'i phreli ba published in 1980 at Paro - we read dbu rna rnam pa Ina rna mthull te ston min 1
char du 'jug mchis te (f. I2Sb4-S). (The authorship of this work is not clear, and one folio is
missing just before the last page of the reprint.)
81
153 The *Prasangika school of Madhyamaka is not known to the bSam gtatl mig sgroll.
Ron zom Chos kyi bzan po (eleventh century) mentions only the mDo sde dbu ma and
the rNal 'byor spyod pa'i dbu ma in his Man liag Ita phreti 'grel pa (f. 28b), ITa ba'i brjed
byan (f IIb-I2a) and Grub mtha' brjed byan (f 5a-6a). However, Klon chen rab 'byams pa
(1308-1363) has ~ecognized the *Prasangika branch, for example in the Grub mtlla' mdzod
(f 54b f. ~ 40a f).
154 Tucci has supposed that KamalaSIla may have been killed by Bon pos (Millar
Buddhist texts, II, p. 45). But since the reference is to executioners from China, could the
mu stegs pa in this case have been Taoists (or even Buddho-Taoists)? On Taoism in Tibet at
this time, see Demieville in M. Soymie (ed.), Contributions aux etudes sur Touw-houallg,
p. 6. Compare for example the Li'i yul luti bstan pa, translated by F. W. Thomas, Tibetan
literary texts and dowments concerning Chinese Turkestan, I (London, 1935), p. 84
82
longer be any mu stegs pas in Tibet, but that instead a split would
develop in the Buddhist Dharma itself.1 55 Bu ston (or possibly a
source of his) may have noticed this discrepancy between the
account that Kamalaslla was murdered by an agent of the mu stegs
pa and what Santarak~ita had foretold in his testament, which he
has recorded in his Ch05 'byuri.156 For according to Bu ston
Kamalasila was murdered by four executioners from China
belonging to the Hva san (hva sari gi rgya'i bJan pa mi bzi,
f. 129b6), a version of events that was later followed by Padma
dkar po (Chos 'byuri, f. 165a6) among others. In other words, Bu
ston (or a source) may have engaged in a kind of rational
reconstruction of events and concluded - given what Santarak~ita
was held to have foretold - that it must have been the Ho-shang's
party that was responsible for Kamalasila's alleged murder. In
these circumstances, fully satisfactory grounds hardly exist for
explaining Bu ston's version (along with that ofPadma dkar po
and others) by simply ascribing to him a desire to denigrate the
Ho-shang.
It is of special interest to observe in this connexion that the
Ch05 'byuri Me tog siiiri po (f. 138a) not only knows nothing ofBu
ston's version mentioning executioners sent by the Hva san as
KamalaSlla's murderers but distances itself from the sBa bzed's
account of a murder by agents of the mu stegs pa by qualifying it
as a report by means of the verb zer ba 'it is said'. And it states
instead that KamalaSlla was killed by an Indian servant Dhanasrl
who was looking for gold.
(5) In the Ch05 'byuri Me tog siiiri po (f. 429b-430a) it is related
that at the start of the Great Debate at the Byan chub glin of
bSam yas the King decreed that the loser should, as the price of
defeat, receive a punishment (chad pa gcod do), which is however
not specified. But no mention is made in this text of the
banishment from Tibet by royal command of the Hva San
Mahayana after he lost the debate. According to this account the
155 sBa bied, G, pp. 66,72; and S, p. 56; Bu ston, Chos 'byun, f. 127b; dPa' bo gTsug lag
phren. ba, mKhas pa'i dga' ston, ja, f. !ISb. Cf. Chos 'byun Me tog siiili po, f. 427b-428a.
156 Bu ston, Chos 'byun, f. 127b. It should however be noted that the teaching of nonactivity which was Mo-ho-yen's, and to which KamalaSi:la refers in his Bhavanakrama (iii,
p. 20), has been connected by Kamalasna with teachings of the mu stegs can (tlrthikas) called
kun tu tshol ba, evidently the Ajlvikas. See below, p. 142.
83
Hva Sail left, apparently voluntarily, not for China but for Bodh
Gaya (rDo rje gdan) in India (f. 436a-b).
In the sEa bzed (G, p. 75 and S, p. 62) it is related that the Hva
Sail returned to China (slar rgya nag tu gsegs) (cf. mKhas pa'j dga'
ston, ja, f. iI9a). The rGyal rabs gsal ba'j me Ion (p. 182) too states
that he departed (bzud pa) for China. For their part, Bu ston
(Chos 'byun, f. I28b2) and Padilla dkar po (Chos 'byun,
f. I65a3-4) report that already at the very start of the Great
Debate the Tibetan monarch decreed that the loser was not to
remain in his realm; and these two sources relate that after losing
the debate the Hva sail was sent off (rdzon ba) - i.e. probably, was
banished - to China (Bu ston, f. I29b6; Padma dkar po, f. r65a5).
Wang Hsi's Cheng-Ii chiieh on the other hand refers neither to a
defeat nor to the banishment of Mo-ho-yen; and it mentions (f.
I29a) an official edict authorizing the practice of his Dhyana
teaching. The Chinese material indicates furthermore that Moho-yen returned to Dunhuang.1 s7
The discrepancies in the accounts of the fate of Ho-shang Moho-yen after the Great Debate have given rise to the hypothesis
that he was involved in more than one debate, and that whereas
he won the earlier debate he was finally defeated in a later one. 1SS
This is of course not impossible, but it would be as difficult to
prove as it would be to disprove on the basis of the available
evidence; it is at least equally possible that the divergence between
the accounts reflects not the different outcome of two or more
debates, but differing views of a single set of events.
(6) Of special interest is the account given in the Chos 'byun Me
tog sfiin po of the fate of the Hva sail Mahayana's teaching in
Tibet.
Both recensions of the sEa bzed (G, p. 75; 2, p. 62) and the
mKhas pa'i dga' ston (ja, f. II9a) relate that the Tibetan ruler
condemned the practice of the Hva sail's Dharma (known as the
157 See Demieville, Candle, pp. 253 and 278, who has identified Mo-ho-yen with a
Chinese Bhadanta living at Dunhuang after the time of the Great Debate and occupying a
high post in the local administration there. Cf. Imaeda, Journal asiatique, 1975, p. 140-L
158 Demieville distinguished at least three distinct sessions; see Toung Pao 56 (1970),
pp. 40-42. See also D. Ueyama, TohO gakuhO 35 (1964), pp. 141-214, with Demieville's
summary in Toung Pao 56, pp. 29-43 (and pp. 43-44 on Z. Yamaguchi's critique of
Ueyama); Imaeda, Journal asiatique 1975, pp. 126, 129, 140.
ston min pa cig char 'jug). And the Zabs btags ma versi:m (S, p. 62),
followed by the mKhas pa'i dga' ston (f. II9a), further specifies
that the King decreed that henceforth the theory of Nagarjuna
should be accepted and that in the sphere of practice the Six
Perfections should be adhered to. This version is followed in
substance by Bu ston (Chos 'byuri, f. 129b) and Padma dkar po
(Chos 'byuri, f. 165a). In the words of Ne'u Pat;lQ.i ta's Chos 'byuri
(f. 22a), after the defeat of the Hva Sail, non-Dharma (chos ma yin
pa) was not to be practised, and Dharma alone (chos 'ba' zig) that is, evidently, the Dharma taught by Kamalasi:Ia following
Santarak~ita and, ultimately, Nagarjuna - was to be practised.
Interestingly, the Blon po bka'i thari yig (f. 28a), a section of the
bKa' thari sde lria which reproduces rDzogs chen traditions (cf.
Pelliot tibetain 116), specifies that it was the Yoga-Madhyamaka
(rnal 'byor dbu ma'i gzuri) that was to be followed, but without
mentioning KamalaSi:la in this connexion; this source in fact states
that the ston mun cig car 'jug pa is the Madhyamaka.
Now the Chos 'byuri Me tog sniri po (f. 435b) relates that, after
the Hva sail Mahayana had conceded defeat in the Great Debate,
. the debate was reconciled on the side of Dharma (rtsod pa chos
phyogs su 'dum par by as so). And the King declared that in
substance there was no disagreement (don la mi mthun pa tsam mi
'dug ste) between the two parties to the debate, and that with
respect to the method of practising the Path the Hva Sail's
Dharma, known as the cig char du 'jug, is a teaching concerning
persons whose faculties are highly developed (dbari po yari rab
sbyaris pa cangyis chos [variant from Ms B: lam] yin la).1 59 But, the
King added, the ten Dharma-practices had been condemned [by
the Hva Sail] starting with the case of those whose faculties are
middling (dbari po 'briri man chad chos spyod bcu la skyon bskal):
Mind thus becomes drowsy (sems ni byiri) , good equipment
(tshogs = sambhara) is not accumulated, and because others' mental training is interrupted the Dharma also declines and is
interrupted. And, the King concludes, you [i.e. the Hva Sail] must
practise meditative realisation (bsgoms fig). Henceforth, in the
sphere of theory (lta ba), the theory of Nagarjuna should be
159 On the highly developed faculties (dba,; po rnon po = tik,,!endriya) required to
penetrate the Hva san's teaching, see also above p. 66.
85
accepted, and in the sphere of praxis (spyod pa) the Six Perfections
should be practised; the ten Dharma-practices (chos spyod bcu)
should be exercised; in the sphere of meditative realization (bsgom
pa) mental training is through the three kinds of discriminative
understanding (ses rab = prajna); Means (thabs = upaya) and
Prajfia should be yoked together (zuri du 'brel bar gyis), and one
should engage thus in meditative realization.
.
This version of the King's decree does not, it is true, actually
disagree with the words found in the sBa bzed, most of which it
contains while omitting only a few. But by including several
additional phrases the Chos 'byuri Me tog sniri po has nevertheless
put another complexion on the Great Debate and the King's
decree following it. Thus, the King's remark that the Hva san
Mahayana's teaching does not disagree in substance with that of
Kamalasi:la's school and that it is the teaching (chos) - or,
according to a variant reading, the Path (lam) - followed by
advanced disciples effaces the radical opposition between the
'Gradualist' and the 'Simultaneist'.160
Nevertheless, in the Chos 'byuri Me tog sniri po there is recorded
the King's decree to the effect that Nagarjuna's theory was
thenceforth to be accepted, and that the practice of the Paramitas
and the yoking together in yuganaddha of Upaya and Prajfia
should be observed.
Interestingly, the bSam gtan mig sgron has quoted (f. 23b-24a) a
verse from the rTen 'brelsniri po, stating that it provides the source
for the meditative realization of the sTon mun, in other words of
the Hva san Mahayana's school. Now this rTen 'brei sniri po must
be the Pratltyasamutpadakarikas ascribed to Nagarjuna; for the
verse quoted corresponds to verse 7 of this text (which is
paralleled by Ratnagotravibhaga i. 154 and Abhisamayala1'/1kara
V.21, and partially by Asvagho~a's Saundaranandakavya xiii.
44).161 It thus appears that an important rDzogs chen text that
160 This view of the matter is similar to the one attested in the bSam gtan mig sgron. And
it is even attested in an intervention in the Great Debate ascribed either to sEa dPal dbyails
or Sail si (see below, p. 86).
161 This verse is quoted also in the dMyigs su med pa tshul gcig pa'i gzuti (Pelliot tib. 116),
[. 164 (cf. F. Faber, Acta Orientalia 46 [1985], pp. 71-72), and in Vimalamitra's Cig car 'jug
pa rnam par mi rtog pa'i bsgom don (D, [. rob). For a recent discussion of the
Pratftyasamutpiidahrdayakiirikiis, see C. Dragonetti, WZKS 30 (1986), pp. 109-22 (where
86
87
PP4 I -4 2 .
167 For a bibliography of this practice, see Tucci, Minor Buddhist texts, II, pp. ro, 284;
E. Lamotte, Traite de la Grande Vertu de Sagesse (Louvain, 1949), p. 740 n. I; J. Filliozat,
Journal asiatique, 1963, pp. 21-51, and Arts asiatiques IS (1967), pp. 65-88; P. Demieville,
Choix d'etudes bouddhiques (Leiden, 1973), pp. xxxviii-xxxix, 264-5. See below, pp. 149-50.
88
Me tog siiin po (f. 429b4) relates that the Hva sail MijMe go died (?
bro dor),168 while his associates - Nail Sa mi, rNog and sNa 169 'contracted illness' (na tsha skyed). The Zabs btags ma version of
the sBa bzed (S, p. 57) also states that the three 'contracted illness'.
Another version of the sBa bzed however specifies that they died
as the result of self-mutilation (G, p. 67--68; cf. mKhas pa'i dga'
ston,ja, f. II6b).
As for the report that the Hva Sail Mahayana departed for
China after leaving his boot behind in Tibet, it is found also in
the 'Alternative Tradition' of the sBa bzed (G, p. 75.8, not
reproduced in the mKhas pa'i dga' ston, f. 120). The motif of the
boot left behind is known also in connexion with another Hva
san (perhaps one connected with sBa gSal snail in the sBa bzed,
G, p. 9'" S, p. 8).170 The association of this motif with
Bodhidharma is of course known from the Chinese Ch'an
tradition; and in Tibetan sources it is found in connexion with
the Dhyana-master Bodhi-Dharmottara in the bSam gtan mig
sgron (f. 12b).171
(8) It is significant that in the Chos 'byun Me tog snin po
(f. 425'a6) the expression dkar po chig thub - a term presumably
borrowed from the vocabulary of Tibetan pharmacology and
denoting in the present context a spiritual sovereign remedy - is
recorded as a description, based on a medical metaphor, of the
Hva san Mahayana's 'spontaneist' and cognitively nativist teaching of face-to-face intuitive confrontation with and comprehension of Mind (sems no 'phrod pa, etc.).
