JIABS

Iournal of the International
Association of Buddhist Studies
Volume 27 Number 2 2004
Katsumi MIMAKI
In memoriam, Yuichi Kajiyama (2.1.1925-29.3.2004) ..................... 255
Oliver FREIBERGER
The Buddhist Canon and the Canon of Buddhist Studies................ 261
Klaus-Dieter MATHES
Tiiraniitha's "Twenty-One Differences with regard to the Profound
Meaning" ........................................................................................... 285
Richard D. McBRIDE, IT
Is there really "Esoteric" Buddhism? .............................................. 329
Jerome DUCOR
Les sources de la Sukhiivati, autour d'une etude recente de Gerard
Fussman .... ... ... ... ..... ................ ...... ............ .... .... .... ....... ....... ... ... ........ 357
David N. GELLNER
Himalayan Conundrum? A Puzzling Absence in Ronald M. David-
son's Indian Esoteric Buddhism....................................................... 411
watermark
The Journal of the International
Association of Buddhist Studies
(ISSN 0193-600XX) is the organ
of the International Association of
Buddhist Studies, Inc. It welcomes
scholarly contributions pertaining to
all facets of Buddhist Studies.
JIABS is published twice yearly, in
the summer and winter.
Address manuscripts (two copies)
and books for review to:
The Editors, JIABS, Section de
langues et civilisations orientales,
Universite de Lausanne,
BFSH 2, CH-1015 Lausanne,
Switzerland.
Address subscription orders and
dues, changes of address, and
business correspondence
(including advertising orders) to:
Dr. Jerome Ducor,
Treasurer lABS,
Section de langues et civilisations
orientales, Faculte des lettres
Universite de Lausanne, BFSH 2
1015 Lausanne-Dorigny
Switzerland
email: iabs.treasurer@orient.unil.ch
Web: www.iabsinfo.org
Fax: +4121 692 30 45
Subscriptions to JIABS are USD 40
per year for individuals and USD 70
per year for libraries and other
institutions. For informations on
membership in lABS, see back cover.
Cover: Cristina Scherrer-Schaub
© Copyright 2004 by the International
Association of Buddhist Studies, Inc.
Printed in Belgium
EDITORIAL BOARD
SCHERRER-SCHAUB Cristina A.
TILLEMANS Tom J.F.
Editors-in-Chief
BUSWELL Robert
COLLINS Steven
Cox Collet
GOMEZ Luis O.
HARRISON Paul
VON HINOBER Oskar
JACKSON Roger
JAINl Padmanabh S.
KATSURA Shoryu
KuoLi-ying
LOPEZ, JI. Donald S.
MAcDONALD Alexander
SEYFORT RUEGG David
SHARF Robert
STEINKELLNER Ernst
ZORCHER Erik
In memoriam
yurem KAJIYAMA
(2.1.1925-29.3.2004)
KATSUMI MIMAKI
Professor Yilichi KAJIYAMA, Emeritus Professor of Buddhist Studies at
Kyoto University, died at 6:09 o'clock in the morning, March 29,2004, at
the age of 79. Because he had practiced martial arts such as Kend6 and Jildo
in his junior high school days, he had always felt confident about his health.
But since his heart operation in the summer of 2000, he had begun to be
anxious about his health. In fact, in January 2004, while taking a walk, he
accidentally fell down and, hitting his face against the ground, broke his
teeth. He said to me: "It was a shock for me, who had practiced so well
the 'safe ways of falling down' (ukemi) of JUdo, to fall down and hit my
face on the ground". The last time I saw him was in a hospital bed two days
before he died. I had been asked by our office to find out whether he would
be able to attend a ceremony to receive an award on May 12. He answered:
"By the time of the ceremony, I will be completely cured and will go by
myself to receive the award." Who could have imagined that he would die
suddenly two days later? The cause of his death was described in the news-
papers as "cardiac insufficiency", but according to the official diagnosis it
was in fact "stomach cancer" that deprived him of life.
This is indeed an irrecoverable loss to the Buddhist Studies world.
Professor Kajiyama treated several topics in Buddhist studies through a
sharp insight and elucidated many unexplored fields through a clear inves-
tigation. We could have expected still many many results from him. Here,
from the depths of profound grief, I would like to pray for the peace of
his soul by presenting a brief survey of his career and academic activi-
ties, even if I am not truly competent to do so.
Born in Shizuoka city on Jannary 2, 1925, Professor Kajiyama attended
Shizuoka Junior High School and Shizuoka High School. He entered Kyoto
Journal of the International Association of Buddhist Studies
Volume 27 • Number 2 • 2004
256 KATSUMI MIMAKI
University in October 1944, studying in the department of Philosophy
(Buddhist Studies) of the Faculty of Letters until his graduation in March
1948. Then he became a special research fellow of the Graduate School
of the same university. He married Hiroko MATSUURA in 1951 and had a
daughter Tomoko in 1952. From April 1953 until March 1956 he contin-
ued his research, while teaching, under the direction of Professors J. Kashab
and Satkari Mookerjee at the Nalanda Pili Institute, in Bihar, India. After
his return to Kyoto, he became an Assistant Professor in the Faculty of Let-
ters of Kyoto University in April 1956. Having received a prize from the
Japanese Academy for his joint-work Ju yong guan ~ , * I J i ~ ~ in May 1959
as well as a prize from the Japanese Association of Buddhist Studies in
October of the same year, he was promoted to Associate Professor in
March 1961. From July 1961 until August 1962 he studied under the guid-
ance of Professor John Brough as a fellow of the British Council in the
School of Oriental and African Studies (SOAS) of London Unversity.
From September to December 1962 he continued his research under the
guidance of Professor Erich Frauwallner at the Indological Institute of
Vienna University. Promoted to Professor of Kyoto University in Novem-
ber 1971 he concentrated on research and education for the sixteen years
until his retirement in March 1988, and in April 1988 he was given the title
of Emeritus Professor of Kyoto University. From April 1988 until March
1997 he was Professor at Bukkyo University (Faculty of Letters, Depart-
ment of Buddhist Studies). There in April 1991 he founded the Compre-
hensive Research Institute, and as its Director he inaugurated its Bulletin.
From April 1997 until March 2001 he was Professor at Soka University.
There in June 1997 he founded the International Research Institute for
Advanced Buddhology, and as its Director he inaugurated the Annual
Report. It was amazing to see how easily he accomplished the difficult tasks
of founding a new Institute and launching a new journal.
The research of Professor Kajiyama was recognized and appreciated not
only inside Japan, but also internationally. He had occasions to teach as
Visiting Professor at the University of Wisconsin (1967-1968), the Uni-
versity of California, Berkeley (1974; 1981; 1997), Harvard University
(1986), Vienna University (1985), and Leiden University (1989).
The formation of his knowledge and his method of guiding his disci-
ples seemed to have been deeply influenced by his own research experi-
YDrcHI KAJIYAMA (2.1.1925-29.3.2004) 257
ence with three foreign teachers, as he often mentioned. The first was
Professor MookerJee of the Nalanda Pali Institute, under whose direction
he produced a Sanskrit reconstruction of the Vaidalyaprakara1J.a of Nagar-
juna. The second was Professor Brough of London University, under
whose direction he completed an annotated English translation of the
Tarkabhii>!ii of (1l-12th c.). As one of the best intro-
ductions to Indian Buddhist epistemology and logic, this work, first
published in 1966 in Memoirs of the Faculty of Letters (Kyoto Univer-
sity No. 10; included also below in Bibl. 1), was reprinted in the series
of the Institute for Tibetan and Buddhist Studies in Vienna (Bib!. 2; cf.
a Japanese translation in Bibl. 8). His third teacher was Professor Frauwall-
ner, under whose guidance he completed a German translation of the first
chapter of the Prajfi.iipradfpa of Bha(va)viveka (ca 500-570) (included in
Bib!. 1).
There is no doubt that his experience of the so-called person to person
method under these three professors deeply influenced his own research as
well as his method of guiding his students. When I was student preparing
a master's thesis, he read my text with me once a week during the summer
vacation. He did not seem to need much preparation, but he corrected my
reading and each time gave me very useful comments and suggestions. It
was indeed thanks to him that I could read this difficult text through. Need-
less to say, I was not the only student to receive the benefit of his method-
ical attention. It was naturally the same not only for other Japanese but
also for foreign students. This is the reason why so many of his disciples
continue now their academic activities in many places all over the world.
The research field of Professor Kajiyama covers several branches of
Indian Mahayana Buddhism. It is almost impossible to present all of his
works in this limited space, but we can at least enumerate the following
five branches as his main SUbjects: (1) Madhyamaka philosophy, (2)
Mahayana sutras, among others the Prajfi.iipiiramitiisittra, (3) Buddhist
logic and epistemology, (4) analysis of the doctrinal positions of several
Buddhist schools, and (5) Chinese and Japanese Buddhist thought, based
on Indian Buddhist philosophy.
As for the first field, we can mention his clear overview of the history
of Indian Madhyamaka philosophy (Bib!. 11), and, in particular, as is
shown by his above-mentioned German translation of the first chapter of
258 KATSUMI MIMAKI
the Prajfiiipradzpa, which elucidates Bha(va)viveka's philosophy, one of
Professor Kajiyama's contributions was to clarify the important role that
the Madhyamika played in the history of Indian logic. Another important
contribution was that, having critically reexamined various theories and
using Bha(va)viveka for important evidence, he established the dates of
Sthiramati (510-570) and DharmapaIa (530-561).
In the second field, we have his Japanese translations of the A!ftasii-
hasrikii-prajfiiipiiramitii-sutra (Bibl. 7) and the Pratyutpanna-buddha-
sarrtmukhtivasthita-samiidhi-sutra (Bibl. 21). He also treated several inter-
esting topics such as the Bodhisattva, transfer of merit, transmigration,
smpa worship, the Buddha-body, the origin of Mahayana Buddhism, Bud-
dhist eschatology, miracles, and supernatural events. In the field of logic
and epistemology we have his clear description of Buddhist logic (Bibl.
16). Besides the above-mentioned translation of the Tarkabhii!fii of
the most important contribution in this field is his trans-
lation of the Antarvyiiptisamarthana of Ratnakarasanti, in which we can
see the final phase in the development of Indian Buddhist logic. This
text of Ratnakarasanti had already been twice translated into Japanese by
Professor Kajiyama (1959, 1989), but its final form was completed in a
new edition of the Sanskrit text, juxtaposed with a collated text of the
Tibetan versions, an English translation, and a facsimile edition of the San-
skrit manuscript (Bibl. 3). In his fourth field, of special note is his study
of Buddhist conceptions of existence and knowledge (Bibl. 12). Finally,
in his fifth field, he studied East Asian Buddhist ideas, including the
thought of Shinran (Bibl. 18). Recently Professor Kajiyama had occa-
sions to discuss contemporary ethical problems (brain death, human
cloning, nuclear weapons, etc.) from the perspective of Buddhist Studies.
The above is only a partial summary of Professor Kajiyama's academic
activities. His works are indeed so numerous and multi-faceted that it is
impossible to mention all of them here. The attached bibliography lists
his books and only his Western-language articles published after 1989. Let
me also mention that a Collected Works of his Japanese publications (an
estimated 8 volumes) is planned by the publisher Shunjilsha. His pre-
1989 Western-language articles have been published by Rinsen Publishc
ers (Bibl. 1). Currently out of print, this collection will be reprinted soon.
Rinsen will also undertake the publication of his post-1989 articles in
Western languages.
YDIcm KAJIYAMA (2.1.1925-29.3.2004) 259
Books
1) Y. Kajiyama, Studies in Buddhist Philosophy, ed. by Katsumi MiMAKI et al.,
Rinsen, Kyoto, 1989.
2) An Introduction to Buddhist Philosophy, an annotated translation of the Tarka-
of Arbeitkreis fUr Tibetische und Buddhistische Stu-
dien Universitiit Wien, Wien, 1998.
3) The Antarvyiiptisamarthana of Ratniikarasiinti, Bibliotheca Philologica et
Philosophica Buddhica II, The International Research Institute for Advanced
Buddhology, Soka University, Tokyo, 1999.
4) Kil no Ronri <ChUgan> (The Logic of the Madhyarnaka School), Coau-
thored with S. UEYAMA, Kadokawa, Tokyo, 1969; Reprint as a pocket-size
book,1997.
5) Bukkyo no Shiso (Buddhist Thought), Coauthored with S. UEYAMA, Chiio-
Koronsha, Tokyo, 1974
6) Ryilju RonshU (Japanese translation of Nagarjuna's works), Daijo Butten 14,
Coauthored with R. URYDZU, ChilO-Koronsha, Tokyo, 1974.
7) Hassenju-Hannyakyo, (Japanese translation of the
piiramitii-siltra), 2 Vols. (The second volume is coauthored with T. TANn),
Chiio-Koronsha, Tokyo, 1974-5.
8) Ronri no Kotoba (The Language of Logic), A Japanese translation of
Chiio-Koronsha, Tokyo, 1975.
9) Seshin Ronshil (Japanese translation of Vasubandhu's works), Daijo Butten
15, Coauthored with N. ARAMAKI and G. M. NAGAO, Chiio-Koronsha, Tokyo,
1976.
10) Hannyakyo - Kil no Sekai (Prajniipiiramitii-siltra - the World of Empti-
ness), Chiio-Koronsha, Tokyo, 1976.
11) ChUkan Shiso (Madhyarnaka Thought), Koza Daijo Bukkyo 7, Coauthored
with R. URYUZU, M. TACHIKAWA, T. EnMA, M. IcmGo, K. MIMAKI, C.
TAMURA, and S. HARADA, Shunjiisha, Tokyo, 1982 (first edition), 1995 (new
edition).
12) Bukkyo ni okeru Sonzai to Chishiki (Existence and Knowledge in Buddhism),
Kinokuniya, Tokyo, 1983.
13) Satori to Eko (Enlightenment and Merit Transfer), KOdansha,Tokyo, 1983.
14) Kil no Shiso-Bukkyo ni okeru Kotoba to Chinmoku (The Idea of Empti-
ness - Language and Silence in Buddhism), Jinbun-Shoin, Kyoto, 1983.
15) Bosatsu to iukoto (On Bodhisattvas), Jinbun-Shoin, Kyoto, 1984.
16) Ninshikiron to Ronrigaku (Epistemology and Logic), Koza Daijo Bukkyo 9,
Coauthored with S. KATSURA, H. TOSAKI, A. AKAMATSU, K. MiMAKI, Y.
MrYASAKA, S. KAWASAKI, and H. NAGASAKI, Shunjiisha, Tokyo, 1982 (first
edition), 1995 (new edition).
17) Budda-charita (Japanese translation of the Buddhacarita), Coauthored with
N. KOBAYASm, M. TACHIKAWA and K. MlMAKI, KOdansha Shuppan-
Kenkyiisho, Tokyo, 1985.
18) Shinran (A Study of the founder of the JOdo-Shin school), Chiio-Koronsha,
Tokyo, 1987.
260 KATSUMI MIMAKI
19) Rinne no Shiso (The Buddhist Concept of Transmigration), Jinbun-Shoin,
Kyoto, 1989.
20) Ku Nyumon (An Introduction to Emptiness), Shunjiisha, Tokyo, 1992.
21) Hanjuzanmai kyo (Pratyutpannabuddha-sarrzmukhiivasthita-samadhi-sutra:
partial translation and interpretation), in JOdo Bukkyo no Shiso (the Idea of
Pure Land Buddhism) II, Kodansha, Tokyo, 1992.
22) Satori to Eko - Daijo Bukkyo no Seiritsu (Enlightenment and Merit Trans-
fer - the origin of Mahayana Buddhism), Expanded version of Satori to Eko
(1983), Jinbun-Shoin, Kyoto, 1997.
Articles in Western languages, published after his collection,
Studies in Buddhist Philosophy (see Bib). 1)
23) "Fundamentals of Buddhist Ethics", Zen Buddhism Today: Annual Report
of the Kyoto Zen Symposium (The Kyoto Seminar for Religious Philosophy),
1990, pp. 61-70.
24) "On the Authorship of the Proceedings of the 2nd Interna-
tional Dharmakfrti Conference, Vienna, 1989, pp. 107-117.
25) "Presuppositions of Amitabha Worship in India", Bericht. Japanisch-
Deutsche Studiengemeinschaft zur Kultur des Buddhismus, Bukkyo-Uni-
versitat, Kyoto, 1990, pp. 11-24.
26) "Sthiramati, Uddyotakara, and Arcafa on Vyabhicara", Asiatische Studien
XL VI. 1, Etudes bouddhiques offertes a Jacques May, Zeitschrift der
Schweizerrischen Asiengeseilschaft, 1992, pp. 212-221.
27) "Prajnaparamita and the Rise of Mahayana", Buddhist Spirituality: Indian,
Southeast Asian, Tibetan, and Early Chinese, ed. by Y. TAKEucm in asso-
ciation with Jan Van Bragt, New York, Crossroad Publishing, 1993, pp. 137-
154.
28) "Buddhist Eschatology, Miracles, and Power other than Self", A Compre-
hensive Review of the Pure Land Buddhism of Honen, International Acade-
mic Symposium, Bukkyo University-L.A. Extension, 1998, pp. 16-21.
29) "Buddhist Eschatology, Supernatural Events and the Lotus Sutra", The Jour-
nal of Oriental Studies, Vol. 8, 1998, pp. 15-37.
30) "Sarp.ghabhadra und Nagarjuna uber die Theorie der zwolfgriedrigen Kausali-
tat", Horin -Vergleichende Studien zur Japanischen Kultur, No.6, Eko-Haus
der japanischen Kultur, Dusseldorf, 1999, pp. 139-149.
31) "AIayavijfiana und Abhangiges Entstehen", Horin, Vergleichende Studien zur
japanischen Kultur, No.7, Eko-Haus der japanischen Kultur, Dusseldorf,
2000, pp. 77-92.
32) "TheSaddharmapuQ.9arika and Siinyata Thought" (tr. by Wayne Yokoyama),
The Journal of Oriental Studies, Vol. 10, 2000, pp. 73-96.
33) "Buddhist Cosmology as Presented in the Yogacarabhiimi", Wisdom, Com-
passion, and the Search for Understanding: The Buddhist Studies Legacy of
Gadjin M. Nagao, ed. by Jonathan A. Silk, University of Hawai'i Press,
2000, pp. 183-199.
THE BUDDIDST CANON
AND THE CANON OF BUDDIDST STUDIES
OLIVER FREIBERGER
In contemporary academia, the concept of a canon is rather unpopular!.
The word 'canon'2 reminds us of elites who use their power to suppress
opposing views by labeling them as non-canonical and heretic. It reminds
us of colonialism and 'Orientalism,' ofthe ways Western scholars 'can-
onized' the knowledge of 'the East.' And it reminds us of a 'classical'
canon in education that conservative instructors and politicians attempt to
save from postmodern randomness. In Buddhist Studies, such general
reservations about the canon appear to become manifest in two demands:
In our research, we must focus on sources other than the Buddhist canon;
and in teaching, we must abandon the inherited 'canon' of class readings,
which again consists mainly of Buddhist canonical texts, and must teach
contemporary Buddhist practice instead.
In this paper, I will reconsider those demands by reflecting upon the
Buddhist canon as a subject of research, and upon our own scholarly
canonizations, the secondary canon of Buddhist Studies. The first part of
the paper examines the role of the Buddhist canon in research and in
teaching, the trend towards non-canonical sources, and the current affec-
tion for contemporary practice. As a textual scholar who works with
canonical texts, I intend to point to some risks that are, in my view, inher-
ent in that general trend. To corroborate my critique and to illustrate what
1 Earlier versions of this paper were presented at the University of Bayreuth and at the
University of Texas at Austin in spring, 2004. I thank the participants of the subsequent
discussions for their responses and, in particular, Janice Leoshko for her comments and
valuable suggestions.
2 The word 'canon' is derived from Greek kanan, "cane", "ruler", figurative: "rule",
"norm", "model". See for the etymology and historical development of the term Gerald T.
Sheppard, "Canon," The Encyclopedia of Religion, ed. Mircea Eliade, vol. 3 (New York:
Macmillan, 1987), pp. 62-69.
Journal of the International Association of Buddhist Studies
Volume 27 • Number 2 • 2004
262 OLIVER FREIBERGER
I consider the value of canonical texts for scholarship, I will, in the second
part of the paper, examine one exemplary issue: the image of the laity in
early Buddhism. I intend to demonstrate that canonical texts are, in con-
trast to the common view, a rich source for current scholarly interests
(such as the issues of religious practice and diversity). I will argue that
the image of the canon as being consistent, one-dimensional, and purely
normative - an image that underlies the current rejection of canonical
texts - is to a large extent the product of a 'canonization' carried out by
earlier generations of scholars. Discussing further implications in the third
part of the paper, I will argue that by excluding the canon, Buddhist Stud-
ies runs the risk of canonizing other sources for research and, at the same
time, enhances particularism in teaching. Rather, the opposite approach
appears to be useful: a roughly 'canonized' introductory education in
Buddhist Studies, spiced with selected data that are suitable for under-
mining simplification, and an attitude in research that is open for all kinds
of sources, including canonical texts.
1. The Buddhist Canon in Research and in Teaching
When scholars of religion apply the term 'canon' to a certain corpus
of texts, they usually wish to emphasize two aspects: its normative, author-
itative character on the one hand, and its fixed and standardized form on
the othez3. The latter feature is the result of a process of canonization. Gen-
erally speaking, this process begins when within a tradition certain insti-
tutions select a limited number of texts and define them as authoritative,
3 Jonathan Z. Smith defmes a canon broadly as "the arbitrary fixing of a limited num-
ber of 'texts' as immutable and authoritative." "Sacred Persistence: Toward a Redescrip-
tion of Canon," id., Imagining Religion: From Babylon to Jonestown (Chicago/London:
University of Chicago Press, 1982), pp. 36-52, here: 44. Gerald Sheppard describes those
two aspects as follows: "On the one hand, [the term 'canon'] can be used to refer to a
rule, standard, ideal, norm, or authoritative office or literature, whether oral or written.
On the other hand, it can signify a temporary or perpetual fixation, standardization,
enumeration, listing, chronology, register, or catalog of exemplary or normative persons,
places, or things. The former dimension emphasizes internal signs of an elevated status.
The latter puts stress on the precise boundary, limits, or measure of what, from some pre-
understood standard, belongs within or falls outside of a specific 'canon. ", Sheppard,
"Canon," p. 64.
THE BUDDIDST CANON AND THE CANON OF BUDDIDST STUDIES 263
that is, 'canonicaI.' Subsequently, those or other institutions will need to
protect and defend this canon
4

Following this deftnition, scholars of Buddhism are used to la,beling
certain Buddhist text collections 'canonical.' I focus in this paper on the
so-called Pili canon of the Theravada school. Although little is known
about the fonnation of the texts after the death of the Buddha, scholarship
holds that for a couple of centuries, Buddhists transmitted the constantly
increasing text collection orally, until, according to the Theravada tradi-
tion, it was written down for the fIrst time in Sri Lanka, in the 151 century
before the Common EraS. From the 5
th
century commentaries onwards,
at the latest, both canonical features are observable: the Pali canon is
regarded as nonnative and authoritative, and its textual contents are deftned
and ftxed.
1.1. Leaving the Canon Behind: Alternative Sources for Research
From its very beginning, Westemschoil,lIship has focused on the Pili
canonical texts, assuming that historically they were the most reliable
source for reconstructing the life of the Buddha, his original teachings,
and the new religion's early development. In recent decades, scholars
have raised new questions and brought up a number of critical method-
ological issues: the problem of the alleged objectivity of the scholar; the
4 According to Aleida and Jan Assmann, we can distinguish three tasks these institu-
tions have to fulfill: censoring; maintaining the text; and maintaining its meaning (Zen-
sur, Textpflege, and Sinnpflege). Censoring means delimitating the text from the extrane-
ous and false; maintaining the text means immunization against change, the orthopraxy of
language within the tradition; maintaining the meaning of the text means compensating the
semantic deficiencies of the orthopraxy of language, a phenomenon which is often mani-
fest in commentaries of canonical texts. Aleida and Jan Assmann, "Kanon und Zensur,"
Kanon und Zensur: Beitrage zur Archaologie der literarischen Kommunikation II, ed.
Aleida and Jan Assmann (Miinchen: Fink, 1987), pp. 7-27. See also the articles in the
same volume by Alois Halm, "Kanonisierungsstile," pp. 28-37; and Carsten Colpe, "Sakrali-
sierung von Texten und Filiationen von Kanons," pp. 80-92.
5 Although it is likely that at this point, the Pili canon was more or less fixed, we can
be fully sure about its actual contents only from Buddhaghosa's commentary in the 5
th
cen-
tury onwards. Cf. K.R. Norman, "Buddhism and Canonicity," id., A Philological Approach
to Buddhism: The Bukkyo Dendo Kyokai Lectures 1994 (London: SOAS, 1997); cf. also
Gregory Schopen, "Two Problems in the History of Indian Buddhism: The Layman/Monk
Distinction and the Doctrines of the Transference of Merit," StlI10 (1985), 9-47, here: 9f.
264 OLIVER FREIBERGER
need for scholarly interpretation and creativity; the issue of evaluative
assessment and normativity; the disputed quest for the original intention
of a text's author; the significance of politics and power; and, in partic-
ular, the role of the written text in general
6
• A considerable number of
scholars have become critical of the traditional, historical-philological
paradigm. Emphasizing the obvious multifaceted character of Buddhism
and the need for adequate ways to deal with its diversity, they suggest
methods other than philological (for example, anthropological, sociologi-
cal, and art historical methods), and approaches other than historical (such
as cross-cultural analysis, feminist criticism; deconstruction, and literary
criticism)7. The critique of the traditional historical-philological approach
concerns, of course, particularly its focus on canonical texts; scholars
tend to look for other meaning-producing forms of sources. In the words
of Jose Cabezon: "There is today a call for the increased investigation
of alternative semiotic forms - oral and vernacular traditions, epigra-
phy, ritual, patterns of social and institutional evolution, gender, lay and
folk traditions, art, archeology and architecture."8
In the course of this trend, the role of the Pilii canon in Buddhist Stud-
ies has been subject to critical examination. Charles Hallisey, for exam-
ple, has analyzed the way Western Buddhologists used to deal with it. He
points to the beginnings of Buddhist Studies and their typical 19
th
cen-
tury historicist approach "with its split between older and later sources
and its positivistic concerns for origins." This approach led scholars to
the attitude that unlike modern Buddhists, only Western scholars, due to
their knowledge of Pilii, have access to 'original' Buddhism.
9
Instead of
6 Jose Ignacio Cabezon has thoroughly examined these methodological issues in his
article "Buddhist Studies as a Discipline and the Role of Theory," JIABS 18 (1995),
pp. 231-268. As he convincingly demonstrates, the debate takes place between the two poles
of positivist/objectivist and interpretivist/subjectivist/constructionist approaches.
7 Cf. Cabezon, "Buddhist Studies as a Discipline ... ," pp. 238f. and 264f.
8 C[ Cabezon, "Buddhist Studies as a Discipline ... ," pp. 262f.
9 Charles Hallisey, "Roads Taken and Not Taken in the Study of Theraviida Bud-
dhism," Curators of the Buddha: The Study of Buddhism under Colonialism, ed. Donald
S. Lopez, Jr. (Chicago/London: University of Chicago Press, 1995), pp. 31-61,here: pp. 34-
38. Trying to apply a one-sided concept of Orientalism to this issue, however, would be
too simple. As Hallisey observes, "there was something like a productive 'elective affin-
ity' between the positivist historiography of European Orientalism and Buddhist styles of
self-representation" (p. 43); see also pp. 47-49. This is true also for the common usage of
THE BUDDIDST CANON AND THE CANON OF BUDDIDST STUDIES 265
cOlitinuing a scholarly tradition that focuses on the Pali canon, scholars
of Buddhism should examine, according to Hallisey, commentaries, sub-
commentaries, and in particular, local contexts and works composed in
vernacular languages. He sketches an "alternative historical paradigm
which will encourage us to expect meaning to be produced in local cir-
cumstances rather than in the origins of the tradition."JO In Hallisey's
view, this local production of meaning is of crucial interest for Buddhist
Studies, because it reveals the interaction between the text and its users
and is thus connected to the 'real life' of Buddhists much more closely
than the Pali canon is. Although not explicitly abandoning the canonical
texts for research, Hallisey discourages from examining them. The Pali
canon appears as one among many representations of Buddhism, but as
a rather unexciting one.
1.2. Teaching Buddhism without a Canon: The Affection for Contempo-
rary Practice
It comes as no surprise that the focus on the Pilii canon in research
caused an identical focus in teaching. Charles Hallisey remarks that "the
study of the Theravada became equated with the study of the Pilii canon,
·and it is still common for a student to finish a graduate program in Bud-
dhist Studies without ever having read a Theravadin commentarial text." 11
the commentaries of the Pali canon; emulating Buddhist tradition, scholars of Buddhism
have used them merely as a tool for understanding the canonical text but not as a religious
expression of its own right that was composed centuries after the canon itself. In addition
to that, several other ways of emulating the Buddhist tradition are observable. Luis Gomez
examines similarities in the philological and the scholastic approaches, in the doctrinal
commitment, and in the notion of history (decay or culmination). Luis Gomez, "Unspoken
Paradigms: Meandering through the Metaphors of a Field," JIABS 18 (1995), pp. 183-230.
For the dialectics of orientalism, occidentalism, 'auto-oc.cidentalism,' and 'auto-orientalism'
in (the study of) Asian religions cf. also the recent volume, Religion im Spiegelkabinett:
Asiatische Religionsgeschichte im Spannungsfeld zwischen Orientalismus und Okziden-
talismus, ed. Peter Schalk, Max Deeg, Oliver Freiberger, Christoph Kleine, Acta Univer-
sitatis Upsaliensis: Historia Religionum 22 (Uppsala: University of Uppsala, 2003); for
an outline of this volume see my note, "Religion in mirrors: Orientalism, Occidentalism,
and Asian Religions," Journal of Global Buddhism 4 (2003), pp. 9-17 (online: http://www.
globalbuddhism.org).
IO Hallisey, "Roads Taken ... ," pp. 50-53.
I1 Hallisey, "Roads Taken ... ," p. 44.
266
OLIVER FREIBERGER
The text selections in 'classical' anthologies of Buddhist Studies confinn
Hallisey's impression. Concerning Theravada Buddhism, those compila-
tions contain passages taken almost exclusively from the Pilii canon
12
.
Moreover, their editors seem to agree upon what the 'significant' topics
were and, correspondingly, what text passages to select
13
• Roughly, the
topics are: the early life of Siddhattha Gotama, his ascetic years, his
awakening, and his death; the 'sermon of Benares', the Four Noble Truths
and the Middle Way, the doctrines of dependent origination, anatta, the
five khandhas, karma and rebirth, meditation and nibbiina, general ethi-
cal principles, and the basic rules of the sangha. These topics reflect
the traditional classification of the 'three jewels' (buddha, dhamma,
sangha); their selection is, certainly not by chance
14
, largely in accor-
dance with the later commentarial tradition of the Theravada school. The
text passages selected for anthologies correspond to these topics. Although
being not entirely identical, the compilations constitute a largely inter-
secting set of texts. The process of selecting topics and texts can be viewed
as a form of canonization.
Taking the anthologies as a means (or an expression) of teaching Bud-
dhism, we may state that the discipline of Buddhist Studies has created
its own teaching canon - a secondary canon, as it were, extracted from
the primary one. This secondary canon possesses the general features:
it is authoritative, and it is, to a certain extent, fixed; the fact that the
anthologies contain very similar text collections indicates that the academic
community has 'observed and protected' the selection
15
.
12 To a much lesser extent, they also contain sections from quasi-canonical works such
as the Milindapafiha or the Visuddhimagga.
13 Examples are: Buddhism in Translations: Passages Selectedfrom the Budhist Sacred
Books and Translated from the Original Pali into English, by Henry Clarke Warren (Cam-
bridge, MA: Harvard University Press, 1915); The Living Thoughts ofGotama the Buddha,
presented by Ananda K. Coomaraswamy and LB. Horner (LondonlEdinburgh: Morrison
and Gibb, 1948); Buddhism: A Religion of Infinite Compassion - Selections from Buddhist
Literature, ed. Clarence H. Hamilton (New York: Liberal Arts Press, 1952); The Teach-
ings of the Compassionate Buddha, ed. E.A. Burtt (New York: Mentor, 1955); Buddhist
Texts through the Ages, ed. Edward Conze et aI. (Oxford: Cassirer, 1954); The Wisdom of
Buddhism, ed. Christmas Humphreys (New York: Random House, 1961).
14 Cf. above, note 9.
15 Below, I point to canonical texts that were omitted in this canonization process.
While my example concerns the image of the Buddhist laity, another example would be
THE BUDDHIST CANON AND THE CANON OF BUDDHIST STUDIES 267
In recent times, the textual focus in teaching has shifted. One example
of a new type of anthology is the voluminous collection Buddhism in
Practice, edited by Donald Lopez, published in 1995
16
• All but one of the
Theravada texts in this volume are non-canonical, some even written by
contemporary Buddhists
17
• Corresponding to the book's title, all texts
concern in one way or the other Buddhist practice; they deal, for exam-
ple, with the consecration ritual of Buddha images, with meditation, or
donation. As an example of anthologies used in teaching, this compilation
shows that the criterion for selecting texts is not their canonical status
anymore but their significance for Buddhist practice.
A recent collection of essays on "Teaching Buddhism in the West"
places strong emphasis on practice, as well
l8
• In its first chapter, Frank
Reynolds criticizes the usual 'Introduction to Buddhism' course. He
describes it as "the kind of survey course that begins with ( ... ) the his-
toricallife and teachings of Gautama Buddha, ( ... ) moves through a
rapid-fire treatment of some 2500 years of Buddhist intellectual and social
history ( ... ), and finally concludes with an equally rapid-fire survey of
contemporary Buddhism in various countries around the world. "19
As examples of an "alternative approach that will be appropriate and
effective within a postmodem liberal arts curriculum," he suggests
three types of courses on Buddhism, two of which are significant for our
the Vinaya passages that I.B. Homer refused to translate in her translation of the
Vinayapi{aka. These passages, which deal with sexual issues, appear, in her words, "unsuitable
for incorporation in a translation designed principally for Western readers," because of "the
outspokenness and crudeness" they contain. The Book of the Discipline (Vinaya-Pi{aka),
trans!. LB. Homer, vo!' 1 (Oxford: Pali Text Society, 1992 [1938]), p. 197. Those passages
have been translated recently; see Petra Kieffer-Piilz, "Parajika 1 and Sanghiidisesa 1:
Hitherto Untranslated Passages from the Vinayapi{aka of the Theravadins," Traditional
South Asian Medicine 6 (2001), pp. 62-84.
16 Buddhism in Practice, ed. Donald S. Lopez, Jr. (Princeton, N.J.: Princeton Univer-
sity Press, 1995). Cf. also John S. Strong, The Experience of Buddhism: Sources and Inter-
pretations, 2
nd
ed. (Belmont, CA: Wadsworth, 2002).
17 The only canonical text in the book, the Gotamf-apadana, is little known and has
been translated into a Western language here for the first time. Buddhism in Practice ... ,
pp. 113-138.
18 Teaching Buddhism in the West: From the Wheel to the Web, ed. Victor Sagen Hori,
Richard P. Hayes, James Mark Shields (London/New York: RoutledgeCurzon, 2002).
19 Frank E. Reynolds, "Teaching Buddhism in the Postrnodern University: Understan-
ding, Critique, Evaluation," Teaching Buddhism in the West ... , pp. 3-16, here: p. 8.
268 OUYER FREIBERGER
considerations
2o
• First, he proposes a concept for an introductory under-
graduate course that deals not with the foundations of Buddhism and its
historical development but with practices of contemporary Buddhists in
Asia and North America. The goal of such a course is "to' introduce
students to a broadly representative variety of the real worlds of real
Buddhists who are involved in real Buddhist practices that generate real
Buddhist experiences. "21 Reynolds' second concept of an advanced course
comprises a canonical text, but focuses on the "ways in which the text
has been received and put to use in the tradition. "22 It is obvious that in
this article, Reynolds does not attach great importance to the Pili canon
for teaching purposes. His introductory course contains no canonical text
whatsoever, and his advanced course focuses not on the content of the
selected canonical text, but on its role in practice.
The approach underlying Reynolds' concepts seems to represent a
general trend. In the volume on Teaching Buddhism in the West, the authors
suggest a number of teaching methods, all of which focus not on canonical
texts but on other religious expressions, particularly on religious prac-
tice. For their class readings, they select either Buddhist texts that deal
with - or are used in - practice, or scholarly articles describing con-
temporary forms of it
23
• Canonical texts are of interest only as far as they
have a role in practice.
20 The third course deals with the establishment and development of Buddhism in the West
and the development of.Buddbist Studies. Reynolds, "Teaching Buddbism ... ," pp. 9-1l.
Reynolds describes his vision of postmodem liberal education in detail in his article "Recon-
;structing Liberal Education: A Religious Studies Perspective." Beyond the Classics? Essays
in Religious Studies and Liberal Education. ed. Frank E. Reynolds. Sheryl L. Burkhalter
(Atlanta: Scholars Press. 1990). pp. 3-18.
21 Reynolds. "Teaching Buddhism ...• " pp. 7-9. here: p. 9. One decade ago. Reynolds'
brief description of an introductory course had included contemporary practice as merely
one among several elements. See his chapter on "Introducing Buddhism" in Teaching the
Introductory Course in Religious Studies: A Sourcebook, ed. Mark Juergensmeyer (Atlanta:
Scholars Press, 1991). pp. 71-77.
22 The introduction of this course includes a "concise consideration of sub-topics" (!)
such as textual criticism. the composition and compilation of the text. its canonical status.
and its form and content. Reynolds. "Teaching Buddhism .... " p. 10 ..
23 See particularly Todd T. Lewis' article. "Representations of Buddhism in Under-
graduate Teaching: The Centrality of Ritual and Story Narratives." Teaching Buddhism in
the West .... pp. 39-56.
THE BUDDHIST CANON AND THE CANON OF BUDDHIST STUDIES 269
This current trend covers up for another approach, the integration of
practice into teaching. The volume Teaching Buddhism in the West also
contains the article "Moving Beyond the 'ism': A Critique of the Objec-
tive Approach to Teaching Buddhism" by O'Hyun Park, Professor of
Religion at Appalachian State University24. Park criticizes what he calls
objective studies of Buddhism, which "are conditioned by occidental or
provincial patterns of thought and arbitrarily limited methodologies.
It is typical of occidentals as well as of many contemporary Buddhists
to wish to teach Buddhism by means of scientific understandings of
Buddhist ideas. These objective studies of Buddhism fail to transmit the
living essence of Buddhism, and in consequence, those whose approach
is purely of this sort may conclude that Buddhism at its best is merely
a form of psychology and has little to do with religious life. "25 Park sug-
gests a different method of teaching Buddhism. The teacher, to begin
with, must be a seeker him- or herself, and the student must be willing
to find a spiritual companion. Then meditation must be included in teach-
ing. Park states: "One must first be still in order to teach and learn
Buddhism. In no other way can its essence truly be known or shown. Seen
from this point of view, teaching and learning Buddhism, if it is not
filtered by meditation, is not worthy of attention. "26 The goal of his
method is "to introduce students to the Buddha's world and to help them
be engaged in the process of moving in that direction themselves. In the
process, the spirit of Buddhism may rub off on them. I personally do not
know what in the process of teaching Buddhism has rubbed off on me,
but I can only hope that whatever it is can be passed on to my students. "27
When using this way of teaching, Park gets rrrixed responses. He admits:
"At times I am informed that my lectures suggest a dogmatic affIrmation."
His reaction to this critique reads: "This is only because I have tem-
porarily been carried away by my deeper bias in the area. However, that
this discussion of non-duality may lead students to re-examine their own
24 O'Hyun Park, "Moving Beyond the 'ism': A Critique of the Objective Approach to
Teaching Buddhism," Teaching Buddhism in the West ... , pp. 57-68.
25 Park, "Moving Beyond the 'ism' ... ," p. 67.
26 Park, "Moving Beyond the 'ism' ... ," p. 59 (italics in the original).
27 Park, "Moving Beyond the 'ism' ... ," p. 68.
270 OLIVER FREIBERGER
approach to their lives and to enlarge their world is for me a sufficient
justification for teaching it. A fair number of students have been very
appreciative. "28 .
Given this attitude, it is particularly interesting to learn about the texts
Park uses for teaching. He writes: "For a text as a proverbial fmger point-
ing to the moon, I have chosen myown translation of a sixteenth-century
Buddhist text written by Xishan, a Korean Zen master. ( ... ) Xishan made
a substantial effort not only to grasp the essence of Buddhism, but also
in most cases to make it relevant to the breadth of human existential
awareness. In my judgement, this text warrants use as an alternative to
most current texts that are based upon a widespread unawareness of the
central thrust of Buddhist religiosity, a deficiency which may be related
to long engrained patterns of dualistic thought. "29 Park selects this text
because in his view, it is suitable for grasping "the essence" of Bud-
dhism. He rejects the canonical texts not because of their normativity
and their distance from current Buddhist practice, nor because they fail
to represent religious diversity, but because for him, they are less suitable
for revealing Buddhism's "central thrust."
It is obvious that this is not an academic or historical, but a religious
criterion. Park's teaching method is thus a form not of academic educa-
tion but of religious instruction. This example points to the important fact
that many Western scholars of Buddhism are committed Buddhists them-
selves, so-called 'scholar-practitioners.'3o Although this is a well-known
fact, there is still too little reflection about its implications for research
and for teaching. In recent years, some scholars came up with ideas for a
new sub-discipline of Buddhist Studies, called "Buddhist Theology." This
discipline, modeled after modem academic Christian Theology, would be
a home for Buddhist scholars who stand normatively in their tradition
and who, by using Western scholarly methods, critically reflect upon this
28 Park, "Moving Beyond the 'ism' ... ," p. 62.
29 Park, "Moving Beyond the 'ism' ... ," p. 63.
30 Cf. Charles S. Prebish, "Buddhist Studies in the Academy: History and Analysis",
Teaching Buddhism in the West ... , pp. 17-36, here: pp. 21-27; Cabezon, "Buddhist Stud-
ies as a Discipline ... ;" Gomez, "Unspoken Paradigms ... ;" Malcolm David Eckel, "The
Ghost at the Table: On the Study of Buddhism and the Study of Religion," JAAR 62 (1994),
pp.1085-1110.
THE BUDDHIST CANON AND THE CANON OF BUDDHIST STUDIES 271
tradition
31
• This interesting development can be a topic for research in itself,
including the question whether it will be possible to retain the distinction
between the "theological" reflections of this new discipline and the. empir-
ical approach of Buddhist Studies as part of the academic study of reli-
gion. Victor Hori supposes that we may encounter a separation into two
disciplines, the "theological" and the academic study of Buddhism
32
• This
would equal the separation of Christian theology and the academic study
of religion (Religionswissenschaft). As the discussions are structurally sim-
ilar (including well-known arguments, for example against alleged neutral
scholarship), this new development may benefit from the long-lasting and
ongoing debate between Christian theology and the study of religion
33