The same expression is found in addition in the 'Alternative
168 See Chos 'byun Me tog sfiin po, f. 426b4, where the expression bra bar is found. In
standard classical Tibetan, bra bar ba and bra dar ba mean 'to swear an oath' (mna' bskyal
bajskyel bal. In the present context, however, not only does this meaning not fit well but
the corresponding passages in the parallel sources indicate that the appropriate meaning is
'to die' (Ii in sBa bied, G, p. 68.1, and in dPa' bo gTsug lag phren ba's mKhas pa'i dga' stan,
ja, f. 116b4; 'das in mKhas pa'i dga' stOI1, ja, f. IIsa5).
169 See above, p. 61.
170 See also 'Gos gZon nu dpal, Deb ther snon po, ka, f. 2Ia3 (Roerich,. p. 41); bSod
nams rgyal mtshan, rGyal rabs gsal ba'i me Ion (ed. Kuznetsov), p. 182. Cf. bSam gtan mig
sgron, f. 12 (in connexion with 'Bodhidharmottara'); mGon po skyabs, rGya nag chos 'byun
(Sichuan ed., 1983), p. 122. See also G. Tucci, Tibetan painted scrolls (Rome, 1949), p. 615
n. 252, and Minor Buddhist texts, II (Rome, 1958), p. 44.
171 Compare Thu'u bkvan Blo bzail Chos kyi iii rna, Grub mtha' sel gyi me lon, rGya
nag Chap., f. IJa.
89
Tradition' of the sEa bzed (G, pp. 72-'75), from where dPa' bo
gTsug lag phren ba has taken it (mKhas pa'i dga' ston, ja,
f. I2oa6-7). And it is known too from Sa skya Pal).Q.i ta's
treatment of the Hva San's teaching in connexion with his
criticism of the Neo-Mahamudra (da lta'i phyag rgya chen po) and
Chinese-style rDzogs chen (rgya nag lugs kyi rdzogs chen)J 72 A
corresponding medical concept is, moreover, found in Wang
Hsi's Cheng-Ii chiieh (f. I46b); there, in the context of Mo-hoyen's own presentation of his teaching on non-reflection and
non-examination, the MahCiparinirvCiIJasutra is quoted on the
subject of the medicine called agada, which is said to heal all
illnesses. 173
This attestation in the Chos 'byuri Me tog sniri po - which as seen
above is not unfavourable to the Hva san Mahayana - of the
expression dkar po chig thub as a description of the 'Simultaneist'
teaching, supported as it is by the presence of a parallel medical
description ascribed to Mo-ho-yen himself in the Cheng-Ii chiieh,
disposes of the suspicion that the comparison with the Hva san's
teaching of certain other doctrines current in Tibet that also made
use of the image of the dkar po chig thub as a spiritual panacea was
arbitrary and nothing but a transparent polemical device used by
Tibetan opponents of these later teachings such as Sa skya Pal).Q.i
ta (see below, Chap. iii).
90
175
9I
92
III
177 Cf. Wang Hsi's Cheng-Ii chiieh, f. 134a If. (translated by DemievilIe, Concile,
pp. 75 If.); Stein 468 (cf. L. Gomez, in R. Gimello and P. Gregory [eds.], Studies in Ch'an
and Hua-yen [Honolulu, 1983], p. 107); Pelliot tibetain 21 (cf. L. Gomez, Studies, p. 124).
94
nirupayato yo 'nupalambhah
yin pa).179
178 Cf. Cheng-Ii chiieh f. 135a f. (Demieville, p. 76 f.); Stein 468 (Gomez, p. 108) and 709
(Gomez, p. 114); Felliot tibhain 823 (Gomez, p. I26). Cf. bSam gtan mig sgron, f. 73a f.
179 KamalaSi:la has taken up this point in his AvikalpapravdadhiiralJ'- Ttkii (P,
f. 156b-I57b). There he observes first that when something is perceived by being
presented in cognition (snari bar 'gyur bas mrion du 'gyur pal, it is something that may then
be removed through non-mentation (amanasikiira). Next he argues that such amanasikiira is
not mere absence of mentation [in the sense of absolute, non-presuppositional and nonimplicative, negation, or prasajyaprati~edha]. For, non-existence being no thing (drios po
med pal, it cannot serve as the cause for anything at all; and without correct analytical
examination (bhutapratyavek~ii) it is impossible not 'to attend (manas-kr-) to the phenomenal
signs (nimitta) of matter (rupa) and the other [skandhas] presented in cognition. Nor,
however, does something other than this mentation constitute amanasikiira [in the sense of
relative, presuppositional and implicative negation, or paryudiisa]; for it would then follow
that some other thing such as rupa and the other [skandhas] too could be amanasikiira,
[amanasikiira] being then not the counter-agent (pratipak~a) against them [as is required by
the theory]. Accordingly, what was intended [when amanasikiira was spoken of in the
Avikalpaprave.!adhiirafJi] is that an amanasikiira that is the characteristic (lak~afJa) of
bhutapratyavek~ii - the contrary of that manasikiira [which is to be counteracted] constitutes amanasikiira.
Alternatively, because amanasikiira is a product (phala) [of analytical examination], it has
been stated that bhutapratyaveklii is to be designated metonymically by the term
'amanasikiira'. That is, by merely indicating its product, it becomes evident by implication
(arthasiimarthya) that [analytical examination as the cause of amanasikiira] is to be effected.
So it is possible fully to remove the phenomenal signs (nimitta). For, granted that the
Yogin thus analytically examines phenomenal signs such as rupa presented in his cognition
[even though] in a form that is erroneous (viparyasta) owing to the force ofmisknowledge
95
where Quieting and Insight operate in conjunction (Samathavipasyanayuganaddhavahi margo ni~pannah = zi gnas dan lhag mthon zun
du 'brei ba'i lam grub pa).185
The concept of the balanced process of Mind in samatha and
vipasyana, as a consequence pf which there is neither drowsiness
(laya = byin ba) nor excitation (auddhatya = rgod pa) of Mind, is
further illustrated by the image of a pair of oxen going along
yoked together (yuganaddhavahibalivardadvaya = glan giiis zun du
'breI ba).186 Crucial in this context is exact analysis (bhutapratyavek~a = yan dag par so sor rtog pa) leading to analysis of the factors
of existence (dharmapravicaya = chos sin tu rnam par 'byed pa) and
discriminative understanding (prajiiii = ses rab), and to Insight
(vipasyanii = Ihag mthon).
According to Kamalaslla, then, there should indeed be a
simultaneous operation (cig car 'jug pa) of Means (thabs = upaya)
consisting in generosity, etc., and ofPrajiia, this being the Path of
their joint processing (yuganaddhavahi margah).1 87
Such practice leads to the Bodhisattva's achievement of objectifying the entirety of things (vastuparyantatalambana = dnos po'i
mtha' Ia dmigs pa), 18 8 to his birth in the Tathagata-Family
(tathiigatakuIa), to his entry into faultless determination (skyon med
pa = niyama),189 and thus ultimately to buddahood.
Bhavanakrama III, p. 9.
186 Bhavanakrama III, p. 10; cf. I, p.
Bhavanakrama II, p. 71.
188 Cf. Bhavanakrama III, p. 30.6.
Bhavanakrama II, p. 77, quoting the Dasabhumikasutra i, p. 12.
207,
97
clear links between Kamalaslla's Bhavanakramas and Avikaplaprave5adharaJJl- Tlka and Mo-ho-yen's views as found in the
Tibetan fragments from Dunhuang and in Wang Hsi's Cheng-Ii
chiieh.
The tw'o tendencies confronting each other have regularly
been described, and clearly contrasted, in the older Tibetan
historical texts by means of the designations rim gyis pa or rim gyis
'jug pa for Kamalaslla's school and (g)cig c(h)ar ba or cig c(h)ar [du]
'jug pa for Mo-ho-yen's school. The term rim gyis pa, a noun
derived from the locution rim gyis 'gradually', corresponding to
Skt. krameIJa, may accordingly be rendered by 'gradualist', the
emphasis in this tendency being on the step-by-step serial
cultivation (bhavana) through reinforcement (abhyasa) of the Path
of Awakening with its successive gradations (rim pa = krama).
And rim gyis 'jug pa can be rendered as 'gradual engagement' or
'gradual process' ('jug pa = pravrH. To render accurately the
second term (g)cig c(h)ar ba is somewhat more difficult. Since in
this tendency emphasis is put on the immediate, instantaneous,
simultaneous and holistic - i.e. the single-moment - nature of
Awakening, and because the term employed is derived from the
expression cig c(h)ar du 'in one instant, simultaneously' which is
used to render the Sanskrit words yugapat and sakrt and which
may be glossed by dus gcig tu 'at one time, at once' (Skt.
ekavaram), the term can be rendered either as 'instantaneous/
instantaneist' or as 'simultaneous/simultaneist'. The frequently
employed renderings 'sudden' for cig c(h)ar and 'Subitist' for cig
c(h)ar ba are of course appropriate also to the extent that these
words - together with Skt. sakrt and ekavaram - are additionally
associated with the idea of suddenness; but only occasionally (see
below) is the word cig c(h)ar actually used to describe actions that
are sudden and abrupt (glo bur) as distinct from those that are
either simultaneous with each other or instantaneous.
In our Tibetan sources 'Gradualism' is in addition often
referred to by the expression (br)tse(n) min/mun, and the 'Gradualist' by the word (br)tse(n) min/mun pa. And 'Simultaneism' is
known as the (s)ton min/mun, and the 'Simultaneist' as a (s)ton
min/mun pa. These words, which are clearly not Tibetan in origin,
correspond respectively to the Chinese expressions chien men
(p'ai) '(school of) gradual entrance' and tun men (p'ai) '(school
of) immediate entrance' .190 Another contrasting pair of expressions in Chinese are chien wu 'gradual Awakening' and tun wu
'sudden Awakening'. It is however not ccertain that simultaneity
and suddenness are in fact totally identical notions in the history
of the Dhyana schools; but the clarification of this point is crucial
for the history of Ch'an rather than for that of the Tibetan
doctrines being considered here. 191
The 'Gradualist' procedure is compared in Tibetan sources
with a progressive, step-by-step ascent toward a mountain peak,
or with a monkey's gradual climbing to the top of a tree from
below (mas 'dzeg). On the contrary, the 'Simultaneist' procedure
is compared with an eagle's sudden or abrupt (gio bur) descent on
to the top of a tree from above (yas babs, yas 'bab) .1 92 This pair of
metaphorical descriptions is thus intended graphically to illustrate
a distinguishing feature of two contrasting procedures.
The Hva sail Mahayana is stated to have said that not thinking
on anything whatever, not conceptualizing anything whatever
and not practising anything whatever constitute an objectifying
that involves simultaneous engagement (dmigs pa gcig char 'jug pa),
so that this is as (one) on the tenth Bodhisattva-stage (sa = bhumi). 193 He is also shown as sometimes asserting that his method is
taught for persons whose faculties are superior (dbari po rnon
po = tT~~1Jendriya), whereas the Dharma-practice based on generosity and so forth (danadi) has been taught rather for those whose
faculties are blunt. 194
Cf. n. 120.
On this terminology cf. Demieville, Concile, pp. ro, I4-I5, I8-I9, 35; n. 50-5I,
74-75, 184, and 279; R. A. Stein, Revue de l'histoire des religions I79 (I97I), pp. 3-30. (See
now also P. Gregory [ed.], Sudden and Gradual [Honolulu, I987].)
192 This comparison is placed in the mouth of the Hva san in the 'Alternative
Tradition' of the sBa b 'ed (G, p. 74), and in Nan Ni rna 'od zer, Chos 'byun Me tog sfiiti po
(ed. R. O. Meisezahl) f. 430b. See however the critique of these examples, as well as of the
corresponding meanings, in the' Alternative Tradition' of the sBa bied (G, p. 74); in dPa'
bo gTsug lag phren ba, mKhas pa'i dga' stan, ja, f. I20b2-3 and f. I22a-b (which refers to
the 'Alternative Tradition'); and in the Chos 'byuti ivle tog sfiiti po, f. 432a. These examples
are also mentioned by Sa skya Pal).,li ta in his sDom gsum rab dbye, f. 25b, and (together
with Kamala!la's critique) in his Thub pa'i dgotis pa rab tu gsal ba, f. 49b.
193 sBa bZed, G, p. 68.20; S, p. 58.7; Chos 'byuti Me tog sfiiti po, f. 430b5.
194 Cf. sBa bied, G, p. 68.I7; S, p. 58.5; Chos 'byuti Me tog sfiiti po, f. 425'b5, 430b3, and
435b3 See pp. 66, 84,93, 117, I4I-2.
190
191
99
24.
IOO
lOr
teaching of non-reflection and non-examination with the agadamedicine mentioned in the MahaparinirvclI}as(ltra, which describes
itself as this antidote that heals all illness.
102
103
10 4
misleading here). This sGom pa Tshul khrims siim po (II16-II69) was the nephew and
disciple of sGam po pa (iia, f.27a-b).
207 C D. Seyfort Ruegg, 'A Karma bKa' brgyud work on the lineages and traditions
of the Indo-Tibetan dEu rna (Madhyamaka)" in: Orientalia 1. Tueci memoriae dieata, III,
pp. 1256, 1258-62.
ro6
with the Tibetan tradition of tathagatagarbha-exegesis that interpreted the Buddha-nature in a 'spontaneist' and 'innatist' fashion.
Now, in several works Sakya mchog- ldan distinguishes between Fixation-Bhavana ('jog sgom) and Inspection-Bhavana
(dpyad sgom) , pointing out that it is necessary to determine
whether Fixation is to be preceded by Inspection or not. 207 < The
dpyad sgom used in dispelling imputation (sgro 'dogs = samaropa) is
connected roughly with Prajiiaparamita-philosophy and the Madhyamaka, the bKa' gdams pa Po to ba being cited as a master of
this method (tsa/I4, f. 24b). As for 'jog sgom, it would correspond
mainly to intuitive awareness (iiams myon), the coincidence of
Bliss and the Empty (bde ston) in Mantrayana and the Mahamudra. And while noting that still other systems of meditation have
been developed based on traditions that are neither Mahamudra
nor Madhyamaka, Sakya mchog ldan points to the fact that they
have been rejected by Sa skya PaI}.Qi ta in his Thub pa'j dgons gsal
(tsa/r8, f. 4a). Sakya mchog ldan furthermore connects analytic
Inspection (dpyad pa) with the scholar-paI}.Qit, i.e. the specialist in
scholastic philosophy (mtshan iiid pa) who engages in pratyavek~a;
Fixation-Bhavana is on the contrary linked by him with that kind
of Yogin who takes everything just as it is, without engaging in
mental construction and analysis, i.e. with the type of practiser
known as the ku sa li pa (tsa/r8, f. 4b; tsa/I4, ff. 15b-r6a; tsa/2I,
f. 7a-b).
With regard to the problem of the kinds ofMahamudra, Sakya
mchog ldan concludes that the theory of the non-duality of
salflsara. and nirva1']a as explicated by Sa skya PaI}.Qi ta and the
Mahamudra-teaching of 'Candraprabha-Kumara' (i.e. sGam po
pa) are, notwithstanding the difference in their names, one in
sense and import (don gcig). This is so in spite of the fact that the
former doctrine as described by Sakya mchog ldan is concerned
with the eradication and stoppage of imputation (samaropa: sgro
207< Siikya mchog ldan, Phyag rgya chen po'i san 'byed ces bya ba'i bstan bcos (Phyag rgya
chen po gsafbar byed pa'i bstan bcos, Tshans pa'i 'khor 10), gSun 'bum, vol. tsa/I4), ff. 8b-I3a
(addressed to a certain Sa skyon mchog, i.e. the Rin spuns pa ruler); Lun rigs giiis kyi phyag
rgya chen po bied tshulla 'khrul pa sel ba'i bstan beos, zun 'jug gi gru chen, gSun 'bum, vol.
tsa/I4), If. 13a-2sa (addressed to Karma dBan phyug dpol); and mKha' spyod dban po'i
spyan drun du 'bul ba'i mol mchid, gSun 'bum, vol. tsa/I8, ff. Ib-sa (addressed to Zva dmar
IV, Chos grags ye ses, 14S3-IS24?); and Replies to the Rin spmis sde pa Siikya rgyal
mtshan and sDe pa gar pa, gSun 'bum, tsa/2I, f. sb If.
107
108
position of the thirteenth-century Sa sky a hierarch on the NeoMahamudra and its connexions with the Hva san's teachings. Go
rams pa's critique of the Hva San's teaching together with the
Neo-Mahamudra may accordingly be considered typical of the
main-line Sa skya school.
Sakya mchog ldan's above-mentioned attempt to harmonize
Sa skya Paf,lc;li ta's teaching with that of sGam po pa in the matter
of the Mahamudra would seem to suggest at all events that he
considered that it was indeed the latter's Mahamudra doctrine
that was the object of the Sa skya hierarch's criticism.
As for the Hva san, according to Sakya mchog ldan his mistake
lay in having failed duly to differentiate between surface-level
saytJVrti and ultimate paramartha, theory and practice, jiiana and
vijiiana, the level oflearning and reflection (thos bsam) and that of
meditative realization (bsgom pa), and the indirect provisional
neyartha and the definitive nttartha, as a consequence of which he
came to believe that mere non-mention (ci yan yid la mi byed pa
tsam: amanasikara-matra) constitutes the essential (tsa/14,
f. rob5-6; tsa/2I, f. 5b). Sakya mchog ldan concludes that in the
true Mahamudra freedom from conceptual construction and
non-mentation are altogether unlike the Hva san's meditation
(tsa/I4, f. rra1-2) and are not dull quietude (ltens po'i gzi gnas,
tsa/18, f. 3b; tsa/2I, ff. 5b, 6b).
Mi bskyod rdo rje (1507-1554) - the eighth hierarch of the
Karma branch of sGam po pa's bKa' brgyud pa school- however
evidently understood Sa skya Paf,lc;li ta's ctiticism of NeoMahamudra as being directed against the non-mentation (yid la
mi byed pa) teaching of Maitripada and also Saraha, sGam po pa's
great predecessors in this Indo-Tibetan lineage. Like sGam po
pa's, their teachings were considered to be linked also with the
Sutra and not solely with the Mantra department ofIndo-Tibetan
thought. 208 Nevertheless, as already noted above, it would seem
that Sa skya Paf,lc;li ta wished rather to distinguish the NeoMahamudra he was criticizing from Nagarjuna's a~d evidently
also from Naropa's and Maitripada's (sDom gsum rab dbye. f. 26a).
With regard to Zan tshal pa as the propagator of the dkar po
208
See D. Seyfort Ruegg, 'A Karma bKa' brgyud work .. .' (n. 203), pp. 1258-62.
109
chig thub in the bKa' brgyud pa schooI,209 Thu'u bkvan BIo bzan
Chos kyi iii ma writes in his Grub mtha' feZ gyi me Zori that up to
the time of Zan Rin po che dkar po chig thub had not been well
known as. a term (tha snad) in that school, but that from his time
onwards it became very well known (bKa' brgyud pa chapter,
f. 19b). And he adds (f. 2Sb) that of the many refutations found in
Sa sky a Pa!).<;ii ta's sDom gsum rab dbye the chief ones turn out to be
directed against the dkar po chig thub teaching of Zan Tshal pa and
the dgoris gcig teaching of 'Bri gun pa; indeed, he notes, many
writers have emulated Sa skya Pa!).<;ii ta and concluded that this
teaching of Zan Tshal pa had the meaning of non-mentation (yid
Iii mi byed pa). Very interestingly, however, BIo bzan Chos kyi iii
ma then observes that if an impartial person considers the sayings
of Zan Tshal pa, it becomes apparent to him that they do not in
fact belong to the position (phyogs) of non-mentation; hence the
refutation in the sDom gsum rab dbye was clearly an over-hasty
statement (thub chod kyi gsuri). As for the fundamental view of
Mar pa, the source of the Dvags po bKa' brgyud, BIo bzait Chos
kyi iii ma describes it as Prasangika-Madhyamaka (f. 17b4).
The fact that the Hva san Mahayana is represented as having
upheld at the Great Debate. the principles of non-mentation and
of not thinking on anything (ci yan mi sems pa) probably accounts
for Sa skya Pa!).<;ii ta's association of him with what he terms the
'almost' Chinese Dharma-system of Neo-Mahamudra, and also
with Chinese-style rDzogs chen. 210 As he has written in his sDom
gsum rab dbye, despite the fact that the Neo-Mahamudra and the
Chinese-style rDzogs chen have different names they are in
substance without difference with respect to their simultaneism
(cig char ba) and subitistic procedure (yas 'bab) (f. 2Sb); and the
Neo-Mahamudra based on the literal wording of the tradition of
the Chinese Master (rgya nag mkhan po'i gzun lugs kyi yi ge tsam)
was for the most part (phal cher) a Chinese Dharma-system
(f. 26a). It is not impossible that in his account Sa skya Pa!).<;ii ta
. was thinking as much in terms of typological strands and family
resemblances between teachings as in terms of direct and immediate historical influences. That he may not have meant to reject all
See above, pp. 103-04.
See also the discussion of Sa skya PaI;tc.\i ta's view in Sog bzlog pa, Nes don 'brug sgra,
if. 117a-I2ob.
209
210
IIO
This passage is quoted from Tson kha pa, Lam rim chen ma (IHa sa ed.), f. Srob.
III
(I059-II09).212 It should nevertheless be noted that Diparpkarasrijiiana's Ekasmrtyupadda (Dran pa gcig pa'i man nag) is
described as an instruction on simultaneous engagement (cig car
'jug pa = yugapadvrtti); both discriminative understanding and
means are involved in this instruction, as is the sequence (krama)
of objectification followed by non-objectification (dmigs pa med
pa) described as effortless and spontaneous (anabhoga).
In his classical treatise on the Path of Awakening, Tsoil kha pa
has devoted much discussion to the relationship in meditative
realization (sgom pa = bhavana) between the settling function of
Fixation-Bhavana ('jog sgom) and the analytical function of
Inspection-Bhavana (dpyad sgom) derived from analysis (dpyod
pa = vicara).213 Typologically speaking, it is no doubt true that
non-mentation and non-analysis as advocated (according to
Tibetan accounts) by 'Simultaneists' such as the Hva sail Mahayana bear a resemblance to Fixation-Bhavana and to Quieting (zi
gnas = samatha). The distinction lies in the fact that alongside
Fixation-Bhavana and Quieting the procedure for philosophical
thought and meditation adopted by Kamalasila and his Tibetan
followers requires in addition the application of the discriminative understanding of investigative analysis (so sor rtog pa'i ses
rab = pratyavek-ia-prajna) and Insight (lhag mthon = vipaSyana).
For Tsoil kha pa the essential point is then that these two forms
of bhavana should be treated as complementary. For otherwise
there might be simple acquaintance (go ba tsam), but there could
be no full realization of the theory (Ita ba) of non-substantiality
and Emptiness, i.e. the 'analytically inspected sense' (dpyad pa'i
don) (Lam rim chen rna, f. 50Ib2). Neither acquaintance (go ba) not
followed by analytical inspection nor the mere assertion (dam bca'
ba tsam) of impermanence and the like can be effective alone;214
212 For Gro luri. pa's understanding of non-mentation (yid la mi byed pa), see his bDe bar
gIegs pa'i bstan pa rin po che la 'Jug pa'i lam gyi rim pa marn par bsad pa (bsTall rim), f. 377a-b
etc. Cf. D. Seyfort Ruegg, in: Oriel/talia I. Tucci memoriae dicata, iii, p. 1257, for Karma Mi
II2
II3
analytical inspection (rnam par dpyad pa: vicara); and Bodhicaryavatara viii. 4 where, following the Prajiiaparamita-Sutras, Santideva
has treated successively the Perfections of Dhyana and Prajiia
(f. 503a-b).
Now, on the ground that by following the Hva san's instructions (as widely understood by the Tibetans) one would be trying
to reach the signless (mtshan med: animitta) and non-construction
(mi rtog pa: akalpa[na]) merely by suppressing all mentation
involved in the proliferating activity of mind (sems 'phro ba : cittalfl
prasarati) without ever engaging in de constructive analysis (rigs
pa'i dpyod pa), Tson kha pa has rejected the Hva san's view as
being contrary to what has been recommended in the Ratnameghasiitra, as interpreted by Kamalasila in his Bhavanakrama II
(PP.29 and 45-46) and III (pp. 3, 7, and IS) (f. 503b-504b).217
In sum, according to Tson kha pa, not only is there to be nonconstruction of any hypostatized entity and of substantiality in
the form of a pudgala and dharmas, but there must be comprehension of non-hypostatization (bden med) and two-fold non-substantiality (bdag med). Accordingly, (negative) absence of construction of a hypostatized entity and twofold substantiality must
be carefully distinguished from (positive) understanding of nonhypostatization and non-substantiality (f. 504b-505a).
Now, the nirvikalpa-jiiana of the Arya is of course immediate
understanding (mrion sum du rtogs pa) of the sense of nonsubstantiality, empty of the cognitive object (yul = vi~aya) that is
falsely hypostatized in the one or other form of substantiality, i.e.
of pudgala or dharmas. Nevertheless, even though it is conceded
that the required meditative realization by means of the postanalytical (dpyad nas) understanding that a hypostatically posited
entity does not exist does in fact involve conceptual construction,
the latter still proves to be an altogether homogeneous cause (sin
tu rjes su mthun pa'i rgyu) for non-constructive Gnosis (nirvikalpa217 Interestingly, however, the bSam gtan mig sgron ascribes to the Hva San Mahayana
the teaching that one should not suppress notions ('du ses dgag par yati mi bya, f. 83a4). This
version of his teaching appears at first sight different from what is usually found in other
sources, both Tibetan and Chinese. It may refer to his rejection of the Sravaka's mere
suppression of notions (sar(1jlia) and feeling (vedita). See Pelliot tibitain 117 and Stein 709 (cf.
L. G6mez in R. Gimello and P. Gregory (eds.), Studies in Ch'an and Hua-yen, pp.
IIo-In); and Demieville, Candle, pp. 63 n. 67, 71 (cf. pp. 75-76, 130, 140); below,
p. 202 f.
II4
218 This highly important theme, mentioned by KamalaSila (Bhavanakrama Ill, p. 20),
is discussed also in the MadhyamakaratnapradTpa, a work ascribed to Bhavya, in Chapter
VII entitled bhavanakrama (P, f. 352b), which also alludes to the question of nonmentation. See above, pp. 94-95 note; below, p. 206.
IIS
elsewhere referred to the account of the Great Debate 219 according to which this very Sutra was contemptuously rejected by the
Hva san (f. 306a).
It is therefore concluded by Tson kha pa that settling FixationBhavana consisting in retention free from dispersal (mi 'phro bar
'dzin pa'j 'jog sgom) and Inspection by prajila consisting in analysis
(so sor rtog pa'i ses rab kyis dpyod pa) should first be made to
alternate in meditative realization. In this way samatha and
vipasyana reach equilibrium (cha milam), where there is an excess
of neither the settling aspect (gnas cha: sthiti) nor of analytical
investigation (f. S09b). Tson kha pa observes that samatha and
vipasyana are realized separately, and are made to alternate (spel
mar byed pa) with each other, there being no rule at this stage that
Inspection and Fixation should be realized [together] in a single
mental continuum (rgyun gcig) (f. SIoas).
But in a later stage there follows the yoking together (zun du
'breI ba: yuganaddha) , or syzygy, of samatha and vipasyana, when
they merge and operate together (samapravrtta: milam du 'jug pa).