O'Hyun Park's approach, however, can hardly be considered an exam-
ple of Buddhist Theology in the sense of critical, academic reflection.
He uses postmodem and postorientalist arguments as a justification for
propagating what he considers the "essence" of Buddhism, which is based
upon his own translation of a sixteenth-century Zen text from Korea.
2. Reconsidering the Value of Canonical Texts
Given this trend of dissociating from the idea of a canon, what is the
future perspective of Buddhist Studies? Should researchers abandon the
31 Cf. the essays in Buddhist Theology: Critical Reflections by Contemporary Buddhist
Scholars, ed. Roger R. Jackson, John J. Malcransky (Richmond: Curzon, 2000), particu-
larly the contributions by John Malcransky, "Contemporary Academic Buddhist Theology:
Its Emergence and Rationale," pp. 14-21, and Jose Ignacio Cabezon, "Buddhist Theology
in the Academy," pp, 25-52.
32 Victor Sogen Hori, "Introduction," Teaching Buddhism in the West ... , pp. ix-xxv. Mal-
colm David Eckel views such a distinction critically; cf. his "The Ghost at the Table ... "
33 I am aware of the fact that differences between academic disciplines are discemable
in normative and programmatic concepts rather than in actual practice. While the individ-
ual scholar could have more in common with one from another 'discipline' than with many
of her or his own profession, it is programmatic concepts prescribing how scholars of a
discipline should work that construct this discipline. Such concepts describe the boundaries
of the subject matter, the theory, and the methods; within one discipline, they tend to be
controversial and to compete with other concepts. Nevertheless, the continuous debate on
a discipline's identity is necessary for self-reflection; inter-, cross-, or transdisciplinarity
is possible only if there are boundaries one can cross. Cf. my "1st Wertung Theologie?
Beobachtungen zur Unterscheidung von Religionswissenschaft und Theologie," Die Iden-
titiit der Religionswissenschaft: Beitrage zum Verstandnis einer unbekannten Disziplin, ed.
Gebhard Lobi' (Frankfurt/M. et al.: Peter Lang, 2000), pp. 97-121.
272
OLIVER FREIBERGER
primary canon of ancient, normative, and standardized texts, and focus
on the local/present productions .of meaning instead? Should instructors
abandon the secondary canon, and focus on the contemporary practice or
the 'central thrust' of Buddhism? Or is working with canonical texts and
using a secondary canon for teaching still justifiable? I think it is. In the
following sections of the paper, I intend to illustrate what I consider the
significance of canonical texts for research and teaching. I start off by
presenting one example: the image of the laity in early Buddhism.
According to the accounts given in 'classical' anthologies and in
most textbooks, early Buddhist laymen and laywomen can be described
as follows. Together with Buddhist monks and nuns (bhikkhus and
bhikkhunfs), male and female laypeople (upiisakas and upiisikiis) form
the fourfold Buddhist community. They provide the former with clothes,
food, lodging, and medicine, and they lead a moral life according to the
paficasflii, that is, they refrain from harming living creatures, from steal-
ing, from sexual misconduct, from false speech, and from consuming
intoxicants. Unlike members of the sangha, lay people are per se inca-
pable of higher spiritual accomplishment. Therefore, they do not strive
for liberation from the cycle of rebirth and will not attain this state.
Their (inferior) goal is rebirth in a heavenly world, and thus they do not
engage in meditative practices but focus on morality and generosity. Par-
ticularly by donating gifts to the "unsurpassable field of merit," that is
the Buddhist sangha, they can accumulate merit that will cause a better
rebirth.
As mentioned before, this roughly sketched image of the laity is preva-
lent in 'classical' textbooks and anthologies. Denying the fact that this
image is common also in many sections of the Pali canon would be absurd.
But claiming that it is the only view in the texts would be equally
incorrect. When examining not only the 'secondary canon' preserved in
modem anthologies but the entire doctrinal section of the Pali canon, the
Suttapitaka, one discovers a number of passages in which the authors
create an image of laypeople that differs immensely from the one sketched
above.
34
Here, the laity appears as a group not only serving the sangha
34 For the following, see my Der Orden in der Lehre: Zur religiOsen Deutung des
SaJigha imfriihen Buddhismus (Wiesbaden: Harrassowitz, 2000), pp. 140-212; an English
THE BUDDHIST CANON AND THE CANON OF BUDDHIST STUDIES 273
but also and controlling the behavior of monks and nuns. Those
accounts portray laypeople as self-confident persons who have the abil-
ity to assess the ethical and 'spiritual' status of the recipient of thei,r gifts.
They do not trust the promise that the best recipient is, by way of ordi-
nation, the Buddhist monk or nun; they reject the concept of the sangha
as a field of merit that is, by definition, unsurpassable. Instead they indi-
vidually select worthy recipients who lead a moral life and who are 'spir-
itually' advanced, because they believe that for gaining religious merit,
the individual 'spiritual' status of the recipient weighs more than the per-
son's status of being an ordained member of the sangha. How to accu-
mulate merit is a serious issue for them, because it may cause rebirth in
a heavenly world. In fact, a close look reveals that in the canonical texts,
rebirth in heaven appears as a Buddhist soteriological goal independent
from nibbiina (Skt. nirviilJa). The two goals rarely appear in the same
context; only a few theological passages link them and declare nibbiina
the superior one. Reportedly, members of the sangha strive for rebirth in
heaven, to0
35
• On the other hand, there are many accounts of laypersons
receiving instructions into the most complex issues of Buddhist doctrine
and attaining certain trance states. Some laypersons, the texts state, have
even gained liberating insight and nibbiina.
Considering these accounts, we must put the clear division between
members of the sangha and laypeople into perspective; members of both
groups strive for - and attain - both goals, rebirth in heaven and nib-
biina. The clear division of the groups appears as an idea belonging to
an institutionalistic concept of the Buddhist sangha. A close view demon-
strates that a different, rather individualistic, concept is just as common
in the canonical texts. Due to the specific scope of earlier generations of
summary of major arguments in Oliver Freiberger, "Profiling tbe Sangha: Institutional
and Non-Institutional Tendencies in Early Buddhist Teachings," Marburg Journal of Reli-
gion 5 (2000) (online: http://www.uni-marburg.de/religionswissenschaft/journallmjr/
freiberger.html). Cf. also Jeffrey Samuels, "Views of Householders and Lay Disciples in
the Sutta Pitaka: A Reconsideration of tbe Lay/Monastic Opposition," Religion 29 (1999),
pp.231-24l.
35 Cf. for this issue Oliver Freiberger, "Salvation fortbe Laity? Soteriological Concepts
in Early and Modern Theravada Buddhism", Stvdia Asiatica 2 (2001), pp. 29-38; id., "'Ein
Vinaya fur Hausbewohner'? Norm und Praxis der Laienanhanger im friihen Buddhismus,"
forthcoming.
274 OLIVER FREIBERGER
scholars, Buddhist textbooks - and scholarship at large - rarely take
notice of this latter concept with regard to the laity.
For the considerations about the canon in Buddhist we can
draw three major conclusions from this example. The first is that the
Pali canon is not homogeneous. A canon's general feature of being author-
itative and normative does not presuppose homogeneity of its contents.
On the contrary, it seems that oftentimes, heterogeneity of the canon
contributes much to the success of a religion; the more views and practices
can be legitimized by passages from the canon, the more worldviews
are represented, and the more people can feel at home in that religious
tradition. It becomes apparent that the secondary canon Buddhist schol-
ars have created is not representative. Already in the early canonical Pali
texts, we discover a broad spectrum of attitudes, beliefs, and practices.
Thus the reference to diversity in Buddhism, a point Frank Reynolds
emphasizes when focusing on contemporary Buddhist contexts, can hardly
be an argument for dismissing the canonical texts.
The second conclusion we can draw is that although canonical texts
are generally normative, they do not only contain theoretical reflections
of elite, ivory-tower theologians but also religious practice. Certainly,
trying to entirely reconstruct social reality in Ancient India would be
hopeless, but to a certain extent, detecting religious practices in nor-
mative texts is possible
36
• Although far from what ethnographic field-
work could achieve, such fmdings show that 'practice' appears frequently
in the canonical texts. Therefore, the distinction between norm and prac-
tice does not necessarily correspond to the distinction between historical-
philological method and socio-ethnographic method, let alone to the dis-
tinction between past and present. Actually, the often-felt rift between
36 One issue of our example was the question whether the laypeople trust in the merit-
promising institution of the sangha and thus act accordingly, or whether they select 'worthy'
recipients and, furthermore, strive for their personal spiritual development and liberation.
Examining the respective passages more closely, we could detect a number of concrete prac-
tices of Buddhist laypeople. For this issue, cf. Oliver Freiberger, "'Ein Vinaya fiir Haus-
bewohner'? .. " Jan Nattier develops methods for extracting historical data from a normative
Buddhist source in her recent book, A Few Good Men: The Bodhisattva Path according
to The Inquiry of Ugra (Ugraparipfcchii) (Honolulu: University of Hawai'i Press, 2003),
pp. 63-69; see also Christoph Kleine, "Der Kampf der Normen und die Suche nach dem
Referenten," forthcoming.
THE BUDDHIST CANON AND THE CANON OF BUDDIDST STUDIES 275
the canon and practice seems to be due largely to our own one-dimen-
sional construction of 'canonical Buddhism. '37 Thus, a focus on practice
in Buddhist Studies need not lead to the abandonment of canonical texts.
A third conclusion we can draw from the example is that although West-
ern scholars have been examining the canon for more than a century, there
is still a lot to discover. We must not believe that the great scholars of our
field have said everything there is to say about the Pili canon
38
• Rather, with
their 'protestant' view on Buddhism, some played down the rather 'catholic'
practice of accumulating merit
39
• Re-reading the same old texts can lead us
to new and surprising insights, which broaden our horizons and, at the
same time, highlight the specific scope of earlier generations of scholars
40

The example shows that canonical texts are heterogeneous and diverse,
that they contain both norms and practices, and that re-reading them helps
us understand our own inherited presuppositions. With this conclusion,
I do not intend to revive the outdated view that the Pilli canon is a source
sufficient for all interests and questions of modern Buddhist Studies. But
it is apparent that there is more to gain from the canon than some Bud-
dhist scholars, who focus on contemporary practice, might expect4
1
.
37 Cf. Martin Southwold, Buddhism in Life: The Anthropological Study of Religion and
the Sinhalese Practice of Buddhism (Manchester: Manchester University Press, 1983),
esp. pp. 202-212.
38 Certainly, this point, just as other arguments in the present paper, refers not only to
canonical texts but to historical sources in general; a discussion of the general value of the his-
torical approach, however, would go beyond the scope of a paper that focuses on the canon.
39 Cf. Gregory Schopen, "Archeology and Protestant Presuppositions in the Study of
Indian Buddhism," History of Religions 31 (1991), pp. 1-23; see also Oliver Freiberger,
"Werke, GuteI: Religionsgeschichtlich," Theologische Realenzyklopiidie, vol. 35 (Berlin/
New York: de Gruyter, 2003), pp. 623-625.
40 Gregory Schopen, who is otherwise well-known for challenging the traditional view
of Buddhologists by referring to other sources, such as inscriptions and archaeological
accounts, has also convincingly shown how we can come to new conclusions by reading
the Pilii canonical texts with a fresh and critical question. Cf. Gregory Schopen, "Monks
and the Relic Cult in the Mahiiparinibbiinasutta: An Old Misunderstanding in Regard to
Monastic Buddhism", From Beijing to Benares: Essays on Buddhism and Chinese Reli-
gion in Honour of Prof Jan Yiin-Hua, ed. Koichi Shinohara and Gregory Schopen (Oakville:
Mosaic Press, 1991), pp. 187-201.
41 Cf. also the considerations in Jonathan S. Walters, "Suttas as History: Four Approaches
to the Sermon on the Noble Quest (Ariyapariyesanasutta)", History of Religions 38 (1998),
pp. 247-284.
276 OLIVER FREIBERGER
3. Implications for Research and Teaching
3.1. The Fundamental Equality of Sources for Research
The discussion about 'Orientalism' has helped Buddhist scholars
develop a greater sensitivity of the fact that some ancestors in the field
had certain presuppositions and motives that were determined by colonial
interests, by the 'protestant' view on Buddhism, by their personal religious
commitments and cultural biases, and so forth. Donald Lopez, Charles
Hallisey, Luis Gomez, Gregory Schopen, Janice Leoshko, and others have
provided substantial analyses of this issue
42
• Some scholars of earlier gen-
erations believed that the only appropriate way of examining Buddhism
is to turn to its most ancient texts while neglecting later, alleged degen-
erate developments. Modern scholarship rightly opposes this implicit
canonization. Presumably, most oftoday's scholars would agree that there
should be no restriction whatsoever as to what texts or religious expres-
sions to select for research - so long as one is able to explain why the
respective source lies within the scope of Buddhist Studies. If we thus
agree that as a matter of principle, all sources have, as religious expres-
sions, the same value for research, then a fixed secondary canon must
not exist. What follows is that a canonical text, as one particular type of
religious expression, has - on principle - no lesser value for research
than contemporary ritual practice has. As a modern ritual handbook in a
local context has its particular significance and meaning within a religious
tradition, ancient canonical texts have theirs, too.
In the "alternative historical paradigm" Charles Hallisey envisages,
"Buddhist scholars are encouraged to "expect meaning to be produced in
local circumstances rather than in the origins of the tradition" (myempha-
sis )43. In her response, Jan Nattier rightly remarks that "meaning is also
42 Cf., for example, the volume Curators of the Buddha: The Study of Buddhism under
Colonialism, ed. Donald S. Lopez, Jf. (Chicago: University of Chicago Press, 1995), which
includes Hallisey's before-mentioned article, and the review essay by Jan Nattier, "Buddhist
Studies in the Post-Colonial Age," JAAR 65 (1997), pp. 469-485; Schopen, "Archeology
and Protestant Presuppositions ... ;" Gomez, "Unspoken Paradigms ... ;" Janice Leoshko,
Sacred Traces: British Explorations of Buddhism in South Asia (Burlington: Ashgate,
2003); cf. also Eckel, "The Ghost at the Table ... "
43 Hallisey, "Roads Taken ... ," pp. 50-53.
THE BUDDIDST CANON AND THE CANON OF BUDDIDST STUDIES 277
pro'duced at the' ?rigins of the tradition (which, of course, had its own
local circumstances), as well as at every other point along the historical
continuum. '>44 Hallisey's "alternative paradigm" is not identical with but
corresponds to an attitude fairly popular in current studies of culture: the
view that local, vernacular, non-canonical, sub-altern, and/or contempo-
rary religious expressions are, as topics of research, of somewhat higher
value than canonical texts. It is beyond doubt that these issues had been
neglected in the past, and that studying them thoroughly is imperative.
But there is a risk of falling into another trap: assuming that now the
single appropriate way of examining Buddhism is the analysis of
contemporary practice in local contexts. It seems crucial not to create a
hierarchy in values that entails an exclusion of canonical texts from
research. An implicit canonization of that kind would resemble the way
earlier generations excluded non-canonical religious expressions
45

Beside the fundamental insight that a scholarly canonization which
excludes the Buddhist canon is methodologically unacceptable, it seems
that the Buddhist canon remains, as I hope to have shown by the example
of the early Buddhist laity, a rich source for the study of religion
46
• Today,
Buddhist Studies has not only expanded its scope in terms of topics for
research, it also approaches its sources (including canonical texts) with
questions different from those of the past. Such questions, and nothing else,
should determine the criteria for selecting appropriate sources and methods
47

3.2. The Significance of Canonization for Teaching Buddhism
Is this fundamental freedom to select sources equally applicable in the
realm of teaching? To begin with, it seems helpful to view introductory
44 Nattier, "Buddhist Studies in the Post-Colonial Age," p, 472.
45 Interestingly, the reservations about scholarly canonization appear to be much less
pronounced when it comes to modem and contemporary texts. Cf. a recent collection of
writings of modem Buddhists, the title and subtitle of which are telling: A Modem Bud-
dhist Bible: Essential Readings from East and West, ed. Donald S. Lopez (Boston: Bea-
con Press, 2002). A "bible" is a canon par excellence, and preserving the "essence" is the
most central objective of canonization.
46 Besides, canon and exegesis are interesting topics also for comparative analysis; cf.
Smith, "Sacred Persistence ... "
47 Cf. Nattier, "Buddhist Studies in the Post-Colonial Age," pp. 483f.
278 OLIVER FREffiERGER
and advanced courses separately. J'he advanced course deals with specific
texts and contexts the inst;ructor is familiar with; in these courses, students
begin to specialize in certain areas of Buddhism. For the in4'oductory
course, I will consider two concepts: the new one suggested by Frank
Reynolds, which deals exclusively with contemporary practice in Asia
and North America; and the conventional survey course with its outline
of Buddhist history.
The strongest point of Reynolds' concept is its significance for liberal
education: undergraduate students, especially those who take only one
course in Buddhism, will get an impression of Buddhism as it is practiced
in the world they live in. I agree that integrating contemporary practice
in the syllabus is of utmost importance. But narrowing the course con-
tents down to current religious practice does not seem to be useful. Luis
Gomez has convincingly argued that in Western education, the focus on
practice in Buddhism is not accompanied by a parallel focus on practice
in Western culture
48
• He states that "an exaggerated inflation of the 'field'
approach to Buddhism that excludes the textual tradition and the canons
that guided that tradition may work in support of the exoticization of
Buddhism, reinforce its alterity, and reinforce the perception among our
students and the public at large that Buddhism is only a curiosjty, and
certainly not comparable to the well ordered and well-demonstrated
products of our own culture. "49 He continues by saying that "the 'method-
ological' exclusion of the textual tradition leads to ( ... ) the questionable
assumption that textual traditions and textual elites are entities separate
from the living traditions and the non-elite groups with which they obvi- .
. ously interact. "50
Moreover, the spread of Reynolds' concept would affect the conformity
and comparability of undergraduate education in Buddhism. The texts he
selects for his introductory course deal with specific practices at certain
48 "The presentation of Buddhism in the classroom as something occurring only in a
practice without canonical benchmarks may be more corrosive than one can perceive on
first blush - after all, this degree of secularization and devaluation of the book is not
accompanied by a parallel secularization and devaluation of the Great Books of our own
culture." Gomez, "Unspoken Paradigms ... ," p. 205.
49 GOmez, "Unspoken Paradigms ... ," p. 205.
so GOmez, "Unspoken Paradigms ... ," p. 206.
THE BUDDHIST CANON AND THE CANON OF BUDDHIST STUDIES 279
places; for example, image consecration and village rituals in Thailand,
the ordination of novices in Korea, or healing rituals in Sri Lanka
51
. Obvi-
ously, Reynolds does not intend to 'canonize' these specific contexts of
practice. In consequence, every instructor of Buddhism will make his or
her own selection of contemporary contexts for their classes. Students of
different universities will gain close insights into those respective contexts
but will lack a common, basic knowledge of Buddhism. Over time, basic
agreements will begin to crumble - a process that reinforces fragmen-
tation and an anything-goes attitude that makes it increasingly difficult to
assess and judge each other's work.
One, already tangible, product of such an anything-goes attitude is
O'Hyun Park's approach. Blaming Western "occidental" scholars for their
"dualistic" view, he uses post-orientalist arguments to justify his teach-
ing method and the selection of texts that allegedly reveal the 'essence'
of Buddhism best. Defining the 'essence of Buddhism' is, however, a reli-
gious act; in Buddhist history, Buddhists have done it in manifold ways.
Due to this historical perspective; academic scholarship cannot determine
the 'essence' of a religion. Thus, undisturbed by the academic commu-
nity, Park gives religious instruction in the cloak of academic teaching,
and his students may get the impression that Xishan's sixteenth century
Zen interpretation is all they should know about Buddhism. The fact that
the editors accepted Park's paper for a volume on Teaching Buddhism in
the West shows that his approach is considered by some a legitimate option
for teaching. Ironically, the current trend of abandoning the idea of a
canon, of emphasizing religious practice, and of stressing the "variety of the
real worlds of real Buddhists" (Reynolds), gives free rein to approaches
that, for their part, reject diversity and reinforce particularism.
Therefore, using a secondary canon in teaching still makes sense. As
it is unlikely that we might agree upon a binding selection of contempo-
rary contexts, and considering the objections raised above, we may be
inclined to return to our conventional, historical survey course
52
• It has
51 Reynolds refers to the (very useful) text collection The Life of Buddhism, ed. Frank
E. Reynolds, Jason A. Carbine (Berkeley et al.: University of California Press, 2000).
52 It is reasonable to follow the historical developments in their proper historical course,
but with caution. G6mez remarks correctly: "The challenge of the future, however, will
be to find a way to retain the obvious pedagogical advantages of a chronological matrix
280
OLIVER FREmERGER
the advantage of providing students with a common basic knowledge of the
beginnings and the historical development of Buddhism, and it also serves
as a basis for further stUdies. This general survey seems to be crucial
for liberal education, for otherwise students would face the complexity of
Buddhism without any point of reference. But as Reynolds rightly
complains, the conventional survey course too has serious weaknesses.
With such a course, one could be tempted to neglect contemporary prac-
tice, to perpetuate a 19
th
century scholarly framework, and to cement a
secondary canon of texts that has proved to be partial and biased. If we
intend to maintain the general concept of the historical survey course, it
needs to be modified. '
Modifications should address what I call the horizontal and the verti-
cal complexities of Buddhism. Making students aware of the horizontal
complexity means to demonstrate, by way of example, that at every point
in history, 'Buddhism' is a complex phenomenon. For instance, to illus-
trate the complexity of practices and beliefs of early Buddhist laypeople,
an instructor could confront the 'classical' readings with other passages
from the PaIi canon that undermine the alleged consistent image of the
laity. Similarly, in other areas the secondary canon can be confronted
with equally undermining data: with passages from non-canonical texts,
with inscriptional records, with artistic accounts, and the like.
Demonstrating the vertical complexity means to make students aware
of the complex nature of historical developments. To begin with, they
must realize that the many forms of contemporary Buddhist practice
represent merely a small portion of the multifarious ways in which Bud-
d1psm has taken shape throughout its history. Since the majority of those
religious expressions is accessible through textual sources
53
, discussing
while we replace the implicit universal linear narrative with a narrative that is neither cul-
minationistic nor atomistic." G6mez, "Unspoken Paradigms ... ," p, 203.
53 In Jan Nattier's words: "The intense and ongoing use of written sources thus will
always be a central part of Buddhist Studies, stemming from the simple fact that most of
the Buddhists who have ever lived are no longer with us today, If we wish to hear their
voices, we must do so through the surviving texts - including, as Gregory Schopen has
so eloquently argued, not just scriptural texts but also inscriptions, archaeological remains,
and artistic data." Nattier, "Buddhist Studies in the Post-Colonial Age," p. 483. For a
defense of the historical-philological method in Buddhist Studies, cf. also Tom J.F. Tille-
mans, "Remarks on Philology," JIABS 18 (1995), pp. 269-277.
THE BUDDIDST CANON AND THE CANON OF BUDDIDST STUDIES 281
the 'conditional nature of texts is crucial. Students must understand that
canonical texts, fbr example, are not a given divine manifestation but a
product of complex historical controversies, which also include power
relations
54
. .
For demonstrating vertical complexity, it is also important to examine
110
W
history has shaped the present. By way of example, I again draw on
the Buddhist laity. Modem Western Buddhism is largely a lay movement,
but Western converts are not very interested in accumulating merit and
rebirth in heaven; they prefer meditation and insight
55
• This preference
is determined by, among other factors, the scope of earlier generations of
scholars and their particular way of presenting Buddhism to Western read-
ers; had they focused on rebirth in heaven, or had they merely presented
the two goals side by side, Buddhism would probably have gained less
attention among seekers in the Wesp6. The observation that the canoni-
cal texts contain various soteriological goals thus reveals the biased per-
spective of our scholarly ancestors; and at the same time, it helps us
understand how that biased perspective has again shaped contemporary
Western Buddhism. On the other hand, realizing that even the ancient
texts advocate the goal of rebirth in heaven may prevent students from
subscribing to the arrogant view that the accumulation of merit, manifested
in daily practices among contemporary Buddhists in Asia and among
immigrant groups in the West, represented a degenerate and diluted
version of Buddhism
57
• If we integrate, which we should, a glance at
contemporary practice in our modified survey course, being aware of the
54 For this, general reflections upon the term 'canon' and comparisons with processes
of canonization in other religions can be useful; cf. Smith, "Sacred Persistence ... " Another
aspect worth mentioning are the circumstances under which canonical texts have first been
made accessible to the Western world and the ways early scholars dealt with them.
55 Cf., for example, the interview with Jack Kornfield, a well-known American teacher
of Theraviida meditation, in: Jack Kornfield, Living Dharma: Teachings of Twelve Bud-
dhist Masters (Boston/London: Shambala, 1996), pp. 287-301. For other references, see
my "The Meeting of Traditions: Inter-Buddhist and Inter-Religious Relations in the West,"
Journal of Global Buddhism 2 (2001), pp. 59-71 (online: http://www.globalbuddhism.org).
56 This Western interest in Buddhism has, of course, its own historical background,
namely a critical stance towards the Christian churches and their alleged authoritarian and
dogmatic claims, and the longing for a rational and individualistic religion that does not
require blind faith, an attitude connected with both Romanticism and Enlightenment.
57 Cf. Southwold, Buddhism in Life ... , pp. 202-212.
282
OLIVER FREIBERGER
vertical complexity, of the complex historical circIffitstances that have
shaped today's Buddhism, is imperative.
Conclusion
This paper has addressed four issues: the role of the Buddhist canon
and the canonization taking place within Buddhist Studies, both for
research and in teaching. Some scholars tend to discourage from study-
ing the Buddhist canon and aim at eliminating its role in teaching. I hope
to have shown that studying the Buddhist canon remains worthwhile in
itself and even more, helps understand better our own presuppositions in
the field. In addition, I have argued that although integrating contempo-
rary practice in the introductory course is necessary, the historical survey
course, which includes readings from the canonical texts, remains the
best option for providing a basis for further studies as well as for liberal
education in general.
Earlier generations of scholars were interested in the origins of Bud-
dhism and thus focused on the earliest, that is the canonical, texts. More-
over, in textbooks and anthologies they presented only a certain selection
of text passages and religious concepts, while excluding others. They
selected texts according to their interests which were determined by their
own culturally and personally, often 'protestant' backgrounds. In such a
way, they again 'canmiized' the sources which for them represented orig-
inal or authentic Buddhism. For overcoming this - still influential....,...
canonization, Buddhist scholars have turned to other expressions of Bud-
ghist religiosity, particularly to contemporary religious practice. I have
argued that following this trend, Buddhist Studies runs the risk to further
a new process of canonization which now excludes the Buddhist canon.
The fact that the issues argued for (diversity and practice) are present in
canonical texts too, is largely ignored, in part due to the inherited ('can-
onized') image of the canon.
The exclusion of the Buddhist canon is notably conspicuous in recent
discussions about teaching. The old, canonized sources for teaching
(anthologies and textbooks) that comprise Buddhist Studies' secondary
canon are abandoned, in favor of accounts of contemporary religious
practice. This trend of rejecting canonization for teaching entails an
THE BUDDHIST CANON AND THE CANON OF BUDDHIST STUDIES 283
anything-goes attitude which encourages the abandonment of common
standards and allows particularistic religious instruction into the classroom.
I have argued that in contrast to research, undergraduate education, par-
ticularly on the introductory level, needs a secondary canon. A common
basis of knowledge gives students some certainty in orientation, a first
frame of reference (which they may criticize and deconstruct at a later
point), and the capability of communicating with, and assessing the work
of other students and scholars in Buddhist Studies. But in addition to that,
I suggest that during the course, every now and then the 'classical' read-
ings should be supplemented and confronted with other, undermining
data. Thereby demonstrating the diversity and the historically conditioned
nature of the sources, an introductory course can make students aware of
the horizontal and the vertical complexity of Buddhism without with-
holding from them a common basis of knowledge.
TARANATHA'S "TWENTY-ONE DIFFERENCES WITH
REGARD TO THE PROFOUND MEANING" -
COMPARING THE VIEWS OF THE TWO aZAN STaN
MASTERS DOL PO PA AND SA.K.YA MCHOG LDANl
KLAUS-DIETER MATHES (UNIVERSITY OF HAMBURG)
1. IDstorical Background
. The distinguishing feature of gzan stan Madhyamaka in the Jonangpa
, school, to which Taraniitha (1575-1634) belongs, is the fact that itnorrnally
restricts the validity of the common Madhyamaka assertion "all pheno-
mena are empty of an own-being" to phenomena on the level of appar-
ent truth. The ultimate, which is inseparably endowed with innumerable
Buddha-qualities, is considered to be not "empty of an own-being" (ran
itan) but "empty of other" (gzan stan), namely accidental stains and so
forth
2
• it was the famous Jonangpa Dol po pa Ses rabrgyal mtshan (1292-
1361) who is said to have gained such an insight during a Kiilacakra
retreat
3
• From the Ri chas nes dan rgya mtsha, which is one of the first
works in which Dol po· pa expressed his new gzan stan understanding, it
becomes clear that the latter's full-fledged gzan stan theory requires includ-
ing even an ultimate sambhagakiiya and nirmii1}£lkiiya within an ultimate
realm of truth, which is equated with dharmatii, or the unchangeable per-
fect nature. This, we are told, is in line with extraordinary Mantrayana
4