The Path being then characterized by this yoked pair functions of
itself (ran gi nan gis 'jug pa = svarasavahin), without effort (mnon
par byed pa = abhisa/flskara) and mental inflection (rtsol ba =
abhoga) (f. S14a-b). Here the force of analytical InspectionBhavana (dpyad sgom) consisting in pratyavek~a makes it possible
to achieve Quieting (Samatha) (f. SI4b6). And whilst inspection
(dpyod pa) is vipasyana, this inspection once brought to perfect
completion (dpyad pa mthar thug pa) is apprehension of Emptiness
(Si:myata) qualified by samatha (f. SIsal).
This Fixation-Bhavana that initially alternates and then finally
coincides with analytical Inspection, in the form first of a regular
sequence and then of a syzygy of samatha and vipasyana, is not,
therefore, to be confused with 'Darkness-Bhavana' (mun sgom)
and with non-construction known as tsom 'jog gi mi rtog pa
(f. 496a6). These last two expressions are used to describe that
one-sided form of totally non-analytical, and practically cataleptic, non-mentation and non-construction so often attributed in
the Tibetan treatises to the Hva san, when asmrti and amanasikara
219 See sBa bied, G, p. 66.I6
ston, ja, f. rr6a2.
II6
The doxographer BIo bzan Chos kyi iii rna (1737-1802) has
stated in his Grub mtha' sel gyi me loti that the Hva san Mahayana
figures not as asTon mun pa (as he does in the sources considered
above) but as a special kind of Tsun men pa. Yet Blo bzan Chos
kyi iii rna observes straightaway that the Hva san Mahayana's
doctrine does not prove to be identical with the general theory
(spyi'i Ita ba) of the Tsun men pa. 220
The same author then remarks that, according to the Tsun men
pa, the results of wholesome and unwholesome karman sustained
by neither release (ries 'byuri = nihsaratJa) nor the bodhicitta can
respectively engender bliss and pain, but that they nevertheless do
not differ in so far as neither turns into the cause of Liberation and
Omniscience. In the same way, black and white clouds are
different in colour; but still they do not differ in their effect of
obscuring the sky.221 Not observing the appropriate distinction,
however, the Hva san Mahayana has mistakenly asserted that
good mental construction (bzari rtog) and evil mental construction (rian rtog) are alike in being fetters. In the instruction on the
meditative practice of real Tsun men pa theory, there is mention
220 Thu'u bkvan Blo bzari Chos kyi ii.i rna, Grub mtha' rei gyi me loti, rGya nag Chapter,
If. rrb-13b. I am indebted to Professor R. A. Stein for pointing out to me that Tib. tsuti
corresponds here to Ch. tsung 'school', i.e. to the Ch'an as the school par excellence.
221 Cf. sBa bied, G, p. 68.18 and S, p. 58.5-6; Stein 709, f. 7a3 (cf. Gomez in Studies in
Ch'an and Hua-yen, p. II4); Chos 'byun Me tog siiin po, f. 430b4; Chos 'byun mKhas pa'i dga'
stan, ja, f. 117a2.
II7
IIS
II9
120
121
233 Cf. D. Seyfort Ruegg in: Tantric and Taoist studies in honour of R. A. Stein (Melanges
chinois et bouddhiques 20 [1981]), pp. 21z-z6; and Acta indolagica 6 (1984), pp. 369-81.
Such antinomianism could, it is true, be derived from 'over-interpretation' of what has
been said in even such classical texts of Buddhism as the Kiisyapaparivarta ( 48-49, on the
wise and skilful Bodhisattva's being untouched by the pain of the passions) and Asanga's
Mahiiyiinasamgraha ( 1o.z8.II-12, on the theory of the destruction of the kldas through
the kldas themselves), which have in fact been cited in just this connexion by the bSam
gtan mig sgran (see below p. I2z). See also Mahiiyiinasutralal'(Jkiira xiii. I I, and Sthiramati,
Madhyiintavibhiiga{fkii ii. 14 (p. 76) on klda as a factor of Awakening (bodhyanga). See in
addition dPa! brtsegs, ITa ba'i rim pa (cf. Tucci, Minor Buddhist texts, II, p. 139); and
passages from Vimalamitra discussed by 1. Gomez in W. Lai and L. Lancaster (eds.), Early
Ch'an in China and Tibet, pp. 403-4.
122
123
In: many Tibetan historical writings, and above all in philosophical and doxographical works, the expressions theory (Ita ba)
of the HvaSari, Dharma-system (chos lugs) of the Hva Sari and
tradition (gzun lugs) of the Hva Sari have come to be used in a
sense that is for all practical purposes dehistoricized and universalized. These expressions have thus come to be widely employed as
generic designations for a type of theory or teaching that is
characterized as quietist, spontaneist, innatist and simultaneist. 236
Since disagreement may exist as to the extent to which Mo-hoyen (and Ch'an) actually adhered exclusively, or mainly, to such
views, this typological use of the expressions is perhaps not
entirely justified historically. Yet it can be convincingly derived
from the view, reliably ascribed to Mo-ho-yen, that all sentient
beings are by their nature buddhas and that in coming to an
awareness of their intrinsic and innate buddhahood - i.e. the
Buddha-nature or tathagatagarbha - any activity or 'reinforcement' of a religious and ethical as well as of an intellectual and
discursively philosophical character is therefore altogether superfluous and irrelevant,237 and may even be a hindrance on the
level at least of the advanced practiser. 238 In this perspective, the
Triple Vehicle (triyana) is set aside in favour of the Unique
Vehicle (ekayana) - or even the N on-Vehicle (ayana) - free from
all verbalizations and conceptualization. 239 This interpretation of
the Ho-shang's teaching is underpinned by his statements that
liberation is achieved in immediate and face-to-face recognition
of Mind free from all discursive and ratiocinative mentation, that
is, in pure tranquillity unaccompanied by analysis and discriminative understanding.
Under this analysis, the Hva sari's doctrine of the Buddhanature and tathagatagarbha would not issue in the eternalist view 236 1. Gomez in Early Ch'an in China and Tibet, p. 428 n. 14, describes the term
quietism to refer to Ch'an as an unfortunate legacy of Demieville's Candle. However, if
not taken as referring specifically to seventeenth-century European thought, the word
does not appear to be unsuitable. The hesychast too does not eschew all activity.
237 See Demieville, Candle, pp. 95, 107-08, 116-19, 151.
238 See above, p. 117.
239 See Demieville, op. dt., pp. 66, II9, 151.
124
240 See Eu ston, Chos 'bytlti, ( I2SaI; compare the 'Alternative Tradition' of the sBa
bzed (G, p. 73) on stan pa iiid la dga' ba. See however, bSam gtan mig sgron, f. 83a-b, on not
falling into annihilation (chad pa = uccheda) and on not suppressing san;jiia and not falling
into absence of san;jiia.
125
'Gradualist' KamalaSIla. And even though the historical documentation available to us of course excludes such an inference, we
still have to bear in mind that the figures in question have come
to exemplify two important, and old, positions that have often
been in te'nsion, either virtual or actual, in the' history of
Buddhism. (Compare the cases of Musi:la and Narada, and of
Mahakotthita/Mahakotthika and Sariputta, in the old Buddhist
canon mentioned below, Chap. iv)
The fact that, for the Tibetan historical and doxographical
traditions, important facts concerning the Ho-shang Mo-ho-yen
were uncertain and wrapped in the mists of time and legend must
have greatly contributed to his becoming a somewhat shadowy
and emblematic figure. For example, it was evidently not clear to
the rNin rna master Tshe dban nor bu (1698/9-1755) whether he
should be placed in the time of Khri Sron Ide btsan or earlier, at
the end of the reign of this king's predecessor Mes Ag tshom(s)
and whether it was he or a disciple of his who debated with
KamalasIla. 241 Furthermore, as already noted, although the Hva
san Mahayana is usually regarded by the Tibetan traditions as a
'Simultaneist', there seems to have been some uncertainty as to
whether he should be identified as aston mun pa or as a tsuri men
pa.
Originally, and historically, 'Simultaneism' was possibly just as
much complementary with as antithetical to 'Gradualism'. Nan
ral's Chos 'byuri Me tog siiiri po has presented the teachings of the
Hva san Mahayana and KamalaSlla as being without difference in
substance, notwithstanding the fact that they were pitched at
different levels (f. 435b).242 And in the bSam gtan mig sgron
ascribed to gNubs Sans rgyas ye ses, simultaneous engagement
(cig car 'jug pa) is a stage that follows on gradual engagement (rim
gyis 'jug pa) and leads on first to Mahayoga (rna I 'byoy chen po) and
241 See Tshe dban nor bu, rGya nag hva san gi byun Ishul, f. 8a-b, where Mahayana is
placed in the latter part of the reign of Mes Ag ,shorn can (i.e. Khri IDe gtsug btsan, the
father of Khri Sron Ide btsan) together with his disciples sEa gSal snan 0) and Myan Tin
ne 'dzin bzan po and is therefore tentatively distinguished from the Hva san who debated
with Kamala,'la and who would then have been Mahayana's disciple. For an account of
ho-shangs in Tibet at the time of Mes Ag tshoms, see for example the sBa bied. And on the
Hva san Me 'go/mgo at that time see Chap. ii above.
242 See above, pp. 84-85. On the similar opinion of dPal dbyans see above, p. 86.
I26
127
246
128
129
II-12;
S, p.
I2.5);
'Gos gZon nu
130
131
253 Cf. S. 1naba in 1. Kawamura and K. Scott (eds.), Buddhist thought and Asian
civilization (Festschrift H. V. Guenther, Emeryville, 1977), pp. 105-13.
254 For the use of this criterion in editing the canon, see D. Seyfort Ruegg, The life of
Bu stan Rin po ehe (Rome, 1966), pp. 27-28.
255 See Demieville, Concile, p. 160 and pp. 25, 39-42.
I}2
I33
that all sentient beings are already buddhas; whilst according to the
'Gradualist' who insists on the need for methodical cultivation the
same teaching signifies that on the 'causal' level (gzi) beings are
all potentially buddhas, that is, that their inborn capacity of
achieving the 'fruit' ('bras bu) of buddha hood is proleptic and still
unfulfilled. Nevertheless, as said in the MahiiparinirviiYJasatra
quoted by Kamalasila in his Bhiivaniikrama (II, p. I9), the Matrix
or Germ (gotra) of the tathiigata (de bzin gsegs pa'i rigs) is perceived
only when Quieting (Samatha) and Insight (vipasyanii) are in
balance, as is the case with tathiigatas or buddhas. For when the
mental tranquillity of concentration (samiidhi) predominates and
the discriminative function of prajiiii is weak, as in the case of an
Auditor (Sriivaka) , one does not see the tathiigatagotra at all;
conversely, if prajiiii predominates and samiidhi is weak, as is the
case with Bodhisattvas still on the Path of Awakening, it is seen
only indistinctly. That KamalaSIla has thus called attention to the
relevance of the tathiigatagarbha and gotra theories to his other
concerns in his Bhiivaniikrama is in keeping with the fact that in his
Madhyamakaloka - which he is reputed to have composed in
Tibet for the benefit of the monarch 257 - he introduced this
theory into the main Madhyamaka tradition.
As regards the Vajrayana, it appears to combine elements of
the rapid way of 'nature' and nativism with a recognition that
means, both ritual and cognitive, have to be brought into play
gradually.
It is of special importance to note that both 'Gradualism' and
'Simultaneism' can find support in the Sutra literature of Buddhism, and that in the Sastras also many traces can be found of the
idea of gnoseological nativism and soteriological spontaneousness
as well as of holism and instantaneousness or 'Subitism'.
In Tibetan literature noteworthy indications are found concerning the manner in which the Tibetans have themselves
regarded the interrelation and classification of the component
elements of their religion and culture in terms of architectural
organization in space.
257 See sBa bied G, p. 77 and S, p. 63;
f. 437a6.
Ni
134
259
135
136
\111
,"'
137
IV
263 In the 'resonance' (dhvani) theory of the Indian poeticians, for example, there is an
opposition between the asal1l1ak~yakramavyaligya and the saI1l1a~yakramavyaligya, the two
divisions of the vivak~itiinyaparaviicya which is abhidhiimula. As for progressiveness as
opposed to sudden immediacy in the spho!a-theory of the philosopher-grammarians, see
K. Kunjunni Raja, Indian theories of meaning (Adyar, I969), p. I24 f.
139
'auditive assimilation' of the sense of the authorless and immemorial scriptural mahCivCikyas inducing knowledge of brahman and
immediate deliverance (sadyomukti, as opposed to kramamukti) has been a subject of discussion. Repeated meditative practice
(prasaytlkhyCina) subsequent to this sravatJa has been held (by
Suresvara) to be superfluous in the face of sCibdajnCina or W ordgenerated knowledge. Another notion germane to the present
enquiry is immediate recognition (pratyabhijnCi) - the 'recollection' as it were of reality 264 - to which may be added the
soteriological concept of the abrupt onset of spiritual realisation
(sCihasa). Mention may also be made of the state of anupCiya
'absence of means' which - in so far as it corresponds to the
ultimate state of the turyCitfta where all means are excluded - is
free from all mediacy, as opposed to the three successive levels
involving means and mediacy of the CitJava (kriyopCiya), the sCikta
(jnCinopCiya) and the sCimbhava (icchopCiya). These terms and ideasthough not identical historically with the questions directly at
issue in the Great Debate of bSam yas, and despite the basic
difference between Buddhists and the Vedanta with regard to
Word (Sabda) and vVord-induced knowledge (SCibdajnCina) as
constituting immediate intuitive gnosis (aparok~ajnCina) - nevertheless provide in their problematics a number of interesting
parallels and points of comparison with the issues considered by
KamalaSi:la and Mo-ho-yen and so merit the attention of the
comparativist.
The fact that the authoritative sources cited by Kamalasila mainly Mahayana Sutras - have a direct or indirect bearing on
this subject indicates at all events that the problems at issue go
back a long way in the history of Buddhist thought. Kamalaslla
was clearly not dealing with issues that had arisen for the first
time during the eighth century in the specific historical context of
the encounter in Tibet between Indian and Chinese ways of
thinking and of a confrontation between Indian and Chinese
masters. That this was so seems moreover to have been recognized in the 'Alternative Tradition' of the sBa bzed,265 where
264
For some references see D. Seyfort Ruegg, Le traite du tathagatagarbha de Bu ston Rin
140
KamalaSila's master Santarak~ita is shown referring to the understanding of Mind known as the Sovereign Remedy (dkar pa chig
thub) as a 'stain in theory' (dr~tika~aya) ,that consists in taking
pleasure in Emptiness (stan pa iiid la dga' ba), and which was to be
found not only in Tibet but also very widely among persons who
are tainted by these 'stains' and take pleasure in the notion of
Emptiness. 266
Since the evidence available to us indicates that this complex of
problems has repeatedly arisen in one form or the other in the
history of the Buddhist traditions in South, Central and East Asia
- none of which has been entirely homogeneous and monolithic
and each of which has included various currents of theory and
practice - contrary to what has been suggested in Demieville's
masterly study 267 it does not seem appropriate to see in Kamalasna's treatment of the issues merely a dehistoricization of the
Great Debate. For Kamalaslla the Great Debate was probably
rather one more occasion when this set of religious and philosophical problems embedded in the history of Buddhist thought
came once again to be focused upon and to raise acute difficulties.
And it was then in the later Tibetan traditions, as already
observed, that the expression 'teaching of the Hva san' was taken
from its specific historical context and came to be employed as a
dehistoricized tapas and as a generic designation for a type of
quietistic and innatist teaching.
What the comparativist has to study here are not so much
abstract entities like 'Indian Buddhism', 'Chinese Buddhism' or
'Tibetan Buddhism' - which are to a certain degree merely
convenient constructs for the scholar - but rather the structural
and typological features subsumed under these designations. This
is of course not to deny that certain features are, at particular
times and places, predominant in a given geographically delimited form of Buddhism, and that they may characterize and
266 On the Ita ba'i sfiigs ma = dr:ftika,aya, see Vasubandhu, Abhidharmakosabha,ya iii.
94ab with Yasomitra's Vyakhyii. It is one of the five ka,ayas, on which see e.g. Lalitavistara
p. 248.13; Saddharmapul!l/artka ii, p. 43.4, 56.8, 58.II; Bodhisattvabhumi i.I7 (p. 252);
Mahavyutpatti 2336-40. These stains characterize especially the last 50o-year period of the
Dharma (cf. sBa bied, G, p. 66.8; S, p. 56.2-3). On the dkar po chig thub see above,
pp. 88-89, lOO ff.
267 DemievilIe, Concile, p. 18.
141
constitute it: if there were no overarching structures and continuities, but only innumerable discrete features, the tenIis 'Buddhism', 'Indian Buddhism' and the like would be mere empty
names of no use to an historian. But the richness and diversity
within the Buddhist traditions militate against taking even such
serviceable terms as names of single and homogeneous individual
entities. In sum, such constructs can fulfil a useful and legitimate
heuristic and descriptive purpose for an historian provided that he
does not reify them in historical and comparative work.
In the following, several themes typical of the ideas at issue in
th~ Great Debate, and ascribed either to the 'Gradualists' or the
'Simultaneists', will be considered with a view to identifying
earlier examples or prefigurations of these themes and to situating
them in the Buddhist (and non-Buddhist) traditions of India.
I. THE GIVING
Up
142
143
274
I44
145
--------------------------- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
146
147
SAMASIsI(N)
The Pali Patisambhidiimagga and Puggalapaiiiiatti recognize a category of persons (puggala) called samaslsi(n) 'equal-headed' because, for them, exhaustion of the impurities (iisavapariyiidiina)
280 See Demieville, Concile, Index s.v. sans pensee. On 'One-practice Samadhi' see
B. Faure in P. Gregory (ed.), Traditions of meditation i" Chinese Buddhism (Honolulu, I986),
pp. 99-128.
281 See MaQ.~anamira, Brahmasiddhi (ed. S. Kuppuswami Sastri, Madras, I937), pp. I3,
26-26, 36, with Kuppuswami's Introduction, p. xxxi if. On the relation between karman
and vidyajjfiana in Sarp.kara, see his Brahmasiitrabha~ya III, iv. 5, II, I3 (with reference
. especially to Brhadara"yakopani,ad IV. iv. 2 and navasyopani~ad 2).
148
282 Patisambhidamagga I 101; Puggalapaiiiiatti, pp. 2, 13. This category of person connected with the naivasalfljiianasalfljiia level - is not accepted by the Sarvastivadins; cf.
A. Bareau, Les sectes bouddhiques du Petit Vehiwle (Paris, 1955), pp. 175, 184, 198, 262.
283 NettippakaralJa, p. 190.
284 Theragatha lIO and Atthakatha I 217 f. (cited by G. Malalasekera, Dictionary of Pali
proper names [London, 1937], p. 439, and W. Rahula, Zen and the taming of the bull
[London, 1978], p. 22).
285 Such cases have been studied by P.-A. Berglie and C. Suneson in E. Kahrs (ed.),
Kalyal]amitraragal]am (Festschrift N. Simonsson, Oslo, 1986), pp. 13-47. To the bibliography cited by them one may add R. Fick, Der indische Weise Kalanos und sein
Flammentod, Nachrichten von der Gesellschaft der Wissenschaften zu Gottingen, Philologisch-historische Klasse, Gottingen, 1938; and K. Bhattacharya, L'atman-brahman dans Ie
bouddhisme ancien (Paris, 1973), pp. 29, 113, 157-9.
-------------------------
149
150
151
294
I52
In other words, questions concerning the immanent, immediate, spontaneous, and sudden were evidently already an issue in
Kasmir and Serindia by the early fifth centu,ry. Indeed, differences
between the tun and chien approaches have been noticed in the
preface to Taisho 6I8 ascribed to Hui-yuan (334-4I6).304
Dharmatrata, to whom the above-mentioned Taisho 6I8 was
wrongly attributed, has often been conflated with a certain
Dharmatala (?), and also with the great Ch'an patriarch Bodhidharma/Bodhidharmatrata. 305 In a Chinese tradition there is a
curious record of the latter's leaving behind one of his shoes as a
token of the future spread of his teaching,306 a motif found also
in the Tibetan traditions which relate inter alia that a certain
Chinese Master (rgya nag mkhan po) - i.e. one of the Ho-shangs
active in Tibet in the eighth century -left one of his boots behind
in Tibet after being defeated in a controversy as a sign that his
teaching would survive and later spread in that country.307
In the Tibetan iconographic and ritual traditions, moreover,
the U pasaka (dge bsiien) Dharmatala/Dharmatrata figures, together with a certain Hva san counted also as an Upasaka,
alongside the sixteen Arhats in a well-known group that thus
consists of a total of eighteen figures. 308 The Hva san in this
group is often identified with the Hva san Mahayana, even
though the latter is usually regarded as a monk; and an identification is made in addition with Pu-tai/Mi-lo (Maitreya) of the
Chinese tradition. 309 Dharmatara/Dharmatala/Bodhidharma
figures in Tibetan sources as an authority of the sTan min pa/Cig
car ba tradition, or of the Tsuri men, that originated with Kasyapa,
See Lin Li-kouang, op. cit., pp. 342-3.
On the Ta-mo ch'arz shih lun discovered in Dunhuang, see P. Yampolsky, op cit.,
p. 21, and in W. Lai and L. Lancaster, Early Ch'an it! China and Tibet, p. 3; B. Faure, Le
traite de Bodhidharma (Paris, 1986).
306 See Yampolsky, Platform Sutra, p. 41.
307 See sBa hzed, G, p. 75.8; G, p. 9 and S, p. 8; and the other sources cited above,
n.17.
308 For the history, iconography and iconology, see S. Levi and E. Chavannes,joumal
asiatique 1916/ii, pp. 189-304, especially pp. 288-90, 297; G. Roerich, Tibetan painting
(Paris, 1925), pp. 29, 31-]2; F. Lessing, Yung-ho-kung (Stockholm, 1942), pp. 35-37;
P. Demieville, in lao Tsong-yi, op. cit., p. 45 f; Z. Yamaguchi, Acta indologica 6 (1984),
pp. 393-422.
309 For Pu-tai/Mi-lo (Maitreya), see Demieville, Concile, p. 12 n.; R. Edwards, 'Pu-tai
Maitreya', Ars orientalis 14 (1984), pp. 5-50.
304
305
153
I54
ISS
tion of discriminative understanding (prajiiaparamita).314 In consequence of this the three components of any of the relevant acts
- viz. the agent, the intended beneficiary and the activity of
giving, etc. - are relativized and cancelled as hypostatic entities
through what is known as triaspectual purification (trimaIJqalavisuddhi).
Haribhadra has next explained that comprehension (adhigama)
results, inter alia,
(i) from 'commemoration' (anusmaraIJa) of the Buddha finding
expression progressively in the factors of the three Paths of
preparation (prayogamarga, i.e. the four nirvedhabhagtyas),
vision (darsanamarga) and meditative realization (bhavanamarga);
(ii) from 'commemoration' bearing on Dharma, i.e. the factors
that are categorized as wholesome (kusala), unwholesome
and undetermined (avyakrta); and
(iii) from 'commemoration' bearing on the Sarp.gha, which
Haribhadra here understands as the community of nonretrogressing (avaivartika) Arya-Bodhisattvas.
Now, very significantly for the purpose of this study, Haribhadra has specified that, in reality (paramarthatas), buddhanusmaralJa
is characterized precisely by non-recollection (asmaralJalak.$alJa).
And comprehension (adhigama) has been stated by him to consist
in understanding that the very nature of all dharmas is precisely
their 'entitylessness' (dharmabhavasvabhava), i.e. their non-substantiality as hypostatized entities. 315
Haribhadra has further explained that anupurvabhisamaya consists in cultivation with a view to stabilization consequent on the
progressive ordering of the matter, which is then understood
both in disconnexion and in connexion (vyastasamastatvenadhiga:"
tan arthan anupurvikrtya sthirikaralJaya vibhavayati).
Such then, according to Haribhadra, is progressive compre314 It is to be noted that - unlike Ratnakarasanti (Saratama, p. 163.19) - Haribhadra
does not here refer to dhyana as comprised in the prajnaparamita; and he speaks (vi. 1,
p. 908) of four Perfections (paramitacatuilaya) being comprised in prajnaparamita. The
significance of this restriction, which distinguishes Haribhadra from KamalaSIla for
example, remains to be clarified in detail.
315 Cf. Bhadanta Vimuktisena, Varttika, P, f. 180a-181a.
I56
hension, the sixth main topic in the Abhisamayalan:ikara's exposition of Prajiiaparamita philosophy.
The next main topic in Prajiiaparamita philosophy, which
makes up Chapter vii of the Abhisamayalarl1kara, is known as
single-moment awareness (ekak~mJavabodha). As comprehension
in a single moment (k~alJenaikenadhigamah), this awareness is
known to Haribhadra in addition as ekak~alJabhisambodha and
ekak~alJabhisamaya. Then, for one who has fully realized this
ekak~alJabhisambodha there ultimately arises, in a second moment
(dvitlye k~alJe), the eighth awareness relating to the dharmakaya as
the culmination of Prajiiaparamita philosophy and practice, i.e.
the final main topic treated in Chapter viii of the Abhisamayalalflkara. 316
In the philosophy of Prajiiaparamita as expounded in the
Abhisamayalalflkara and in Haribhadra's V~tti and Aloka - works
which are rightly regarded as veritable monuments of the
'Gradualist' current in Buddhist thought - what exactly is meant
by single-moment awareness or understanding?
Following Chapter vii of the Abhisamayalalflkara, Haribhadra
has explicated the concept under the following four headings. 317
(i)
316
317
318
(Tib. srion gyi smon lam gyi 'phen pa'i Jugs dan [has kyi dbyiris kyi mthus).
157
simultaneously and at once (sakrt = cig car, glossed as ekavaram = dus gcig tu) by a single impulse of energy. In this way,
in the first form of ekak~alJabhisambodha, the single pure
Gnosis (arzasravajnarza) 'presents' (abhimukhlkarayati), in one
single 'moment, all that is in its scope as homogeneous
(sajatlya = ri[g]s mthurz pa). It is defined by Haribhadra as
characterized by the single-moment comprehension of all
pure, non-fruitional factors (avipakarzasravadharma).
(ii) The next aspect of ekak~alJabhisambodha is described as
consisting in the fact that, once all obstacles have been
removed in the Bodhisattva's meditative realization of the
appropriate counteragents (pratipak~a), there arises the state
of 'reality of fruition' consisting in the aspect of total
purifica tion
(s akalavy av adarzap ak~ avip akadharmatav astha) .
And through comprehension of all arzasravadharmas that
have therewith reached, in one single moment, this state of
fruition, there arises the Gnosis that corresponds to prajnaparamita. This second aspect of ekak~alJabhisambodha has
accordingly been defined by Haribhadra as characterized by
the single-moment comprehension of all pure factors in the
state of reality of fruition. Haribhadra's forerunner Arya
Vimuktisena (sixth century) had earlier specified that fruidon (vipaka) through understanding (abhisamaya) in a single
moment arises and ceases all at once (cig car du = sakrt or
yugapad?).319
(iii) The following aspect of ekak~alJabhisambodha is defined as
being characterized by the single-moment comprehension
of all dharmas as devoid of characteristic marks (alak~alJasar, vadharmaikak~alJalak~alJa), this marklessness also being
known in one single moment. Indeed, as had been explained
by Arya Vimuktisena, were dharmas on the contrary differentiated from each other by distinct characteristic marks,
their required inclusion within one single dharma would be
impossible; and it would then wrongly follow that there
could be no realization (abhisamaya) in a single moment. 320
According to the later commentator Ratnakarasanti
(c. rooo), because of the absence of characteristic mark,
319
320
-----------
158
I7I; =
P, f I2ob).