1 The present article is an enlarged version of a paper read at the Eighth Seminar of
the International Association for Tibetan Studies in Bloomington (USA) in July 1998. Thanks
to a scholarship from the Gennan Research Council (DFG) I have been able to continue
my research on tathtigatagarbha during the last three years and can now rest my original
study of Tliranatha's comparison on a much broader basis. Improvements to my English
by Philip H. Pierce (Nepal Research Centre, Kathmandu) are gratefully acknowledged.
2 Mathes 2000: 195-6.
3 Dol po pa's disciple Lha'i rgyal mtshan (1319-1401) infonns us that his master's real-
ization was counected with the Ktilacakratantra (see Stearns 1995: 829-31).
4 Dol po pa: Ri chos iles don rgya mtsho, 343, 11. 19-21 & 344, 11. 8-9: "As to the
two aspects of the fonn-ktiyas, they are here the commonly known sambhogaka)ia and
nirmtilJaktiya of the apparent [truth]. As to the ultimate sambhogakiiya and nirmtilJaka)ia,
Journal of the International Association of Buddhist Studies
Volume 27 • Number 2 • 2004
286 KLAUS-DIETER MATHES
But as a commentator of non-Tantric texts, such as the Ratnagotravibhaga,
Dol po pa explains that the sambhogakaya and nirmalJakaya are brought
forth by a fortified potential which arises from virtuous deeds being newly
adopted with effort
5
, and it is only in texts such as the Ri ch;s nes don
rgya mtsho that we are informed that the created kayas are merely the ones
pertaining to apparent truth. In view of this hermeneutic strategy6, the
differences between the Ri chos nes don rgya mtsho and the Ratnagotravi-
bhaga commentary appear to be so fundamental that Rookham wonders
if the latter is by Dol po pa at all and not rather by the Third Karmapa
Rail byuil rdo rje (1284-1339)1.
Still, Dol po pa to some extent reads his originally KaJacakra-
based gzan ston into the Siltras and such non-Tantric treatises as the
they are completely [contained] in the dharmatii, perfect [nature] and suchness. [.0.] There-
fore the ultimate sambhogakitya and nirmiiIJakiiya are known by way of the extraordinary
Mantra[yana]." (de la gzugs sku rnam pa gfiis ni kun rdzob kyi Ions spyod rdzogs pa dan
sprul pa'i sku ste thun man du rab tu grags pa'o / / don dam pa'i Ions spyod rdzogs pa dan
sprul pa'i sku ni chos fiid yons grub de bZin fiid la tshan ste / [ ... J des na don dam gyi Ions
spyod rdzogs pa dan sprul pa'i sku ni thun man ma yin pa snags kyi tshulla grags pa'o I).
5 Dol po pa: "Ni ma'i 'od zer", 986, 1. 6-987,1. 3: "For example, in the same way as
the inexhaustible treasure underground is naturally present, not newly brought about by
effort, while the tree with its fruits gradually grows in a garden by bringing about [the nec-
essary conditions] with effort, the Buddha-potential, which has the ability to bring forth
the three kiiyas, should be known to be twofold as well. It is both the natural potential,
[namely] the pure dharmadhiitu (which latter is intimately present as the nature of [one's]
mind throughout beginningless time), and the fortified potential [which is] supreme in terms
of virtues (which are conducive to liberation). [The latter potential] arises from [virtuous
deeds] being newly adopted with effort, [namely by] something being done, such as focus-
ing on [the naturally present potential] and studying." (dper na 'bad rtsol gyis gsar du ma
bsgrubs sin Ions spyod zad mi ses pa dan ldan pa'i gter chen sa'i 'og na ran Min gyis gnas
pa dan 'bad rtsol gyis bsgrubs pas 'bras bu dan bcas pa'i sin ljon sa skyed mos tshal du
rim gyis skye baji Ita ba biin du sku gsum 'byun du run ba'i sans rgyas kyi rigs de yan
rnam pa gfiis su ses par bya ste / thog ma med pa 'i dus nas sems kyi ran biin du fie bar
gnas pa'i chos kyi dbyins mam par dag pa ran biin gyi rigs dan / de la dmigs te thos pa
la sogs pa byas pas 'bad rtsol gyis gsar du yag dag par blans pa las byun ba'i dge ba thar
pa'i cha dan mthun pas mchog tu gyur pa rgyas 'gyur gyi rigs fiid do).
6 The possibility that Dol po pa wrote his Ratnagotravibhiiga commentary before
achieving his insight into gian stan can be ruled out, for he also refers to ultimate qualities
in his "Ni ma'i 'od zer" (911, 11. 3-4).
7 She reinforces her view with the assertion that the text was copied by KOIi sprul BIo
gros mtha' yas nearly verbatim (Hookham 1991: 173-4). But such an assumption is unlikely,
since the text is signed by "One Endowed with the Four Reliances" (rton pa bii ldan), which
was the most common pseudonym used by Dol po pa in his works (Steams 1999:201).
TARANATHA'S "TWENTY-ONE DIFFERENCES" 287
Tathagatagarbhasiltras and the Maitreya works. The hermeneutic princi-
pIes according to which he interprets the Buddhist teachings are laid out
in his "bKa' bsdu bii pa" (i.e., his own "fourth council"S), in which the
whole of Buddhist doctrine is "reckoned" by dividing the teaching into
four epochs. Alongside the four epochs of varying quality which make
up a cosmic age, Dol po pa uses a lesser set of four epochs to refer to the
qualitatively different periods of the teaching. He thus allocates philo-
sophical doctrines to epochs (yuga) according to purely dogmatic criteria
9

The teachings transmitted by Siikyamuni and also the Maitreya works, for
example, belong to the Krtayuga of doctrine, while other works, such as
the ones by Arya Vimuktisena and Haribhadra, represent the teachings of
the inferior Tretayuga. The common interpretation of the Y ogacara works
of Maitreya, Asailga and Vasubandhu as mere cittamatra itself reflects for
Dol po pa the historical degeneration of the Dharma. The Maitreya works
are only "Krtayuga" Dharma when they are explained as "Great Madhya-
maka" (dbu ma chen pO)lO.
The theories of tathiigatagarbha ("Buddha-nature") and trisvabhiiva
("three natures", i.e., the imagined, dependent and perfect natures) in the
Maitreya works offer good canonical support for a distinction between ran
ston and gzan ston, and it is thus no surprise that an interpretation which
supports such a distinction is a major concern for the Jonangpas. Dol po
pa takes the ultimate to be absolutely unconditioned, and it is the Ratna-
gotravibhiiga among the Maitreya works which is adduced as the best
support for this stance. Thus Dol po pa comments RGV I.5a in his Ri chos
nes don rgya mtsho in the following way:
Even though [the verse RGV I.5a]: "[Buddhahood] is unconditioned and spon-
taneously present"l!, and other [passages] teach that the ultimate Buddha is not
conditioned, the underlying intention is that he is [also] free from moments
l2
.
8 Virtually the entire Buddhist tradition accepts only three great councils in India held
for the purpose of consolidating the teaching after the Buddha's nirviiJ;za.
9 Kapstein 2000: 115-6.
IO Kapstein 1992:24-5.
1J Cf. ROVV 7, 11. 14-5: asmhsk[tarn anabhogarn aparapratyayoditarn I buddhatvam
jiianakarUl:zyasaktyupetam dvayarthavat II.
11 Dol po pa: Ri chos nes don rgya rntsho, 97,11. 15-7: 'dus rna byas sin lhun gyis
grub Ices pa la sags pas rnthar thug gi sans rgyas 'dus rna byas su gsuns pa yan skad cig .
dan bral ba la dgons pa yin no /.
288 KLAUS-DIETER MATHES
For Dol po pa's disciple Sa bzail Mati PaI! chen (1294-1376) the ulti-
mate or Buddhahood is thus permanent in the sense of being beyond the
three times (i.e., past, present and future), as becomes clear in Mati PaI!
chen's Ratnagotravibhaga commentary on RGV I.6cd:
Buddhahood is unconditioned, since in the beginning, middle and end it has
the nature of being free from conditioned phenonema which arise, abide
and pass out of existence, as has been said in the [MahiiparilnirviilJasutra:
"A phenomenon that abides in permanence does not belong to the three times.
Likewise, the Tathagata does not belong to the three times, and is therefore
permanent. "13
It is obvious that in this case the perfect nature of the Y ogacar-a must
be restricted, as Tar-anatha has done in his final summary of the trisva-
bhiiva theoryl4, to its unchangeable aspect (nirvikara), since in an
absolutely permanent and atemporal Buddhahood or Buddha-nature (both
are ontologically the same for the Jonangpas) there is no room for an
unmistaken (aviparyiisa) wisdom cultivated on the path, namely -
according to MA V III. I led - the perfect in terms of being unmistaken.
Already at the time of Dol po pa, the Third Karmapa Rail byun rdo
rje (1284-1339) was propounding a different "gian ston position", or
rather a position that was eventually called gian ston by a few later Kagyu-
pastS such as Karma Phrin las pa (1456-1539)16. Rail byun rdo rje bases
his distinction between the true nature of mind or Buddha-nature and
that from which it is free on Mahayanasamgraha 1.45-9, in which an
impure aZayavijfiiina is strictly distinguished from a "transmudane
rnind,,17. In this context Rail byun rdo rje stresses the need to distinguish
13 Sa bzail Mati pal,! chen: "Theg pa chen po rgyud bla ma'i bstan bcos kyi mam par
bsad pa iles don rab gsal snail ba", 55, 11. 2-3: sans rgyas fiid thog ma dan dbus dan mtha'
mar 'dus byas kyi chos skye ba dan gnas pa dan 'jig pa rnams med pa'i ran Min can yin
pa'i phyir 'dus rna byas pa ste I my a nan las 'das pa'i mdo las I rtag tu gnas pa'i chos ni
dus gsum la (text: las) ma gtogs te I de biin gsegs pa yan de dan 'dra bar dus gsum la ma
gtogs pa de bas na rtag pa' 0 ies gsuns pa Itar roo
14 See Mathes 2000:219-220.
15 It should be noted that the term gian ston is found nowhere in the works of Rail byuil
rdo rje.
16 Karma 'Phrin las pa: "Dris Ian yid kyi mun sel ies bya ba lcags mo'i dri Ian Mugs
so" 91,11. 1-4.
17 This is clear from Rail byuil fdo rje's autocommentary on the Zab mo nan gi don
(9b4-10bl) and commentary on the Dharmadhtitustotra (l2bl-13b6).
TARANATHA'S "TWENTY-ONE DIFFERENCES" 289
"ground-consciousness" (Skt. iilayavijfiiina, Tib. kun gii rnarn ses) from
"ground" (kun gii) in terms of suchness
18
. Referring to this passage, Koit
sprul Bla gras mtha' yas (1813-1899) proceeds in his commentary on the
Zab rno nan gi don to use the gian ston term kun gii ye ses for the trans-
mundane mind of the Mahiiyiinasamgraha
19
• Koit sprul' s use of the term
kun gii ye ses 20 does not imply, though, that he took Rail byuit rdo rje's
position to be the same as Dol po pa's. It rather suggests that Koit sprul
himself maintains a gian ston whose "basis of emptiness" (ston gii) is
defined in accordance with Rait byuit rdo rje's Zab don ran 'grel, which
in this crucial point follows not the Ratnagotravibhiiga but the Mahii-
yiinasamgraha. Rail byuit rdo rje is a gian ston pa for Koit sprul, but one
who explains that which remains in emptiness in a way different from Dol
po pa. And indeed, in the ninth chapter of his Zab rno nan gi don Rail
byuit rda rje takes the stainless Buddha-nature (which is liberated from
everything else (Tib. gian grol) - i.e., the basis of emptiness) as being
endowed with the two truths
21
. From the autocommentary it is clear,
however, that it is not the normal apparent truth which is included in
Buddha-nature here, but only a pure aspect of the latter, namely the "non-
existence of the stains [or] delusions in the eight consciousnesses"22.
18 Ran bymi. rdo rje: Zab mo nan gi don gsaZ bar byed pa'i 'gre! pa, fol. 8a6-7: "In
this regard, if 'ground' (kun gii) is not mentioned [together with] the word 'consciousness',
'ground' may refer to suchness. Therefore, consciousness is mentioned [together with it]."
Cdi la kun gii zes bya ba rnam par ses pa'i sgra ma smos na de Min fiid la yan kun gii
brjod du run ba 'i phyir rnam par ses pa smos so !.)
19 Kmi sprul Blo gros mtha' yas: Zab mo nan gi don gyi 'greZ pa, 17b4-6.
20 A term thought to be newly coined by Dol po pa.
21 Ran bymi. rdo rje: Zab mo nan gi don, 22b6: "The [Buddha]-element in sentient
beings, the stainless Buddha-nature, is endowed with the two truths." (/ sems can khams
ni sans rgyas kyi / / sfiin po dri med bden gfiis ldan I).
22 Rail bymi. rdo rje: Ran 'grel, 62a7-62b2: "What exists ultimately? It is the mind
beyond every net of thought, the naturally pure element of sentient beings, [and] the Bud-
dha-nature (sans rgyas kyi sfiin po). Because these two exist, they have been expressed by
way of these [terms]. Therefore it is stated: "as for the element of sentient beings, the stain-
less Buddha-nature is endowed with the two truths." In this regard, the Buddha-nature is
simply the non-existence of stains [or] delusion in the above-mentioned eight accumula-
tions [of consciousness]." (don dam par gaJi zig yod na / rtog pa'i drva ba thams cad las
'das pa'i sems ran biin gyis dag pa'i sems can gyi khams sans rgyas kyi sfiin po dag ni
yod pas de'i tshul brjod pas / sems can khams ni sans rgyas kyi / sfiin po dri med bden
giiis ldan ies smos so / / de la sans rgyas kyi sfiin po ni snar smos pa'i tshogs brgyad kyi
'khrul pa dri ma med pa kho na yin mod kyi .. . ).
290 KLAUS-DIETER MATHES
What this latter term exactly refers to is explained a little further down
in the Zab don ran 'grel, where the use of the word "truth" in the term
"apparent truth" is justified on the grounds that one cannot deny mere
appearance as such, even though its interpretation as a perceived object and
perceiving subject is not true:
What has been imagined as the duality of a perceived and a perceiver does
not exist at all, given the pronouncement [in MA V I.3] by the Venerable
Maitreya: "A consciousness arises which has the appearances of objects,
sentient beings, a self and perceptions. It does not have a [corresponding
outer] object, and since [such] an object does not exist, it (i.e., a perceiving
subject) does not exist either. "23 Thus it has been said that no perceived
[objects] or perceiving [subjects] of the imagined [nature] exist at all. Well
then, how can it be presented as a truth? [The answer is:] Even though it
does not exist, [something] appears. This is what is called apparent truth, for
it has the nature (ran gi no bo Rid) of not being deceptive
24

In response to the objection that these mere appearances would then be
the ultimate truth, since the latter is defined as not being deceptive in the
treatises on logic, Rail byuil rdo rje further clarifies his understanding of
the ultimate truth as follows:
These [mere appearances] are presented as the expressible ultimate (paryiiya-
paramartha), while the ultimate truth [here] is that which
25
is related to the
reasoning of dharmata, [namely] the natural emptiness previously men-
tioned during the presentation of the eighteen [types of] great emptiness
26
.
In other words, the Buddha-nature or the pure mind includes "mere
appearances" in the form of the expressible ultimate truth, and it is only
23 MA VBh, 18,11.21-2: arthasattvatmavijfiaptipratibhiisam prajayate Ivijfianam nasti
casyarthas tadabhiivat tad apy asat. My additions in brackets are accordings to Vasuban-
dhu's bhii0'a.
24 Rail bymi. rdo rje: Ran 'grel, 63a3-5: I gzun ba dan 'dzin pa gfiis su kun btags (text:
brtags) pa ni mampa thams cad du med pa dag yin te I 'phags pa byams pa'i tal sna nas
kyanl [MA V 1.3]1 ces kun btags (text: brtags) pa'i gzun ba dan 'dzin pa thams cad mam
pa thams cad du med pa fiid du gsuns so II '0 na bden pa ji ltar Mag ce na I med btin du
yan snan ba tsam de ni kun rdzob kyi bden pa zes bya ste I bslu ba med pa'i ran gi no bo
fiid yin pa 'i phyir ro !.
25 The use of the plural particle dag should be noted here. It indicates that there is
more than one truth related to dharmatayukti.
26 Ibid., 63a5-6: 'di yan mam grans kyi don dam par bzag pa yod mod kyi I chos fiid
kyi rigs pa'i rjes su 'breI pa dag ni stan pa fiid chen po beo brgyad kyi mampar b§ad pa'i
ran Min stan pa fiid snar smos pa de fiid don dam pa'i bden pa yin no!.
TARANATHA'S "TWENTY-ONE DIFFERENCES" 291
the iatter which is taken as apparent truth here. That it is different from
what is ordinarily 'included in apparent truth is clear from Ran byun rdo
rje's Dharmadhiitustotra commentary, where the two aspects (nirvikiira
and aviparyiisa) of the perfect nature in MA V IIUl cd are explained in
the following way:
The two [aspects of the perfect], the unchangeable and unmistaken, are taken
[respectively] as the defIning characteristics of the two truths. Acceptance
by common consent (lokaprasiddha) and by reason (yuktiprasiddha) are
varieties of the apparent truth27.
In other words, the unchangeable perfect is taken as the ultimate, and
the perfect in terms of being unmistaken as a restricted form of apparent
truth, which does not include acceptance by common consent and the like.
It should have become clear by now that Ran byun rdo rje, in contrast
to the Jonangpas, fully accepts the Yogacara theory of trisvabhiiva,
. that is, two aspects of the perfect nature. This entails that the latter pos-
sesses moments, in the Zab rno nan gi don the true nature of mind which
is free from everything else (gzan grol) being consequently equated with
dependent arising
28
• In this context, it is of interest that Kon sprul BIo gros
mtha' yas, who otherwise strictly follows Dol po pa's Ratnagotravibhiiga
commentary, deviates from the latter's gzan ston understanding of the
term "unconditioned" (asarhs!q:ta) in his commentary on RGV 1.6. Refer-
ring to Ron ston Ses bya kun rig's (1367-1449) explanation of four ways
of understanding "unconditioned", Kon sprul states that the dharmakiiya
only shares this quality of being unconditioned to a certain extent, inasmuch
as it does not appear to disciples. If one claimed that it is completely
27 Rail bymi rdo rje: dBu ma chos dbyins bstod pa'i rnam par Mad pa, 7b1-2: bden
pa gfiis kyi rali gi mtshan fiid kyis 'gyur ba med pa dan phyin ci ma log pa gfiis so / 'jig
rten pa dan rigs pa'i grags pa ni / kun rdzob bden pa'i bye brag ste !.
28 Rail byuil rdo rje: Ran 'grel, IOb3-4: "As to the 'beginningless [mind-essence]"
since a beginning and end of time is a [mere] conceptual superimposition, [the cause of
everything] is here [taken as] the true nature (ran gi no bo) of both the stainless [mind]
and the [mind] mingled with stains - it is precisely this dependent origination; and it is
completely liberated (i.e., free) from [all] else. Since there is no other beginning than it,
one speaks of beginningless time." (thog med la ies bya ba ni / dus kyi thog ma dan tha
ma ni rtog pas sgro btags pa yin pas 'dir ni dri ma med pa dan dri ma dan bcas pa'i ran
gi no bo ni rten cin 'brei bar 'byun ba de fiid dan gian las rnam par grol ba ste / de las
thog ma gzan med pa'i phyir thog ma med pa'i dus ies bya ste I),
292
KLAUS-DIETER MATHES
unconditioned, tbis would contradict the fact that it possesses knowledge,
compassion and power.
29
To sum up, whether one wants to call RaiL bymi rdo rje's "free from
other" (gzan las grol ba) "empty of other" (gzan ston) or not, there is an
alternative way of defining how the pure mind or Buddha-nature is free
from or empty of other (i.e., adventitious stains), and some Kagyupas
decided to call tbis gzan ston, too. It should be noted that with an ulti-
mate that still possesses moments a distinction founded on gzan grol
(or gzan ston) can be better brought into line with mahiimudrd teachings
30
,
and tills is exactly what RaiL byun rdo rje did
31
. It is thus no longer so puz-
zling how Situ paJ). chen ehos kyi byun gnas (1699/1700-1774) "blended
the seemingly irreconcilable gzan ston and mahdmudrd positions"32.
Another famous scholar whose gzan ston differs from the Jonang posi-
tion is gSer mdog paJ). chen Sakya mchog ldan (1428-1507). Georges
Dreyfus (1997:29) has observed that Sakya mchog ldan fully endorses a
gzan ston view only in works written after his first meeting with the
Seventh Karmapa (1454-1506) in 1484, and it is thus interesting that Sakya
mchog ldan's gzan ston differs from the Jonang position in a way similar
to Ran byun rdo rje's gzan grol (or gzan ston). Kon sprul Blo gros mtha'
yas notices in bis Encyclopedia of Knowledge (Ses bya kun khyab mdzod)
that Sakya mchog Idan has bis own views on what is exactly empty of
what, or, to use the technical terms, how the negandum (Tib. dgag bya),
the basis of negation (Tib. dgag gzi) and the mode of being empty (Tib. ston
tshul) are defined.
Kon sprul says that gzan ston pas take the perfect nature as the basis
of negation, the imagined and the dependent natures as the negandum, and
the mode of being empty as the absence of these two neganda in the basis
of negation
33
. He adds, however, that Sakya mchog Idan holds a view
29 Kon sprul BIo gros mtha' yas: rGyud bla ma'i b§ad sral, fol. 12bl-5; see also
Tsultrim Gyamtsho & Fuchs 2000: 103-4. .
30 A dharmakaya that possesses moments it not entirely transcendent and can be expe-
rienced as the true nature of thoughts and the like.
31 For further information on this topic I refer to my forthcoming habilitation thesis.
32 Smith 1970:34.
33 In the subchapter on ascertaining the view (7.3.), Kon sprul (Ses bya kun khyab
mdzad, vol. 3, 61, ll. 19-24) defmes the tradition of the gzan stan Madhyamaka as follows:
"The basis of negation is the perfect, the sphere (dhatu), suchness, what is beyond [any]
TARANATHA'S "TWENTY-ONE DiFFERENCES" 293
different from this, and illustrates this by quoting from Sakya mchog ldan's
' ~ Z a b ii spras bral gyi Mad pa":
As to the basis which is empty, it is the dependent, the entire "mind" (§es pa),
which takes on various forms of a perceived object and perceiving subject.
The negandum is the imagined. Given its division into perceived and perceiver,
it [consists of] two. [ ... J34
As to in what sense it is empty, the basis of negation is empty of the negan-
dum by virtue of being "empty of other", not by virtue of being "empty of
self", for the following reason: The negandum, namely the duality of a per-
ceived and a perceiver, is an "other-being" with regard to the basis of nega-
tion, [namely] the mind" (§es pa), which appears as two, [duality] not being
taken as its own-being. What is then the own-being of this mind which appears
as two? It is non-dual wisdom, namely mere awareness and luminosity that
experience knowable objects35.
Kon sprul continues his presentation of gzan ston along this (namely
Sakya mchog ldan's) line of thought, elaborating it on the basis of the
sixteen fonns of emptiness in the Madhyantavibhiiga. Kon sprul's posi-
tion on gzan ston still needs further clarification, but it is at least note-
worthy that while in the sixth chapter of his Ses bya !run khyab mdzod he
quotes nearly the entire dbu ma chen po paragraph of Tiiranatha's gian
mentally fabricated object. The negandum is the two defining characteristics of the imag-
ined and the dependent. The mode of being empty is: 'empty of these two neganda in the
basis of negation' Only the perfect, therefore, is empty of other. Thus says the Y ogacara,
the proponents holding the tradition of gzan ston." (dgag gzi yons grub dbyins de Min fiid
spros pa'i yullas 'das pa / dgag bya kun btags (text: brtags) dan gian dban gi mtshan fiid
gfiis / ston tshul dgag bya de gfiis kyis dgag gzi la ston pas yons grub fiid gzan gyis ston
pa yin ces mal 'byor spyod pa ste gzan ston gi srol 'dzin pa rnams smra'o I).
34 Kon sprul: ibid., vol. 3, 62, 11. 1-3: gan ston pa'i gzi ni gzan dban ste gzun 'dzin
gfiis snan can gyi ses pa mtha' dag go / dgag bya ni kun btags (text: brtags) pa ste gzun
ba dan 'dzin pa'i dbye bas gfiis yin la /. This is a literal quote from Sakya mchog ldan's
work "Zab ii spros bral gyi Mad pa ston fiid bdud rtsi'i lam po che", 114, 11. 3-4.
35 Kon sprul, op. cit., vol. 3, 62, 11. 4-10: ji ltar ston pa'i tshul ni / dgag bya des dgag
gzi de gzan ston gi tshul gyis ston pa yin gyi / ran ston gi tshul gyis ni ma yin te / dgag
bya 'kun btags (text: brtags) kyi' gzun 'dzin gfiis po de ni' dgag gii gfiis snan gi ses pa
gfiis po de la ltos pa'i gzan gyi no bo yin gyi / de'i ran gi no bor mi 'jog pa'i phyir / gfiis
snan gi ses pa de'i ran gi no bo gan ie na / gfiis med kyi ye ses ses
b
bya myon ba rig cin
gsal tsam de fiid do. (= "Zab ii spros bral gyi biad pa", 114,1. 7 -115,1. 2).
, Added by Kon sprul.
b The edition of the collected works of Sakya mchog Idan reads zes, the editors prob-
ably having thought that zes had changed into ses according to an old sandhi rule.
294 KLAUS-DIETER MATHES
ston sfiin po, he skips the last part, where - against the purport of the
Y ogacara works - the trisvabhava theory is brought into line with that
of tathagatagarbha by restricting the perfect nature to its unchangeable
aspect
36
• .
To sum up, from the time of Dol po pa it is possible to trace, parallel
to the Jonang position, another "gian grol" or gian ston which distin-
guishes the basis of negation from the negandum in a different way.
Whereas for the Jonangpas the basis of negation is a perfect nature which
is restricted to its unchangeable aspect and thus transcendent and doctrinally
mainly based on the tathagatagarbha theory, Sakya mchog Idan, Rail bymi
rdo rje and some other Kagyupas adhere to a distinction based on Y ogacara,
that is, mainly the Mahayanasamgraha and the Madhyantavibhiiga.
2. Taranatha's Twenty-One Differences with regard to the Profound
Meaning
37
For a short but brilliant analysis of the positions of Dol po pa and Sakya
mchog ldan we are very much indebted to the Jonang master Taranatha,
who is considered to be a follower and proponent of Dol po pa's doctrine.
In each of the Twenty-one Differences with regard to the Profound M ean-
ing a fictive initial statement of Sakya mchog ldan is followed by a sim-
ilarly fictive reply of Dol po pa, Taranatha being, of course, well aware
of the fact that this is all ahistorical
38
. To be sure, it is not possible to estab-
lish Sakya mchog ldan's or Dol po pa's views on the basis of this short
text alone, but it does sharpen our awareness of the subtle aspects of
gian ston when studying the bulky and often not very systematic works
of these masters. Furthermore, critically evaluating these doctrinal dif-
ferences against the background of pertinent Indian texts in such traditions
as the Madhyamaka, Y ogacara and Tathagatagarbha promises to be a second
interesting task. Both are, however, beyond the scope of this paper. Such
an evaluation will, however, be undertaken with regard to the different
presentations of trisvabhava as an example of how one might proceed.
36 See Kon sprul: ibid., vol. 2, 546-9.
37 Tib. zab don khyad par fier gcig pa, which is the title according to the colophon
(Taranatha: "Zab don fier gcig pa", 795, 1. 5.
38 Taranatha: "Zab don fier gcig pa", 792, 1. 4.
TARANATHA'S ''TWENTY-ONE DIFFERENCES" 295
-Taranatha begins his somewhat delicate task of comparing the two
masters Dol po pa and Sakya mchog ldan in a conciliating manner, by
explaining that both supposedly see what is profound reality and hence
should not have different thoughts about it. It is only in order to accom-
modate the different needs of their disciples that they enunciate variant
views. Evert though the essential gzan ston view and meditation practices
of both masters are the same, there are a lot of minor differences regard-
ing tenets (grub mtha ') that arise when formulating the view on the level
<if apparent truth
39

- -The first four of the twenty-one points address differences in the exe-
gesis of the Madhyamaka and Maitreya texts which are considered to be
20mmentaries on the Buddha's intention underlying the second and third
tUrnings of the "Wheel of the Dharma" (dharmacakra)40. Points 5-8
e ~ b o d y Sakya mchog Idan's and Dol po pa's different understanding of
non-dual wisdom. In points 9-16, their views on the trisvabhiiva theory
lite distinguished. In a related topic, Taranatha also elaborates the differ-
ent understandings of self-awareness (point 11), entities and non-entities,
and conditioned and unconditioned phenonema (all in point 13). Next,
oUr attention is -drawn to different ways of relating the four noble truths
. With the apparent and ultimate (point 17). The last four points deal with
the two masters' views on the Buddha-nature.
2.1. Translation: The Twenty-One Differences
Difference No.1
Sakya mchog Idan
41
: All the views of the Prasailgika- and Svatantrika-
[Madhyamaka] are logically correct [and accurately represent] the Buddha's
39 These remarks should not be taken too seriously, though. 'Ba' ra ba rOyal mtshan
dpal bzail (1310.1391) launches into his "Chos rje mam gfiis kyi dgOIis bsad iii ma'i 'od
zer" (496-8) by stating, in a similar way, that Dol po pa and Bu ston (sic) are both omnis-
cient and must see the same reality, but teach it in various ways with hidden intentions.
40 The Indo-Tibetan exegetical traditions summarize the teachings of the Buddha in three
circles or "[turnings of the] Wheel of the Dharma" (dharmacakra). See Mathes 1996:155.
41 Lit. "The one named Sakya claims that .... " Here and in the following sentences
simply the proponent's name is given in bold letters.
296 KLAUS-DIETER MATHES
intention in the middle turning and the corpus of analytical works
42
. The
explicit teaching of the middle [turning], in addition, [has to] be taken
_ literally, and the corpus of analytical works is not in accordance with the
explicit teaching of the last turning.
43
.
Dol po pa
44
: Even though [the ran stan pas] are proud that these
Prasailgika and Svatantrika views [represent] the intention of the middle
turning and the corpus of analytical works, [their interpretation of this]
intention is not free from mistakes. Although the explicit teaching of the
analytical works generally appears to be consistent, it is not so in a great
number of cases. Since many passages
45
of the treatises of the middle
turning clearly teach gian stan, the explicit teaching of the middle turn-
ing and the analytical works [should] not be [taken] literally. The explicit
teaching of most passages of the middle turning and the analytical works
contradicts neither the Prasailgika and Svatantrika nor the gian stan.
Nevertheless, for those appealing to the extraordinary tenet known as ran
stan, it has become a cause of confusion. On the other hand, given that
[these texts] do not teach different tenets, that they contradict other tra-
ditions, and that there are [in fact] many extraordinary passages which only
teach gian stan, even the middle turning and the analytical works [can be
said to] teach gian stan Madhyamaka. From these texts [of the middle
turning], however, the extraordinary points of gian stan - namely [those
reached by] following only the lines of commentary on the intention of
the last turning - have not clearly or extensively emerged. They are the
extraordinary tenets of the Prasailgika and Svatantrika. What is nowa-
days known as the ran stan view was not taught [in the middle turning];
nevertheless, this ran stan [interpretation of] the intention of the Buddha
and his sons is taught in detail [nowadays]46.
42 E.g., the analytical works of Nagarjuna, such as the Millamadhyamakakarika.
43 Taranatha: "Zab don fier gcig pa", 782, 11. 3-5: de la sakya'i mtshan can ni I thai
ran gi Ita ba 'di kun 'khar la bar pa dan I rig tshags kyi dgans dan 'thad Idan yin cin I
bar pa'i dnas bstan sgra ji Min pa yan yin I rig tshags dan 'khar la tha ma'i dnas bstan
mi mthun par bied la I.
44 Lit. "the great omniscient one from Jonang".
45 Lit. "words".
46 Taranatha: ap. cit., 782, 1. 5 - 783, 1. 6: kun mkhyen ja nan pa chen pa ni I thai ran
gi Ita ba 'di 'khar la bar pa dan rig tshags kyi dgans par rlam pa yin kyan I dgans pa rma
med pa ma yin la / rig tshags dnas bstan gyi tshig phal cher la 'byar ba Itar snan yan /
TARANATHA'S ''TWENTY-ONE DIFFERENCES" 297
Difference No.2
Sakya mchog Idan
47
: With regard to the fact that the Abhisamayiilarhkiira
teaches both the tenets of ran ston and gzan ston, [Maitreya] considered
the necessity of gzan ston in tenns of a meditation practice, and that of
Prasiuigika and Svatantrika, [which are at the same time] the ran ston of
the subsequent three works
48
, when it comes to cutting through mental fab-
rications with the help of the view. The remaining four Maitreya works
49
teach only gzan ston
50
• With regard to these [latter four] there are two
types: In the Ratnagotravibhiiga ultimately only one single path is taught
and the possibility of a cut-off potential refuted. In the other three
[Maitreya] works (i.e., Mahiiyiinasutriilarhkiira, Madhyiintavibhiiga and
Dharmadharmatiivibhiiga) ultimately three paths and a cut-offpotential
51
.are explained
52

Dol po pa: There are no different tenets in the five Maitreya works
at all. The tenet of the so-called ran ston is not explained even in the
mi 'byor ba yan man du yod la / 'khar 10 bar pa 'i giun tshig man pos gian ston gsal bar
ston pas 'khor 10 bar pa dan rig tshogs kyi dnos bstan sgra ji biin pa rna yin no / de la
'khor 10 bar pa dan rig tshogs kyi tshig phal cher gyi dnos bstan ni thaI ran dan gian ston
giiis ka la mi 'gal yan / ran ston par grags pa'i thun mon ma yin pa'i grub mtha'i khuns
la 'dren pa mams ni / de mams iiid la 'khrul gii byun ba yin gyi / grub mtha' de dan de
mi ston cin / de las gian de'i lugs dan 'gal iin / gian ston kho na ston pa thun mons rna
yin pa'i tshig kyan du ma yod pas / bar pa dan rig tshogs kyis kyan gian ston dbu rna iiid
ston no / 'on kyan de dag nas / 'khor 10 tha ma dgons 'grel dan bcas pa tsam du gian ston
thun mons ma yin pa mams gsal iin rgyas par 'byun ba min la / thaI ran gi thun mons rna
yin pa 'i grub mtha '/ den san ran ston gi Ita bar grags pa de mi ston kyan / rgyal ba sras
bcas kyi dgons pa'i ran ston de rgyas par bstan te / ces gsun so!.
47 From here on, Sakya mchog ldan and Dol po pa are referred to as "the former" and
"the latter".
48 le., the Mahayanasutralamkara, Madhyantavibhliga, Dharmadharmatavibhliga.
49 le., the Mahayanasutralamkara, Madhyantavibhliga, Dharmadharmatavibhliga, and
Ratnagotravibhliga.
50 This does not obviously exclude recourse to a Prasatigika view when cutting through
mental fabrications in the Mahayanasutralamkara etc.
51 Usually a cut-off potential and the potentials for entering on the paths of the Sravakas
and Pratyekabuddhas and on the Mahayana are distinguished.
52 Taranatha: op. cit., 783, 1. 6 -784, 1. 3: mnon rtogs rgyan gyis / ran ston gian ston
gi grub mtha' gnis ka ston pa ni / Ita bas spros pa gcod pa la thaI ran giun phyi gsum gyi
ran ston dan / sgom pas nams len gian ston dgos pa la dgons la / byams chos Ihag ma biis
gian ston kho na ston mod / de la' an rigs giiis te / rgyud blar mthar thug theg gcig bsad /
rigs chad bkag giun gian gsum du mthar thug theg gsum dan / rigs chad bsad gsun !.
298
KLAUS-DIETER MATHES
Abhisamayiilarhkiira. A real cut-off potential and three ultimate paths are
not explained in the [Mahiiyiinalsutriilarhkiira and so forth
53
.
Difference No.3
Sakya mchog ldan: Ran ston is considered to be more profound when j
it comes to cutting through mental fabrications with the help of the view.
When it comes to the practice of meditation, however, it is said that gian
ston is more profound. The ran ston
54
of the latter in turn, namely Pra-
sangika and Svatantrika, is acknowledged in the tradition of the subse-
quent three works (Le., the Mahiiyiinasutriilarhkiira etc.)55.
Dol po pa: The view of ran ston as taught by the Buddha and his sons
is superior in cutting through mental fabrications. Nevertheless, it is
contained in gian ston, and therefore view and practice are not opposed
to each other5
6
• To maintain that the ran ston, [namely,] the Prasangika and
Svatantrika - as it is known nowadays - is the view of the subsequent
three works, [thinking that according to the latter] nothing exists ultimately,
is wrong. [Such a ran ston] is therefore not better in cutting through men-
tal fabrications with the help of the view, for this would be a false denial
57
.
Difference No.4
Sakya mchog ldan: Even though gian ston goes beyond Cittamatra and
is thus acceptable to Madhyamaka, ran ston is superior to it with regard
53 Ibid., 784, 11. 3-4: byams chas sde Ina la grub mtha' so so ba ye med / ran stan par
-grags pa'i grub mtha'mnon [rtags] rgyan nas /cyan ma bsad/mda [sde] rgyan sags nas
/cyan / gtan nas rigs chad pa dan mthar thug theg gsum ma Mad / zes gsun no /.
54 The gzan stan view includes a ran stan that refers to the negandum, but not to what
is left over in emptiness.
55 Tiiranatha: ap. cit., 784,11.4-5: Ita bas spros pa gcad pa la ran stan zab / sgam pas
fiams su len pa la gzan stan zab ces te / de'i ran stan yan thai ran gzun phyi gsum gyi lugs
Ia nos 'dzin (text: 'dzi).
56 As would be the case if the former were strictly ran stan and the latter strictly gzan
stan.
57 Tiiranatha, ap. cit., 784, 11. 5-7: rgyal ba sras bcas bzed pa'i ran stan gi Ita ba de
spros pa gcad byed mchag yin /cyan / gzan stan du 'dus pas Ita grub logs logs pa min Ia
/ den san grags pa'i ran stan thaI ran gzun phyi gsum gyi Ita ba don dam bden med du
'dod pa ni nor ba yin pas / Ita bas spros pa gcad pa la bzan ba min te / skur 'debs su 'gyur
ba'i phyir yin zes gsun /.
TARANA.THA'S "TWENTY-ONE DIFFERENCES" 299
. to the view. Still, the former (i.e., gian stan) is not wrong, for it accords
with the experiential object of meditation
58

Dol po pa: Ran stan, too, goes beyond Cittamlitra, and thus falls. under
Madhyamaka within the system of the four tenets. It is not the pure ulti-
mate, however, the highest view being gian ston alone
59

Difference No.5
Sakya mchog Idan: For this reason non-dual wisdom is not analyzed when
following the Maitreya works. When critically analyzing it, after having
excluded [its] opposite, [wisdom] cannot withstand such analysis. Therefore,
since it cannot withstand a critical analysis [aimed at] ascertaining the ulti-
plate, ran stan is more profound in terms of the view. Even though it does'
not withstand analysis, this wisdom is experienced uninterruptedly. There-
fore it abides like the experiential object of meditation, namely gian ston
60

Dol po pa: Non-dual wisdom does withstand critical analysis61. There-
fore, this very analysis itself is self-delusion
62

Difference No.6
Siikya mchog Idan: Non-dual wisdom is momentary awareness (rig pa),
not permanent, and has no chance to abide
63