159
the objective (grCihya) factors of existence, beside (iii) a preliminary consequent receptive perseverance in knowledge (anvayajnCinak~Cinti) and (iv) a full consequent knowledge (anvayajnCina)
bearing on the subjective (grCihaka). This set of four factors is
brought into relation with each of the four Principles or Realities
of the Nobles (Ciryasatya).
Quoting one view of this matter in his AlokCi,323 Haribhadra
has explained that, in the exact moment when III (duhkha, i.e. the
first Ciryasatya) is eliminated following on its recognition, the
other three satyas also are all simultaneously involved, viz. in the
form of elimination of the origin of III (i.e. the second Ciryasatya) ,
the realization of the cessation of III (i.e. the third Ciryasatya), and
the practice of the Path (i.e. the fourth Ciryasatya). And the same
applies mutatis mutandis in the case of each of the following three
satyas. Hence, according to this view, the Mahayanist intuition of
the Darsanamarga is to be regarded in this respect as a singlemoment intuition (ekak~a1'JCibhisamaya) with respect to its intuition
of a single effect (ekakaryCibhisamaya).324 According to another
view of the matter also mentioned by Haribhadra, intuition of
the Darsanamarga is an ekak~ar;Cibhisamaya because there here
arises a pure knowledge (anCisravajnCina) making known the
nature of all modes (sarvCikCirasvarupaprativedhakCirijnCina); and this
knowledge has within its scope the totality of all factors (sarvadharmavi~aya), thus bearing a certain (at least formal) resemblance
to the two aspects of the ekak-iar;Cibhisambodha at the very end of
the Path.
324
I60
161
162
r63
164
327 On these two terms, see H. V. Guenther, Tibetan Buddhism in western perspective
(Emeryville, 1977), p. 151; G. Tucci, Religions of Tibet (London, 1980), pp. 85-87, 131;
R. A. Stein, Revue de l'Histoire des Religions 179 (1971), pp. 23-28; and M. Broido,jollrnal
of the International Association of Buddhist Studies 8 (1985), p. 35 (concerning the types of
person known as cig car ba, thad rgal ba and rim gyis pal. Thad rgal has often been regarded
as an essentially gradual, even though accelerated, process as opposed to khregs gead. But in
some cases thad rgal too can be extremely rapid and for all intents and purposes sudden.
328 In the Bod rgya tshig mdzod chen mo (Zang-Han daicidian, Peking, 1985), khregs chad is
defined as (I) stia 'gYllr ba'i lhag mthoti gi braa chad (i.e. an expression used by the rNiIi. ma
pas for lhag mthoti = vipasyana), and (2) ka dag khregs chad kyi bsdus tshig.
329 In Sakaki's edition of the Mahavyutpatti the spelling thad rgyal is found under no.
1496.
330 Cf. Bod rgya tshig mdzod chen mo, S.V.
331 It should be noted that ava-skand- means not only 'to jump down' but also 'to
assault, storm'. The latter meaning would be appropriate when the avaskanda(ka)
technique designates an accelerated and very rapid process.
As for the term vi~kanaa(ka), the word vi~kanda is attested in the meaning of 'dispersing,
moving away'.
165
------------
166
r67
336
r68
the Buddhist traditions as being worthy of honour (dakkhiIJeyya);337 he is characterized by the faculty of concentration
(samadhindriya).338 By extension (pariyayena) the name Kayasak~in is used for one who has 'contacted' the four Dhyanas and the
four .i\rupyas; and strictly (nippariyayena) this designation applies
to one who has attained the Sarp.jiiavedayitanirodha, in which
case the impurities (asava) are exhausted by discriminative
knowledge (panna, which is to say that the Kayasak~in is then an
Arhat).339 When his mental attention (manasikara) focuses on III
(dukkha) the samadhindriya predominates in him; and his realization, or 'attestation' (sacchikar-), is described as being by bodily
contact, for he first 'contacts' Dhyana and then Stoppage and
Nirval).a (jhanaphassartl pathamaytl phusati, paccha nirodhaytl nibbanalfl sacchikaroti).340 Moreover, the designation Kayasak~in
applies both to a person who attains the sotapattimagga by the
power of the samadhindriya and to the persons who by the power
of this faculty of concentration attain the sotapattiphala, the Paths
and Fruits of the Sakadagamin and the Anagamin, the Path of
Arhathood and the Fruit of Arhathood; while the samadhindriya
predominates in this type, the other four faculties (saddha, panna,
etc.) nevertheless playa subordinate part in his spiritual constitution. 341 The Kayasak~in is described as one who abides having
contacted 'in the body' the tranquil Vimok~as - the Arupyas
beyond the Rupas - and some of whose asavas are exhausted by
discriminative knowledge; the restriction 'some' (ekacce) is also
applied to the asavas in the cases of the spiritual types of the
Dntiprapta, the Sraddhavimukta and the Dharmanusarin, but
significantly it is not so applied in the cases of the Ubhayatobhagavimukta (who also abides in bodily contact with the tranquil
Arupya Vimok~as) and the Prajiiavimukta (who however does
not abide in bodily contact with the Arupya Vimok~as).342 In the
169
344
170
I7I
patti = mthar gyis gnas pa'i snoms par 'jug pa) and then of the
Avaskanda-Samapatti (v. 24-25):
kCimCiptam avadhikrtya vijnCinam asamCihitamj
samCipattir gatvagamya nava dvidhCij j
sanirodhCi~
ekadvitricatu~panca.)atsaptCi.)tavyatikramCitj
348
172
vyutkrantakasamapattr 350
The procedure (prayoga) in question consists, according to the
Bha,)ya, in the practiser's first passing over in ascending and
descending sequence eight sasrava stages, and then seven anasrava
stages. 351 Next, he attains the third sasrava Dhyana from the first,
350
173
174
175
- - - - - - --
I76
--------------------------------------------
sarvaiva madhya, k,antir yadaikam eva k,a'1arr tad adhimatrhij k,a1]ika casau, na prakar,ikt.
366 Abhidharmakosabha,ya vi. 19C. Compare Vasumitra's view mentioned below,
P179
177
vimuktimarga. 372
The Darsanamarga proper has been described by Vasubandhu
as consisting in fifteen moments (k.$a/Ja) beginning with this
dharmajfianak.$anti concerning III and culminating in the anvayajfianak.$anti concerning the Path (marga). For the sixteenth and
final moment - i.e. consequent knowledge concerning the Path
As with the Dharmaguptas, according to Yasomitra.
Abhidharmakosabha,ya vi. 27a. On the notion of abhipraya, see D. Seyfort Ruegg,
Jot/mal of Indian Philosophy 13 (1985), pp. 309-25, and 16 (1988), pp. 1-4, with the
literature cited there.
369 Abhidharmakosabha,ya vi. 27bc.
370 On the notion of abhisartJdhi, see D. Seyfort Ruegg, 'Allusiveness and obliqueness in
Buddhist teO\ts', in C. Caillat et al. (ed.), Formes dialectales dans les litteratures indo-ilryennes
(Paris, 1989), p. 299 ff.
371 Abhidharmakosabha,ya, vi. 27bc.
372 Abhidharmakosabha,ya vi. 26a.
367
368
178
k~iintis
Abhidharmakosabhii~ya
vi. 28a.
374 Abhidharmakosabhii~ya
priikar~ikl
as the antonym of
k~al)ikii
in
376
179
concerning the gradualness as against the simultaneity of comprehension. These differences figure prominently among the
doctrines the doxographers of these schools have ascribed in
particular to the Mahasarp.ghikas and their branches, the Lokottaravadins and the Ekavyavaharikas. 378 And a connexion between
the Mahasarp.ghikas (phal chen sde) and the Simultaneous Engagement (cig car 'jug pa) of Kasyapa as transmitted by the school of
(Bodhi-)Dharmottarala (sic) is seemingly suggested in a chapter
of the bKa' thari sde lria, the Blon po bka'i thari yig. 379
According to V asumitra' s Samayabhedoparacanacakra, in the
view of the Mahasarp.ghikas it is by a single thought that [a
buddha] knows all (sems gcig gis chos thams cad rnam par mkhyen to);
and it is through discriminative knowledge (prajfia) conjoined
with single thought-moment that [a buddha] fully knows all
dharmas (sems kyi skad cig ma gcig dari mtshuris par ldan pa'i ses rab
kyis chos thams cad yoris su mkyen to).380 A branch of the
Mahasarp.ghikas, the Ekavyavaharikas, are so named according to
Bhavya's Nikayabhedavibharigavyakhyana because they have in this
way accepted one single procedure (ekavyavahara).381
According to Vinltadeva's *Samayabhedoparacanacakre Nikayabhedopadariananama-salf/graha, moreover, in the view of the Lokottaravadin-Mahasarp.ghikas, the four Principles of the Nobles
(aryasatya) are realized all at once (bden pa ni cig car mthori riO).382
And in the view of the two subschools of the Mahasarp.ghikas in
Cf. Lamotte, Le traite de la Grande Vertu de Sagesse, V (Louvain, I980), pp. 2328-35.
lHa sa edition of the bKa' than sde Ina, f. I9a: 'ad sruns cig car 'jug pa phal chen sde:
mkhan po dharmotta ra la'i brgyud pa ni ... 'the Simultaneous Entry of Kasyapa,
Mahasarrtghika: the line of the Master Dharmottarala .. .'. The standard Tibetan
equivalent of Mahasarrtghika is dge 'dun phal chen po('i sde), and Tucci's translation in his
Minor Buddhist texts (II, p. 8I) differs from the one offered here. There is no correspondence in the parallel passage of the bSam gtan mig sgron, f. 8.
. 380 Vasumitra, Samayabhedoparacanacakra (Tibetan translation ed. E. Teramoto and
T. Hiramatsu, Kyoto, I935), p. 5. For ekacittak,alJasamayukta prajna, see Mahavastu (ed.
E. Senart), I, p. 229 and II, p. I33, 285, 4I6. Cf. Lalitavistara (ed. Lefmann), p. 350.I3-I4;
Sik,asamuccaya, p. 278.I2.
381 Bhavya, Nikayabhedavibhangavyakhyana (Tibetan trans!', ed. E. Teramoto and
T. Hiramatsu), p. I9: sans rgyas beam Idan 'das mams kyi chos thams cad thugs gcig gis mam par
378
379
mkhyen cin skad cig gcig dan IdOl! pa'i ses rab kyis chos thams cad yons su mkhyen to zes tha snad
'dogs tef des na tha snad gcig gis pa zes bya' o. On the interpretation of the term ekavyavahara
see A. Bareau, Les sectes bouddhiques du Petit vehicule (Saigon, I955), p. 78.
382 VinItadeva, *Samayabhedoparacanacakre Nikayabhedopadarfana-nama-salVgraha (Tibetan translation, ed. E. Teramoto and T. Hiramatsu), p. 41.
180
384
--------------------
181
'Dharma-eye', i.e. the knowledge that whatever originates (samudayadhamma) ceases (nirodhadhamma).388 A closely related point
is made in Kathavatthu i. 4, where there is a discussion as to
whether Candidates for the four Fruits of the Nobles eliminate
defilement (kiZesa) piecemeal (odhisodhiso), through their vision of
the aryasatyas. (According to the Atthakatha [po 43] the doctrine
discussed in i. 4 was that of the Sammitiyas and some others.)
And according to another section of the Kathavatthu (xxii. 8) and
its Atthakatha, two branches of the Mahasarp.ghikas - the
Pubbaseliyas and the Aparaseliyas - maintained that all dhammas
belong to a single mental moment (ekacittakkhalJika). The same
two sources further mention (xi. 6) the opinion - ascribed to the
Sabbatthivada and Uttarapathaka - that samadhi pertains to a
single mind-moment (ekacittakkhalJiko samadhi).389
According to Asanga's Abhidharmasamuccaya,390 the Streamwinner (srotaapanna) - that is, the first of the four Nobles (arya) at
the pivotal point of the sixteenth moment of the Darsanamarga
and the first of the Bhavanamarga - may be either one who
obtains release gradually (rim gyis pa) or one who obtains release
all at once (sakrnnairyalJika = cig carries par 'byin pa). The first type
is said to be of the kind described earlier (p. 89). The sakrnnairyalJika is on the contrary defined as one who, having achieved the
comprehension of the four satyas, takes the threshold-meditation
(anagamya = mi lcogs pa med pa)391 as his base and eliminates all at
once (sakrt = cig car) all the defilements/afflictions of the three
182
levels (traidhatukavacara~ sarvakleSa~) by means of the transmundane Path (lokottaramarga). In his case only two (rather than
the usual four sequential) Fruits are achieved, namely those of
the Stream-winner and the Arhat. The person in question is then
stated mainly to reach full liberating knowledge (Cijiia, of the
Arhat) in the present existence (dnta-dharma) or at the time of his
death. 392
It thus emerges clearly that the sakrnnairyalJika type of Srotaapanna can achieve his goal rapidly, without acquiring successively
all four Fruits of the aryamarga.
4.
One of the most important points repeatedly made by Kamalasi:la in his Bhavanakramas is that Quieting (Samatha) and Insight (vipasyana) should be conjoined (yuganaddha) , that they
must operate so to speak in conjunction like a pair of oxen
teamed together (yuganaddhavahibativardadvayavat). The perfect
Path is accordingly described as operating as a syzygy of
Quieting and Insight (Samathavipasyanayuganaddhavaht margo
ni.)panna~).393
r83
184
399 See below, p. 202-03. Compare sBa hied, G, p. 69; S, p. 58-59; Nan Ni rna 'od zer,
Chos 'byuti Me tog siiili po, f. 431a-b; dPa' bo gTsug lag phren ba, mKhas pa'i dga' ston,
f. II7a.
400
401
402
403
404
Bhavanakrama
Bhavanakrama
Bhavanakrama
Bhavanakrama
185
406
I86
187
188
189
190
191
192
s.
ABSENCE OF NOTION
193
I94
195
The supreme degree of Samapatti and Vimok~a, the salfljiiavedayitanirodha which is recognized in classical Buddhist literature as
belonging to the Arya alone,431 is moreover very clearly distinguished both from the attainment of unconsciousness (asalfljiiisamapatti, asalfljiiasamapatti) and from the unconsciousness of the
asalfljiiika state, that is, from two states that are not coun.ted as
Vimok~as forming part of the aryamarga and which are not
cultivated by the Buddhist Arya as components of his consecutive
stages (anupurvavihara) of spiritual practice. 432
As for the Bhavanamarga, in Buddhist soteriology it may be
either mundane (laukika) or transmundane (lokottara), quite
unlike the Darsanamarga which is always transmundane and
pure. 433 The transmundane Bhavanamarga is of course the pure
(anasrava) one practised by the Arya, which includes the four
Arupya Samapattis culminating in the 'peak of existence' (bhavagra) and then issuing in the nirodhasamapatti. On the contrary, for
the practiser of the mundane Bhavanamarga, detachment from
the bhavagra is not possible because he has no access to a state
higher than it on the basis of which he could so detach himself. 434
This mundane and impure (sasrava) Bhavanamarga is accordingly
one that is not specific to the Arya, though it may once have been
practised by him too; it can precede the Arya's Darsanamarga and
does not have as its object the four Noble Principles (aryasatya) as
such (Bha~ya, vi. 1). An Arya may have acquired detachment
(vairagya) previously by means of this laukikamarga, but the
acquisition of such detachment is then a mundane one (vi. 46ab).
The fruits of asceticism (SramafJyaphala) of a Sakrdagamin and an
Anagamin can even be obtained by this laukikamarga (vi. 53Cd).
According to Yasomitra, Quieting (Samatha) is characteristic of
this laukikamarga, full liberating knowledge (ajiia) being on the
Abhidharmakosa ii. 43.
See Abhidharmakosa ii. 41-42; cf. E. Lamotte, Le traite de la Grande Vertu de Sagesse,
iii, p. 1299; below, p. I96 f.
433 Abhidharmakosabhii,ya Vi.I, 45C; and vii. 22. Cf. Asailga, Abhidharmasamuccaya,
pp. 68-69; 1. de La Vallee Poussin, L'Abhidharmakosa, ii, p. II7; viii, pp. I44-6; Lamotte,
Le traite de la Grande Vertu de Sagesse, ii, p. 1027; iii, p. 1274. See however Kathiivatthu i. 5,
which denies that the worldling (puthujjana) eliminates kiimariigavyiipiida. It is to be
recalled also that the naivasan;ljiiiiniisan;ljiiiiyatana is not counted as aniisrava and lokottara; see
above, note 351 and below, p. 200.
434 Abhidharmakosabhii,ya and Vyiikhyii vi. 45; cf. viii. 20.
431
432
196
435 Abhidharmakosavyiikhyii ii. I6d and vi. 46ab. For samatha and manaskiira on the
laukika level according to Asailga, see Abhidharmasamuccaya, p. 68.
436 Abhidharmakosabhii,ya vi. 33; cf. 11.16 and viii. 14 on the iinupiirvika.
437 Abhidharmakosabhii,ya ii. I6cd, vi. 30cd, 55. Cf. L. de La Vallee Poussin, L'Abhidharmakosa, ii, pp. 117, 134-6, 180,205; iii, p. 196 note 3; vjvi, pp. vi-ix, II9, 194,233,243,
266, 288; Melanges chinois et bouddhiques 5 (1937); p. 192 If; Vibhii,iiprabhiivrtti on
Abhidharmadfpa ii. 2 [92] (ed. Jaini, pp. 57-58).
438 Cf. L. de La Vallee Poussin, Melanges chinois et bouddhiques 5, pp. 197 note I,
219-22; Lamotte, Le traid de la Grande Vatu de Sagesse, ii, p. I035 n. I.
439 In the Mahavyutpatti (Sakaki's ed., 104, nos. 1987-9), the iisarrjiiika, together with
the asarrjiiisamiipatti and the nirodhasamiipatti, appears in the list of caitasika-dharmas. This
may be an error of redaction, for they are preceded by the viprayuktasarrskiiras priipti and
apriipti and followed by jfvita and nikiiyasabhiiga, etc. However, the question as to what
197
the function of stopping both citta and the caittas (ii. 42). The
difference between these two forms of unconsciousness is that the
asan~iiiika, as fruition (vipaka), is neutral (avyakrta), whereas the
aSaJniiiisamapatti is wholesome (Subha = kusala, ii. 42). The latter is
cultivated by ordinary worldlings (nthagjana), who take it to be
release (nihsaralJa) and liberation (mok~a), whereas the Aryas
consider it a vinipatasthana (ii. 42d). his furthermore described as
being the product of great mental effort (mahabhisan:zskarasadhya,
ii. 42d). These two states are in the Buddhist tradition clearly not
thought of as being characteristic of the Buddhist Path.
These two forms of notionlessness are accordingly carefully
distinguished from the cittaviprayuktasan:zskara already mentioned
above termed 'attainment of cessation of notions and feelings'
([ saJniiiaved( ay )ita] nirodha-samapatti) - the ninth of the Samapattis
which follows on the four Arupyas after the 'peak of existence'
(blzavagra) and the eighth Vimok~a - which also has the function
of stopping both citta and the caittas. 440 It differs from the two
forms of notionlessness just mentioned by being cultivated only
by the Arya. Following on the fourth Arupya - the naivasan:zjiiaIlasat[ljiiayatana sphere where notions are so subtle that it can be
described as neither with nor without notions - it.is defined as
'born of the peak of existence' (bhavagraja) , and as good (Subha
= krlsala, ii. 43 bc). This stage is attained through a mental act
relying on the notion of residence in quietude (Santaviharasan:zjiiapiirvaka manasikara, ii. 43b). However, even though it is described
as a simulacrum of Nirva1).a (nirvalJasadrsa, vi. 43Cd),441 one can
still fall away from the nirodhasamapatti (ii. 44d); for it is acquired
by effort (prayogalabhya) rather than by pure dispassion (vairagya,
kind of consciousness may subsist in the nirodhasamiipatti is an old one (see n. 440).
The Mahiivy"tpatti also evidently counts (no. 2297) the asalfljfiisattvas in the ninth
sorr"ii"iisa - i.e., apparently, on the level of the salfljfiiivedayitanirodha - rather than in the
fifth sattviiviisa pertaining to the fourth Dhyana of the Riipavacara (cf Dighanikaya III
263). See also Abhidharmakosa iii. 6. (The Mahiivastu (I, p. I27.5) seems in addition to
imply a criticism of the salfljfiiivedayitanirodha. See Edgerton, Buddhist Hybrid Sanskrit
DictitllIary, s.v.)
440 Abhidharmakosa ii. 43a; cf ii. 44d, vi. 43cd, 64a and viii. 33. On the persistence of
subtle thought in this nirodhasamiipatti, see L. de La Vallee Poussin, L'Abhidharmakosa viii,
p. 207 n. 6, and ii, p. 21 I n. 3; Vijfiaptimiitratiisiddhi, La Siddhi de Hillan-tsang, pp. 204 If.,
+00 If.
441 Cf. Vibhii,iiprabhiivrtti on Abhidharmadipa, p. 93.5.
198
ii. 44a).442 And only in the case of a buddha - for whom there is
nothing produced through effort (prayogika) - is the nirodhasamapatti acquired in virtue of Awakening. (bodhilabhya, ii. 44a) ~
According to the Vaibha~ikas, furthermore, because in the
nirodhasamapatti there is no citta, the Non-returner (anagamin)
Arya who attains this Samapatti takes a 'body-aggregate' (kaya)
as support; and he is then termed a kayasak:)in in so far as he
realizes this Nirva1).a-like factor through a 'body-aggregate'
(kayena, vi. 43 cd). 443
The nirodhasamapatti has also been defined in Asanga's Abhidharmasamuccaya (pp. IO-II), where it is distinguished from the
asalfljnisamapatti on the ground that the latter issues from a mental
act relying on the notion of release (nihsaraYJasalfljnapurvaka
manasikara) on the part of one free from passion (vttaraga) on the
Subhakrtsna level of the third Dhyana, but not yet free from
passion above this level; whereas the nirodhasamapatti issues from a
mental act relying on the notion of residence in quietude
(Santaviharasalfljnapurvaka manasikara) for a Vitaraga on the level
of the Akirp.canyayatana. As for the asalfljnika, it differs according
to the Abhidharmasamuccayabha~ya (p. 9) from both the asalfljnisamapatti and the nirodhasamapatti in so far as it lacks manaskara;
whereas the latter are both specified with respect to several factors
one of which is manaskara. 444
442 In the Cii!asunnatasutta (Maiihimanikaya III I07-08), the animitto cetosamiidhi which (like the nirodhasamiipatti) follows on the nevasanniiniisanniiyatana and is the object of
manasikiira - is said to be deliberately constructed (abhisankhata) and intentionally formed
(abhisancetayita), so that it is impermanent (anicca) and subject to cessation (nirodhadhamma). But it nevertheless leads to pacification and stabilization of citta, and finally to
the freedom of citta from the kiimiisava, bhaviisava and avijjiisava, and thus to liberation.
Compare the Atthakaniigarakasutta (Maiihimanikaya I 350-2) on the contemplation of the
successive stages up to and including the iikincaniiiiyatana as abhisankhata and abhisaiicetita,
and accordingly as anicca and nirodhadhamma.
The Mahiimiilunkyasutta (Maiihimanikaya I 436-7) lists neither the nevasaiiiiiiniisaiiniiyatan a nor the saiiiiiivedayitanirodha as a basis for liberating knowledge; and it mentions the
amata dhiitu as the final goal.
.
In Maiihimanikaya I 333, one who has entered this state of saiiiiiivedayitanirodha is said to
look like one who is dead (kiilakato). Cf N. Hakamaya, Journal of Indian and Buddhist
Studies (IBK), 23/2 (1975), p. ro83.
443 See above, pp. 168-70, 191-2, 194.
444 On the nirodhasamapatti in the Vijiianavada, see Hakamaya, loc. cit., pp. I081-ro74,
where attention is called (following Asanga's Mahiiyiinasarttgraha i. 7) to the absence of the
manovijiiana and the kli,!amanas, and to the presence of the iilayavijiiiina, in this samiipatti.
199
200
aiiiiapativedho).448
'
This principle has been specifically invoked in Asanga's Abhidharmasamuccaya and its Bha-!ya, where it is concluded that the
naivasa11Jjiianasa11Jjiiayatana - viz. the fourth Arupya described as
neither involving the total absence of all notions nor as comprising (distinct) notions - is mundane (laukika) rather than transmundane (lokottara), and that the aryamarga is not to be found at
its level. While the naivasa11Jjiianasa11Jjiiayatana is thus not regarded
as trans mundane, the sa11Jjiiavedayitanirodha is classified as lokottara
in so far as it is the outcome of the aryamarga involving liberating
knowledge. 449
According to the Abhidharmasamuccaya, the Anlpya-attainments are totally infused with Quieting (Samathaikarasa). Yet,
according to this same source and its commentary, a set of seven
mental acts (manaskara) makes for the attainment not only of the
four Rupa-Dhyanas but also of the four Anlpyas up to and
including the sphere where ideation subsists in a form that can be
described as neither total absence of notions or as containing
(distinct) notions (naivasa11Jjiianasa11Jjiiayatana). The Abhidharmasamuccayabha-?ya moreover specifies that the second of these seven
forms of mental act, the adhimok-?ika manaskara, transcends learning (Sruta) and reflection (cinta) and realizes both Quieting
(Samatha) and Insight (vipasyana) having as object the phenomenal sign of the characteristic of gross quiet (audarikasantalak-?a1Janimittalambana).45o In the relevant summary verse (uddana) of the
Sravakabhumi, the Dhyanas and Anlpyas are connected with
mental acts; and the vqjropama-samadhi is identified as the sixth
kind of mental act, the prayogani-?tha manaskara. 451 This placing of
vipasyana as well as samatha in the four Arupyas is noteworthy.
448 Ari.guttaranikaya IV 426. For the Sanskrit version of this Siitra-text, see Asari.ga,
Abhidharmasamuccaya, p. 69, and Abhidharmasamuccayabha,ya, p. 81. And for a discussion of
this, and of the meaning of the compound aiiiiapativedha, see L. Schmithausen, lac. cit., p.
224 and p. 229.
449 See Asari.ga, Abhidharmasamuccaya, p. 69; Abhidharmasamuccayabha,ya, p. 8 I.
450 See Abhidharmasamuccaya, p. 68; Abhidharmasamuccayabha,ya, p. 80.
451 Sravakabhumi, ed. Shukla, p. 5ID; ed. Schrnithausen in: L. Hercus et al. (ed.),
Indological and Buddhist studies G. W. de Jong Felicitation vol., Canberra, 1982), p. 472.
201
(Samaikayana) .