58 Ibid., 784, 1. 7 -785,1. I: gian stan sems tsam las 'das pas dbu ma go chad po yin
kyan / Ita ba'i nos nas de bas kyan ran stan mtho / 'on kyan sna ma de nor bar (text: par)
ni mi 'gyur te / sgom don dan mthun pas so gsun /.
59 Ibid., 785, 11. 1-2: ran stan yan sems tsam las 'das pas / grub mtha' bii'i dbu mar
bsdu ba tsam yin kyan / TTlam dag mthar thug min la / Ita ba'i mtho sos gian stan kho na
yin no gsun.
60 Ibid., 785, 11. 2-4: de'i rgyu mtshan du / byams chos rjes 'bran dan bcas par / gfiis
med kyi ye ses la dpyad pa mi byed pa yin la / spyi Idog nas de yan rig pas dpyad na dpyad
mi bzod pas / don dam gcod byed kyi rig pas dpyad bzod mi srid pas Ita ba ran stan
zab / dpyad ma bzod kyafl ye ses de iiams myon rgyun mi 'chad pas / sgom don gian stan
Itar gnas gsun /.
61 For it is beyond one and many. Moreover, wisdom is omnipresent, in that the Bud-
dhas embrace with their non-dual wisdom the all-pervading suchness of all phenomena
(cf. 2.2. below).
62 Taranatha: op. cit., 785, 1. 4: gfiis med ye ses rigs pas dpyad bzod yin pas / de la dpyod
pa de ran gi 'khrul gsun.
63 Ibid., 785, 1. 5: giiis med ye ses de rig pa skad cig ma yin / rtag pa min / gnas pa'i
go skabs med pa cig yin gsun.
300 KLAUS-DIETER MATHES
Dol po pa: This [wisdom] is not momentary, but p e ~ a n e n t and stable,
in that it is beyond the three times (i.e., past, present and future)64.
Difference No.7
Sakya mchog ldan: Likewise, given that it is knowledge (ses pa), wisdom
[can be] taken to be an entity/existent (dftos pO)65.
Dol po pa: And it [can be] taken to be beyond both [the state of] an
entity/existence and a non-entity/non-existi.mce
66

Difference No.8
Sakya mchog ldan: Likewise it [can be] taken to be conditioned
67

Dol po pa: It [ can be] taken to be unconditioned, to0
68

Difference No.9
Sakya mchog ldan: If one isolates its specific aspects (raft Zdog), all
knowledge is - as generally accepted in Tibet - only clarity and aware-
ness, and here an entity of the dependent [nature]. The isolation of the
specific aspects of mere dualistic appearances which arise in this [clarity
and awareness] results in the imagined nature. When viewed under its
aspect of being accompanied by these dualistic appearances, clarity and
awareness constitute the dependent nature. From the perspective, however,
that it is unstained by these dualistic phenomena throughout beginning-
less time, this clarity and awareness constitute the perfect nature. Based
on that, dependent entities as such are by nature the same as the perfect
nature, even though they are different as isolates (ldog cha) and different
in terms of their respective defining characteristics
69
.
64 Ibid., 785, 11. 5-6: de skad Gig ma ma yin / dus gsum las grol bas rtag pa brtan pa
yin gsun.
65 Ibid., 785, 1. 6: de biin du ses pa yin pa'i phyir dnos por bied pa dan!
66 Ibid., 785, 1. 6: dnos po dnos med giiis ka las grol bar bied pa ste!
67 Ibid., 785, 11. 6-7: de biin du 'dus byas su bied pa dan!
68 Ibid., 785, 1. 7: 'dus ma byas su bied pa yan ste!
69 Ibid., 785, 1. 7 -786,1. 3: bod spyi dan mthun run du ses pa thams cad kyi ran ldog
gsal rig {sam 'di ka gian dban gi dnos po yin la / de iiid la sar ba'i giiis snan kho na'i
TARANATHA'S "TWENTY-ONE DIFFERENCES" 301
Dol po pa: The imagined aspect, which is imputed by the mind's multi-
tude of thoughts; and its appearances in the form of external objects, is
the perceived. The isolation of its specific aspects is the imagined nature.
The isolation of the specific aspects of the mind and mental factors results
in the dependent nature, namely knowledge constituted by knowledge or
consciousness of apparent truth. Clarity and awareness, by nature free
from mental fabrication, is the perfect nature. Thus the imagined and the
dependent are substantially the same; their defining characteristics are
very different, however. Not only are the perfect and the dependent dif-
ferent as isolates and in terms of their defining characteristics, but they
are also not the same by nature (flO bo gcig pa)70. The previous presen-
tations of this [trisvabhava-theory] were mainly in line with Cittamatra,
but [Dol po pal thinks that the tradition of Madhyamaka is only this [tri-
svabMva]71.
Difference No. 10
Sakya mchog ldan: The imagined nature fully pertains to what is not true,
the perfect to what is true, and the dependent to both72.
ran ldog nas kun btags (text: brtags) yin no I gsal rig gfiis snan de dan bcas pa'i cha nas
gzan dban dan I gsal rig de la gfiis chos gdod nas ma gos pa 'i cha nas yons grub yin pas
I gzan dban yons grub ldog cha nas tha dad cin I mtshan fiid kyi cha nas tha dad kyan
gzan dban gi dnos po de fiid dan I yons grub no bo gcig par bzed do !.
70 This negation of identity h!ls been often misunderstood and misleadingly represented.
(Cf. Newland, who writes that for Dol po pa the two truths are different entities (no bo
tha dad pa). Instead of referring directly to the Jonang material, however, he quotes Sey-
fort Ruegg, Hopkins and Thurman (Newland 1992: 30 & 260). In fact, Dol po pa negates
not only identity but also difference. In his "bDen gfiis gsal ba'i iii rna", 23, 11. 2-3) he
explains that "the two truths should be called neither identical (de fiid) in terms of their
nature nor different (gzan) [in terms of their nature].' Cf. also Mathes 1998:465-6.
, For Tib. de fiid dan gian, Skt. tattviinyatva, see MA VBh, 23, L 10.
7l Taranatha: op. cit., 786, 11. 3-6: blo rnam rtog sna tshogs pas brtags pa'i btags cha
dan I phyi don du snan ba'i snan cha ste I gzun ba'i ran ldog kun btags (text: brtags)
dan I sems sems byun ses pa'i ran ldog kun rdzob pa'i ses pa'am rnam ses kyis bsdus pa'i
ses pa gian dban dan I spros pa dan bral ba'j ran biin.gsal rig yons grub ste I des na kun
btags (text: brtags) ni gzan dban las rdzas tha dad du med kyan I mtshan fiid kyi sga nas
ni sin tu tha dad do I yans grub dan gian dban ni I [dog cha dan mtshan fiid tha dad par
ma zadl no bo gcig pa yan ma yin no I sna ma'j rnam gzag ni sems tsam dan mthun sas
che la I dbu ma'i lugs ni 'di kho na'o ies dgans so!.
72 Ibid., 786, 11. 6-7: kun btags (text: brtags) la bden med kyis khyab lyons grub la bden
yod kyis khyab I gian dban la cha gfiis mdzad !.
302
KLAUS-DIETER MATHES
Dol po pa: The imagined and the dependent both fully pertain - that is
to say, through and through - to what is not true
73

Difference No. 11
Sakya mchog ldan: All self-awareness - understood as the isolation of
its specific aspects - [belongs] exclusively to the ultimate [truth]14.
Dol po pa: Given that the self-awareness
75
of consciousness [belongs]
exclusively to the apparent [truth], self-awareness, too, has both an appar-
ent and an ultimate aspect1
6

Difference No. 12
Sakya mchog ldan: The perfect [nature] is emptiness. The imagined
[nature] is not emptiness, even though it is purely empty77. Emptiness fully
pertains to the ultimate [truth]18.
Dol po pa: Everything, phenomena and their true nature, can only be
called emptiness. Emptiness does not pertain to (lit. "is not pervaded by")
the ultimate truth. It is not counted unambiguously among the synonyms
[of the ultimate] : [emptiness] is related to [the ultimate only] in a general
sense
79

Difference No. 13
Sakya mchog ldan: The works on valid cognition, the Abhidharma etc.
are mostly [written] in accordance with general Dharma terminology.
73 Ibid., 786, 1. 7: kun btags (text: brtags) gian dban giiis ka la mtha' gcig tu bden med
kyis khyab par mdzad do /.
74 Ibid., 786, 1. 7 -787, 1. 1: ran rig thams cad ran rig gi ran ldog nas don dam kho
na yin gsun la /.
75 The Tibetan uses the plural: "moments of self-awareness".
76 Ibid., 787, II. 1-2: kun rdzob roam ses kyi ran rig roams kun rdzob kho na yin pas
/ ran rig yan kun rdzob don dam giiis yod par bied do /.
77 I.e., it is the negandum and nothing more.
78 Ibid., 787, 1. 2: yoils grub ston pa iiid yin / kun btags (text: brtags) ston pa tsam yin
kyan ston pa iiid min / ston pa iiid la don dam gyis khyab par bied /.
79 Ibid., 787, II. 2-3: chos dan chos iiid thams cad la ston iiid tsam du brjod dgos / ston
iiid la don dam gyis ma khyab / min gi roam grans la khyab mtha'i rna rtsi / gtso bo'i don
du sbyor gsun /.
TiillANATHA'S "TWENTY-ONE DIFFERENCES" 303
This being the case, the attainment of pacification fully pertains to both
categories, those' of entities/existence and non-entities/non-existence
8o
;
knowledge (§es pa)81 [only] to that of entities
82
• The ultimate is not an
entity. Since it is not conditioned, it is a non-entity, [like] the sky and so
forth. There are different aspects of the unconditioned - suchness not
being conditioned by causal defilements, or mere clarity and awareness
not being newly produced, etc. Therefore, when one enumerates cate-
gories, these are designated as unconditioned. They are, however, not the
[real] unconditioned as opposed to the conditioned ('du byed) and the
defining characteristics (mtshan ilid); therefore, they are unconditioned
only in a metaphorical sense
83

Dol po pa: Explanations along the lines of Pramal),a or Abhidharma
belong to traditions that mainly ascertain the apparent truth. With regard
here to definitive meaning, when it is mainly the ultimate truth that is
being ascertained, entities and non-entities fully pertain to the apparent
[truth] and vice versa
84
• The ultimate truth is neither an entity nor a non-
entity; therefore, the attainment of pacification
85
certainly does not per-
tain to the ultimate. [If it did,] wisdom would not be an entity, while
being knowledge at the same time. Therefore, knowledge would not per-
tain to [the category of] entities, while to maintain that the ultimate truth
80 Pacification, or cessation, falls under this latter category by virtue of being uncon-
ditioned.
81 In the context of the Abhidharma: usually the "knowledge of the destruction [of
passions etc.]" (Skt. Tib. zad pa ses pa) and the "knowledge of no further
occurrence [of passions etc.]" (Skt. anutpiidajiiiina, Tib. mi skye ba ses pa).
82 In the following Tib. dizos po is rendered as "entity", even though the Sanskrit equiv-
alent bhava also means "existence".
83 Taranatha: op. cit., 787, II. 3-6: tshad ma'i giuiz dan mizon pa sogs spyi skad daiz
phaL cher mthun par / / ii grub pa La dizos po dnos med gaiz ruiz gis khyab / ses pa La dnos
pas khyab / don dam dnas pa min /' dus rna byas (text om. byas) pas ni dizas med nam mkha'
sags yin/ de biin iiid La las iion gyi 'dus rna byas daiz/ gsaL rig tsam gsardu 'dus rna byas
sags / 'dus rna byas pa'i cha re yod pas / mam grans kyi sga nas 'dus rna byas su btags
pa yin gyi / 'dubyed daTi mtshan iiid 'gal ba'i 'dus rna byas rna yin te / des na 'dus rna
byas btags pa ba yin gsuiz !.
84 This means that the totality of entities and non-entities is exactly identical with the
apparent truth.
85 Dol po pa restricts the ultimate truth to the actual cessation, which exists throughout
beginuingless time. The actual pacification attained thus still forms, together with suffer-
ing, a dualistic concept yet to be transcended.
304
KLAUS-DIETER MATHES
is a non-entity would be improper Dharrna
86
. To maintain that the ulti-
mate is an entity [is in accordance with] the tradition of maintaining the
. [ultimate existence of] entities. All non-entities like the sky etc. which
the Abhidharrnikas take to be unconditioned, are there considered to be
conditioned, and for this reason, both entities and non-entities fully per-
tain to the conditioned. The ultimate is the real unconditioned. The sky
etc. are thus unconditioned [entities] only in a metaphorical sense
87

Difference No. 14
Sakya mchog ldan: The "very face" (ran no) of the dependent, being
empty of the imagined, that is, the negandum, is the basis of emptiness.
It may be taken as the ultimate being empty of the apparent
88

Dol po pa: The perfect is the basis of emptiness. It is empty of the two
neganda, the dependent and the imagined, in that the ultimate is empty
of the apparent. [The explanation of] the dependent as being empty of the
imagined applies only when ascertaining mere apparent truth
89

Difference No. 15
Sakya mchog ldan: Even though the "pure dependent" is widely known
in Tibet, it is in reality not the dependent but rather what is "perfect in
86 This reductio ad absurdum presupposes the inclusion of wisdom under the ultimate
truth.
87 Taranatha: op. cit., 787, L 6 -788, L 3: tshad mnon sags su bSad pa de I kun rdzob
gtso bar gtan la 'bebs pa'i lugs yin I don dam gtso bar gtan la 'bebs pa'i nes don gyi skabs
'dir I kun rdzob fa dnos po dan dnos med kyis khyab cin I dnos po dnos med la'an kun
rdzob kyis khyab I don dam dnos po yan min I dnos med (text: mod) kyan min pas ii (text:
gii) grub fa nes ma khyab lye ses dnos po ma yin La ses pa yin pas I ses pa fa dnos pas
ma khyab dnos med don dam du 'dod pa chos mi rigs lal don dam dnos par 'dodpa dlios
smra ba'i lugs so I mnon pa ba rnams 'dus ma byas su 'dod pa'i nam mkha' sags I dnos
med thams cad kyan skabs 'dir 'dus byas yin pas I dlios po dnos med la 'dus byas kyis khyab I
don dam 'dus ma byas dnos yin I nam mkha' sags 'dus ma byas btags pa ba yin gsun!.
88 Ibid., 788, L 4: stan gii gian dbmi gi ran no de dgag bya kun btags kyis stan pa ste I
de fiid kun rdzob kyis stan pa'i don dam du mdzad do!.
89 Ibid., 788, 11. 4-6: stan gii yons grub I dgag bya gian dban kun btags (text: brtags)
gfiis kyis ston pa I don dam kun rdzob kyis stan pa 'i don yin fa I gian dbali kun btags (text:
brtags) kyis stan pa ni I kun rdzob bden pa kho na gtan la 'bebs pa'i skabs kho na yin par
bied do I.
TARANATHA'S "TWENTY-ONE DIFFERENCES" 305
termS of being unmistaken." The latter is the actual perfect nature. Since
the origin and usage of the conventional [term] "pure dependent" is not
clear, it is not good to use it
90

Dol po pa: Even though the usage of the conventional [term] "pure
dependent" is not clear - the term is not found in the treatises - its
meaning is fully established [in them], and therefore it is proper to use
it. This follows from the fact that the Buddha's teaching is based on mean-
ing [rather than words proper], and that in olden times in Tibet all agreed
on such a convention. Therefore, it is appropriate not to find any fault
in the transmitted pith-instructions deriving from Maitreya. Even though
some [parts of the] wisdom of the noble [path of] learning are [called]
"perfect in terms of being unmistaken" in [certain] passages of the
[Maitreya works], the presentation of its conventional [term] (i.e., the
term "pure dependent") is good. This is because of [instances] where
some [phrases] such as "for those who have attained the [Bodhisattva]-
levels the ground appears as gold" are also [taken as denoting] "perfect
In terms of being unmistaken"91.
Difference No. 16
Sakya mchog Idan: The perfect in terms of being unmistaken fully per-
tains to the actual perfect nature
92

Dol po pa: This [being unmistaken] is only taught as being the express-
ible perfect, in the same way as the twelve limbs of the Buddha's speech
90 Ibid., 788, n. 6-7: dag pa gzan dban ies bod spyi la grags kyan / don la gian dban
min cin phyin ci ma log pa'i yons grub yin la/ phyin ci ma log pa'i yons grub kyan yons
grub mtshan Hid pa yin cin / dag pa gian dban gi tha sHad pa'an khuns gsal med pas /
de'i brda 'chan byed pa mi legs par dgOlis so I.
91 Ibid" 788, L 7 -789,1. 3: dag pa gzan dban zes pa'i tha sHad gzun las gsal po ma
byun yan / don tshan bas tha sHad byar run ste / sans rgyas kyi bstan pa don la brtan pa
yin pa dan / bod sna ma thams cad tha sHad de Ita bu mthun par byed pas / byams pa nas
brgyud pa'i man ilag ma nor ba yin du run bas so / de'i nan tshan 'phags pa slob pa'i ye
ses 'ga' zig phyin ci ma log pa'i yoils grub yinkyan / sa thob pa mams la sa gii gser du
snail ba sags 'ga' iig phyin ci ma log pa'i yoils grub yin pa'an yod pas / de'i tha sHad
rnam giag legs par dgoils so I.
92 Ibid" 789, L 4: phyin ci ma log pa'i yans grub la/yons grub mtshan Hid pas khyab
parbzedl.
306
KLAUS-DIETER MATHES
have been also said to be the perfect [in terms of being unmistaken].
Thus the latter, incontrast to the unchangeable perfect nature, is in real-
ity something that belongs to the pure dependent and is the perfect only
in a metaphorical sense. The unmistaken perfect which is the same as
the unchangeable [perfect] is called the "ultimate perfect in terms of
being unmistaken." It is purely unchangeable
93
• Therefore, when one
ascertains the true state of being, it is only this "[ultimate] perfect". When
one explains in detail [its] synonyms, both types [of the perfect] (i.e.
the unchangeable and the perfect in terms of being unmistaken) are pre-
sented
94
.
Difference No. 17
Sakya mchog Idan: The [noble] truth of the path also [belongs to] the
ultimate truth
95
.
Dol po pa: Among the four noble truths the truth of cessation is the ulti-
mate, and the other three are the apparent truth. To be more precise, only
the actual cessation, which exists throughout beginningless time, [is called]
ultimate [truth]. The other three [noble] truths and the analytical cessa-
tion fully pertain, in reality, to the apparent truth. Hence the actual [noble]
truth of the path fully pertains to the apparent [truth], and the actual truth
of cessation to the ultimate [truth]. This follows from the fact that the
[noble] truth of the path in its ultimate aspect is one with the beginningless
93 The distinction between "perfect in terms of being unmistaken" and "ultimate perfect
in terms of being unmistaken" reflects the Jonangpas' view that wisdom; like all other Bud-
dha-qualities, mainly pertains to the unchangeable ultimate truth and only to limited extent
to the apparent truth.
'In the Yogacara, "the perfect in terms of being unmistaken" usually refers to non-con-
ceptual wisdom cultivated on the path.
94 Taranatha: op. cit., 789, 11. 4-7: de ni mam grails kyi yoils grub tu bstan pa tsam ste /
gsuil rab yan lag bcu gfiis kyail yoils grub tu gsuils pa dail 'dra'o / des na / 'gyur med
yoils grub kyi zlas drails pa 'i phyin ci rna log pa de ni gian dbail dag pa pa yin ciil / yoils
grub btags pa ba yin / 'gyur med fiid dail gcig pa 'i yoils grub phyin ci ma log pa de la /
don dam pa 'i phyin ci ma log pa'i yoils su grub pa ies bya ste / 'gyur med kho na' 0/ des
na / gnas tshul gtan la 'bebs pa'i skabs su yoils grub de fiid kho na yin la / mam grails
rgyas par bsad pa 'i skabs su gfiis kyi mam giag byed do gsuil /.
95 Ibid., 789,1. 7 - 790, 1. 1: lam bden yail don dam par mdzad la /.
TARANATHA'S "TWENTY-ONE DIFFERENCES" 307
[ultimate truth]. Because it is [in reality] the [noble] truth of cessation, [this
ultimate aspect] is the [noble] truth of the path [only] in a metaphorical
sense
96

Difference No. 18
Sakya mchog ldan: There is no Buddha-nature in the mind-stream of
sentient beings. The natural luminosity of the mind of sentient beings is
JIlerely the cause of the Buddha-nature and [its] "basic element" (khams).
Therefore, there is a Buddha-nature or basic element as a cause in all
. ordinary sentient beings, but it is not like the actual [Buddha-nature], which
is rather the [same as] Buddha-wisdom
97

Dol po pa: The actual Buddha-nature is nothing else than [the Buddha-
nature] of the mind-stream of sentient beings, and if it is the actual
[Buddha-nature] of a Buddha, then it is established that sentient beings
possess it, precisely because it is the dharmata of sentient beings. This is
proven, in particular, by a number of canonical passages. The explana-
tion [of the Buddha-nature] as the basic element and cause [refers to] a
cause different from the sphere/element (dbyihs)98, given that the latter is
neither an efficient cause nor an efficient sphere
99

96 Ibid., 790, n. 1-3: bden pa biir phye ba'i 'gog bden don dam dan / bden pa gian
gsum kun rdzob tu bied / iib mor na / 'gog bden mtshan fiid pa gdod ma'i 'gog bden la
don dam kho na dan / gian bden pa gsum dan / so sor brtags 'gog la kun rdzob kyis khyab
par don la gnas pas / lam bden mtshan fiid pa la kun rdzob kyis khyab / 'gog bden mtshan
iiid pa la don dam gyis khyab / don dam pa'i lam bden ni gdod ma dan gcig pa 'i phyir /
'gog bden iiid yin pas / lam bden btags pa ba yin gsun!
97 Ibid. 790, II. 3-5: sems can kyi rgyud la bde gsegs siiili po med sems can kyi sems
ran biin 'od gsal de / bde gsegs siiin po'i rgyu dan khams tsam yin pas / rgyu bde gsegs
siiin po 'am khams bde gsegs siiin po sems can thams cad la yod kyan / de ni de 'dra
mtshan iiid pa min / sans rgyas kyi ye ses bde gsegs siiin po' 0 !
98 The Buddha-nature with all its qualities is already present in one's mind-stream
and thus does not need to be generated. Here, Buddha-nature as cause means that focus-
ing on the true nature of mind, which is the Buddha-nature, causes the removal of all defIle-
ments.
99 Tliranatha: op. cit., 790, II. 5-7: sems can gyi rgyud kyi de ka bde gsegs siiin po
mtshan iiid pa yin te / sans rgyas de kyi mtshan iiid pa yin na / de iiid ka sems can gyi
chos iiid yin pas / sems can bde gsegs siiin po can du grub la / khyad par lun mtha' yas
pas grub bo / khams dan rgyur Mad pa yan / skyed rgyu dan skyed khams min par dbyins
dan bral rgyu la dgons pa' 0 gsun !
308 KLAUS-DIETER MATHES
Difference No. 19
Sakya mchog ldan: [Passages which] state that the Buddha-nature is
endowed with essentially inseparable qualities refer only to the fruit [of the
path]. On the level of the cause, the qualities still have to be developed.
Having this capability, the Buddha-nature exists only as seeds [in ordinary
sentient beings]loo.
Dol po pa: The essentially inseparable qualities are naturally present.
They exist even in the Buddha-nature of the basis, since [firstly] some-
thing that arises, [in the sense of being] newly [acquired], may possibly
be not naturally present; [secondly], the division of basis, path and fruit
applies only to the level of "phenomena" (dharmin)lOl [or] apparent truth;
and [thirdly], there is, [in terms of] the nature [ of phenomena], only one
Buddha-nature. It must henceforth be the Buddha-nature, adorned with all
the qualities of the ultimate
102

Difference No. 20
Sakya mchog ldan: The major and minor marks and the like [of a Bud-
dha] do not belong to the qualities of the dharmakaya
103

104
100 Ibid., 790, 1. 7 - 791, 1. 1: sfiin po la yon tan no bo dbyer med ran biin fiid ldan du
gsuns ba 'bras bu kho na'i skabs yin la I rgyu'i skabs su yon tan 'byun run gi nus pa sa
bon tsam yod par bied !.
101 In the context of phenomena (dharma) and their true being (dharmatii) , dharmin
refers to the dharmas which possess dharma{tii} (see Mathes 1996: 185).
102 Ibid., 791, ll. 1-4: yon tan no bo dbyer med I ran biin fiid ldan de gii bde gsegs
sfiin po la yan yod de I gsardu byun bas ran biin gyis ldan par mi 'gro ba 'i phyir dan I
gii lam 'bras bu gsum kun rdzob chos can gyis phye ba ma gtogs I ran biin bde gsegs sfiin
po gcig kho na yin pa'i phyir I bde gsegs sfiin po yin phyin chad don dam pa'i yon tan
thams cad kyis brgyan pa yin dgos par bied do I.
103 The Y ogacaras use the term dharmakiiya in an exclusive as well as in an inclusive
sense. In its exclusive sense, the term dharmakiiya is defined as the transformation of the
basis which results in the non-conceptual wisdom of the purity of suchness. In its inclu-
sive sense, it refers to the totality of Buddhahood including all kttyas. The interpretation
of the dharmakiiya as a distinct kiiya, which does not include the sambhogakiiya, is also
found in the particular four-kiiya theory of Haribhadra's commentary on the Abhisama-
yiilamkiira. Haribhadra qualifies the dharmakiiya for the first time with the compound
jfiiiniitmaka "the dharmakiiya consisting of wisdom", and understands it as conditioned
jfiiinas on the level of apparent truth. It is only the sviibhiivikakiiya which encompasses
suchness, or the ultimate emptiness of all phenomena. Contrary to this, Arya Vimuktisena
TARANATHA'S "TWENTY-ONE DIFFERENCES" 309
Doi pO pa: With regard to all types of Buddha-qualities, there is an aspect
of them that pertains to the ultimate qualities of the dharmakiiya, and
appears only to the Buddha himself, and another aspect that pertains to
the apparent qualities of the form-kiiyas, and appears to others, namely
the disciples [of the Buddha]. As for the explicit teaching of the Ratna-
diirikiisiitra
105
and the Uttaratantra [Ratnagotravibhiigaj, in general it is
necessary to explain them in terns of the qualities of the two kiiyas106.
On the whole, with respect to the major and minor [marks of a Buddha],
[Dol po pal only makes analogies in accordance with what is generally
accepted. When taken as taught in other Siitras, and the Tantras, both [the
dharmakiiya and the form-kiiyas] share aspects common to all oftheml07.
What is different [from the Ratnagotravibhiiga] when Mantra[yana is'taught]
with regard to gian ston is precisely this [inclusion of all qualities in the
. ultimate F08. 109
Difference No. 21
Sakya mchog ldan: Only the seeds of the fruit are inherently present in
the form of the natural luminosity of mind. [Their] improvement is
achieved by meditating on the path, until the fruit is finally actualized 110.
takes dharmakaya as a synonym of svabhiivikakaya, and as such the totality of Buddha-
hood comprising all kayas (see Makransky 1997:9-13 & 39-41).
104 Taranatha: op. cit., 791, I. 4: chos sku'i yon tan la mtshan dpe sogs mi bied /.
105 The explanation of the sixty-four qualities in the third chapter of the Ratnagotra-
vibhaga is based on the Ratnadarikasiltra (see Takasaki 1966: 14).
106 I.e., the thirty-two qualities of the dharmakaya and the thirty-two qualities of the
form-kayas.
107 Which means that the dharmakaya possesses aspects of the major and minor marks,
and the form-kayas aspects of the thirty-two qualities of the dharmakaya.
108 This sentence is the beginning of the following paragraph, but refers to the previous
one.
109 Taranatha: op. cit., 791, 11. 4-7: sans rgyas kyi yon tan gyi rigs thams cad la / chos
sku'i yon tan don dam pa sans rgyas ran snan re dan / gzugs sku'i yon tan kun rdzob pa
gdul bya gian snan gi cha re thams cad de yod pa yin la / bu mo rin chen gyis ius pa'i
mdo dan / rgyud bla ma'i dnos bstan ni spyir sku giiis ka 'i yon tan' chad dgos pa la gtso
che chun gi dban las / grags pa spyi dan mthun pa dper brjod tsam du mdzad la / mdo
gian rnams dan rgyud sde sogs las gsuns pa [tar na / giiis ka la thams cad kyi char yod
pa yin no gsun / [ ... J de ni snags kyi skabs su gian ston la mi 'dra ba ni 'di yin te /.
110 Ibid. 791,1. 7 -792,1. 1: sems ran biin 'od gsalla 'bras bu'i sa bon tsam ran chas
su yod pa / lam bsgoms pas gon 'phel thob / mthar 'bras bu mnon gyur du 'byun bar bied /.
310
KLAUS-DIETER lViATHEs
Dol po pa: Throughout beginningless time wisdom is effortlessly perfect
in the form of the ultimate malJq.aia. On the path, stains are removed by
meditating on it, and [this ever-present wisdom] is actualized
lll
.
2.2. Tiiraniitha's Conclusion
Having elaborated Siikya mchog ldan's and Dol po pa's twenty-one
differences with regard to the profound meaning, Taranatha' concludes
by pointing out one fundamental difference, to which all the other ones
basically refer:
PIlI). chen Sakya mchog ldan takes non-dual wisdom to be non-abiding and
impermanent in every moment, in that it is not something single but multi-
ple. [For the] omniscient Jonangpa (i.e., Dol po pa) it is in reality neither
one nor many; provisionally he accepts it as reasonable when [wisdom] is
presented as being single, and takes it as being permanent, impartible, all-
pervading, free from mental fabrication and ineffable. In view of this, the
[main] difference is, in short, that [the former] takes [wisdom] as being
impermanent, and [the latter] takes it as being permanent
I12

We are further informed that Dol po pa infers from the omnipresence
of non-dual wisdom that all qualities of a Buddha are already present in
ordinary beings. For the same reason, non-dual wisdom cannot be
impaired by reasoning, such as that it is neither one nor many, and hence
withstands analysis. On these grounds the tenets of the Prasailgika and
Svatantrika, which assert the destruction of non-dual wisdom by analy-
sis, is wrong, and hence these Prasailgika and Svatantrika views are
. impure. One comes to know this by way of analysis which makes use of
reasoning without distorting the original intention of the middle turning.
Dol po pa and Siikya mchog ldan agree, however, that ultimately the Bud-
dha-nature is beyond words and'thoughts, and the unmistaken object of
III Ibid., 792, 11. 1-2: gdod rna nas ye ses don dam pa'i dkyil 'khor du lhun grub tu
rdzogs pa de lam bsgoms dri rna bsal te milon du gyur pa yin par bied do I.
112 Ibid., 792, 11. 5-6: palJ chen sakya'i mtshan· can ni I gfiis med kyi ye ses de gcig pu
ma yin pa du ma'i tshul can I skad cig gis mi gnas pa mi rtag par bied pa dan I kun
mkhyen jo nan pa ni I de dnos gnas la gcig min du ma yan min par nes mod I gnas skabs
su gcig pu'i (text: bu'iJ mam gzag 'thad ldan du mdzad cin I de rtag pa I cha med kun
khyab I spros bral brjod bral yin par bied pa 'i dban las te I mdor na mi rtag pa dan rtag
par bied pa'i khyad par ro I.
TARANATHA'S "TWENTY-ONE DIFFERENCES" 311
wisdom. Taranatha concludes by explaining at length that
Dol po pa's "perinanent" is not the ordinary opposite of impermanent:
This ["pennanent"] is free from mental fabrications. It is the unchangeable
sphere which is free from both the impennanence of an established entity
and the pennanence of a negated non-entity. It is free from the characteristic
signs of pennanence1l3.
3. Taranatha's Presentation of Dol po pa's and Sakya mchog ldan's
Positions
A comparison of Taranatha's summary of Dol po pa's position with
what we fmd in the latter's pertinent works, such as the Ri chos ftes don
rgya mtsho, shows that the subject matter is correctly presented. Of par-
ticular interest are Taranatha's elaborations on difference no. 20, where he
confirms my own observation that Dol po pa explains the Buddha-nature
more in line with general Mahayana when commenting the Ratnagotra-
vibhiiga114. Also, the presentation of the trisvabhiiva theory is in accor-
dance with the Ri chos ftes don rgya mtsho, in which Dol po pa explains:
The basis which is empty of the imagined is the dependent, and the basis
which is empty of even the dependent is the true nature of phenomena, the
perfect1l5.
A little further down Dol po pa further explains:
It has been taught that phenomena which [belong to] the imagined [and] the
dependent do not really exist, and that the true nature of phenomena, [namely]
the perfect really does. The meaning of ran ston and gzan ston is taught in
these two statements
1l6
.
113 Ibid., 794, 11. 6-7: 'di ni spros bral te I dnos po sgrub pa mi rtag pa dan I dnos med
dgag pa rtag pa giiis ka las grol ba'i dbyins 'gyur med de yin I rtag pa'i mtshan ma las
grol iin /.
1I4 See my introductory remarks on Dol po pa in the fIrst paragraph of this paper.
liS Dol po pa: Ri chos nes don rgya mtsho, 148,11. 3-4: ... kun btags (text: brtags) kyis
ston pa'i gii ni gian dban no I I gian dban gis kyan ston pa'i gii ni chos iiid yons grub
bo (text: po) /.
1I6 Ibid., 149, 11. 8-10: ... kun btags (text: brtags) gian dban gi chos rnams ni yan dag
par med pa dan I chos iiid yons grub ni yan dag tu yod par bstan te 'di dag gis kyan ran
ston dan gzan ston gi don bstan pa yin no /.
312 KLAUS-DIETER MATHES
These two quotes clearly show that it is only the perfect which really
exists as the basis which is empty of the dependent (and thus also the
imagined). Moreover, the following passage from the Ri chos nes don
rgya mtsho confirms Taranatha's observation in difference no. 14 that
"[the explanation of] the dependent as being empty of the imagined
applies only when ascertaining mere apparent truth":
As to the lack of an own-being in the imagined, the [imagined] does not exist
in terms of its own defining characteristics. Being established 'as the mere
apparent, [or rather as] the mere mistaken apparent, it is established neither
as apparent truth nor as the correct apparent. As to the lack of an own-being
in the dependent, even though [the dependent] exists on the level of apparent
[truth] as an own-being which arises from something else, it does not exist
as an own-being which arises from itself, and is not in the least established
in reality. In this way, the two lack an own-being, because they are ran
stan. As to the true nature of phenomena, the perfect, or the basis of the
non-existence of these two (i.e., the imagined and dependent), even though
it is not the case that it lacks an own-being, it is the basis for the lack of an
own-being in the phenomena of apparent [truth], which are different from
[this basis]. Therefore it is the own-being of the ultimate truth, or the "body
belonging to the own-being" (sviibhiivikakaya)ll7.
That the ultimate basis of emptiness is restricted to the unchangeable
perfect becomes clear in the following passage where the perfect in terms
of being unmistaken is equated with the form-kayas of the apparent truth:
Thus the ultimate Buddha is the kiiya of the five self-arisen wisdoms. He
abides permanently in the form of [these] five wisdoms, which are suchness
and the unchangeable perfect. The form-kiiyas of the apparent [truth] pos-
sess correct wisdom, namely the perfect in terms of being unmistaken, and
[thus] the wisdom of the Mahayana[ -path] of no more learning which is not
beyond moments
ll8
.
Il7 Ibid., 319,11. 16-24: lam btags no bo med pa ni ran gi mtshan iiid kyis med pa ste
kun rdzob tsam mam log pa'i kun rdzob tu grub pa tsam las kun rdzob bden pa'am yan
dag kun rdzob tu yan ma grub bo / gian dba/i no bo fiid med pa ni kun rdzob tu gian las
skye ba'i no bar yod kyan ran las skye ba'i no bar med cin yan dag par na cir yan ma
grub pa ste de ltar de gfiis ni ran stan yin pa 'i phyir ran gi no bo med pa' 0 / / de gfiis med
pa'i gii chos fiid yons grub ni ran gi no bo med pa ma yin yan de las gian pa kun rdzob
kyi chos rnams kyi no bo fiid med pa 'i gii yin pa 'i phyir don dam bden pa'i no bo ste no
bo iiid kyi sku ....
118 Ibid., 356, L 22-357, L 2: de [tar don dam pa'i Sa/is rgyas ni ran byun ye ses lna'i
sku yin la / de yan de Min fiid dan 'gyur med yons grub fiid ye ses lnar rtag tu biugs pa'o
TARANATHA'S "TWENTY-ONE DIFFERENCES" 313
For Dol po pa, the ultimate is beyond moments and the three times
(i.e., the past, present and future). The permanence of the ultimate wis-
dom is thus not an ordinary permanence as opposed to impermanence, but
one that is, as Taranatha puts it, beyond these latter two categories. To be
sure, for Dol po pa all kayas have an ultimate aspect that is beyond the
three times:
That the permanent Buddha and the liberation of the Buddha are form, that
even space is the form of the Buddha, and so forth - the meaning of such
statements must be understood in the context of forms etc. being explained
[on the level] of suchness or as forms etc. which are beyond the three times
and the threefold world1!9.
Such an extreme form of transcendence explains Dol po pa's sharp
distinction between the ultimate and apparent truths - which he defines
with the phrase: "a difference in terms of a negation of an identity" (gcig
pa bkag pa'i tha dad pa). From this it does not follow, though, that the
two truths are different entities 120 , but simply that the ultimate exists and
the apparent does not (negation of identity). To be sure, since there is
only one essence for Dol po pa, namely that of the ultimate, it does not
make sense to speak of an essential difference, since this would require
the existence of another essence from which it differs. This also means
that Dol po pa's distinction between ultimate and apparent kiiyas does not
entail the absurd ontological view that there really are two different sets
of kiiyas
121
• It is rather that only the ultimate kiiyas exist ontologically.
The kiiyas of apparent truth, which are equated with the perfect in terms
of being unmistaken, do not really exist, any more than the apparent world
does. Still, on the level of apparent truth they are produced to the same
extent as the accidental stains of the apparent truth are removed (which
/ / kun rdzob gzugs kyi sku ni yan dag ye ses phyin ci ma log pa'i yons grub dan ldan pa
ste theg chen mi slob pa'i ye ses skad cig las ma 'das pa dan [dan pa'o /.
119 Ibid., 142,11. 17-9: sans rgyas rtag pa dan sans rgyas kyi thar pa gzugs yin pa dan
nam mkha' yan sans rgyas kyi gzugs yin / zes pa la sogs pa'i don ni f. .. J de biin fiid kyi
gzugs sogs dan/ khams gsum dan dus gsum las 'das pa'i gzugs sogs ies pa la sogs pa 'chad
par 'gyur pa'i skabs su rig par bya ....
120 See Dol po pa: "bDen gills gsal ba'i iii rna", 23, 11. 2-3: bden pa gfiis ni no bo de
fiid dan gian du brjod du med pa gcig pa bkag pa 'i tha dad pa yin te I.
121 See also Broido (1989:88), who has made the same observation with regard to two
sets of skandhas in the Ri chos nes don rgya mtsho.
314 KLAUS-DIETER MATHES
enables the ultimate to manifest on the level of apparent truth), and in this
sense there are accumulations of merit and wisdom. Taranatha's restric-
tion of the perfect to its unchangeable aspect is thus perfectly in line with
the position of Dol po pa. .
Things become a bit more complicated in the case of Sakya mchog
ldan. Even later Tibetan thinkers had difficulties in pinning down his
position
l22
. To give an example, in his short presentation of the Buddha-
nature which was written in 1474
123
, Sakya mchog ldan endorses Bu ston
Rin chen grub's (1290-1364) and Sa skya paIJQ.ita's (1182-1251) Madhya-
maka hermeneutics
124
of ascribing a provisional meaning (neyiirtha)
to the tathiigatagarbha theory. But in difference no. 21 Sakya mchog
ldan is said to hold that the seeds of the fruit (Buddhahood) are naturally
present in the form of the natural luminosity of mind. This is strikingly
similar to the position of 'Gos Lo tsa ba gZon nu dpal (1392-1481), who
speaks of "subtle qualities" or "seeds of qualities" in the mind-stream
of sentient beings. By explaining a natural growth of qualities, gZon nu
dpal is able to read the Ratnagotravibhiiga as a teaching with definitive
meaning (nftiirtha), without being forced to accept the ontological con-
sequences of the tathiigatagarbha theoryl25. It is likely that Sakya mchog
ldan later adopted such a stance, Taranatha being right in this point,
but only a careful study of Sakya mchog ldan'sworks written after 1484
will tell.
The notion in points 1-4 that ran ston is more profound when mental
fabrications are cut through with the help of the view finds support in
Sakya mchog ldan's Dharmadharmatiivibhiiga commentary, in which
the commentator shows that the teaching of a transformation of the basis
(iisrayapariv[tti) does not contradict the Svatantrika and PrasaIigika
views 126.
122 For a short description of Sakya mchog ldan' s position see also Tillemans & Toma-
bechi (1995: 891-6).
123 Slikya mchog ldan: "Saris rgyas gyi sifui. po'i mam bsad mdo rgyud kyi sifui. po", 136,
1. 3.
124 See Seyfort Ruegg 1973:29-33.
125 See Mathes 2002:88-9. In his Ratnagotravibhiiga commentary, gZon nu dpal refers
to these qualities in sentient beings as "seeds" (see Mathes 2003: 121, 11. 6-7).
126 Slikya mchog Idan: "Byams chos ma'i lam gyi rim pa", 154,11. 1-7. See also Mathes
1996:176.
TARANATHA'S "TWENTY-ONE DIFFERENCES" 315
A comparison with the "Zab ii spros bral gyi Mad pa"127 shows that
Taranatha also got the main points of Siikya mchog Idan's definition of
the basis of emptiness and the negandum right. Tiiranatha's difference
no. 14, though, which has Siikya mchog ldan equating the emptiness of
the dependent from the imagined with the ultimate which is empty of the
apparent, is problematic. In fact, Siikya mchog ldan takes the dependent
to exist ultimately only when describing Cittamatra. Thus he says in his
"Sill rta srol gfiis mam dbye"128:
, ~
Outside objects and what is explained as general characteristics are the imag-
ined, and empty of an own-being. What appears as [the imagined] through
mental imprints is the dependent and truly established. Emptiness which [is
taken as] the basis of negation (the dependent), empty of the negandum (the
imagined), is the perfect. Therefore it is the ultimate truth
129