In the old canon, the ekayano maggo was of course the way of
the four Applictions of mindfulness (satipatthana = smrtyupasthana) that, uniquely, leads to the realization of Nirva1).a. 452 In some
places, however, the term samaikayana came to be used in
connexion with a class of persons who seek a more or less
cataleptic calm, that is, with persons attached, in terms of the
Three-Vehicle (triyana) theory, to the sravakagotra considered as a
'genus' fundamentally different from the bodhisattvagotra. A
Sravaka of this particular class would therefore be unable ever to
attain the supreme Awakening (anuttarasamyaksambodhi) of a
buddha, unlike the type of Sravaka who on the contrary turns
towards bodhi. 453
Now, according to the Ratnagotravibhaga (ii. 58-59), those
persons who, being established on the path of calm, conceive the
notion that they have achieved Nirva1).a (praptanirvaIJaSalfljiiin
= myari 'das thob 'du res can) are deflected from their earlier
postulation (purvagraha) by the teachings of the SaddharmapuIJ4arlkasutra; and being thus made to mature in the supreme Vehicle
(uttama yana, viz. the Mahayana), they receive the prophecy
(vyakaraIJa) that they are to achieve supreme bodhi. Furthermore,
according to a text quoted by Haribhadra in his commentary on
the section of the Abhisamayalalflkara that treats of the Single
Vehicle (ekayana) - a text close to the BodhicittavivaralJa ascribed
to Nagarjuna - those persons whose minds are tormented by
Sarp.saric existence (bhavad uttrastamanasah), and who conceive the
notion that they have achieved Nirva1).a (praptanirvaIJaSalfljiiin)
once their life-span is spent, have in fact not really achieved
Nirva1).a but merely the cessation of birth in the three realms of
452
453
202
458
204
461
206
6.
207
208
opponent and critic of Bhavaviveka, the author of the Prajfiapra'dtpa. Moreover, it quotes a verse (f. 354a3-4) to be found in the,
Apabhramsa Dohako~a of Saraha, and it describes the author of
this doha as 'teacher's teacher' (bla ma'i bla ma);470 in other words,
the author of the Madhyamakaratnapradtpa may have been the
grand-pupil of Saraha, the teacher (also known as Rahulabhadra)
of Arya-Nagarjunapada, who lived perhaps in the seventh
century.471 Finally, beside many other texts often placed at the
earliest in the seventh century such as the Bodhicittavivaral'}a (also
ascribed to Nagarjuna), it quotes (f. 36Ib) a 'prophecy' on
Nagarjuna from the Mafijusrtmulatantra, whereas the RajavyakaraI).aparivarta of the Mafijusrtmulakalpa in its form now known
to us contains - in addition to a differently worded vyakara1Ja on
Nagarjuna - another 'prophecy' relating to King Gopala who
founded the Pala dynasty in Bengal (rg. c. 770-810 or
775-812).472 In other words, it is possible that the Madhyamakaratnapradtpa was composed by an author who lived after the
seventh century, and perhaps as late as the ninth century, and that
he was either a contemporary or perhaps even a successor of
Santarak~ita and KamalaSIla. 473 It was indeed in the eighth
century that the bhavanakrama-theme attracted special attention
among Madhyamikas, as is demonstrated not only by Kamala470
P, mi, f. 77a2. This has been noted too by Lindtner, Wiener Zeitschrijt ... 26 (1982),
p. 175
471 See Seyfort Ruegg, in Indological and Buddhist studies, p. 5 I 1. If this is so, the author
of the Madhyamakaratnapradipa could have been a disciple of the Deutero-Nagarjunapada
since the latter was a disciple of Saraha = Rahulabhadra.
472 MafijuSrimiilakalpa (ed. T. Gal).apati Sastri), liii. 628, 816 (=ed. Rahula Samlqtyayana, verses 683 and 883, in K. P. Jayaswal, An imperial history of India, Lahore, 1934). The
Mafijusrimiilakalpa was translated into Chinese by T'ien hsi tsai at the end of the tenth
century, and into Tibetan in the eleventh century by Kumarakalasa and Silkya blo gros at
the command of Byan chub' od at Tho 00. On the date of the Mafijusrimiilakalpa, and on
earlier Chinese versions, see Y. Matsunaga, Melanges chino is et bouddhiques 20 (Melanges
R. A. Stein, iii, Brussels 1985), pp. 882-93.
It is of course possible that the version of the Tantra quoted in Bhavya's Madhyamakaratnapradipa did not contain the vyakarat}a relating to King Gopala, which could be a later
interpolation; and the date of Gopala is not therefore necessarily a terminus d quo for dating
the Madhyamakaratnapradipa. But the references to other texts, such as Saraha's Doha and
the Bodhicittavivarat}a, as well as the citation of both Dharmakirti and Candrakirti as
authorities would seem to suggest a very late seventh-century date at the earliest, and
more probably a date in the eighth or even ninth century.
473 This is the opinion of Y. Ejima, as quoted by Lindtuer in Wiener Zeitschrijt for die
Kunde Sudasiens 26 (1982), p. 183.
209
SILENCE
2IO
itself inexpressible discursively (anabhilapya, nirabhilapya) and conceptually unthinkable (acintya). Silence is thus so to speak the only
adequate way of signifying reality. As said by Candrak"irti (Prasannapada i. I, p. 57), ultimate reality (paramartha) corresponds to the
silence of the Nobles (aryar;alfl tu~r;ibavah). And according to the
Tathagataguhyasiitra, between 'the night of his awakening to supreme and perfect Awakening and his Parinirval)a, no syllable
(ak~ara) is uttered by the Tathagata. 477 This principle of in express ibility and silence is mentioned in a number of texts such as the
Samadhirajasiitra and Nagarjuna's Madhyamakakarikas. 478
In the old canon, Noble Silence (ariyo tur;hibhavo) is praised on
the same level as speech relating to Dhamma (dhammi katha);479
either one preaches the Dhamma, or one enquires of another, or
again one does not disdain Noble Silence (ariyalfl va tur;hibhavalfl
ncltimaiiiiati).480 This Noble Silence - placed on the level of the
second Dhyana and described as resulting from the cessation of
reflection and investigation (vitakkavicarar;am viipasama) - involves internal quiet (ajjhattalfl salflpasadanam), concentration of
mind (cetaso ekodibhavo), freedom from reflection (avitakka) and
investigation (avicara), and origination from samadhi. 481 Moreover, the Dharma to which the Buddha awakened is so subtle and
profound that it can barely be communicated, as a consequence of
which the Buddha at first hesitated to teach it until requested by
Brahma for the sake of people. 482
Not altogether unconnected with this principle of the conceptual and verbal inexpressibility - that is, the non-discursiveness of reality may have been the idea that it was by a single sound
only that the entire Dharma was communicated. The thesis of the
'univocality' of the Buddha's speech was maintained by all
477 See the Tathagataguhyasiitra quoted in Candraklrti's Prasannapada xviii. 7 (p. 366)
and xxv. 24 (p. 539). Cf. Prajiiakaramati, Bodhicaryavatiirapafijikii ix. 36.
478 Samiidhiriijasiitra, Chap viii and xxxii; Nagarjuna, Madhyamakakarikiis, Chap. xviii;
cf. Niraupamyastava 7 and Acintyastava 23. See also Latikavatiirasiitra iii, p. 142-4;
VimalakfrtinirdeSasiitra (ed. Oshika) , iii, p. 24 (Lamotte, L'enseignement de Vimalakfrti,
p. 147); viii, p. 75 (Lamotte, p. 317); x, p. 86 (Lamotte, p. 342); Candraklrti, Prasannapadii
i, p. 57; Santideva, Bodhicaryiivatara ix. 35-36.
479 Udana, p. II.
480 Anguttaranikaya IV 153.
481 Sarp.yuttanikaya II 273.
482 Mahiivagga pp. 4-5; Majjhimanikaya I 167-168; Sarp.yuttanikaya I 136; Anguttaranikaya II 131; Mahiivastu III, p. 314. Cf. Nagarjuna, Miilamadhyamakakarika xxiv. I2, and
Ratniivalr ii. 18.
branches of the Mahasarpghika school according to Vasumitra;483 it was rejected, however, by the Sarvastivadins,484 who
also (unlike the Mahasarpghikas who held the Buddha's speech to
be in accordance with reality)485 did not accept that all the
Buddha's Siitras are definitive in sense (nitartha).486 In a similar
context, the idea of a single sound as 'expressive' of the Buddha's
teaching is attested in the Bhadracariprat]idhay}araja (verse
ekasvara) and in the Dasabhumikasutra (ix, p. 79: ekagho~odahara). In
the Vimalakirtinirddasutra we find gsuri gcig 'single utterance'.487
This notion of 'univocality' probably stands to that of silence,
and to that of the inexpressibility or ineffability of absolute
reality, as does the notion of the Single or Unique Vehicle
(ekayana) - so often alluded to by Mo-ho-yen - to that of the
ultimate non-existence of any Vehicle at all (ayana) in the
Larikavatarasutra488 - also emphasized by Mo-ho-yen. 489
In his preference for Noble Silence and for the Single Vehicle
or even the N on-Vehicle, therefore, Mo-ho-yen clearly stands in
one major line of Buddhist thought attested in older Siitras and
then stressed by the Mahayana. 49o Assessments may differ as to
30:
212
492
adikarmika II7
adhara 19. 45
anupurvika 196
abhiprayika 10. 22. 26-35. 38. 52
arya 48-49. 158. 160. 178. 195. 200. 202.
209
ekakkha!'}ika 185
ekakkha!'}ika-yugatladdha 185
ekakia!'}a/ekakkha!'}a 161-3. 179. 186
ekakia!'}abhisamaya 154. 158. 160. 178
ekakia!'}abhisambodha 154. 157. 160. 207
ekakia!'}avabodha 154.
ekaghoia 2II
ekacaratva 162
ekacitta 186
ekacittakkha!'}ika 181
ekacittakia!'}a 167. 179
ekanaya 184
ekapativedha 185
ekapraptilabha 178
ekayana.49-50. 123. 201. 2II
ekarasa 158. 161
ekasvara 2Il
ekabhisamaya 177. 185
obhasa (vipassanao) 186. 188-9
karu!'}a 10
karmakiaya 142. 144
karman 93. 138. 141-7
kalpana 45. 100. IlO. II3
kaya I68f, 191-2. 194. 198. 206
214
kiiyasiinti 169-70
kiiyasiik,in 167-70, 192, 198
kiiryiibhisamaya 159, 177-8
kiilacakra 42
Kiisyapaparivarta 95, II4, 121,204,206
kusala(caryii) 93, 142
krama 4,12,97,120,138-9,176,180, IBI,
nirgrantha 142f.
nirvii1]a 6, 40/42, 43, 106, 163, 169, 194,
197,201, 203
nirvikalpa(ka) II2-I3
nftiirtha 10-II, 31-35, 37-38, 52, 66,
107-oB,2II
nairiitmya 25, 26, 28, 33, 35, 37, 38, 41, 43,
48-49,52, 110, II2, 183
IB~
pariIJiimanii 7
parivriijaka 22f.
paryudiisa 41
prakrtisthagotra IB-I9, 45, 164
prakrtiprabhiisvaratii (of citta) IB-19, 151,
162
jaina 142f.
}iiiinakarmasamuccaya 138, 147
prajFiapita 23, 34
prajFiapti 37
prajFiiijpaFiFiii 6, ro, 25, 64, 85, 94[, III,
tathiigatakula 96
tathiigatagarbha 4-7, 10-II, 17-55,
209-12
darSana 3, 206
dar.ianamiirga 155, 158-60, 163, 170, 177,
178, 181, 195f.
pratyabhijFiii 139
pratyavek,ii (bhutaO) 64, 94, 96, IIO, II4
pratyekabuddha 46, 48-49
prapaFica 41, 45, 48, 183, 193
prayogamiirga 158, 163, 176, 177
prayojana 22, 29-32, 38
prasajyaprati,edha 41, 66
priithamakalpika 173
*priisa/1gika 80-81, 109
brhatphala 196, 202, 205
brahmacarya 24
brahman 20-21, 26, 39-40, 54, 138
bhiivaniimiirga 9, ISS, 17of, 178, 181, 194,
195f, 205
205, 212
dhruvajdhuva 19, 43
dhvani 30, )2, 33, 138
nitya 19,21,43
Mahabherfsutra 36
mahamudra 12--':13, 72, 89, 102f, II7. See
also under phyag rgya chen po.
mahasarrghika 179-81, 2II
mfmarrsa 3 I
mukhyartha(badha) 29-32
murkhajana 93, 142
mulasarvastivada 130
yatna 189
yatharuta 37, II 8
yuganaddha 6, 85, 96, lIS, 182-92
yugapad 4,12,97, III, I20, 162, 163, 178
yoga-madhyamaka 84
yogaciira-madhyamaka/madhyamika 56, 93,
153, 164
vineyajana 34
viparyaya 25, 37
vipasyana/vipassana 4,6, 10-II, 64, 96,
105, IIOf, II4f, 126, 133, 13 8f, 164,
182-90, 200, 206
vimok~a 9, 99, 168, 192f, 202
vivak~a 28-29
vivada
212
vi~kanda(ka) 164f.
vi~1!u 40
santakara 203
sasvata
19
samathayanika 186
samasfsi(n) 1471
samapalti 9, 99, 17of, 192f, 197f, 202
sarvakaravaropeta 184, 190
sak~in 57, 77
kha tshom pa 10 5
khregs gcod 164
'khrid tshul II6, II9
bot! po 75, 81
Blon po bka'i t!Lati yig 67,84,118,137,153,
179
(g)Jags 57, 77
bJad brgyud/pa 117, I34, I9I
INDEX OF NAMES
Abhinavagupta 30-3 1,33
Anandavardhana 30-3 1,33
Aryadeva 25, 34, 86, II7, 145-6, 152
Asanga 33, 34, II6, 170, 181, 198,200
AtiSa DlpaI]1karasrljiiana 66,76, ID3, 104,
III, II6, 121
A valokitesvara 40
sBa 59f, 68f, 79, 86, 89-91, 126
Bhartrhari 29
Bhavaviveka 21, 40-41, II2, II6, 207 (cf.
Bhavya)
Bhavya II4, II9, 179, 206-09, 212 (cf.
Bhavaviveka)
Bodh Gaya 83, 86
Bodhidharma (etc.) 66-67, 88, I17f,
152-3, 179
107-08, 137
dGe lugs (pa) 36, 37
mGon po skyabs II7f, 153
loO,