The Yogacaras, on the other hand, who according to Siikya mchog
Idan
130
belong to the Madhyamaka, are not said to claim the ultimate exis-
tence of the dependent nature. Ultimate truth is equated rather with the
unchangeable perfect nature:
The Yogaciiras explain: "The imagined is empty of an own-being, and the
dependent empty of an other-being. The remainder left over as something
which does not lack an own-being is precisely the nature of the dependent.
or the so-called perfect. [ ... ] When both Madhyamikas (i.e., the Yogaciiras
and the Asvabhavavadins) postulate what must be meditated upon or expe-
rienced in the meditative equipoise of the Noble Ones, their perception is
in mutual accordance: both explain it as the wisdom of dharmadhiitu. When
labelling it after rising from meditative equipoise, they differ: The Yogaciiras
label it [the experiential] wisdom of dharmadhiitu or non-dual wisdom,
which goes by the name "unchangeable perfect" - the actual ultimate truth,
127 I.e., the two passage (114, 11.3-4 & 114,1. 7 -115,1) quoted above in the first para-
graph of this paper.
12' According to Dreyfus (1997:29) this work was written in 1489.
129 Sakya mchog ldan: "Sill rta srol gfiis mam dbye", 476, II. 3-5: [sems tsam pa'i
mnam biag gi ltaba ni ... des drans pa'i rjes thob kyi grub mtha' niJ phyi Tal gyi don dan
spyi mtshan du b§ad pa rnamsni kun btags dan ran gi no bas stan pa'a II bags chags kyis
der snan ba ni gian dban dan bden par grub pa' a II dgag gii gian dban de dgag bya kun
btags kyis stan pa'i stan pa nid ni yans grub dan I des na don dam pa'i bden pa'a I.
130 The Y ogaciira treatises on the Maitreya works, for example, are taken to go beyond
Cittamatra and thus to be in accordance with the intention of the Madhyamaka. See Sakya
mchog ldan: "Byams chos ma'i nes don rab tu gsa! ba", 6, 11.3-7; and Mathes 1996:174.
316 KLAUS-DIETER MATHES
the supreme self, the permanent, stable, quiescent, steady and truly estab-
lished.l3l
In other words, Sakya mchog ldan restricts the ultimate truth in the
same way as the Jonangpas to the unchangeable perfect nature, which is
also equated with wisdom. Against the background of this passage, the
quintessence of Taranatha's comparison of Dol po pa with Siikya mchog
ldan, namely that they take wisdom to be resepectively per:manent and
impermanent, appears questionable. The main difference is rather that
Sakya mchog ldan does not define gzan stan as the ultimate being empty
of the apparent, but includes the dependent nature within the basis of
negation. This is also clear in the following passage from the "Zab ii
spros bral gyi Mad pa", where Sakya mchog ldan disagrees with a pop-
ular gzan stan position:
The apparent [truth], [comprising] all conditioned entities, is empty of an
own-being (raft stOli), while the ultimate, everything!32 unconditioned, is
empty of other (gian stoft). This explanation is the assertion of the great
Madhyamika Vasubandhu, for this is how it is explained in the Brhartfhi.
Such an explanation does not hold true, since it is not in accordance with
the basic Maitreya works, and contradicts the clear gian stoft teachings of the
indisputable works of Asanga and his brother as well as the text tradition
of Dignaga and his disciple133.
131 Sakya mchog ldan: "Sill rta sro! gfiis mam dbye", 483, 1. 7 -484,1. 4: mal 'byor
spyod pa pas ni / kun btags ran gi no bas stan pa dan / gzan dban gzan gyi no bas stan
pa dan / de 'i suI du ran gi no bas mi stan par Ius pa ni / gzan dban gi no bo' am yons grub
ces bya ba de iiid do lZes 'chad f. .. ] dbu rna pa de giiis ka yan 'phags pa'i mfiam gzag
gis bsgom bya'am mfiam su myon bya zig khas len pa'i tshe nos 'dzin tshul ni mthun pa
yin te / gfiis kas kyan chos kyi dbyins kyi ye ses la 'chad pas so / mfiam gzag de las lans
pa'i rjes thob tu tha sfiad 'dogs tshul ni mi mthun pa yin te / mal 'byor spyod pa pas ni
'gyur ba med pa 'i yons grub ces bya ba'i min can / chos dbyins ye ses sam gfiis su med
pa'i ye ses de fiid don dam pa'i bden pa dnos dan / bdag dam pa dan / rtag brtan zi ba
g.yun drun dan / bden par grub pa iiid du tha sfiad 'dogs par byed la!.
132 The plura! particle shows that there is more than one unconditioned element.
133 Sakya mchog ldan: "Zab ii spros bra! gyi biad pa", 117,11.1-3: kun rdzob 'dus
byas kyi dnos po thams cad ran ston dan / don dam 'dus rna byas mams gzan ston du
'chad pa ., di ni dbu rna pa chen po dbyig gfien gyi bied pa yin te / yum gyi gnod 'jams las
de ltar Mad pa'i phyir / zes 'chad pa ni rigs pa rna yin te / / rje btsun gyi gzun rtsa ba
dan mi mthun zin / thogs med sku mched kyi gzun rtsod med mams dan / phyogs glan yab
sras kyi gzun lugs las gian ston gi 'chad tshul gsal par gsuns pa mams dan' gal ba 'i phyir
ro !.
TARANATHA'S "TWENTY-ONE DlFFERENCES" 317
Sakya mchog ldan continues by presenting his own definition of gian
stan based on the Madhyantavibhiiga, namely that the dependent is empty
of the imagined, and explains:
Just as in the lines: "False imagining [equated with the dependent nature]
exists. Duality is not found in it"134 the dependent is "phenomena" (dharmin),
or the basis of emptiness, and both [aspects of the] imagined, the perceived
object and the perceiving subject, are the neganda, or that of which [the
dependent] is empty. There is an of the wisdom beyond the dual-
ity of a perceived object and the perceiving subject as an entity which is
empty, but [the latter] is not taken as the subject, or the basis of emptiness 135.
It should be noted that for Sakya mchog ldan the dependent nature, or
false imagining, exists in terms of its own-being, specific marks and
its own nature, but not truly, on the level of ultimate truth, or in reality.
It is like an illusion
136
• This presentation is based on the reasonable inter-
pretation of the Y ogacara works as implying that the dependent nature
only exists on the level of apparent truth. Sakya mchog ldan is, of course,
well aware that in the Ratnagotravibhiiga and the B[ha{tfka the perfect
.. nature is taken to be empty of the imagined 137.
134 MA VBh, 17, L 16 (MA V Dab): abhiltaparikalpa 'sti dvayan tatra na vidyate i.
135 Sakya mchog Idan: "Zab ii spros bral gyi biad pa", 117, L 5: ji skad du / yan dag
ma yin kun rtag yad / / de la gnis po yad ma yin / ies 'byun ba ltar / gian dban stan gii'i
chas can dan / gzun 'dzin kun btags (text: brtags) gnis po gan gis stan pa'i dgag bya dan /
gzun 'dzin gnis med kyi ye ses la stan pa'i dnas par Mad pa ies bya ba zig yad pa yin gyi /
stan gzi'i chas can du 'jog pa ma yin te I.
136 Ibid., 115, L 3: de ltar na gian dban yan dag pa ma yin pa'i kun tu rtag pa zes bya
ba ran biin kyis yad pa dan / ran gi mtshan nid kyis dan / ran gi no bas yad pa ni yin la /
bden pa dan / don dam par dan / de kha na nid du yad pa ni ma yin te sgrub byed go rim
Min du / dgag bya dag gnis kyis stan pa'i phyir dan / don dam pa'i bden pa ma yin pa'i
phyir dan / de kha na nid ma yin pa'i phyir dan / sgyu ma biin no I.
137 Sakya mchog Idan: "Sin rta chen po'i srol gfiis mam dbye", 520, 11. 2-3: rgyud
bla ma dan yum gyi gnad 'jams su chos nid yans grub dgag bya kun btags kyis ston par
Mad pa' o. In other words, here again the dependent nature is not included in the negan-
dum. The Brha{flkii (Karmapa Tanjur (T6hoku no. 3808), ses phyin, pha, 572, 1. 5) supports
the Jonangpas, however, in that the ultimate, or the perfect nature, is defined as "that
which is free from these names (roughly referring to the imagined nature in the Brhattlkti)
and the forms of mistaken appearances (i.e., the dependent nature)," and that which is inef-
fable and the form of signlessness." (gan min dan / phyin ci log tu snan ba 'j mam pa de
dan bral ba brjod du med pa / mtshan ma med pa 'i mam pa gan yin pa de ni don dam pa
yans su grub pa'i no bo nid de I).
a The expression "forms of mistaken appearances" defInes the dependent nature: "The
forms, which, under the sway of ignorance and so forth, appear to the consciousness in a
318 KLAUS-DIETER MATHES
From what has been said till now, it is clear that the way Taranatha sum-
marizes Sakya mchog ldan's view on trisvabhava is not strictly accurate.
Even though the dependent nature is undoubtedly taken to be the basis
of negation, Siikya mchog ldan describes it as existing only in
the Cittamatra, but not in the Y ogacara. And it is the presentation of the
trisvabhava in the Yogacara which reflects his own gzan stan view. It is
also questionable whether wisdom is really only a conditioned entity for
him; as we have seen above, Siikya mchog ldan explains the unchange-
able (!) perfect nature as being non-dual wisdom.
4. The Theory of trisvabhiiva in the Madhyiintavibhiiga and Its Com-
mentaries
The trisvabhava theory of the Madhyantavibhaga plays an important
role not only for those gzan stan pas who define the basis of emptiness
in line with Yogacara, but also the Jonangpas, whose main doctrinal sup-
port otherwise is the tathagatagarbha theory. How is it possible, though,
that such different positions on emptiness can be doctrinally supported by
one and the same text?
The main focus for the proponents of both "Yogacara gzan stan" and
"Tathagatagarbha gzan stan" lies on the initial two stanzas of the first
chapter of the Madhyantavibhaga, in which the Middle Path is defined
by three philosophical propositions: (a) false imagining exists; (b) sub-
ject-object duality, though created by false imagining, is not found in
the latter itself; (c) false imagining is found in relation to emptiness in
. the sense that emptiness is found in false imagining as its true nature
138

mistaken way as phenomena, are the dependent nature." (rna rig pa la sags pa'i dban gis
mam par ses pa la chos mams su phyin ci log tu snan ba'i mam pa gan yin pa de ni gian
dban gi no bo fiid do I, ibid. 572, ll. 4-5).
138 While (a) and (b) are the padas MA V 1.1a and I.lb, proposition (c) reflects the
double locative relationship between false imagining and emptiness in the second part of
MA V 1.1 (But emptiness is found there (i.e., in false imagining) and [false imagining] is
found in relation to it (i.e., emptiness) as we/I). If the second locative (tasyam, i.e., siiny-
atayam) is taken literally in the sense that x is found in y, and y in x, x would be y. Total
identity, however, of false imagining and emptiness can be excluded on the grounds that
the fIrst is considered to be conditioned and the latter not (cf. MA VBh on I.2). I therefore
suggest the preliminary translation "in relation to" for the two locatives.
TARANATHA'S "TWENTY-ONE DIFFERENCES" 319
One has to bear in mind that the root text, which does not make much
use of trisvabhiiva 'terms in the initial stanzas, equates the perceived object
with the imagined nature, false imagining with the dependent nature, and
the absence of duality, or emptiness, with the perfect nature in MA V I.S.
As I have already noted in my paper on Taranatha's "gZan ston sfiiIi
po" 139, the relationship between false imagining and emptiness can be
variously defined along the lines of two different trisvabhiiva models, in
. the Madhyiintavibhiiga and its Indian commentaries. The central focus
of the fIrst model, which is mainly based on the first section of the first
chapter (MAV I.l-11), lies on a false imagining or dependent nature
which at times is taken to exist ultimately, though not by Maitreya and
Vasubandhu. Duality and emptiness are just two different aspects of
false imagining, namely the way it appears and the way it really is. In the
second section (MA V 1.12-22) a positively understood emptiness (com-
parable to suchness or the Buddha-element in the RGV) replaces false
imagining at the centre of the old equation. It is now emptiness, defined
as natural luminosity, which can appear in two modes, either as being
. accompanied by adventitious stains (under which false imagining is
included) or free from these stains (see below). This results in two tris-
vabhiiva models which come close to what Sponberg (1981 :99) calls the
pivotal and progressive exegetical model of trisvabhiiva. The first model
is centred on the dependent nature as a bearer of the perfect, which lat-
ter is understood as something abstract, like the state of suffering or imper-
manence. In the progressive model the focus lies more on an emptiness
which pervades or transcends all phenomena of the dependent nature.
This all-pervading emptiness possesses positive qualities and can exist,
contrary to the fIrst model, in its own right. The three natures represent
three levels, each revealing a progressively deeper degree of reality140,
This leads to the question whether the Madhyiintavibhiiga takes the
dependent nature as existing on the level of ultimate truth 141, One might
139 Mathes 2000: 195-223.
140 See Mathes 2000:204-14.
141 This is what TSOIi kha pa (1357-1419), for example, claims on the basis ofMA VT I.1,
where the verse abhiltaparikalpo 'sti is glossed as svabhavatal:z. A little further down Sthi-
ramati does not object to an opponent's claim of its ultimate existence: "[Opp.:] If thus
duality was entirely non-existent, like a hare's horn, and false imagining existed ultimately
320 KLAUS-DIETER MATHES
argue that the Y ogacara does not distinguish existence on two levels of
truths, its trisvabhava theory being rather an alternative to the apparent
and ultimate truths of the Madhyamaka
142
• Many passages in the Madhy-
antavibhiiga support this. This becomes particularly evident in the third
chapter (on reality), where older concepts relating to truth/reality, such
as the four noble truths of early Buddhism or the apparent and ultimate
truth of the Madhyamaka, are explained in terms of the new trisvabhava.
Even the noble truth of cessation is subsumed under the scheme of the
imagined, dependent and perfect natures. A continuity between main-
stream Buddhist thought and Y ogacara is thereby established. It is note-
worthy, however, that in the case of the ultimate truth of the Madhyamaka
only the perfect nature is accepted as a fit candidate for it, the dependent
nature, or false imagining, being dismissed as something to be ultimately
given up. If one applies this to the definition of the madhyama pratipat
in MA V 1.1-2, it would be safe to say that the propositions "the existence
of false imagining" (MA V l.la) and "the non-existence of duality" (MA V
I.1b) refer to the level of apparent truth, while "the mutual existence of
false imagining and emptiness" (MA V LIed) defInes the relation between
apparent and ultimate truth. Resorting to two levels of truth not only
explains the initial stanzas in a meaningful way, but also resolves some of
the tensions between the two parallel trisvabhiiva models mentioned above.
And this is exactly what did when he explained the theory of
trisvabhava in terms of his favoured Yogacara-Svatantrika-Madhyamaka
143

The first chapter of the Madhyantavibhaga is divided into two sections,
one on false imagining and the other on emptiness. While the latter is in
perfect harmony with the Ratnagotravibhaga, the former seems to draw
on older strands of more conservative Y ogacara material. Vasubandhu
(and to some extent also the author of the root text) nevertheless man-
aged to harmonize the originally unbalanced strands. In MA V l.1 false
imagining and emptiness are said to mutually exist in each other, and
based on this Vasubandhu defines emptiness in his as
in its own right. ... " (Sanskrit in brackets reconstructed: [yadi evam dva jyam
sarvathii nasti / abhUtaparikalpas ca svabhiivato 'sty ... , MA Vr, 10,11. 17-9).
See Thurman 1989:226-8.
142 See Boquist 1993: 17-22.
143 See Lindtner 1997: 193.
TARANA.THA'S "TWENTY-ONE DIFFERENCES" 321
"the state of this false imagining being free from the relation of a perceived
object and perceiving subject"l44.
Whereas emptiness is simply taken here as a property of the dominant
"false imagining", the latter hardly matters in the definition of emptiness
in the second part of the first chapter, where emptiness is not only the
absence of something in false imagining, but something more positive, the
own-being of non-duality, which is associated with positive attributes
such as the natural luminosity of the mind. In fact, in MA V 1.22 empti-
ness is defined in the same way as in the Ratnagotravibhaga:
[Emptiness is] neither defiled nor undefiled, neither pure nor impure.
(MA V I.22ab) How is it that it is neither defiled nor impure? It is because
of the natural luminosity of mind (MA V I.22c). How is it that it is neither
undefiled nor pure? It is because of the adventitious nature of defilements
(MA V I.22d)145.
It is obvious that the natural luminosity of the mind has taken the place
offalse imagining here
146
• That the latter cannot truly partake of the lumi-
nous nature is clear from a passage in the Siigaramatiparip[ccha quoted
in RGVV 1.68, in which the example of an ever-pure vai¢urya stone
drawn out from mud is taken to illustrate the relation between the lumi-
nous mind and accidental stains:
In the same way, 0 Sagaramati, the Bodhisattva knows the naturallumi-
nosity of the mind of sentient beings. He also perceives that it is defiled by
adventitious defilements. Then the Bodhisattva thinks as follows: These
defilements would never penetrate into the natural luminosity of the mind
of sentient beings. These adventitious defilements have sprung from false
imagining147.
144 MAVBh, 18, II. 2-3: sunyatii tasyiibhutaparikalpasya griihyagriihakabhiivena
virahitii.
145 MAVBh, 27, 11. 5-9: na klilf!ii niipi viiklilf!ii suddhii 'suddhii na caiva sii I katham
na klilftii niipi casuddhii I prak[tyaiva I prabhiisvaratviic cittasya I katham niiklilffii na sud-
dhii I kleiasyiigantukatvatalJ I.
146 What is defmed as all defilements (samklda) in MA V 1.10-1 can here only be the
adventitious defilements.
147 RGVV, 49, 11. 9-12: evam eva sagaramate bodhisattvalJ sattviiniim prak[tipra-
bhiisvaratiim cittasya prajiinati I tiim punar iigantukopakleSopaklilf!iim pasyati I tatra
bodhisattvasyaivam bhavati I naite kldiif:z sattviiniim cittaprak[tiprabhiisvaratiiyiim pravilftiif:z
I iigantuka ete kldii abhiltaparikalpasamutthitiif:z /.
322 KLAUS-DIETER MATHES
It is now luminosity which is centred on and occurs in two modes, one
of which is being stainless and thus even free from the false iinagining
which causes these adventitious stains. That the natural luminosity of the
mind refers to an originally pure nature of the mind in the Madhyantavi-
bhaga, too, becomes clear in stanza I.16, on the differentiation of emptiness:
How should the differentiation of emptiness be known? As being defiled as
well as pure (MA V I.16a). Thus is its differentiation. In which state is it defiled
and in which is it pure? It is accompanied as well as not accompanied by
stains. (MA V I.16b) When it occurs together with stains it is defiled, and
when its stains are abandoned it is pure. If, after being accompanied by
stains it becomes stainless, how is it then not impermanent, given that it has
the property of change? This is because its purity is considered to be like
the one of water, gold and space. (MA V I.16cd) [A change is admitted]
in view of the removal of adventitious stains, but there is no change in terms
of its own-being
I48

It should be noted how the terms "defiled" and "pure" of the first sec-
tion are explicitly equated with the iinported terminology "accompanied
by stains" and "stainless". The latter doubtlessly stem from the Ratna-
gotravibhaga, where the Buddha-nature is defined as suchness accompa-
nied by stains (samala tathata) and the transformation of the basis as
stainless suchness (nirmala tathata).
To sum up, the Madhyantavibhiiga combines the traditional Y ogacara
formula "the perfect is the dependent empty of the imagined" with strands
from the tathiigatagarbha theory, according to which an unconditioned
149
Buddha-element is empty of adventitious stains, but not of the inseparable
Buddha-qualities 150.
148 MA VBh 24, 11. 4-13: katham siinyataya/:t prabhedo jiieya/:t I samklilitii ca visuddhii
ca / ity asyii/:t prabheda/:t / kasyam avasthiiyam samklilita kasyam visuddhii I samalii nir-
malii ca sii / yada saha malena varttate tada / yada prah17:zamala tada visuddhii I
yadi samala bhiitva nirmala bhavati katham vikaradharmilJftvad anitya na bhavati / yas-
mad asya/:t abdhiitukanakiikiisasuddhivac chuddir iliyate II agantukamalapagaman na tu
tasya/:t svabhiivanyatvam bhavati /.
149 See ROY 1.5-6, where Buddhahood is taken to be without beginning or end and thus
unconditioned (ROVV, 7, 1. 14-8,1. 1: "Buddhahood is unconditioned. [ ... ] As having
neither beginning, middle nor end by nature, it is unconditioned." asamsk[tam [. .. J bud-
dhatvam £ ... J 1/ anadimadhyanidhanaprak[tiatvad asamsk[tam I).
a Johnston reads -prak[ta-.
150 ROVV, 76, 11. 3-4 (ROY I.155): "The [Buddha]-element is empty of adventitious
[stains], which have the defining characteristic of being separable; but it is not empty of
TARANATHA'S "TWENTY-ONE DIFFERENCES" 323
-Ilowever one wishes to combine these two formulas, a consistent read-
ing of the MadhYiintavibhiiga requires, as I already pointed out in my
paper on Taranatha's "gZan sfiiIi po", operating with the Ma!ihya-
maka distinction of two truths, and following MA V III. I 0 in accepting
. only the perfect nature as the ultimate truth. In doing so, one should
not overlook the fact of two models of trisvabhiiva which reflect varied,
hot yet completely harmonized strands of thought. In this respect, the
Madhyiintavibhiiga does not differ from other texts of the early Yogacara
school in not only drawing on early Mahayana thought but also featuring
a rich background of Abhidharma analysis. Sthiramati's uncertainty about
'the ontological status of false imagining may thus reflect the Abhidharmic
background of this early Yogacara material. Thus, it is generally asserted
in the HInayana schools that conditioned, dependently arising entities
really exist
151
• On the other hand, such a stance would of course be incom-
ilatible with a Madhyamaka understanding of the Yogacara, which is at
least attempted in some passages.
is. Conclusion'
Both Sakya mchog ldan and Dol po pa profit from the tensions between
. different trisvabhiiva models in the pertinent passages of the Madhyiinta-
vibhiiga and its commentaries, and follow the exegetical solution by
restricting the ontological status of false imagining to the level of appar-
ent truth. But from this point onwards the two masters depart from each
other. Sakya mchog ldan remains more faithful to the Yogacara, in tak-
ing the dependent nature as being empty of the imagined. What remains
in emptiness is thus not only an unchangeable perfect nature, but also the
perfect in terms of being unmistaken. This is similar to Rail byuIi rdo
unsurpassable qualities, which have the defining characteristic of not being separable."
(siinya iigantukair dhiituly. / a§unyo 'nuttarail; dharmair avinirbhii-
/1).
151 See v. Rospatt (1995:69ff.), who observes that in the early Yogaciira the contradiction
between Abhidhanna and Mahayana ontology was solved by more or less incorporating
the doctrine of the existence of momentary caused entities into the description of the depend-
ent nature. The Mahayana stance that the momentariness of the dharmas means nothing
other than their mere non-existence could then be comfortably brought into line with the
imagined nature of the trisvabhiiva doctrine.
324 KLAUS-DIETER MATHES
rje's "mere appearance", which corresponds to the apparent truth included
in the Buddha-nature. Following the Y ogacara definition of emptiness in
such a way, the Ratnagotravibhiiga must be interpreted in t e r m ~ of a Bud-
dha-nature which is inside time and thus consists of moments. This allows
for a theory of seeds which naturally grow into the qualities of a Buddha.
For Sakya mchog ldan, the basis of emptiness is thus not the ultimate
truth alone. In other words, his Yogacara-based gian ston is not defined
along the lines of an ultimate being empty of the apparent. .
Dol po pa, on the other hand, follows more the Ratnagotravibhiiga when
defining his gian ston: an unconditioned Buddha-element interpreted as
being completely transcendent (beyond the world and time)lS2 is taken to
be empty of adventitious stains. Such a tathiigatagarbha-based gian ston
requires reinterpreting the trisvabhiiva theory by taking a perfect nature
restricted to its unchangeable aspect as the basis of negation. Given the
Ratnagotravibhiiga elements in the Madhyiintavibhiiga, such an interpre-
tation is not completely out of question. One could argue in support of
Dol po pa, that Sakya mchog ldan's gian ston interpretation of the first
part of the first chapter in the Madhyiintavibhiiga which is centered on
false imagining or the dependent nature defInes in a first step the empti-
ness of the correct apparent from the false apparent. From that one still has
to go one step further, though, and explain the emptiness of the ultimate
from the correct (and false) apparent in line with the Ratnagotravibhiiga.
Bibliography
INDIAN TEXTS
B[hatrika (Tibetan translation)
Quoted from the Kannapa Tanjur (= Tohoku no. 3808). Rumtek/Delhi 198?
(no date).
MA V: MadhyantavibhCiga. See MadhyantavibhCigabhCi.rya
MA VBh: MadhyantavibhCigabhCi.rya
Ed. by Gadjin M. Nagao. Tokyo: Suzuki Research Foundation, 1964.
152 It is not the case though, that all parts of the Ratnagotravibhiiga explain Buddha-
nature or Buddha-hood in such a way. Thus the explanations of the three kayas in the second
chapter rather suggest that the latter constantly remain in samsiira - and thus in time -
as long as sentient beings need help (see Takasaki 1966: 331-5).
TARANATHA'S "TWENTY-ONE DIFFERENCES"
MA VT: Madhy{mtavibhiiga{lkii
Ed. by S. Yamaguchi. Nagoya: Librairie Hajinkaku, 1934.
See also NGMPP reel no. A 38/10.
RGV: Ratnagotravibhiiga Mahayiinottaratantrasiistra
325
Ed. by Edward H. Johnston. Patna: The Bihar Research Society, 1950. (fucludes
the Ratnagotravibhiigavyiikhyii)
RGVV: Ratnagotravibhiigavyiikhyii. See Ratnagotravibhiiga
TIBETAN TEXTS
Karma 'Phrin las pa
"Dris Ian yid kyi mUll sel ies bya ba Icags mo'i dris Ian biugs so". The Songs
of Esoteric Practice (mGur) and Replies to Doctrinal Questions (Dris Ian)
of Karma 'Phrin las pa, 88-92. Reproduced from prints of the 1539 Rin chen
ri bo blocks. New Delhi: Ngawang Topgay 1975.
Karma pa Ran byun rdo rje (the Third Karmapa)
dBu ma chos dbyins bstod pa'i mam par bsad pa biugs so. 52 fols., dbu med,
unpublished.
Zab mo nan gi don ies bya ba'i giun biugs so (block print). Published
together with the rNam ses ye ses 'byed pa and the bDe bar bsegs pa'i siiin
po bstan pa. Rumtek Monastery: 1970.
Ran 'grel: Zab mo nan gi don gsal bar byed pa'i 'grel pa biugs so (block
print). No place, no date. (The work itself was composed at the 0 rgyan kyi
mkhan po padma 'byuil gnas kyi sgrub gnas in 1325 (fo1. 92b6)).
Kon sprul BIo gros mtha' yas:
rGyud bla ma'i bsad srol: Theg pa chen po rgyud bla ma'j bstan bcos siiin
po'i don mnon sum lam gyi Mad srol dan sbyar ba'i mampar 'grel paphyir
mi ldog pa sen ge'i na ro ies bya ba biugs so. Rumtek Monastery: no date.
Zab mo nan gi don gyi 'grel pa: rNal 'byor bia na med pa'i rgyud sde rgya
mtsho siiin po bsdus pa zab mo nan gi don iiun nu'i tshig gis mam par 'grel
ba zab don snan byed. Rumtek Monastery: 1970.
Ses bya kun khyab mdzod. Beijing: Mi rigs dpe skrun khan, 1982.
'Gos Lo tsa ba gZon nu dpal
Theg pa chen po rgyud bla ma'i bstan bcos kyi 'grei Mad de kho na iiid rab
tu gsal ba'i me Ion. Ed. by Klaus-Dieter Mathes (Nepal Research Centre Pub-
lications 24). Stuttgart: Franz Steiner Verlag, 2003.
Taranatha:
. "gZan ston sfiiil po", rJe btsun tiiraniitha'i gsun 'bum biugs so, vol. 4, 491-514.
Collected Works. Leh: Namgyal and Tsewang Tarn, 1982-5.
~ "Zab don fier gcig pa bZugs so", op. cit., vol. 4, 781-95
Dol po pa Ses rab rgyal mtshan:
"bKa bsdu bii pa'i don gtan tshigs chen po", Kun mkhyen dol po'i gsun 'bum, .
Vol. ka, 363-418. Delhi: Jamyang Khyentse, 1984.
Jo nan ri chos nes don rgya mtsho. Beijing: Mi rigs dpe skrun khan, 1998.
326 . KLAUS-DIETER MATHES
[Ni ma'i 'od zer:] "Theg pa chen po rgyud bla ma'i bstan bcos legs Mad
iii ma'i 'od zer". The 'Dzam-thang Edition of the Collected Works of Kun-
mkhyen Dol-po-pa Shes-rab rgyal-mtshan, vol. 4 (rna), 883-1161. Delhi:
Shedrup Books, 1992.
"bDen giiis gsal ba'i iii ma", Kun mkhyen dol po pa'i gsun 'bum, 1-45.
Published by Jamyang Khyentse. Kathmandu: Shechen Publications, no date.
'Ba' ra ba rGyal mtshan dpal bzail
"Chos rje mam gills kyi dgons Mad iii ma'i 'od zer". A TIbetan Encyclopedia
of Buddhist Scholasticism. The Collected Writings of 'Ba' ra bq rgyal mtshan
dpal bzan, vol. 11,496-557. Dehra Dun: Ngawang Gyaltsen and Ngawang
Lungtok, 1970.
Sa bzang Mati paIJ. chen 'Jam dbyangs Blo gros rgyal mtshan
"Theg pa chen po'i rgyud bla ma'i bstan bcos kyi mam par Mad pa nes don
rab gsal snail ba". Sa skyapa'i mkhas pa rnams kyi gsun skor, vol. 4,1-520.
Kathmandu: Khenpo Abbey, 1999.
gSer mdog paIJ. chen Sakya mchog ldan
"Byams chos ma'i nes don rab tu gsal ba ies bya ba'i bstan bcos biugs so",
gSer mdog pan chen siikya mchog ldan gyi gsun 'bum legs bsad gser gyi bdud
rtsi glegs bam, vol. 11, 1-38. Collected Works. Thimphu: Kunzang Topgey,
1975.
"Byams chos ma'i lam gyi rim pa gsal bar byed pa'i bstan bcos rin chen sgron
gyi sgo 'byed ces bya ba bzugs so", op. cit., vol. 11,39-155
"Zab ii spros bral gyi biad pa ston fiid bdud rtsi'i lam po che zes bya ba biugs
so", op. cit., vol. 4, 107-207 .
"SiD. rta chen po'i srol giiis kyi mam par dbye ba Mad nas nes don gdg tu
sgrub pa'i bstan bcos kyi rgyas 'grel biugs so", op. cit., vol. 2, 471-619.
"Sails rgyas gyi sfiiiJ. po'i mam Mad mdo rgyud kyi sfiiiJ. po", op. cit., vol. 13,
124-136.
OTHER WORKS
Boquist, Me (1993). Trisvabhiiva: A Study of the Three-nature-theory in Yogiiciira
Buddhism. Studies in African and Asian Religions 8. Lund: Department of
History of Religion, University of Lund. .
Broido, Michael M. (1989). "The Jo-nang-pas on Madhyamaka: A Sketch". TIbet
Joumal14, no. 1,86-91.
Dreyfus, Georges B.J. (1997). Recognizing Reality (SUNY Series in Buddhist
Studies). New York: SUNY
Hookham, Susan K. (1991). The Buddha Within (SUNY Series in Buddhist Stud-
ies). New York: SUNY.
Kapstein, Matthew T; (1992). "Introduction", in The 'Dzam-thang Edition of the
Collected Works of Kun-mkhyen Dol-po-pa Shes-rab rgyal-mtshan. Delhi:
Shedrup Books.
TARAN.ATHA'S "TWENTY-ONE DIFFERENCES" 327
(2000). The Tibetan Assimilation of Buddhism: Conversion, Contestation and
Memory. Oxford: Oxford University Press.
Lindtner, Christian (1997). "Cittamatra in Indian Mahayana until KamalasIla",
WZKS 41, 159-206.
Makransky, John J. (1997). Buddhahood Embodied. New York: SUNY.
Mathes, Klaus-Dieter (1996). Unterscheidung der Gegebenheiten von ihrem
wahren Wesen (Dharmadharmatavibhtiga) (Indica et Tibetica 26). Swisttal-
Odendorf: Indica et Tibetica Verlag .
...,-. (1998). "Vordergriindige und hochste Wahrheit im gZan ston-Madhyamaka".
Anniiherung an das Fremde. XXVI. Deutscher Orientalistentag vom 25. bis
29.9. in Leipzig. Ed. by H. Preissler and H. Stein. Zeitschrijt der Deutschen
Morgenliindischen Gesellschaft 11, 457-468.
_. (2000). "Taranatha's Presentation of trisvabhtiva in the gZan ston sfiin po".
JIABS 23, no. 2, 195-223.
-. (2002). "'Gos Lo tsa ba gZon nu dpal' s Extensive Commentary on and Study
of the Ratnagotravibhtigavyakhya". Religion and Secular Culture in Tibet,
79-96. Tibetan Studies n. Proceedings of the International Association of
Tibetan Studies 2000, vol. 2/2, 79-96. Ed. by H. Blezer with the assistance
of A. Zadoks. Brill's Tibetan Studies Library. Leiden: Brill.
-. (2003). See 'Gos Lo tsa ba gZon nu dpal: Theg pa chen po rgyud bla ma'i
bstan bcos kyi 'grel Mad de kho na fiid rab tu gsal ba'i me Ion.
Newland, Guy (1992). The Two Truths in the Mtidhyamika Philosophy of the Ge-
luk-ba Order of Tibetan Buddhism. Ithaca, N.Y.: Snow Lion Publications.
v. Rospatt, Alexander (1995). The Buddhist Doctrine of Momentariness. Alt- und
Neu-Indische Studien 47. Stuttgart: Franz Steiner Verlag.
Seyfort Ruegg, David (1973). Le Traite du Tathtigatagarbha de Bu ston rin chen
grub (Publications de I'Ecole d'Extreme-Orient 88). Paris: Ecole
d 'Extreme-Orient.
Smith, E. Gene (1970). "Introduction", in Kongtrul's Encyclopaedia of Indo-
Tibetan Culture, vol. 1, 1-28. Ed. by Lokesh Chandra (Sata-Pitaka-Series 80).
New Delhi: International Academy of Indian Culture.
Sponberg, Alan (1981). "The Trisvabhava Doctrine in India & China: A Study
of Three Exegetical Models", Bukkyo bunka kenkyujo kiyo 21 (1981),97-119.
Steams, Cyrus (1995). "Dol-po-pa Ses-rab rgyal-mtshan and the Genesis of the
gian ston Position in Tibet". Asiatische Studien 49, no. 4, 829-852.
-. (1999). The Buddha from Dolpo: A Study of the Life and Thought of the Tibetan
Master Dolpopa Sherab Gyaltsen (SUNY series in Buddhist Studies). New
York: SUNY.
Takasaki, Jikido (1966). A Study on the Ratnagotravibhtiga (Uttaratantra) Being
a Treatise on the Tathtigatagarbha Theory of Mahayana Buddhism (Rome
Oriental Series 33). Rome: Istituto Italiano per il Medio ed Estremo Oriente.
Thurman, Robert, A.F. (1989). The Speech of Gold: Reason and Enlightenment
in the Tibetan Buddhism. Delhi: Motilal Banarsidass.
328 KLAUS-DIETER MATHES
Tillemans T. and Tomabechi T. (1995). "Le dbu ma'i byuli tshut de Sakya mchog
ldan", Asiatische Studien 49, no. 4, 891-918.
Tsultrim Gyamtsho Rinpoche & Fuchs, Rosemarie (2000). Buddha Nature: The
Mahayana Uttaratantra Shastra with Commentary. Ithaca, N.Y. : Snow Lion
Publications.
Zhang, Yfsiln (1985). Bod rgya tshig mdzod chen mo: Zing-Han Dacfdian. 3 vols.
Published by Zhang Yfsiln. Beijing: Mi rigs dpe skrun khaIi..
IS THERE REALLY "ESOTERIC" BUDDHISM?
RICHARD D. MCBRIDE, II
In his seminal book The Order of Things, Michel Foucault makes an
important observation on a methodological problem associated with clas-
sification:
Historians want to write histories of biology in the eighteenth century; but
they do not realize that biology did not exist then, and that the pattern of
knowledge that has been familiar to us for a hundred and fIfty years is not
valid for a previous period. And that, if biology was unknown, there was a
very simple reason for it: that life itself did not exist. All that existed was
living beings, which were viewed through a grid of knowledge constituted
by natural history.!
In contemporary Western scholarship, the concept of "Esoteric Bud-
dhism" has become part of a three-fold grid of knowledge deployed to
describe the history of Buddhism. For instance, The Encyclopedia of Reli-
gion presents three general essays titled: "Hfnayana," "Mahayana," and
"Esoteric Buddhism."2 In this case the classification "Esoteric Buddhism"
is clearly a euphemism and replacement term for "Buddhist Tantrism"
or "Tantric Buddhism," a problematic classification repeatedly shown
by scholars in recent years to be largely a product of nineteenth-century
Western Orientalist imagination.
3
For this reason, perhaps, many scholars
now favor the category of "Esoteric Buddhism." It also may be because
1 Michel Foucault, The Order of Things: An Archeology of the Human Sciences [A trans-
lation of Les mots et les choses] (New York: Pantheon Books, 1970; rpt. New York: Van-
tage Books, 1994), 127-128.
2 See the book Buddhism and Asian History, ed. Joseph M. Kitagawa and Mark D.
Cummings (New York: Macmillan, 1989), 195-256, which is comprised of selections from
The Encyclopedia of Religion, Mircea Eliade, editor in chief, 15 vols. (New York: Macmil-
lan, 1987).
3 See, for instance, Hugh B. Urban, "The Extreme Orient: The Construction of 'Tantrism'
as a Category in the Orientalist Imagination," Religion 29 (1999): 123-146; and Donald
S. Lopez, Jr., Elaborations on Emptiness: Uses of the Heart Satra (Princeton: Princeton
University Press, 1996), 78-104.
Journal of the International Association of Buddhist Studies
Volume 27 • Number 2 • 2004
330 RICHARD D. MCBRIDE, II
the classification suggests a relationship with the Sanskrit word guhya
(secret, esoteric) and implies that this form of Buddhism was taught
secretly and is for and understoo,d by the initiated only. Many scholars
in Asia and the West are confident that there was a separate "Esoteric
Buddhism" that was known by various names in East Asia and that it is
related directly to the teachings of three masters: SubhakarasiIpha (Shan-
wuwei 637-735), Vajrabodhi (Jin'gangzhi 671-741), and
Amoghavajra (Bukong 705-774).4 However, this classification is
equally problematic because the interpretation of the category or classifi-
cation "Esoteric Buddhism" (Jap. mikkyi5 *t3l:) may in reality be a product
of Japanese sectarian Buddhism, the influence of which on Western schol-
arship on Buddhism cannot be understated.
My purpose here, however, is not another Foucaultian deconstruction
of a problematic scholarly category. Bob Sharf's essay "On Esoteric Bud-
dhism in China" accomplishes this purpose nicely,S My aim here is to explore,
in a more nuanced way, how Buddhists in the Sinitic cultural sphere
from the fifth to the eighth centuries C.E. and beyond, including some
figures whom historians want to categorize as the earliest "Esoteric" or
4 The classic example is Omura Seigai :k#irn'L:!!, Mikkyo hattatsushi (History
of the Development of Esoteric Buddhism) (Tokyo: 19l8;!pt. Tokyo: Daito Shuppansha,
1972). For a discussion of Omura's work and other Japanese scholarship on Esoteric Bud-
dhism see Robert H. Sharf, "On Esoteric Buddhism," in Coming to Terms with Chinese
Buddhism: A Reading of the Treasure Store Treatise (Honolulu: University of Hawaii
Press, 2002), 263-266. See also Taganoo Shoun Himitsu Bukkyoshi
(History of Esoteric Buddhism) (1933; Rpt. Tokyo: Ryfibunkan, 1981); Chou Yi-liang,
"Tantrism in China," Harvard Journal of Asiatic Studies 8 (1945): 241-332; Lti Jianfu
§M!Wrll, Zhongguo Mijiaoshi (History of Esoteric Buddhism in China) (Beijing:
Zhongguo Shehui Kexue Chubanshe, 1995); Abe Ryfiichi, The Weaving of Mantra: Kilkai
and the Construction of Esoteric Buddhist Discourse (New York: Columbia University
Press, 1999); Charles D. Orzech, "Seeing Chen-yen Buddhism: Traditional Scholarship
and the Vajrayana in China," History of Religions 29/2 (1989): 87-114; "MaI,1qalas on the
Move: Reflections from Chinese Esoteric Buddhism Circa 800 C.E." Journal of the Inter-
national Association of Buddhist Studies 19/2 (1997): 209-244; or his Politics and Tran-
scendent Wisdom: The Scripture of Humane Kings in the Creation of Chinese Buddhism
(University Park, PA: The Pennsylvania State University Press, 1998),135-136 n. 1; and
Michel Strickmann, Mantras et Mandarins: Ie bouddhisme tantrique en Chine (paris: Edi-
tions Gallimard, 1996) and Chinese Magical Medicine (Staoford: Staoford University Press,
2002).
5 See Robert H. Sharf, Coming to Terms with Chinese Buddhism (Honolulu: University
of Hawai'i Press, 2002), 263-278.
IS THERE REALLY "ESOTERIC" BUDDIDSM? 331
"Tantric" Buddhists in China, deployed the ideas of "esoteric teaching"
(mijiao and its mate "exoteric teaching" (xianjiao and related
concepts to describe and classify Buddhist teachings.
The grid of knowledge established by intellectual Buddhist mollks in
medieval China was openly biased toward the Mahayana. Based on the
rhetoric of Mahayana scriptures, such as the Lotus Sutra, they understood
the teachings of Buddhism as being comprised of three vehicles: 1) the
Sravakayana (shengwensheng the vehicle of the disciples; 2) the
Pratyekabuddhayana (bizhijosheng .!l¥:SZ:fll!l*), the vehicle of the solitary
buddha; and 3) the Bodhisattvayana (pusasheng ifill*), the vehicle of
the bodhisattvas. The first two vehicles (ersheng =*) were conceptual-
ized as inferior; hence, they were labeled with the pejorative title Hina-
yana, the Lesser Vehicle (xiaosheng IJ\*). The vehicle of the bodhisattvas
was conceived of as superior; hence it enjoyed the designation Mahayana,
the Greater Vehicle (dasheng **).6 This polemical dualism presents an
interesting irony since the prevailing mode of doctrinal discourse projects
a view of reality that is ultimately non-dual or indivisible. Although the
. duality is ultimately transcended, it is still fundamentally polemic. The idea
of ultimate non-duality is projected onto these differentiated Buddhist
teachings through the concept of the One Vehicle (yisheng -*, Skt.
Ekayana) or the Buddha-vehicle (josheng fll!l*, Skt. Buddhayana), which,
though not really different than the bodhisattva vehicle mentioned above
portrays the Mahayana as subsuming, comprehending, and transcending
the Hinayana. The most famous explication of this approach to the Bud-
dhist teachings is the famous "Parable of the Burning House" in the Lotus
Sutra'? Furthermore, other Mahiiyana sutras proclaimed the superiority
of the Mahayana in ways that influenced Sinitic Buddhist exegetes' concep-
tualizations of the development of the Buddhist doctrines. The Sa1[ldhi-
nirmocana Sutra, for instance, explains that the Buddha "turned the wheel
of the dharma" (zhuan falun three times: the first being the Hina-
yana teaching of the Four Noble Truths in Deer Park, the second being
6 See, for instance, Miaofa lianhuajing 1, T 262, 9.8a, fasc. 2, T 262, 9.18b;
cf. Leon Hurvitz, trans., Scripture on the Lotus Blossom of the Fine Dharma (The Lotus
Sutra) (New York: Columbia University Press, 1976),34,95.
7 See, for instance, Miaofa lianhuajing 2, T 262, 9.13c; cf. Hurvitz, trans., Scripture
on the Lotus Blossom of the Fine Dharma (The Lotus Sutra), 63-64.
332 RICHARD D. MCBRIDE, II
the early Mahayana teaching of "emptiness" (kong @, Skt. sunyatd) of
the Prajfiaparamita sfitras, and the third and final teaching being the
advanced Mahayana teaching that "all dharmas lack substantial marks
(tixiang 1l1if'§, are neither produced nor destroyed but
are in quiescence and that their self-nature is nirvliI).a."8
The concepts of "esoteric" and "exoteric" must be understood as func-
tioning within this polemical context. They refer not only to the ideas of
being secret, hidden, or concealed versus being explicit, evident, or man-
ifest, but also implicitly to the inherent superiority of the Mahayana teach-
ings to the Hfuayana. However, in the second sense, esoteric also refers
to a fundamentally transcendent kind of knowledge that represents the
bodhisattva's comprehension of ultimate reality, the emptiness of all
dharmas, their fundamental lack of self-nature and marks and their orig-
inal quiescence - "the acquiescence to the non-production of dharmas"
(wushengfa ren Skt. - but also
that the Buddha employed skillful means (updya) to lead aspirants to
understand the esoteric teaching. In other words, esoteric teachings are,
by definition, advanced Mahayana teachings suited to bodhisattvas.
Northern and Southern Dynasties Period (317-589)
The most explicit examples that employ the polemical rhetoric that
the Mahayana is esoteric and the Hfuayana is exoteric are found in the
single most important document for understanding Buddhism in medieval
China: The Treatise on the Great Perfection of Wisdom (Dazhidu lun
*'&.§tffilll, T 1509).10 There is nothing in Indian Mahayana literature that
8 Shenmi jietuo jing 2, T 675, 16.673c; cf. John Powers, trans., Wisdom of
Buddha: The Sarrzdhinirmocana Mahayana Sutra (Berkeley: Dharma Publishing, 1995),
138-141. See also Paul Williams, Buddhist Thought: A Complete Introduction to the Indian
Tradition (London: Routledge, 2000), 153-154.
9 Obtaining the "acquiescence to the non-production of dharmas" (wushengfa Ten
i!\li§:.$;'2.) is the phrase commonly used in the Mahayana teachings to reflect an adherent's
awakening to the ultimate truth of reality, the way things really are; see Weimoji suoshuo
jing 1, T 475, 14.539a, 540c; fasc 2, 14.546a.
10 See Etienne Lamotte, trans., Le traite de la grande vertu de sagesse de Nagarjuna
(Mahiiprajfiaparamitasastra), 5 vols. (Louvain: Institut orientaliste, Universite de Louvain,
1944-1981). On many different names by which this text was known in medieval China
IS THERE REALLY "ESOTERIC" BUDDHISM? 333
remotely approaches the authority this work enjoyed in medieval Sinitic
Buddhism. It is a' large compendium of Mahayana views and practices
attributed to the monk-scholar Nagarjuna (Longshu ca. 150-200).u
It was translated into Chinese between 402 and 406 by Kumarajlva (Jiu-
moluoshi N,W;J**i1t, 344-413), the famous Central Asian translator and expli-
cator of Buddhism to the Chinese and founder of Madhyamaka philosophy
in China. 12 The recent dissertation of Chou Po-kan presents a strong case
for a "partly Chinese" authorship of the work, since the hand of Kumara-
jlva's editor and scribe Sengrui 11f1J:Z (352-436) can be seen in the trans-
lation and because some subjects treated by Kumarajlva appear to be
responses to questions by Sengrui and the project's sponsor Yao Xing
(365-416), sovereign of the Later Qin dynasty.13 Nagarjuna's views
on the ideas of esoteric and exoteric teachings provide the original context
for the discussion of this issue in medieval Sinitic Buddhist exegesis:
and on the attribution of the text to Nagarjuna see Paul Demieville review of the second
volume of Lamotte's translation (originally published in 1950), in Choix d'etudes boud-
diques (1929-1970) (Leiden: E. J. Brill, 1973), 470, n. 1,475-476.
11 There is a great debate as to whether Nagilljuna actually existed or whether he is a
literary creation concocted by Mahayana writers. This is unimportant to our discussion
because he existed to the Chinese. In India Nagarjuna is referred to variously as the author
of one or another particular essay. However, in China, when a Buddhist exegete says
"Nagarjuna" he is alluding almost invariably to the Dazhidu lun. For the problem of
Nagarjuna's existence and dating in Indian literature see Joseph Walser, "Nagarjuna and
the Ratnavalf: New Ways to Date an Old Philosopher," Journal of the 1ntemationalAsso-
dation of Buddhist Studies 25/1-2 (2002): 209-262.
12 For the biography of Kumara]1va see Gaoseng zhuan jljij{jf-f,\!): 2, T 2059, 50.330a-333a;
see also Kenneth Ch'en, Buddhism in China: A Historical Survey (Princeton: Princeton Uni-
versity Press, 1964),81-83.
13 Some of the most notable evidence provided by Chou is that the Dazhidu [un's com-
mentary on the Mahiiprajfiaparamita Satra follows Chinese word order rather than Indian
and that the whole of the commentary is in the form of a dialogue. Dialogue was not only
commonly employed in Sarvastivadin commentarialliterature, with which Kumarajlva was
familiar, but also in contemporary Chinese "Neo-Daoism." (This is a misleading transla-
tion of xuanxue Z"0l'l, "dark learning" or "learning of the arcane/mysterious," which is to
be preferred.) Questions appear to be written into the text and answered as the text pro-
ceeds. Furthermore, Sengrui appears to have written down everything that Kumarajlva
said and perhaps, due to other involvements, did not edit out old translations of technical
terms; hence, both old and new Buddhist terms remain in the Dazhidu lun. Thus, the
Dazhidu [un seems to reflect the work-in-progress nature of this translation. See Chou Po-
kan, "The Translation of the Dazhidulun: Buddhist Evolution in China in the Early Fifth
Century" (ph.D. dissertation, University of Chicago, 2000), 62, 68,74-77,78,80,81-84.
I would like to thank James Benn for referring me to this recent dissertation.
334 RICHARD D. MCBRIDE, II
There are two ldnds of Buddhadharrna (fofa 1) esoteric (mimi
and 2) exoteric (xianshi :ij/,7f-;). In the exoteric [form], the Buddha, pratyeka-
buddha, and arhat are' all fields of merit since their defilements have been
exhausted without residue. In the esoteric [form], it is explained that bod-
hisattvas obtain the acquiescence to the non-production of dharmas, the
absolute cessation of defIlements, and all of the six spiritual penetrations (liu
shentong 1\jfr$ill[, Skt. $arj-abhijiiiilJ)14 to benefit sentient beings. According
to the exoteric dharma (xianshi fa fJ!7f-;$), the arhats are mentioned first
[in the sutra] and the bodhisattvas are mentioned after.15
To Nagarjuna/Kumarajlva, the "exoteric dharma" is simply the teach-
ing of the Two Vehicles, the Sravakayana, the goal of which is becoming
an arhat, and the Pratyekabuddhayana. The "esoteric dharma" is the
totality of the Mahayana approach to the "three teachings" of Buddhism:
morality, meditation, and wisdom. The wisdom aspect as explained as
the bodhisattva's enlightened comprehension of the non-production of
dharmas, the emptiness of all conceptualizations and the non-dual nature
of reality; the morality aspect corresponds to the complete cessation of
defilements; and the meditation aspect matches up with the acquisition of
the six spiritual penetrations, which are thaumaturgic powers putatively
acquired as a by-product of the cultivation of meditative absorption
(samadhi).
In some editions of the text the compound xianshi is written using the
character we are more familiar with in later discourse. The two charac-
ters, both pronounced xian mean the same thing and are often used
14 The six spiritual penetrations (Ch.liu shentong t::mIfIJ1!!; Skt. earj-abhijfiiif:z) are 1) psy-
chic power (rddhi-vidhi-jfiiina, shenzu tong magical power; 2) heavenly ear (divya-
srotra-jfiiina, tianer tong supernormal hearing; 3) cognition of others' thoughts
(para-citta-jfiiina, taxin tong ftj:J"L.,J1!!), the ability to read minds; 4) recollection of past lives
(pilrva-nirviisiinusmrti-jfiiina, suming tong 5) heavenly eye (divya-cakeus-jfiiina,
tianyan tong the ability to discern the previous lives of others; and 6) cognition
of the extinction of outflows (iisrava-keaya-jfiiina, loujin tong a state in which
one is no longer plagued by any fonn of defilement. See Apidamo da piposha lun 102,
T 1545, 27.530a18-b1O; and Dazhidu lun 28, T 1509, 25.264a-266b; cf. Etienne Lamotte,
trans., Le traite de la grande vertu de sagesse de Niigiirjuna (Mahaprajiiiipiiramitiisiistra),
5 vols. (Louvain: Institutorientaliste, Universite de Louvain, 1944-1981),4: 1809-1838.
By means of the spiritual penetrations a bodhisattva purifies his buddhakeetra; see Mohe
zhiguan 2a, T 1911, 46.14a-b.
15 Dazhidu lun 4, T 1509, 25.84c-85a; cf. Lamotte, Le traite de la grande vertu de sagesse
de Niigiirjuna, 1:235.
IS THERE REALLY "ESOTERIC" BUDDHISM? 335
interchangeably. The writer employs the idea of the "esoteric dharma" a
'. few more times iater in his translation and, due to the foregoing expla-
nation, we can understand that he is referring generally to advanced. teach-
ings of the Mahayana.
16
At the end of this long exegesis the writer attempts
to be more explicit with respect to what about the Mahayana is and is not
"esoteric." However, the odd construction of the passage leaves room
for much interpretation. The passage is as follows: "The Prajiiiipiiramitii
is not an esoteric dharma (mirrLifa), and yet all the sfitras, such as the Lotus
Siltra, explain to the arhats that they will become buddhas."17 Based on
i:his statement, some later exegetes understood this passage to mean that the
Lotus Satra is "esoteric" but that the Perfection of Wisdom literature was
"
The idea of an esoteric dharma or esoteric teaching referring to Maha-
yana teachings and techniques was employed in several seminal trans-
lations of Buddhist scriptures in the first quarter of the fifth century.
Faxian (d. after 423), the famous Chinese Buddhist pilgrim who
traveled to India during the years 399A14,18 in his translation of the
Mahiiparinirviil}a Satra (Daban nihuan jing T 376, com-
pleted ca. 417-418) consistently utilizes the idea of an "esoteric teaching"
.. (mijiao to refer to the central teaching of the Mahayana sutras,
including the Buddha's stratagems or "skillful means" (Skt. upiiya; Ch.
fangbian 1J1J!). In several places the translation refers to the Mahayana
. variously as "the esoteric teaching of the Tathagata's skillful means"
(rulaifangbian mijiao "the Buddha's esoteric teaching"
(fa mijiao "the esoteric teaching of the Vaipulya [scriptures]"
. (fangdeng mijiao and "the esoteric teaching of the Vaipulya
Mahayana [scriptures]" (fangdeng dasheng mijiao
16 See Dazhidu lun 76, T 1509, 25.597b16, for the usage "the Buddha's esoteric
dhanna" (fo mimifa {9Ilff,zWl!).
17 Dazhidu lun 100, T 1509, 25.754b20-21.
18 For the biography of Faxian see Gaoseng zhuan 3, T 2059, 50.337b-338b; see also,
James Legge, trans. A Record of Buddhistic Kingdoms (Oxford: Clarendon, 1886; rpt.
New York: Dover, 1965), 1-8; and Ch'en, Buddhism in China, 89-91.
19 Dahan nihuanjing 1, T 376, 12.853b, 853c; fasc. 2, T 376, 12.867c-868a, 868b; fasc.
3, T 376, 12.870a, 872a; fase. 4, T 376, 12.879c; fasc. 5, T 376, 12.884a-b, 886a, 887a,
890a; fasc. 6, T 376, 12.894b, 898c, 899a.
336 RICHARD D. MCBRIDE, IT
A few good examples of the way "esoteric teaching" is used in this
text to refer to the various skillful means learned by the bodhisattva
according to the Mahayana are as follows:
Since the Tathiigata entices sentient beings to make [spiritual] progress,
for the sake of sentient beings he explains all dharmas and cultivates prac-
tices of no-self. When one cultivates no-self one eradicates and forsakes
views of the self. Having eradicated views of the self one enters nirv3J;J.a.
Forsaking the worldly self, for this reason, is not my esoterio teaching of
skillful means (fangbian mijiaa). Nevertheless, I have explained that the
nature of the Tathagata is called "the self of the truth of leaving behind the
world."20
In other words, here, the NirviilJa Sutra says that although all of the
Buddha's teachings are skillful means, the ones that entice people to leave
the mundane world and forsake the self are not esoteric teachings. This
being said, teachings that inspire people to seek the bodhisattva path are
the "esoteric teaching" because these expedients lead them to understand
the true nature of reality, the emptiness of all dharmas. The scripture is
more explicit in another passage:
The Buddha told [me] Kasyapa that the fIrst thing that I [the Buddha]
preached was namely the esoteric teaching of the Tathagata' s skillful means
(rulaifangbian mijiaa). He said that all the sravakas did not obtain nirv3J;J.a.
For this reason you should know all [things] by means of this Mahiipari-
nirviil,la Sutra and that "parinirv3J;J.a'.' is merely a Buddha-sphere (fa jingjie

This second passage from the scripture supports an important issue for
;Sinitic Buddhist exegetes, particularly those associated with formulating
doctrinal classification taxonomies (panjiao that immediately after
the Buddha became enlightened he taught the Mahayana initially and
only later taught the Sravakayana because his disciples could not under-
stand his superior teaching. Since the greater path of skillful means
employed by the bodhisattva was not understood by the sravakas it was
called the esoteric teaching because it was comprehended by bodhisattvas
only.
20 Dahan nihuanjing 5, T 376, 12.883c.
21 Dahan nihuanjing 6, T 376, 12.895b.
IS THERE REAlLY "ESOTERIC" BUDDIDSM? 337
. (Tanwuchen 385-433)22 translation of the
Mahiiparinirvar;a'Satra (Dahan niepanjing T 374, completed
414-421), though better known for its support of Daosheng's (ca.
360-434) views about all beings possessing Buddha-nature,23 also uses
."esoteric teaching" in the same manner as Faxian's translation but employs
a few new compounds such as "the fine esoteric teaching of the Vaipu-
. lya Mahayana [scriptures]" (fangdeng dasheng weimi zhi jiao
lfzii&), "the Buddhas and Tathagatas' fme esoteric teaching" (fo rulai
.' weimi zhi jiao and "the esoteric teaching of the Maha-
yana Vaipulya [scriptures]" (dasheng fangdeng mijiao
Furthertnore, Buddhabhadra's (Fotuobatuoluo 359-429)25
translation of the Buddhiivatarrzsaka Satra (Dafangguang fo huayan jing
. T 278, completed 418-422) uses "esoteric teaching" to
refer to its presentation of the advanced Mahayana teachings. In this
scripture we find such constructions as "the Tathagata's deep and broad
esoteric teaching" (rulai shenguang mijiao "the Buddha's
fine esoteric teaching" (fo weimi jiao a.nd "the dharma of the
Buddha's esoteric teaching" (fo mimijiao fa The Buddhii-
vatarrzsaka, which claims to be the sutra taught ftrst by the Buddha after
. his enlightenment and which was not understood by the srav'akas, was
considered to be an "esoteric teaching" along with the Lotus and Nirvar;a
Satras. Buddhist sutras and treatises contain more examples of the ideas of
."esoteric" and "exoteric" deployed as polemical interpretive devices.
They often appear in lists.27
22 For the biography of see Gaoseng zhuan 2, T 2059, 50.335c-337b;
see also Ch'en, Buddhism in China, 88, 114.
23 See, for instance, Young-ho Kim, Tao-sheng's Commentary on the Lotus Sutra
(Albany: State University of New York Press, 1990), 17-18,24,34-38,61,65-66; Whalen
Lai, "Tao-sheng's Theory of Sudden Enlightenment Re-exantined" in Sudden and Gradual:
Approaches to Enlightenment in Chinese Thought, edited by Peter Gregory (Honolulu:
University of Hawai'i Press, 1987), 169-200.
24 Daban niepanjing 1, T 374, 12.366a-b, 368c; fasc. 8, T 374, 12.415c; fasc. 9, T 374,
12.417a; fasc. 10, T 374, 12.427a; fasc. 11, T 374, 12.431a.
25 For the biography of Buddhabhadra see Gaoseng zhuan 2, T 2059, 50.334b-335c;
see also Ch'en, Buddhism in China, 107, 109.
26 Dafangguangfo huayanjing 1, T278, 9.395b; fasc. 4, T 278, 9.419b; fasc. 6, T 278,
9.434a; fasc. 14, T 278, 9,621a; fasc. 51, T 278, 9.720a; fasc, 58, T 278, 9.773c, 774b,
27 translation of the Bodhisattva-bhUmi (completed between 414421,
or 426) briefly employs an "hidden esoteric explanation" (yinmi zhi shuo Jlil1f;;tim) and an
338 RICHARD D. MCBRIDE, II
Although few Chinese scholastic works from the Northern and South-
ern dynasties period remain, there is evidence to demonstrate that monas-
tic intellectuals made use of "esoteric teaching" to refer to Mahayana
ideas. A natural place to begin is the Collected Exegesis on the Maha-
parinirviil}a Siitra (Dahan niepan jing jijie T 1763)
compiled by the monk Baoliang JfTc; (444-509) and others early in the
Liang period (502-557). At this stage in the development of Chinese
Buddhist scholarship, monks understood the "esoteric teachmg" to refer
to things that can only be understood if one grasps the Mahayana perspec-
tive. For instance, in one passage that explains "the esoteric teaching of
the Tathagata's skillful means," the monk Sengzong fflffi (438-496)
explains: "[The Buddha] manifests the Three Vehicles by skillful means;
the One Vehicle (yisheng -*, Skt. Ekayana) is the practice of reality
(shixing .11'). The sixteen-foot [body of the Buddha] is for sarp.sara
(youwei the Dharmakaya is for nirvfu).a (wuwei "28 Although
complicated by its use of "matched meaning" (geyi terminology, a
problematic system of translation drawing upon concepts from native
Chinese Daoists and scholars of Dark Learning (xuanxue in an
attempt to make Buddhist ideas more intelligible to a Chinese audience,
Sengzong understands "esoteric teaching" to refer to what is ultimate
and real behind what is manifest to ordinary beings. What is "esoteric"
is the fact that the skillful means employed by the Buddha to get his mes-
sage across to sentient beings is merely a shadow of underlying reality.
In other words, although the Buddha taught the Three Vehicles
(Sravakayana, Pratyekabuddhayana, and Bodhisattvayana), there is only
really the One Vehicle of the Buddhayana. This same understanding of
exoteric "lucid explanation" (ming shuo see Pusa dichijing i'l'iii:l1flM'ffill 3, T 1581,
30.905a. In a list of 27 upaya for teaching the Buddhadharrna in his translation of the Yogii-
ciira-bhUmi (completed between 646-648) Xuanzang 1r!!R (ca. 600-664) refers to an eso-
teric method (yinmi shuo fa and an exoteric method (xianliao shuo fa
as numbers 15 and 16; see Yuqie shidi [un 37, T 1579, 30.497a. Prajiia's (Bore
i!l1:;Ei, fl. late 8th century) forty-fascicle version of the BuddhiivataYf1saka Siitra (presented
to the Tang emperor in 798) includes a list of various comparative and opposing terms in
which an exoteric "lucid explanation" (mingliao shuo and "esoteric explanation"
(yinmi shuo are two, see Dafangguang fo huayan jing 5, T 293, 1O.683c. There are
certainly other examples of this kind of usage; the foregoing are the most representative
examples.
28 Daban niepanjing jijie 2, T 1763, 37.386b29-c3.
IS THERE REALLY ''ES01ERIC'' BUDDIDSM? 339
"esoteric teachilig" is deployed elsewhere in this text to explain why the
Buddha gave the appearance of being sick although it is known from the
Mahayana point of view that the Buddha is permanently free from :UIness
and that the Buddha seemed to die although it is known that it is impos-
sible for the Buddha to die.
29
Sui and Early Tang Period (589-712)
, In the late sixth century the great Mahayana scholiast Jingying Huiyuan
(523-592?O inherited his understanding of "exoteric teaching"
and "esoteric teaching" directly from the Treatise on the Great Perfection
o/Wisdom and used it often in his works on seminal Mahayana scriptures.
Although the words "esoteric teaching" are not found in the Vimala-
· kirtinirdda Satra (Weimoji suoshuo jing T 475), many
· exegetes beginning with Huiyuan found the polemical categories to be a
useful heuristic device they could employ to explain why crucial Maha-
yana teachings were taught to sravakas and to evaluate the respective
· merit of the competing systems of Buddhism. In his Record on the Mean-
· ing of the Vimalakirtinirdda Satra (Weimo yiji T 1776) Huiyuan
. grapples with the issue of why sravakas are listed before the bodhisattvas·
in the scripture:
Summarizing differences in merit is also called summarizing differences in
teachings. As Niigiirjuna explained there are two kinds of teachings: 1) the
exoteric teaching (xianshijiao which displays marks and conceals
reality, and 2) the esoteric teaching (mimijiao which rejects marks
and manifests reality. ill the exoteric teaching, it is proclaimed that arhats
and pratyekabuddhas, like the Buddha, have exhausted [all their] outflows.
It proclaims that a bodhisattva manifests .[karmic] actions, that his defilements
have not been cut off and that they all bear fruit. With respect to this
[esoteric] teaching, if one hears of bodhisattvas they are superior people
with reference to the arhats and many are astonishing and extraordinary.
For this reason [the scriptures] first list the sriivakas and afterward list the
bodhisattvas. If one relies on this extremely deep teaching that manifests real-
ity, the virtuous honor of the bodhisattva, in principle, should come first
29 Daban niepanjing jijie 8, T 1763. 37.4l1c, and fasc. 26, T 1763, 38.476c.
30 For the biography of Jingying Huiyuan see Xu gaoseng zhuan 8, T 2060,
50.489c-492b.
340 RICHARD D. MCBRIDE, II
and the inferior practices of the sravaka should come afterward just as in the
Buddhavatarrzsaka Sutra.
31
Instead of saying that the reason arhats are listed before bodhisattvas
in the scripture is mere literary convention, as many modem scholars
would suggest, Huiyuan uses this as an opportunity to differentiate
between the Mahayana and the llinayana. From an exoteric llinayana point
of view, he says that the adherents of the Two Vehicles (arhats [disciples]
and pratyekabuddhas) are listed first because they are presumed to be
more advanced than bodhisattvas. However, from the esoteric Mahayana
point of view in which bodhisattvas are superior to the Two-Vehicle
adherents, they are listed afterwards because they are of higher and more
extraordinary quality. It is merely the case of listing teachings in the order
of increasing significance. The important issue for us is that the "esoteric
teaching" is clearly associated with the path of the bodhisattva. How-
ever, this does not mean that Chinese Buddhist exegetes did not invert
or play with these categories from time to time in formulating their argu-
ments for the superiority of the Mahayana path. Notice how the ideas of
exoteric and esoteric are first affirmed than then reversed in the follow-
ing passage from Huiyuan's Mahayana Compendium (Dasheng yizhang
T 1851):
Some say that the HInayana is intelligible and that the Mahayana is unin-
telligible. The HInayana is rough and exoteric (cuxian so it is said to
be intelligible. The Mahayana is secret and esoteric (mimi so it is said
to be unintelligible. Some say that the Mahayana is intelligible and that the
HInayana is unintelligible. [Since] the Mahayana manifests (xian reality
it is said to be intelligible. [Since] the HInayana obscures (ju fi) reality it
is said to be unintelligible.
32
From the enlightened perspective of the advanced Mahayana teach-
ings, the bodhisattva comprehends reality the way it really is - this is the
great secret of Buddhism.
Tiantai Zhiyi (538-597),33 in his Literary Passages of the Lotus
Sutra (Miaofa lianhua jing wenju T 1718), says that the
31 Weimo yiji lA, T 1776, 38.426a.
32 Dasheng yizhang 11, T 1851, 44.679b.
33 For the biography of Zhiyi see Xu gaoseng zhuan 17, T 2059, 50.564a-568a; see also
Leon Hurvitz, Chih-i I''AA (538-597); An Introduction to the Life and Ideas of a Chinese
IS THERE REALLY ''ESOTERIC'' BUDDHISM? 341
Two Vehicles taught before the Lotus Sutra are an "exoteric explanation"
(xianshuo of the Buddhadharma and that what was taught to the
. 'aSsembly in the Lotus Sutra was an "esoteric explanation" (mishuo
.As is well known, Zhiyi employed the concept of a "secret" or "esoteric
. teaching" (mimijiao in his multi-tiered doctrinal classification
system.
35
In his Arcane Commentary on the Vimalakirtinirdda Sutra
(Weimojing xuanshu T 1777) he outlines the four teachings
but gives a different order than is usually presented in scholarship:
1) sudden teaching, 2) gradual teaching, 3) indeterminate teaching, and
4) esoteric teaching.
36
Zhiyi's explanation of "esoteric teaching" portrays
an attempt on his part to express the meaning of "esoteric teaching" in
its transcendent sense. It is the cuhnination of his classification of the four
teachings:
4) The esoteric teaching. The Treatise on the Great Perfection of Wisdom says
that [when] the Buddha first achieved enlightenment he turned the wheel of
the Dhanna of the Four Noble Truths in Deer Park.
37
In the teaching he pre-
sented on the road he clarified [the understanding] of the 5 people who saw
the Truth and obtained the fruit of the srota-tipanna [stream-winner] and
80,000 people obtained the Purity of the Dhanna-Eye.
38
[As for] the eso-
teric teaching, immeasurable bodhisattvas heard him explain the Mahiiyana
and obtained the acquiescence to the non-production [of dhannas]. After
Buddhist Monk, Melanges chinios et bouddhiques (Bruxelles: l'Institut BeIge des Hautes
Etudes Chinoises, 1962), and Ch'en, Buddhism in China, 303-313.
34 Miaofa lianhuajing wengou 4A, T 1718, 34.48a.
35 Due to the evidence that follows, I am confused by Bob Sharf's statement that "the
place of a 'secret teaching' within the T'ien-t'ai tenet-classification is a matter of some com-
plexity and debate." See Sharf, Coming to Terms with Chinese Buddhism, 340, n. 21.
36 Weimojing xuanshu 6, T 1777, 38.561c29-562a1. The traditional order is: 1) sudden
teaching, 2) gradual teaching, 3) secret teaching, and 4) indeterminate teaching. See Hurvitz,
Chih-i (538-597), 247; Ch'en, Buddhism in China, 308.
37 Cf. Dazhidu lun 2, T 1509, 25.62a; fasc. 7, T 1509, 25.109b-c, fasc. 22, T 1509,
25.225c; cf. Lamotte, Le traite de la grande vertu de sagesse de Niigiirjuna, 1:49-51,
1:415-421,3:1405-1406.
38 The Purity of the Dhanna-Eye (fayanjing Skt.
has different connotations in the non-Mahayana and Mahayana traditions. In the non-
Mahayana or Sriivakayana tradition it refers to attaining the first of the four attainments
of the sravaka, the fruit of the srota-iipanna, or the stage of the stream-winner; see Za ahan
jing 15, T 99, 2.104c. In the Mahayana it refers, as the quote above, to obtaining
the acquiescerice to the non-production of dhannas(wushengfa ren see Weimoji
suoshuo jing 1, T 475, 14.539a, 540c fasc. 2, 14.546a.
342 RICHARD D. MCBRIDE, II
that, from his frrst moment of enlightenment until his nirvfu:1a he constantly
explained the Prajfiii at night, [but only] some were able [to com-
prehend] its meaning. This scripture (VimalakirtinirdeSa Sutra) says: "The
Buddha preaches the Dharma by means of one sound / Sentient beings,
according to their class, are each able to obtain liberation. "39 This is also a
mark of the esoteric teaching. If there is a time in which all the masses are
unable to hear or see [understand the Buddha's teaching], then this is an
esoteric teaching.
40
Zhiyi's explanation of "esoteric teaching" is inextricably tied to his
understanding of the chronological classification of siitras, and yet it .
still refers directly to the advanced teachings of the Mahayana. It is an
esoteric teaching because even though people may hear it they cannot
understand unless they have the spiritual capaCity of a bodhisattva.
Although not stated explicitly, the "exoteric teaching" is what is heard
by people of limited spiritual capacity and refers to the Four Noble Truths.
An adherent of the Two Vehicles can attain the stage of a stream-winner,
a benefit of the exoteric meaning of the teaching, This is contrasted to
the bodhisattva who is able to comprehend the esoteric meaning of the
teaching that causes the acquiescence to the non-production of dharmas,
the defining characteristic of an advanced bodhisattva's wisdom on the
Mahayana path.
Jizang alii (549-623),41 the famous scholar-monk of the Chinese Madhya-
maka tradition (Sanlun also wrestles with the concept of "esoteric
teaching" in many works and, like his colleagues, draws inspiration from
the seminal exegesis attributed to Nagarjuna. To him, an esoteric teach-
ing was a dharma entrusted to bodhisattvas only. In his Commentary on
ihe Vimalakfrtinirde.sa Sutra (Weimojing yishu T 1781) he
problematizes the matter in the form of a dialogue:
Question: The Treatise on the Great Perfection of Wisdom says that the
Lotus Sutra was the first esoteric dharma (mifa because it was entrusted
to bodhisattvas. The Prajfiii (bore ?1l!l:E') [literature] is not an esoteric dharma
because is was entrusted to sravakas.
42
If this scripture was not yet clear to
39 Weimoji suoshuo jing 1, T 475, 14.538a.
40 Weimojing xuanshu 6, T 1777, 38.562a.
41 For the biography of Iizang see Xu gaoseng zhuan 11, T 2059, 50.513c-515a; see
also Ch'en, Buddhism in China, 132-134.
42 Cf. Dazhidu [un 100, T 1509, 25.754b20-21.
IS THERE REALLY "ESOTERIC" BUDDIDSM? 343
sravakas when they received prophesies of [their future] attainment of Buddha-
hood then it is not an esoteric dhanna (mimifa Wt'i'B'l$;). Why was it entrusted
to bodhisattvas?
Answer: There are two types ofPrajna [literature]: 1) [that preached to the
people] of the Three Vehicles together and 2) that explained to the bodhi-
sattvas only. Having searched through the Large [Prajiiiipiiramitii Siara]
(dapin 7cd"o), [1 found that] it is taught to the people of the Three Vehicles
together because it was entrusted to sravakas. Even though this scripture
(Vimalakfrtinirde.sa Sidra) is not esoteric it is only understood by bodhisattvas
as an inconceivable approach to the Dhanna. It is not what can be known
[by people] of inferior ranking [viz. spiritual capacity]. How can the [teach-
ings] of the Two Vehicles be fathomed? - because they are entrusted to
bodhisattvas; they do not bother with sravakas.
43
In the foregoing passage we can see that Jizang attempts to refine the
definition of the concept of an "esoteric teaching" to include teachings
in the Prajfia literature understood only by bodhisattvas. Jizang under-
stands that not all Mahayana scriptures are subsumed in the "esoteric
teaching" because many scriptures contain both Mahayana and
Mahayana teachings, but more importantly because, according to his
understanding, the Treatise on the Great Perfection of Wisdom suggests
that the Prajfia literature is not esoteric and that the Lotus Siitra is. This
should not be surprising because intellectual concern with doctrinal clas-
sification systems reached its zenith during the seventh century in China.
We have seen how previously Zhiyi incorporated "esoteric teaching" into
his doctrinal classification system. According to Jizang's defInition, since
the Vimalakfrtinirdda SLitra was entrusted to both bodhisattvas and
sravakas it is not an "esoteric teaching." However, he says that since the
principles taught in the scripture are fundamentally "inconceivable" to
sravakas it should be viewed as in the same general class as the Lotus Sidra,
which he takes to be a prototypical "esoteric teaching" entrusted to and
understood only by bodhisattvas.
Elsewhere in his writings, however, Jizang attempts to explain the
esoteric teaching in causative terms as what transforms sravakas into
bodhisattvas. For instance, in his Commentary on the Lotus SLitra (Fahua
yishu T 1721), Jizang says: "[To] Sravakas [who rely on the]
43 Weimojing yishu 6, T 1781, 38.990c.
344 RICHARD D. MCBRIDE, II
esoteric teaching (mijiao shengwen the Buddha, as a Dharma
King, appears as a self-transfonning bodhisattva and now he causes the
Hfuayana to enlarge into the great path (dadao them
into bodhisattvas. '># Although it is beyond the scope of this paper, in his
Arcane Discourse on the Lotus Satra (Fahua xuan lun T 1720)
Jizang labors to demonstrate that both the Lotus Sutra and Perfection of
Wisdom literature have exoteric and esoteric components. It should not
be a surprise that he would attempt to portray the "exoteric teaching" of
the Prajiia literature as just as important as the "esoteric teaching" since
it serves as the scriptural basis for the Chinese Madhyamaka tradition.
He understands the ideas of exoteric and esoteric polemically since the
"esoteric teachings" of the Lotus and Nirviif}a Sutras were taught "exoter-
ically" - explicitly, openly - to bodhisattvas.
45
Also in his Commentary
on the Lotus Sutra, Jizang uses "esoteric teaching" to explain the supe-
riority of the Mahayana to the Hfuayana, which he calls the "Vehicle of
Men and Gods" (rentiansheng A7(*). The "esoteric teaching" fits nicely
into this heuristic role representing the Mahayana in the "ten teachings
in five pairs" (wushuang shijiao sudden [Buddhiivata1'{tSaka
Sutra] and gradual [Vehicle of Men and Gods up to the Lotus Sutra] (dun-
jian mundane [Vehicle of Men and Gods] and supra-mundane
[Lotus Sutra] (shi-chushi t!:!:1±It!:!:); great and small (da-xiao */N; exo-
teric and esoteric (xian-mi and self and others (zi-ta EHif!).46
In the second half-of the seventh century, Kuiji (632-682),47 a close
disciple of Xuanzang (ca. 600-664) and the founder of the Ci'en
"Mf,Jf!J or Faxiang 1:t;f:§ (Dharma Characteristics) school of Chinese Yoga-
also elucidated his views on the dual ideas of esoteric and exoteric
dharmas. In his commentary, Hidden Praise of the Heart Sutra (Bore
poluomiduo xinjing youzan T 1710) he makes the
following observation:
Only those hastening in the Mahayana tum the wheel of the hidden and
secret [esoteric dhanna] (zhuan yinmi [un which explains that all
44 Fahua yishu 7, T 1721, 34.552a.
45 Fahua xuan [un 3, T 1720, 34.383b.
46 Fahua yishu 7, T 1721, 34.552b-c.
47 For the biography of Kuiji see Song gaoseng zhuan *iiii{!Jj$ 4, T 2061, 50.725b-726c;
see also Ch'en, Buddhism in China, 320-321.
IS THERE REALLY "ESOTERIC" BUDDIDSM? 345
'dharmas, each and every one, is devoid of self nature, is neither produced
nor destroyed, 'and is originally nirvfu}.a. Even though [dharmas] are still
curious and miraculous, they are as if their meaning is not intelligible. Also,
all [dharmas] have been demonstrated as existing in a state of complete tran-
quility. Now, those propounding all vehicles (yiqie cheng -t7J*) [tum] the
wheel of the exoteric [dharma] (xianliao fun !.'l!i7fffil), which is unexcelled
and featureless, and in its surpassing meaning it is the teaching that mean-
ing is truly intelligible (zhenliaoyi jiao 7 [In this teaching] not all
[dharmas] are explained as existing in a state of complete tranquility.48
Kuiji's definition of these ideas also suggests that what makes a par-
ticular teaching esoteric or exoteric depends on the way that it views
dharmas. If the view of dharmas coincides with the advanced Mahayana
doctrine that they lack self nature, then it is an esoteric teaching because
the great secret is that all dharmas are originally in a state of quiescence.
This is the transcendent approach of the One Vehicle, the BUddhayana.
When viewed from the perspective of the teachings of all the various
vehicles, however, if dharmas are explained as existing and as being intel-
ligible, then it is an exoteric teaching - no matter how profound.
The influential Buddhist scholiast W5nhyo JGIlji (617-686),49 from the
Korean state of Silla deploys the polemical concepts of esoteric and
exoteric as a: heuristic device to indicate superior and inferior approaches
to the Buddhadharma. In his Thematic Essentials of the Sukhiivatfvyuha
Sutra (Muryangsu-gyong chongyo T 1747), he uses the
polemical ideas to explicate the passage in the Sukhiivatfvyuha Sutra that
encourages the practice of buddhiinusmrti (yombul, Ch. nianfo the
"ten recollections" (simnyom, Ch. shinian He explains that there
is both an "esoteric meaning" (iinmil iii, Ch. yinmi yi and an
"exoteric meaning" (hyoZZyo iii, Ch. xianliao yi mi7R) to this practice.
51
To describe the esoteric meaning of the ten recollections W5nhyo gives
a list of ten practices drawn from a now-lost text titled the Scripture O'n
48 Bore poluomi duo xin jing youzan 1, T 1710, 33.523b,
49 For the biography ofWonhyo see Song gaoseng zhuan 4, T 2061, 50.730a-c; Samguk
yusa 4, T 2039, 49.l006a-c; see also Robert E. Buswell, Jr., The Formation of
Ch'an Ideology in China and Korea: The Vajrasamadhi-Siitra, A Buddhist Apocryphon
(princeton: Princeton University Press, 1989),41-73.
50 Wuliangshoujing 1, T 360, 12.268a.
51 Muryangsu-gyong chongyo, T 1747, 37.129a.
346 RlCHARD D. MCBRlDE, II
Questions asked by Maitreya (Mile fawen jing Being con-
cerned with compassion toward other beings, not slandering the teachings
and practices of others, and cultivating the thought of enlightenment
(bodhicitta), there is nothing "secret" about any of the ten practices. 53
Wonhyo explains that they are descriptive of bodhisattvas above the first
stage (bhitmi). The point is that Wonhyo calls them "esoteric" because
they are superior and are practices of the bodhisattva. The exoteric mean-
ing of the "ten recollections" refers to the practice of chanting the name
of the Buddha Amitayus ten times or for ten thought-moments, paying
homage to the Buddha, as explained in the Book on the Visualization
of the Buddha Amitayus (Guan Wuliangshou jing T 365),54
52 The Mile fawen jing has a complicated genealogy that discussed briefly by Matsub-
ayashi Koshi t-,,;f!:'lL.::Z:, "Chosen Jodokyo no kenkyii: Miroku somon no jiinen 0 meguru
gimon" : <' (Study of Korean Pure Land: Focus-
ing on the issue of the ten recollections of the Mile sowen). Ryukoku Daigaku Bukkya
bunka kenkyu kiyo 6 (1967): 82-85; Matsubayashi had published
similar findings earlier as "Shiragi JOdokyo no ikkosatsu: Gangyo no jodokyo shiso 0
megutte" (An inquiry into the Pure Land
teachings of Silla: With an emphasis on Wonhyo's Pure Land teachings and thought), IBK
15/1 (December, 1966): 196-198. Later scholarship on Wonhyo's ten recollections derives
from Matsubayashi's work: for instance, see Chong Hakkwon "Gangyo Taishi no
jiinengi ni tsuite" '?"l (On Master Wonhyo's view of the ten recol-
lections), IBK25/l (December, 1976): 269-271; and Kakehashi Nobuaki t>IliBlI!Ii, "Shiragi
JOdokyo no hatten (1): jiinen-ron ni chakuganshite" (-) : (The
development of Silla Pure Land teachings (1): Looking at the theory of the ten recollec-
tions), IBK 42/2 (March, 1994): 650-653.
53 Muryangsu-gyong chongyo, T 1747, 37. 129a-b. The ten recollections according to
the Mile fawen jing are as follows: "1) Constantly arouse thoughts of compassion toward
all beings. Do not slander the practice of all beings, for if you do slander their practice, you
will never be reborn in the Pure Land. 2) Deeply arouse thoughts of sympathy toward
all beings. Forsake remaining hannful intentions. 3) Arouse the thought of protecting the
dharma without begrudging your body or your life. Do not slander any of the dharmas.
4) Produce a mind of assurance with regard to everything you are enduring. 5) With your
mind profoundly pure, do not covet profit and gain. 6) Arouse a mind of omniscient wis-
dom. Constantly reflect (yom) on these day after day without faltering. 7) Arouse thoughts
of honor and respect toward all beings. Forsake all sentiments of self-conceit and be hum-
ble when you speak. 8) Do not take pleasure in worldly gossip. 9) Stay close to the thought
of enlightenment and deeply arouse all the conditions of wholesome faculties. Stay far away
from thoughts that are trOUbled, tumultuous, scattered, and chaotic. 10) Visualize the Bud-
dha with the correct recollection (yom) and forsake all doubts." The Yu simallak to
(Traveling the Path to Mental Peace and Bliss), a later text attributed to Wonhyo, which is
based on the Muryangsu-gyong chongyo, contains this same passage; see T 1965, 47.114c.
54 Guan Wuliangshoujing 1, T 365, 12.346aI2-22.
IS THERE REALLY "ESOTERIC" BUDDIllSM? 347
which, from the context, Wonhyo considers an inferior practice to becom-
ing a bodhisattva:
In his Thematic Essentials of the Nirviil}a Siitra (Yolban chongyo
T 1769), Wonhyo deploys these concepts heuristically to discuss
the meaning of nirvfu;la:
The concept "nirvfu}.a," verily, contains two meanings: that which we may
call an "esoteric expression" (mira, Ch. miyu and an "exoteric expres-
sion" (hyallyoa, Ch. xianliaoyu Relying on the exoteric expression,
it is the straight translation "liberation [through the] extinction [of out-
flows]" (myalto, Ch. miedu .... If we rely on the esoteric expression,
it contains many instructions.
55
Wonhyo then proceeds to discuss a few aspects of the meaning of
nirvaIJa. For instance, he treats nirvfu;la's association with the concept of
death (samyol, Ch. simie as an "exoteric expression" and as belong-
ing to the "esoteric expression" of nirv1iIJa he includes the idea of "a lack
of suffering" (mugo, Ch. wuku In this case, esoteric and exoteric
do not refer to the polemical distinction between the Mahayana and the
Hinayana, but instead suggest the literal distinctions between "hidden"
and "apparent." Although not used frequently in this manner, there is ample
evidence of "esoteric" and "exoteric" being used this way in Buddhist
scriptures.
57
Also, Jiacai 3tm::t (fl. 645), for instance, employs the com-
pound yinxian to refer to hidden and manifest interpretations in his
Pure Land Treatise (Jingtu lun T 1963).58
55 Yolban chongyo, T 1769, 38.240c-241a.
56 Yolban chongyo, T 1769, 38.241a.
57 translation of the Bodhisattva-bhumi (completed between 414-421,
or 426) briefly employs an "hidden esoteric explanation" (yinmi zhi shuo p';wzm) and
an exoteric "lucid explanation" (ming shuoa)lm); see Pusa dichijing 3, T 1581, 30.905a.
In a list of 27 upiiya for teaching the Buddhadhanna in his translation of the Yogiiciira-
bhumi (completed between 646-648) Xuanzang (ca. 600-664) refers to an esoteric
method (yinmi shuo fa p.;wml*) and an exoteric method (xianliao shuo fa !!Il:7ml*) as num-
bers 15 and 16; see Yuqie shidi lun ftl:iih1l+ftllg,jff 37, T 1579, 30.497a. PrajfHi.'s (Bore i!9:!i,
fl. late 8th century) forty-fascicle version of the BuddhiivataT(lSaka Sutra (presented to the
Tang emperor in 798) includes a list of various comparative and opposing terms in which
an exoteric "lucid explanation" (mingliao shuo a)l7m) and "esoteric explanation" (yinmi
shuo I\iWm) are two, see Dafangguang fa huayanjing 5, T 293, IO.683c. There are cer-
tainly other examples of this kind of usage; the foregoing are the most representative
examples.
58 Jingtu lun 2, T 1963, 47.90b.
348 RICHARD D. MCBRIDE, II
The concepts of "esoteric" and "exoteric" were understood as polem-
ical terms by medieval Sinitic Buddhist exegetes. While they were deployed
most commonly to explain the supreme Mahayana teaching on the true
nature of dharmas as being empty of self-nature, they were also used to
promote the superiority of the bodhisattva. Thus, for three hundred years
the polemical heuristic device known as the esoteric teaching or esoteric
dharma had been employed regularly by Buddhist exegetes to promote the
superiority of the advanced Mahayana teaching of the emptiness of all
dharmas and the acquiescence to the non-production of dharmas, the tran-
scendent knowledge acquired through skillful means cultivated by adher-
ents to the bodhisattva path. Furthermore, the Buddhavataf(lsaka, Lotus,
and NirvalJa Sutras were held to embody the esoteric teaching.
High Tang through the Late Tang (712-907) and Beyond
The putative first true "Esoteric" master, SUbhakarasiIpha (Shanwuwei
637-735)59 arrived at the Tang capital early in the reign of the
Tang Emperor Xuanzong 1r* (r. 712-756), probably around the year
716, and, with the help of the brilliant Buddhist polymath Yixing -ff
(673-727),60 translated the Sutra on Mahavairocana 's Attaining Buddha-
hood (Da Piluzhena chengfo jing Skt. *Mahiivairo-
canaabhisaf(lbodhi Siitra, T 848), which they completed in 725. Before
Yixing's death in 727, and probably in the process of translating the siltra,
they composed the first "Esoteric" Buddhist exegesis: the Commentary
to the Scripture on Mahavairocana's Attaining Buddhahood (Da Piluzhena
chengfo jing shu T 1795).
The siltra describes rituals and practices for the use of mantra (zhenyan
l'ij) in a new role as helping generate the three esoterica (sanmi =.W, Skt.
*tri-guhya) or three mysteries: the body, speech, and mind of the Buddha.
61
59 For the biography of SubhakarasiIpha see Song gaoseng zhuan 2, T 2061, 50.714b-
716a; see also Chou Yi-liang, "Tantrism in China," Harvard Journal of Asiatic Studies 8
(March, 1945): 251-272.
60 For a study on Yixing see Jinhua Chen, "The Birth of a Polymath: The Genealogical
Background of the Tang Monk-Scientist Yixing (673-727)," Tang Studies 18/19 (2000-2001):
1-39.
61 The sUtra introduces the term "mantra vehicle" (zhenyansheng Skt. man-
trayiina) to describe this approach to the Dharma; see Da Piluzhena chengfo jing 1, T 848,
IS THERE REALLY "ESOTERIC" BUDDIDSM? 349
These are new developments that the commentators develop further in the
exegesis.
62
However, in each of the six times "esoteric teaching" is used
in the commentary it is deployed in a way not fundamentally different than
the preexisting tradition as representing advanced Mahayana teachings.
63
For certain, the idea of secrecy is emphasized, but they do not claim that
the point of the secrecy is somehow different than obtaining the acqui-
escence to the non-production of dharmas.
For example, after providing a list of code words, some gendered and
some not, in which the concept of "female" or "woman" (nuren -J7:.A)
is explained as meaning "prajiia, the mother of the buddhas" (bore fomu
the commentators say: "This esoteric teaching cannot be pro-
claimed directly since there is much secret language (yinyu Scholars
presently consider it to be coarse. "64 In this passage, since other" esoteric
teachings" of the Mahayana have been proclaimed directly previously,
only because this new esoteric teaching contains gendered language offen-
sive to contemporary Buddhist scholars is it taught secretly. The way the
category of "esoteric teaching" is used resembles that of the earlier intel-
lectual tradition. Note the following example:
You should dwell in non-profligacy (bufangyi and you will inherit
and take up the previously [ -mentioned] text (in other words, you will be
entrusted with the sutra), this esoteric teaching of the Mahayana (ci dasheng
mijiao and you will be an inheritor of the mark of the Dharma
(jaxiang cheng i*i'i3J¥i:).65
I8.5c; fasc. 7, T 848, I8.51b (uses dasheng zhenyansheng "mantra vehicle of
the Mahaylina"), and 54c. For the "approach of the three esoterica" (sanmi men
see Da Piluzhena chengfo jing 7, T 848, 18.51c, 52b.
62 For example, the concept of the "three esoterica" is discussed 19 times, beginning
with Da Piluzhena chengfo jing shu 1, T 1796, 39.579b-c. It is also interesting to note that
the term "mantra vehicle" (or mantrayana) is deployed three times; see Da Piluzhena
chengfojing shu 4, T 1796, 39.625c25, c27 (here dasheng zhenyansheng), and fasc. 9, T 1796,
39.671aI2; "mantra teaching" is used four times (usually zhenyanjiaoJa zhenyan-
jiao once); see Da Piluzhena chengfo jing shu 7, T 1796, 39.651a5, b26; fasc. 10,
T 1796, 39.688a25; and fasc. 14, T 1796, 39.724a17 (here zhenyanjiao); and "vajra vehi-
cle" (jin'gangsheng Skt. vajrayana) is found once; see Da Piluzhena chengfo jing
shu 5, T 1796, 39.629al1.
63 Da Piluzhena chengfo jing shu 1, T 1796, 39.579c29; fasc. 3, T 1796, 39.614aI9;
fasc. 4, T 1796, 39.616c27; fasc. 5, T 1796, 39.627a26; and fasc 20, T 1796, 39.787alO, 13.
64 Da Piluzhena chengfo ling shu 1, T 1796, 39.579c-580a.
65 Da Piluzhena chengfo jing shu 20, T 1976, 39.787a9-I2.
350 RICHARD D. MCBRIDE, II
There is no attempt on the part of SubhakarasiIpha or Yixing to differ-
entiate their "esoteric teaching" from the advanced Mahayana teachings;
rather, they emphasize that it is an "esoteric teaching of the Mahayana."
Nevertheless, it is difficult to transmit and they are clear that to receive
and observe their esoteric teaching requires special spiritual capacities.
Aspirants and adherents are sternly warned not to give in to profligacy
and lust because the ritual practices introduced in the scripture use the
senses to overcome the senses. Seen from this perspective, dne can see
how to many Chinese Buddhists, the esoteric teachings of the SiUra on
Mahdvairocana's Attaining Buddhahood, which emphasize recreating the
body, speech, and mind of the Buddha directly as the "esoteric teach-
ing," are no more esoteric than the teachings of the Buddhavataf[lsaka
Sutra or the Lotus Sutra, because one could easily understand that acqui-
escence to the non-production of dharmas means fundamentally the
same thing as acquiring or reproducing the body, speech, and mind of the
Buddha.
Amoghavajra (Bukong 705-774),66 the third of the three "Eso-
teric" masters, deployed the concepts of esoteric and exoteric teaching in
a short essay he composed on the meaning of dharaI)I titled Encomia on
a General Interpretation of the Meaning of DharalJl (Zongshi tuoluoni
yizan T 902), which was probably written sometime
between 762 and 774. At the beginning of the exegesis, after listing
four types of dharaI)I he says: "Relying mostly on the exoteric teaching
(xianjiao) they are explained in the Mahayana teaching (dashengjiao
At the end of this short piece, after detailing four kinds of Bud-
dhist vocative devices: dharar:n (tuoluoni W8*im), true words (zhenyan
J1:l.: § ), esoteric words (miyan W § ) - both are Chinese translations of the
word "mantra" - and vidya (ming 8)3), he says that they are all based
on words in the Indian language and that they "have been explained
repeatedly in the siitras of the exoteric teaching (xianjiao)." He then goes
on to say that "true words" have been "explained in the esoteric teach-
ing (mijiao)" using the above four designations and that they resonate with
the "approach of the three esoterica" (sanmi men '=:WF5).67 At fIrst glance
66 For the biography of Amoghavajra see Song gaoseng zhuan 1, T 2061, 50.712a-
714a; see also Chou, "Tantrism in China," 284-307.
67 Zongshi tuoluoni yizan, T 902, 18.898a-b.
IS THERE REALLY "ESOTERIC" BUDDHISM? 351
Amoghavajra's statements seem to support the idea of a separate esoteric
teaching, but since he does not explain what he means by either esoteric
teaching or exoteric teaching we are left to conclude that his deployment
of these terms follows the standard intellectual interpretation. He
nizes that dhar3l)I and several other related terms for spells and codes are
found commonly in Buddhist literature; probably a tacit reference to their
deployment in Prajfiapatamita literature, which was held by many to be an
, exoteric teaching. "True words" occupy a special place in his "esoteric
teaching" but it is not conceptualized as anything more than an advanced
Mahayana teaching, resonating with the approach of the three esoterica,
the same as with the commentary by SubhakarasiIpha and Yixing men-
tioned previously.
Although scholars have become accustomed to describing Sinitic Bud-
dhism around the time of the An Lushan 3i:ffrU<lli rebellion (ca. 755-763)
as generally dominated by "Tantric" or "Esoteric" Buddhism in the cap-
ital and Chan (Zen) in the outlying areas, we must remember that Bud-
dhist exegetes throughout the Sinitic cultural sphere continued to digest
Xuanzang's translations and retranslations of Yogacara materials and that
the Yogacara and Huayan *D intellectual traditions were still influen-
tial. In the Yogacara literature, the concept of "esoteric" was, as before,
connected to the understanding of the true nature of reality. For example,
the writings of Tullyun (a.k.a. Toryun d.u.), a Yogacara monk
'in the Korean kingdom of Silla who lived during the eighth century,
suggest that "esoteric teaching" continued to refer to the Mahayana doc-
trine that dharmas lack self-nature. In his Record of the Yogiiciira-bhUmi
(Yuga-ron ki T 1828) Tullyun says: "The esoteric meaning
(mirui, Ch. miyi is explained summarily that all dharmas, each and
everyone, has no self-nature and is neither produced nor destroyed, etc.
[Hence,] they are called scriptures of unintelligible meaning. "68
Later, during the ninth century, Pei Xiu (797-870), the famous
Buddhist layman and Chan advocate explains the term xian-mi IIW, lit-
erally "exoteric-esoteric," in his "Preface to the Annotated Commentary
to the Book of Perfect Enlightenment" (Dafangguang yuanjue xiuduoluo
liaoyi jing lueshu zhu shu T 1795),
68 fuga-ron ki 20B, T 1828, 42.776a.
352 RICHARD D. MCBRIDE, II
which was composed by his friend Zongmi *W (780-841), the well-
known Huayan and Hezei1ITw Chan patriarch.69 Pei Xiu, however, defines
this concept in a straight-forward Mahayana way: "exoterically explained
and esoterically preached" (mi shuo er xian yan
Bob Sharf has demonstrated that it was not until the late tenth century,
during the Song * period (960-1279), that Chinese Buddhist exegetes first
began to group together particular ritual practices and the monks who
promoted those practices - which is somewhat close to preserit-day acad-
emia's "Esoteric Buddhism." The earliest evidence is from Zanning's
.$ (919-1001) Lives of Eminent Monks compiled in the Song (Song
gaoseng zhuan T 2061), which was commissioned in 983 and
completed in 988, and further reedited. In a brief editorial comment fol-
lowing the hagiography of the three "Esoteric" masters he classifies them
as among those who promulgated the "Wheel of Instruction and Com-
mand" (jiaoling lun which Sharf identifies as "one of the ear-
liest known expressions used to characterize the teachings and practices
of these prelates." Perhaps more importantly, Zanning also says that they
"claim to teach the great doctrine of Yoga" (yuqie dajiao ffj;r(J]ojdtz).7!
Perhaps this explains why Buddhist thaumaturges, usually thought to be
"Esoteric Buddhists," are often referred to as "Yoga monks" (yuqie seng
filrJ11Jo11W) in Buddhist literature compiled during the Song and succeeding
periods.
72
The fact that Zanning coins a new classification and does not
employ the idea of "esoteric teaching" is circumstantial evidence that
"esoteric teaching" still simply referred to the advanced teachings of the
Mahayana.
The earliest conceptualization of "esoteric teaching" in exegetical mate-
rials that seems to support the ideas commonly-held by scholars today
regarding "Esoteric Buddhism" is found in a work titled Anthology on
the Essentials of the Heart of Attaining Buddhahood and the Peifect Inter-
69 For more on Zongmi and Pei Xiu see PeterN. Gregory, Tsung-mi and the Sinifica-
tion of Buddhism (princeton: Princeton University Press, 1991),27-90.
70 Dafangguang yuanjue xiuduoluoliaoyijing lueshu zhu shu, T 1795, 39.523b28-c2.
71 Sharf, "On Esoteric Buddhism in China," 269-270; Song gaoseng zhuan 1, T 2061,
50.714aI5-18.
72 See Lii lianfu, Zhongguo mijiaoshi, 432-513. For an example in Korean Buddhist
hagiographicalliterature compiled about 1285 by !ryan -!f& (1206-1289), see Samguk yusa
2, T 2039, 49.972b.
IS THERE REALLY "ESOTERIC" BUDDHISM? 353
penetration of the Exoteric and Esoteric (Xianmi yuantong chengfo xinyao
ji T 1955). This text was also treated briefly by
Sharf.1
3
It was compiled by a Liao monk named something like Daoe
(a.k.a. DaozhenlDaochen fl. 1085-1096), who lived at Jinhesi
on Mt. Wutai E:::' in Shaanxi Province in North China.
74
Since his name is so uncommon he may have been non-Chinese, but since
he resided at this famous pilgrimage site he probably knew Indian and
Tibetan Buddhists, and he may have been somewhat familiar with tantras.
Nevertheless, Daoe's work is full of allusions to the writings of Fazang
(643-712), referred to here as "Xianshou" .Ii§[ (Worthy Head), and
especially to Huayan doctrine. After discussing the "mantra teaching"
(zhenyanjiao explaining it in Huayan terms, Daoe says:
As for the essentials of the heart of the second esoteric teaching (er mijiao
xinyao zhe they are said to be the commentarial documents
on spiritual transformations 75 and the commentarial documents on m3I.J.qala,
both of which demarcate the teaching of dharaJ.1I, which is [the tradition of]
esoteric perfection (miyuan The previous perfect tradition of the exo-
teric teaching (xianjiao yuanzong viz. Huayan) necessarily pre-
cedes awakening in the dharma-sphere of Vairocana. Thereafter, depending
on his awakening [the practitioner] cultivates the whole ocean of practices
of Samantabhadra,76 is able to abandon [the cycle of] rebirth and death, give
73 Sharf, "On Esoteric Buddhism in China," 273-275.
74 For more on Daoe see Lii Jianfu, Zhongguo mijiaoshi, 472, 485-489; and Nogami
Shunj6 Ryo Kin no Bukkyo (Kyoto: Heirakuji Shoten, 1953),42-45,
70, 108, 165, 166, 169. The TaishO and Japanese secondary sources use the character
but Lii, probably following the Xinxiu gaoseng zhuan (compiled in 1884) uses another
rare character IRt which suggests his name may be pronounced either Daozhen or Daochen.
See Yu Qian (aka Yu Mei'an ui!ii$l<Jl;l;), Xinxu gaoseng zhuan siji (Taibei:
Liuli jingfang, 1967),4: 19a-b (modem edition in four volumes; Daoe's biography is found
in 1: 237-238).
75 Spiritual transformations (shenNan refer to miraculous powers displayed by the
Buddha in teaching and converting sentient beings. It is often interchangeable with "spir-
itual penetrations" (shentong /iiljijffi), the six supernormal powers obtained by Buddhist
adepts as a by-product of meditation (samadhi). See Apidamo jushe lun Il"fEE.U{J'!*IDfu 27,
T 1558, 29.l43c-144a; and Miaofa lianhuajing 7, T 262, 9.60a.
76 The practices of Samantabhadra (Puxian xing commonly refer to all the expe-
dient means cultivated by the bodhisattva in the 52 stages of the Huayan conceptualization
of the bodhisattva path. See Dafangguang fo huayan jing 33, T 278, 9.607a-
611a; Dafangguangfo huayanjing 49, T 279, 1O.257c-262a; cf. Thomas Clearly, trans. The
Flower Ornament Scripture, one-volume ed. (Boston: Shambhala, 1993),952-970.
354 RICHARD D. MCBRIDE, II
evidence to the completion of the ten bodies,?7 and [acquire] the unhindered
fruit of buddhahood.
78
Although this description of "esoteric teaching" begins to approach
what academics conceptualize as "Esoteric Buddhism," it is in a late work
and one not held to be of particular significance in the academy's imag-
ined "Esoteric" tradition because it is so thoroughly mingled with Huayan
doctrine. But what is more interesting is that Daoe separates !he practice
of zhenyan mantras and dharaJ)l into two different types of "esoteric
teachings," the essentials of both he explains using Huayan doctrine. Cer-
tainly, monks in the Sinitic cultural sphere did not conceive of "Esoteric
Buddhism" the same way that scholars of the present-day do.
Some Concluding Remarks
In medieval Sinitic Buddhist exegesis "esoteric teaching" is not a
descriptive term or titular designation of what academics presently call
"Esoteric Buddhism." The polemical concepts of "esoteric" and "exo-
teric" are hermeneutical devices employed generally to laud the merits
of the Mahayana over the so-called Hfuayana. Although there is not com-
plete uniformity in the explanations of "esoteric teaching" and "exoteric
teaching" in the writings of the Sinitic Buddhist exegetes from the fifth
through the eighth centuries C.E. there is a definite congruity of meaning
to the point that we can state confidently that through the eighth century
the object of the interpretation of such terms as "esoteric teaching" or
"esoteric dharma" was the advanced Mahayana teachings and that "exoteric
teaching" or "exoteric dharma" referred to the non-Mahayana tradition
and usually also to the Perfection of Wisdom literature. The evidence
also suggests that despite some scholars attempts to refine the defInition
77 The ten bodies (shishen +!l) are 1) the sentient being body (zhongsheng shen
*1:!l), 2) the [Buddha-]land body (guotu shen 3) the karma-reward body (yebao
shen *¥&!l), 4) the sravaka body (shengwen shen §flIJ!l), 5) the pratyekabuddha body
(bizhiJo shen !l'F:3'z:i9ll!l, dujue shen 6) the bodhisattva body (pusa shen i§'i\I!l), 7)
the tathagata body (rulai shen 8) the wisdom body (zhishen 9) the dharma
body (jashen lti:!l),'and 10) the emptiness body (xukong shen See Dafangguang
fo huayan jing 26, T 278, 9.565b; Dafangguang fo huayan jing 38, T 279, 1O.200a.
78 Xianmi yuantong chengfo xinyao ji 1, T 1955, 46.993c.
IS THERE REALLY "ESOTERIC" BUDDHISM? 355
of "esoteric teaching" it remained basically synonymous with "advanced
Mahayana teaching" throughout the medieval period.
An even more significant point is that the so-called "Esoteric" mas-
ters did not attempt to redefine the terms "esoteric teaching" and "exo-
teric teaching." The esoteric teaching they advocated, conceptualized as
the body, speech, and mind of the Buddha, though not readily accepted
or understood by contemporary scholars and most individuals, was on all
counts harmonious with the general message of the Mahayana. This sug-
gests that they used "esoteric teaching" in the same way it was deployed
in earlier Mahayana literature and Sinitic Buddhist exegesis: that "eso-
teric" means "the best," that it refers to advanced Mahayana teachings,
that it is only intelligible to individuals with the spiritual capacity of
bodhisattvas, and that it employs updya (skillful means) as a means of
causing the practitioner to obtain "the acquiescence to the non-produc-
tion of dharmas" - which seems to be the great secret of the Mahayana.
Thus, the three "Esoteric" masters did not presume to establish a new
teaching that was fundamentally different than the advanced Mahayana,
they merely claimed a privileged place within the expansive Mahayana
teaching for their ritualized approach to overcoming duality and desire to
achieve buddhahood. That this was conceptualized as "esoteric" along
with other advanced Mahayana teachings is axiomatic and the polemics
of the Mahayana demanded it. This may be a reason why there is no clear
documentary evidence for a separate or distinct "Esoteric School" in the
Sinitic cultural sphere during the Tang.
The grid of knowledge deployed by Buddhists in medieval China and
Korea always conceptualized the idea of "esoteric teaching" as referring
generally to the advanced teachings of the Mahayana, particularly to the
way the bodhisattva understands the nature of reality. Furthermore,
Buddhists in the Sinitic cultural sphere did not begin to develop a tenta-
tive classification for "Esoteric Buddhism" (not to mention the classifi-
cation "Tantric Buddhism") until at least the tenth century and they
never deployed a classification comparable to that which developed in
either Japan or Tibet. We scholars need to be careful about either apply-
ing or projecting models developed from Indian, Tibetan, or Japanese
sectarian developments backward onto Buddhist history in China and
Korea.
356 RICHARD D. MCBRIDE, II
So, in conclusion: Is there really "Esoteric" Buddhism? There are two
possible answers: 1) Yes., it is the advanced teachings of Mahayana Bud-
dhism, and 2) No, it just means the advanced teachings of Mahayana
Buddhism. Observed from this perspective, "Esoteric Buddhism" may
not be any better than "Tantric Buddhism" as a category. In medieval
times it never really meant what sectarian scholars impute to it but was
instead employed as a polemical device to claim that what it represented
- the advanced teachings of the Mahayana - was the best or most supe-
rior form of Buddhism. All of the Buddhist exegetes who used the term
imputed an esoteric quality to the teachings they held to be superior in their
analyses. While classifications are indispensable tools to facilitate under-
standing, they may often cause us to overlook the complex relationship
between dynamism and continuity that is a defining characteristic of
medieval Sinitic Buddhism.
mROME DUCOR*
LES SOURCES DE LA SUKHAVATi',
AUTOUR D'UNE ETUDE RECENIE DE G. FUSSMAN
Le quart de siecle ecouIe a vu un renouvellement important de l' etude des
sources indiennes relatives a la Bienheureuse (Sukhavatf), ce champ de
buddha du Buddha Arnitiibha/Amitayus, dont Ie culte se
developpera ulterieurement dans la tradition sino-japonaise de la Terre Pure.
Ce renouveau est d'autant plus significatif que Ie sujet a longtemps ete
considere comme relevant d'un bouddhisme de seconde zone par les india-
nistes occidentaux. Certes, on n'en est plus aux invectives de F.M. Mi.iller!,
mais Ie temps n'est pas loin ou Andre Bareau pouvait encore ecrire a pro-
pos du culte d'Arnitiibha: "A une religion construite pour des surhommes,
heros doubles d'ascetes et de penseurs, se substitue une religion destinee
a la masse des gens depourvus d'energie, de moralite et d'intelligence"2.
Cependant, ce regain d'interet temoigne aussi de la difficulte pour bien des
specialistes a maitriser des sources aussi diverses que l' epigraphie
et la litterature canonique indiennes, les traductions chinoises avec leurs repe-
res chronologiques ou les etudes de leurs collegues orientaux. En temoigne,
notamment, Ie long article publie par Gerard Fussman sur "La place des
Sukhiivati-vyuha dans Ie bouddhisme indien" (JA 1999). Celui-ci souleve
plusieurs points litigieux, tant sur les sources que sur leur interpretation, et
on trouvera ici quelques observations a leur sujet. Sans pretendre a l'ex-
haustivite, elles rassemblent des informations souvent difficiles d'acces.
* Ma vive gratitude s'adresse a Mme Helen Loveday (Universite de Geneve) et a M. Jacques
May (Universite de Lausanne) pour leurs precieux conseils, ainsi qu'a M. Paul Harrison
(University of Canterbury) pour la communication de ses analyses recentes.
1 V. Miiller 1881, p. 29-30.
2 Bareau 1966, p. 177. La Vallee Poussin n'etait guere plus tendre: "Le nom de Grand
Vehicule devrait etre reserve a des speCUlations mystiques d'un interet beaucoup plus
grand" (LVP, Dogme, p. 71). Aucun texte lie ala SukhavatI n'apparalt dans Le boud-
dhisme, belle anthologie publiee par Lilian Silbum (1977) et reeditee sous Ie titre de Aux
sources du bouddhisme (1997).
Journal of the International Association of Buddhist Studies
Volume 27 • Number 2 • 2004
358
mROME DUCOR
I. Sources epigrapbiques
Les donnees epigraphiques sont particulierement maigres, puisqu'elles
ne comprennent que deux documents tres courts: l'inscription de Govind-
nagar et celie de Sanc!, qui datent respectivement du IIe siecle et de la fin
du VIIe siecle.
(A.l) L'inscription de Govindnagar se trouve sur Ie socle d'une statue
cassee d'un buddha debout, decouvert en 1977 a Govindnagar et conserve
au Mathura Museum (n° 77.30). Forme de quatre lignes en ecriture brahmI,
Ie texte dit
3
:
"Le 26 du 2" mois de l'an 26 du Maharaja En ce jour,
pere du marchand Sacika (?), petit-fils du marchand Balakatta, Ie fils de
Buddhapila (?), a etabli une image du Bienheureux Buddha Arnitabha (bha-
gavato buddhasya amitiibhasya) en offrande a tous les buddha. Puisse cette
racine de merite faire obtenir a taus les etres la connaissance insurpassable
d'un buddha!"
Si l'on accepte comme hypothese de travailla date de 78 ap. I.-C. pour
la 1 ere annee de cette inscription remonterait aI' an 104
de notre ere. Tant la mention du nom du buddha que la derriiere phrase
du texte en font la plus ancienne reference datee non seulement a Amita-
bha mais, aussi bien, au Grand Vehicule, voire a un "proto-Mahayana"5.
(A.2) L'inscription de Sanc! figure sur un fragment de stele provenant
du monastere 43 de Sanc! et conservee au Sanc! Museum (nO C 833).
Mais elle est d'une importance bien moindre que l'inscription de Govind-
nagar, puisqu'il s'agit d'un hymne en sanskrit a Lokanatha/Avalokites-
;vara ou Amitabha n' est que mentionne comme figurant dans la coiffe de
ce bodhisattva ("yo=mitiibhan=ca dhatte")6.
3 Editions et traductions: Schop. 1987, p. 101-111 (cf. id. 1985, p. 39); Fus. 1999, p. 541-
542. Reproduction: Schop. 1987, p. 135-137. Cf. Huntington 1988, p. 86ab et PI. 9.1. Shanmi
1984, p. 231-232, n. 169, et fig. 151; id. 1989, p. 313b-314b; id. 1995, p. 214-215, fig. 146 .
. Autres references in Fujita 1996b, p. 9, n. 10, id. 1996c, p. 43-44, n. 25; id. 2001, p. 119-
120, n. 7.
4 Resume de cette bouteille a encre dans Ie bel et utile ouvrage de Pierre Guenee, p. 187.
5 Schop. 1985, p. 39-40.
6 N.G. Majumdar, in The Monuments of sand, Vol. I, p. 394-395, inscription no. 842;
Vol. ill, PI. CXXXIX, 64, 1. 2. Cf. Schop. 1987, p. 99, 117.
LES SOURCES DE LA SUKHAYATI 359
Les deux temoignages mentionnes ci-dessus constituent les seules
sources epigraphiques certaines concernant Amitabha. Mais certains
specialistes se sont efforces de l' identifier dans plusieurs autres sculptu-
res. Trois d'entre el1es ont plus particulierement inspire leur sagadte.
(B.l) La premiere est la celebre stele de Mohammed Nari conservee
au Central Museum de Lahore (n° 1135/G. 155), qu' Alfred Foucher tenait
pour une illustration des miracles de Sakyamuni a SravastI
7
. L'interpre-
tation qui voit en elle une representation de la SukhavatI ne s'est deve-
loppee qu'a partir de la tMorie publiee en 1980 par John C. Huntington
8
,
qui a sus cite depuis bien des reactions contradictoires. Si cette hypothese
a notamment ete reprise par Mario Bussagli9, Gregory Schopen a cepen-
dant montre que la sculpture pourrait tout aussi bien - sinon rnieux -
representer Abhirati, Ie champ de buddha Plus recemment,
Jacques Gies et Monique Cohen ont aussi estime qu'elle etait "loin d'illus-
trer la Terre Heureuse de 1'Ouest, telle qU'elle se dessine dans les deux
Sukhiivatfvyuha-sutra"; ils y voient plutat la manifestation d'un Sakya-
muni mahayaruste, "peut-etre" Ie Vairocana de l'AvataJ?1saka
ll
, mais non
pas Ie Sakyamuni du "Sutra du Lotus" comme Ie pensent les specialis-
tes japonais12.
Par contre Fussman juge toujours que notre stele constitue "une repre-
sentation pas trop maladroite de la SukhavatI", ce qu' i1 fait en se r€ferant
notarnment a la "remarquable mise au point" d'Anna Maria Quagliotti
13

Cette derniere conclut en effet son etude de 1996 par une identification
avec Ie "SukhavatI Paradise" (p. 287-288). Mais son interpretation est des
plus douteuses, au moins sur trois points. Quagliotti se fonde principale-
ment sur la scene figurant a l' extreme droite du 2
e
registre en partant du
7 Foucher 1917, p. 171-175, PL XXVIll, 1; JA 1909-1, p. 5-78. Cf. lngholt, n° 255;
Kurita 1, n° 395. Photo couleur: Bussagli 1984, p. 140; id. 1996, p. 245.
8 Y. aussi Huntington 1988, p. 86a; l'auteur annonce la prochaine publication de "Stu-
dies in Sukhavatf Art and Literature" (ibid., n. 5).
9 Bussagli 1984 et 1996, loc.cit. Pour Paul Harrison aussi, notre stele represente bien
Amitabha en SukhavatI (communication personnelle).
10 Schop. 1987, p. 117, n. 50 (p. 130-131).
11 Serinde, Terre de Bouddha, p. 341-343; la stele y est attribuee au IVe_ye siecle. Pour
Fussman (1999, p. 550), sa date n'est "pas determinable a trois siecles pres (ly_ye)", tan-
dis que Tissot la situe au "lye s. ap. I.-C. au plus t6t" (1987, p. 99).
12 Cf. Quagliotti 1996, p. 281(-282), n. 7.
13 Fus. 1999, p. 550 et 548; cf. Fus. 1987, p. 73b et 74b.
360 JEROME DUCOR
haut de la stele
14
: elle l'identifie comme la visite rendue par Indra - i.e.
Sakra
15
- a Sakyamuni meditant dans une grotte (Indrasailaguhii) et elle
juge que cette image est "highly relevant for interpretation of the stele"
(p. 276). Neanmoins, on re1evera, tout d'abord, qu'il est par trop artifi-
ciel de vouloir se fonder quasiment sur cette seuIe scene pour interpreter
une stele qui en compte tant d'autres. Ensuite, rien ne prouve qu'il s'agisse
de la visite d'Indra a Sakyamuni, bien au contraire. En effet, toutes les
representations bien attestees du meme theme montrent, en toute logique,
Ie Buddha faisant Ie geste de la meditation, d'une part, et abrite sous une
grotte, d'autre part
16
• Or, dans la scene excipee par Quagliotti, Ie Buddha
a la main droite levee. En outre, il y est figure non pas dans une grotte
mais sous un arbre. Quagliotti elle-meme devance cette derniere objection
en affirmant que "the point is relatively unimportant smce the Buddha is
in fact shown seated on a rock" (p. 284). Mais cet argument, bien leger
en un point si crucial de la demonstration, ne convaincra que son auteur.
Enfin, et surtout, on ne voit vraiment pas Ie rapport entre la Sukhavatl et
la visite d'Indra. Quagliotti imagine que cette demiere aurait ici une
dimension symbolique renvoyant a un enseignement d'ordre superieur:
Sakyamuni en nirmal}akdya guiderait vers une contemplation d' Amitabha
en safTlbhogakaya qui serait reservee a des inities; et Ie moine agenouille
au pied de Sakyamuni representerait symboliquement ces inities recevant
l'enseignement ultime du Buddha, qu'il s'agisse de l'Ananda du "Grand
Sukhavatlvyuha", ou du Sariputra du "Petit Sukhiivatfvyuha"17 (p. 284-
285,287). Or, non seulement il reste a prouver que la doctrine du trikdya
ait ete maitrisee par l'auteur de la stele, mais, surtout, on doit relever que
rien dans Ie texte des deux sutra n'autorise une telle hypothese a tendance
esoterique. Au contraire, la manifestation du champ de buddha d'Amita-
bha telle qu'elle est decrite dans Ie "Grand Sutra" n'est pas reservee au
seul auditoire du Buddha sur Ie Mont Grdhrakil!a: elle est vue par tous les
14 Photo: Huntington 1980, PI. ill, fig. 4; reproduite in Quagliotti 1999, PI. ill, fig. 4.
15 Cf. Lamotte, "VajrapiiI).i en Inde", p. 116, 118.
16 Voir les exemples in Ingholt nOI28-134; Kuirta 1, nO 330-340,629-630.
17 Les deux Sukhtivativyuha-sutra seront mentionnes ci-apres comme Ie "Grand Siitra"
et Ie "Petit Siitra". Le "Grand Siitra" sera cite selon les pages et les lignes de I 'ed. Ashikaga
(Ash.), accompagnes de la division commode en 155 paragraphes de la traduction libre de
G6mez (p. 61-111). Pour Ie "Petit Siitra", on suivra la division habituel1e en 20 paragraphes
(Ducor 1998, p. 141-149; p. 100, n. 345); cf. son edition H!Cente in Fujita 2001, p. 79-88.
LES SOURCES DE LA SUKHAVATI 361
religieux, nonnes et hucs des deux sexes du champ de buddha de Sakya-
muni tout entier, sans parler de ses dieux,niiga et autres divinites
18
.
Plus recemment encore, Ie meme detail de la stele a ete repris par
Chhaya Haesner qui, suivant l'hypothese de Huntington, l'identifie non
pas avec la visite d'Indra, mais avec cette manifestation de la SukhavatI
que nous venons d'evoquer19. Cependant, meme si nous adoptions cette
hypothese, rien n'attesterait que la stele tout entiere constitue une illus-
tration du theme de la SukhavatI avec Amitabha en son centre. Deux .
details au moins prouvent meme Ie contraire. En effet, Ie buddha central
est assis sur un lotus flanque par un homme et une femme, chacun debout·
sur un lotus s' enracinant dans les memes eaux que Ie lotus du buddha.
Haesner les identifie comme des "reborn souls in SukhavatI,,20. Or, ce der-
nier point est impossible, puisque Ie 35
e
vreu d' Amitabha dans Ie "Grand
Sutra" assure que les femmes ne naissent pas en son champ de buddha
sous une forme feminine. Et il en va de meme des poissons nageant dans
les eaux d' ou poussent les lotus
21
: leur presence en SukhavatI serait tout
aussi incongrue, puisque Ie 1 er vreu affirme que les trois mauvaises des-
tinees - dont celie des animaux - sont absentes du champ de buddha
d' Amitabha. En resume, rien dans cette stele de Mohammed Nari ne per-
met d'y reconnaitre positivement une representation de la SukhavatI.
(B.2) La deuxieme sculpture a avoir ete identifiee a notre theme est la
Triade Buddhamitra, conservee au John and Mable Ringling Museum of
Art, a Sarasota (n° MF94.8.5). Ce fragment montre un buddha avec un
bodhisattva sur son cote gauche, tandis qu 'un autre bodhisattva devait
figurer en pendant sur sa droite dans Ie morceau manquant. Les specia-
listes datent cette piece avec des ecarts considerables, allant des environs
de 78-178 au tournant des rne_IVe siecles, voire meme des ye_ YIe sie-
cles
22
. Sur Ie socle se trouve une ligne d'inscription gandhan redigee en
18 Ash. 55:20-56:3, §125.
19 Haesner, p. 432-433 et fig. 2a Cf. Huntington 1980, p. 658-659.
20 Haesner, p. 435 et fig. 5-5a; photo: Huntington 1980, PI. VI, fig. 7-8. Suivant Hun-
tington (iti., p. 660, 680), Quagliotti y voit des figures de donateurs (1996, p. 275).
21 Haesner, p. 436, et p. 450, fig. 12; photo: Huntington 1980, PI. XV, fig. 23, et PI. XVI,
fig. 24.
22 Respectivement Fus. 1999, p. 543; Salomon & Schopen, p. 4; GvM, p. 229b. Brough
disait "probably of the second century" (1982, p. 70).
362 .JEROME DUCOR
Fondee sur une photo mediocre, I' edition Brough, reprise par
Fussman, pouvait etre rendue it peu pres en ces termes:
"De Buddhamitra: Avalokitesvara (oloispare), en don. De Buddhamitra:
AmWibha (amridaha) ... "23.
A partir de cette identification epigraphique, la triade a ete reconnue
par Brough comme une representation d' Amitabha et de ses deux bodhi-
sattva assistants, tels qu'ils sont mentionnes dans Ie "Grand Slitra": Ie
bodhisattva conserve serait donc A valokitesvara, tandis que son pendant
disparu aurait ete Mahasthamaprapta
24
. Pour Fussman, la Triade Buddha-
mitra constituerait ainsi "Ie seul temoin sUr de la devotion it Amitabha au
Gandhara"25.
Cependant, il semble bien qu'il n'en soit rien. En effet, Richard Salomon
et Gregory Schopen ont recemment publie cette inscription sur nouveaux
frais et sont arrives it une lecture qui peut se rendre ainsi:
"Don de Dhamitra [sic] (Dha<*rma>mitra7), a (au lieu-dit) Oloispara [7],
pour l'immortalite (amridae > amrtaya) [i.e. nirva(1a] de Buddhamitra ... "26.
Les auteurs livrent cette interpretation avec plusieurs reserves, notam-
ment sur Ie terme "oloispare"; en outre, ils n'abordent pas la question
iconographique et ne proposent pas d'identification de la piece. Neanmoins,
leur analyse me parait suffisamment solide pour que I'on souscrive it leur
conclusion categorique: l'inscription "definitely contains no reference to
_ 23 Brough 1982, p. 66-67; Fus. 1987, p. 73a, et fig. 4; Fus. 1999, p. 543-544, n. 49. A
propos de "amridaha", Fussman s'est demande "si Amitabha n'est pas une fausse sans-
kritisation d'un terme moyen-indien correspondant it sanskrit Amrtiibha" (Fus. 1994, p. 37;
cf. id. 1999, p. 543-544; cf. Brough 1982, p. 67-68; GvM, p. 228b). Ogiwara Unrai (1867-
1937) avait aussi avance l'hypothese qu'Amida venait d'Amrta, mais celle-ci a ete refutee
par Fujita (p. 287-306). Ce dernier refute aussi l' opinion d 'Ogiwara selon qui "Amita" aurait
precede Ie doublet Amitabha/Amitayus (v. Nakamura, p. 201, 202; cf. Mbdj. 1, p. 72a).
V. Yamada Isshi, I, p. 189-196; Machida, p. 3-8.
24 Brough, p. 66. Cf. Ash. 49:9-10, §102; Hobi5girin 7, p. 999b-1000a.
25 Fus. 1999, p. 550 (cf. id. 1987, p. 73b-74a; 1994, p. 36-37). Quagliotti estirne aussi
que, "in the lack of decisive confutation", Ie Buddha peut etre identifie comme Arnitabha
(1996, p. 282, en note).
26 Salomon & Schopen, notarnment p. 9-18, 27 (cf. p. 5, et n. 3). Les crochets sont des
auteurs, tandis que j'ajoute les parentheses en fonction de leurs commentaires. Sur l'equi-
alence amrta = nirviiIJa, v. Lamotte, Vimalakfrti, p. 430.
LES SOURCES DE LA SUKHAVATI 363
Amitabha"27. D'ailleurs, ce constat se confirme sur Ie plan iconogra-
phique. En effet, Brough (p. 65-66) etayait son interpretation notamment
par la presomption de la presence d'une representation d'Amitabha dans
Ie turban du suppose A valokitesvara. Or, les photographies recentes ·mon-
trent qu'il ne s'agit pas d'un buddha mais d'un simple element conique
28
.
Tant est si bien que rien n'interdit d'identifier Ie bodhisattva subsistant
avec, par exemple, Maitreya, comme Ie faisait LeRoy Davidson
29
.
(B.3) Enfin, Ie troisieme des plus celebres candidats it une hypotMtique
sukhavatIsation est Ie fameux "Buddha de Bruxelles", sculpture qu'il
serait plus juste d'intituler "Pentade Buddhananda", du nom de son dona-
teur
30
. Publiee pour la premiere fois en 1973
31
, elle porte une inscription
qui, la meme annee, fit l'objet de communications de James C. Harle
et de Fussman
32
. Entree dans la collection Claude de Marteau (Bruxel-
les), cette piece a - elle aussi - suscite des commentaires tres contra-
dictoires, tant sur sa datation que sur son identification, et meme des doue
tes sur son authenticite
33

La sculpture represente un buddha assis sous un arbre et flanque de deux
bodhisattva debout, avec deux dieux sur l'arriere. Le bodhisattva sur la
gauche du buddha porte dans sa coiffe un buddha assis en meditation, ce
qui l'identifie facilement it Avalokitesvara; tandis que l'autre bodhisattva
ne presente aucun signe distinctif. Lesoc1e porte une inscription, qui fait
entrer cette piece dans Ie cerc1e tres ferme des sculptures du Gandhara qui
27 Salomon & Schopen, p. 28. Cf. aussi les doutes de Fujita 2001, p. 120, et n. 8, qui
mentionne les etudes japonaises recentes.
28 L'absence de tout buddha etait deja mentionnee par Lee, p. 315, n. 25. Cf. LeRoy
Davidson, p. 23, fig. 23 (photo reprise in GvM, p. 229, fig. 4); Salomon & Schopen, fig. l.
29 LeRoy Davidson (loc.cit.). Selon Mitterwallner, cependant, les personnages subsis-
tants sont bien Amitabha et AvalokiteSvara, tandis que c' est Ie pendant de ce dernier qui
aurait ete Maitreya (GvM, p. 228b, 239a).
30 Lecture "Buddhananda" selon Harle (p. 128), Czuma (p. 198) et Schopen (1985, p. 35;
1997, p. 62); mais "Buddhananda" selon Quagliotti (1977, p. 137), et "Buddhanada" selon
Fussman (1974, p. 54; 1999, p. 547).
31 "Oriental Art", N. S., Vol. XIX, no. 1 (1973), p. 24. Cf. Harle, pI. 71; Fus. 1974,
pI. XXXI, fig. 27; Quagliotti 1977, Tavola 1. Photo couleur: Czuma, p. 199; Bussagli 1984,
p. 107; id. 1996, p. 188; Kurita 1, P3-VIII.
32 Cf. Harle (2
nd
South Asian Archeology Conference, Amsterdam 1973), et Fus. 1974
(XXIXe Congres des Orientalistes, Paris 1973).
33 Cf. Guenee, p. 198, qui renvoie a Tanabe Katsurni (1988).
364 JEROME DUCOR
soient datees
3
4, ou plutot, dans Ie cas present, quasi datees. Redigee en
"exceptionally well-written and legible"35, elle est en prakrit
"fortement sanskritise"36, mais sans teneur mahayana contrairement ace
qu' avait cm Fussman
37
:
"Le 5 du mois de Phalguna de I'an 5, don de Buddhananda verse-dans-Ie-
Tripitaka (trepirJakasa), en offrande pour feu son pere et sa mere".
En l'absence d'un nom d'ere ou de souverain, cette "date". en dit trop
ou pas assez, au point qu' on a pu la traduire par des datations extreme-
ment divergentes. Partisan convaincu de la datation la plus longue, Fuss-
man adopte l'ere de debutant en 78 et il en condut que la sculp-
ture remonte a l'an 83 de notre ere, ou, "au plus tard, 130 de n.e."38.
Avec la date de 83, la Pentade Buddhananda et son inscription cumule-
raient les superlatifs. Ce du Gandhara constituerait en effet
Ie plus ancien temoignage du Grand VehicuIe, puis que precedant d 'une
vingtaine d'annees l'inscription de Govindnagar; ce serait aussi Ie plus
ancien exemple d'un buddha a I'epaule droite denudee, d'une part, et fai-
sant Ie geste de l'enseignement, d'autre part
39
• En outre, son inscription
serait Ie plus ancien texte date en sanskrit mixte bouddhique et eIle serait,
a 2-3 ans pres, run des plus anciens temoignages du terme "Tripitaka"40.
A l'autre extreme, Gritli von Mitterwallner attribue l'inscription a l'ere
de Khirigila, ce qui donnerait une date suivant de peu ran 453 ap. J._C.
41

Son hypothese s'appuie sur de nombreux details, tant stylistiques qu'ico-
nographiques, Ie plus frappant etant la presence de l'image d'Amitabha
dans la coiffe caracteristique qui n'apparaitrait qu'a
partir de la seconde moitie du IIIe siede
42
. On relevera cependant que
34 Harle, p. 131; Czuma, p. 198, n. 5.
35 Harle, p. 128, qui en donne I'edition et la traduction etablies par John Brough.
36 Fus. 1974, p. 54-55, qui en donne sa propre edition ayec sa traduction; cf. Fus. 1999,
p.546-547.
37 Schop. 1985, p. 35-38. Cf. Fus. 1974, p. 56; et id. 1999, p. 547, n. 56.
38 Fus. 1999, p. 547, n. 56. Cf. id. 1974, p. 54, n. 2; id. 1987, p. 72b.
39 GyM, p. 217b-219b. Cf. Tissot 1985, p. 73, n. 34, et p. 202-203.
40 Elle ne serait precedee que par I'inscription de I'image de KausambI offerte en
I'an 2 de par la nonne Buddhamitra, ainsi que par I'inscription de Samath datant
de ran 3 (Quagliotti 1977, p. 137, n. 4; Schop. 1985, p. 24, n. 6; id. 1997, p. 243, 247).
• 41 GyM, p. 213b, 222b, 239b.
42 GyM, p. 215b-216a. Cf. Kurita 2, n° 171.
LES SOURCES DE LA SUKHAVATI 365
Marie-Therese de Mallmann avait signale une tete de ce bodhisattva omee
de la figurine d'un buddha assis, au Field Museum de Chicago, tout en
avan<;:ant la date de la seconde moitie du lIe siecle de notre ere; en outre,
R.C. Sharma a aussi presente une tete similaire d' A valokitesvara au
Mathura Museum (n° 2336), qu'il attribue a la fin du lIe siecle
43

A vancee des la premiere publication de cette sculpture, l'identification
avec Ie Miracle de SravastI fut reprise sans autre par les differents com-
mentateurs, dont Fussman et Harle. Pour ce demier, Ie bodhisattva sur la
gauche du buddha est evidemment A valokitesvara, tandis que son pen-
dant est Maitreya, identifie "almost certainly" par la forme de sa cheve-
lure
44
• Cependant, la publication par Brough de la Triade Buddharnitra
amena ensuite Fussman a reconnaitre la Pentade Buddhananda comme
constituant, elle aussi, une representation de la Sukhavatl, en
l'identification de Maitreya par celle de Mahasthamaprapta
45
• Plus nuance,
Fussman estime prudemment aujourd'hui que cette pentade "est Ie possible
temoin de l'existence d'un culte rendu a Amitabha", son buddha central
etant "probablement identifiable a Amitabha,,46.
Mais Mitterwallner, tout en excluant que notre sculpture puisse figu-
rer Ie Miracle de Sravastl, se refuse a y voir une representation d'Ami-
tabha en raison de la presence des deux dieux Brahma et Indra, representes
derriere Ie buddha, respectivement sur sa droite et sur sa gauche
47
• On
objectera pourtant que les dieux ne sont jamais tres loin d'un buddha,
y compris d' Amitabha: c' est ainsi que Ie 4 e va:u du "Grand Sutra" atteste
que les dieux, aussi bien que les hommes, peuvent mitre dans son champ
de buddha
48
• D' autre part, Mitterwallner continue a identifier les deux
bodhisattva comme etant A valokitesvara et Maitreya, tout en affir-
mant que ce demier temit une cruche dans sa main gauche aujourd'hui
43 Respectivement: Malhnann, p. 120-121, pI. I-a et XXI-a; et Shanna 1984, p. 232,
fig. 155 (Huntington 1988, PI. 9, II). Cf. Kurita 2, nO 168-169. Pour des fragments de bud-
dha assis en meditation qui ont pu se trouyer dans la coiffe de statues d' AvalokiteSvara,
y. Ingolt n° 242; Kurita 2, n° 173-174, 884.
44 Harle, p. 132-133; cf. Guenee, p. 206.
45 Fus. 1987, p. 73b-74a; id. 1994, p. 38.
46 Fus. 1999, p. 546 et p. 547, n. 56 (c'est moi qui souligue). Cf. Kurita 2, p. 311a.
47 GyM, p. 239ab. La position des deux dieux est classique: Lamotte, "Vajrapfu)i", p. 124,
130.
48 Ash. 11:9-10. Cf. Dantinne, p. 264-265.
366 JER6ME DUCOR
disparue
49
. En resume, la Pentade Buddhananda representerait, selon elle,
Sakyamuni entoure d' Avalokitesvara et Maitreya, ainsi que de Brahma et
Indra
50
. .
Or, ce programme iconographique correspond a celui d'une autre pen-
tade, qui est conservee au Museum of Art and Archeology (MAA) de
I'University of Missouri-Columbia. Fussman y reconnait les memes per-
sonnages que ceux identifies par MitterwaIlner dans la Pentade Buddha-
nanda
51
et il evoque Ie "Sutra du Lotus" en guise d'identification
52
. Mais
quel est donc l'indice lui permettant de differencier ainsi ces deux pen-
tades quasi identiques? Pour lui, tout ne tient qu'a un seul detail: l'Ami-
tabha de Buddhananda se trouverait sous une voute de fleurs, tandis que Ie
Sakyamuni du MAA serait assis sous un arbre pipal
53
. On en conviendra,
l'argument est mince, d'autant que la dendrologie de la Pentade Buddha-
nanda n'est pas sure: il pourrait s'agir aussi bien d'un manguier en fleurs
que d'une simple stylisation elaboree, si ce n'est meme d'un "jewel-tree"54.
Mais puis que nous en sommes au domaine des hypotheses, la Pen-
tade du MAA pourrait tout aussi bien illustrer la predication du "Grand
Sutra" par Sakyamuni. En effet, Maitreya/Ajita et Avalokitesvara sont
aussi les interlocuteurs du Buddha dans ce sutra
55
; et si Brahma et Indra
n'apparaissent pas nommement dans Ie prologue de sa version sanskrite,
49 GvM, p. 215b-216a; meme supposition deja in Harle, p. 133. La cruche peut etre un
accessoire cornmun aux bodhisattva Maitreya, Mahasthamaprapta et A valokitesvara; cf.
Hob6girin 3, p. 267a; et 7, p. 1013a; Tissot 1985, p. 231-232.
50 Mitterwallner ne propose pas d'interpretation globale de la scene et se contente d'avancer
que la presence de Braluna et Indra serait Jiee a I'episode ou les dieux, apres I'eveil du Bud-
dha, l'invitent a precher (GvM, p. 215a, 238a-239b). Czuma adopte la meme identification
des personnages et date Ie relief de 182 ap. J.-c. (p. 35 et 198). Bussagli, qui opte pour 149
ap. J.-c., ne reconnait qu'Avalokitesvara (1984, p. 137, 159; 1996, p. 188-190, cf. p. 223).
51 Avec une hesitation pour Brahma. Sans buddha dans sa coiffure, Avalokitesvara est
identifie, sur la gauche du buddha, par Ie lotus et Ie rosaire qu'il tiendrait en mains (Fus.
1987, p. 75ab, fig. 5). Cf. GvM, p. 216b, 239a.
52 Fus. 1987, p. 75a; cf. Fus. 1999, p. 545, n. 51. Mais on s'attendrait a voir Ie "Siltra
du Lotus" plutot illustre par les deux Tathagata Sakyamuni et Prabhiltaratna; cf. l'identi-
fication suggeree par Jacques Gies pour la stele de Loriyan Tangai a I'Indian Museum de
Calcutta (n° 5090): Serinde, Terre de Bouddha, p. 352-353, fig. 257; Kurita 1, n° 398.
53 Fus. 1987, loc.cit.; cf. Fus 1999, p. 546.
54 Harle, p. 133, et n. 13; GvM, p. 214a, 239a. Pour Bussagli, il s'agit de "l'arbre de
la bodhi" (1984, p. 106; 1996, p. 189).
55 Respectivement: Ash. 2:12, §4 et 56:17, §129; Ash. 46:3, §96-13. Cf. Hob6girin 7,
p. 1000a.
LES SOURCES DE LA SUKHAVATI 367
la presence des dieux est bien attestee par la conclusion du meme texte
56

Cette identification pourrait egalement s'appliquer a la Pentade Buddha-
nanda, car rien ne permet d'affirmer que Ie bodhisattva sur la droite du
buddha ne serait pas Maitreya. D'ailleurs, Ie Peshawar Museum conserve
deux autres pentades extremement semblables a celIe de Buddhananda,
jusque dans Ie decor de l'arbre abritant Ie buddha. Or, l'une d'elles (n° 158)
se distingue par la coiffe du bodhisattva sur la droite du buddha, qui porte
un element ressemblant au smpa symbole de Maitreya
57
; dans la seconde
pentade (nO 1527), Ie bodhisattva situe a la meme place a la tete cassee
et ne peut etre identifie, mais celui sur la gauche du buddha tient une
cruche pouvant l'identifier, lui aussi, a Maitreya
58
• Et, a tout prendre, on
pourrait meme adopter un raisonnement similaire pour la Triade Buddha-
mitra - quoi qu' il en soit de son inscription - et pour bien d' autres tri-
ades encore
59
• En tout cas, notre tres hypothetique identification d'un
Sakyamuni exposant Ie "Grand Sutra" aurait au moins l'avantage de mar-
quer une transition entre les representations purement consacrees a S a k y a ~
muni dans Ie monde indien et celles reprenant exclusivement Ie theme de
la Sukh1lvatl telles qu'elles apparaitront en Asie Centrale.
En conclusion de ce survol des sources epigraphiques, force est de
constate{ non seulement leur rarete, mais aussi leurs faiblesses: a l'ex-
ception notable de l'inscription de Govindnagar, leur interpretation est
si aleatoire qu'aucune d'elles ne peut servir de temoignage probant sur
Ies origines et Ie developpement de la tradition de la Sukh1lvatI. Ce cons-
56 Ash. 66:25, §155. Les dieux des differents paradis, dont ceux de Brahma et Indra,
sont bel et bien detailles dans Ie prologue de l'une des traductions chinoises archaIques du
sutra (T. 12,361, k. 1, p. 279b). Quant au prologue de la traduction du "Petit Sutra" par
Kumarajlva, elIe ne cite nommement qu'Indra (cf. Fus. 1999, p. 530, n. 11, qui mentionne
Ducor 1998, tout en attribuant, par erreur, cette traduction du chinois a J. May).
57 Reproduite in Ingholt n° 254; et Fisher, qui la date du Ill-IVe s. (p. 48, fig. 35);
Kurita 1, n° 403.
58 Fus. 1974, p. 58, n. 3. V.les reproductions in Ingholt, n° 253; Harle, plate 74; Qua-
gliotti 1977, p. 139, n. 11, et Tavola Ill; Czuma, p. 35, fig. 15; Tissot 1987, p. 97, fig. 96;
Kurita 1, nO 404 (cf. nO 405-407, 409-413, 632).
59 Cette identification pourrait ainsi etre attribuee a une triade du Lahore Museum: dans
ce cas, c'est Ie bodhisattva sur la gauche du buddha qui manque, tandis que son pendant
conserve tient une cruche; cf. Fus. 1987, p. 75b-76a (fig. 7). Fussman reproduit aussi (ibid.
fig. 6) une piece similaire, au Musee d'ethnographie de Geneve (nO 38665); sur cette der-
niere, v. Erac1e 1987, p. 54-55. Cf. aussi Kurita 1, nO 406.
368 JEROME DUCOR
tat s'applique tout particulierement a la Pentade Buddhananda, que Fuss-
man, malgre les incertitudes de sa datation et de son identification, consi-
dere neanmoins comme uri "terminus ante quem pour l'existence du culte
d' Amitabha", jugeant meme inutiles les reperes chronologiques pour-
tant bien plus significatifs - fournis par les traductions chinoises
60
• Ce
demier sujet merite par ailleurs quel'on s'y attarde, parce qu'il traduit
bien Ie danger de references repetitives qui tarderaient a se renouveler.
II. Sources Iitteraires
1. Sources chinoises
Face a la nebulosite de la chronologie indienne, les indianistes se sont
en effet rabattus depuis longtemps sur les sources chinoises pour essayer
de donner un terminus ante quem a la version sanskrite du "Grand Sutra"
et de cemer ainsi la date de I' origine du culte d' Amitabha
61
• De Mallmann,
notamment, recourut a cette methode; elle Ie fit en prenant la precau-
tion de faire reviser ses dates par Paul Pelliot, ce qui n'etait pas rien
62

Mais si cette methode reste toujours valable et meme necessaire, malgre
ses limites
63
, ses criteres ont entre temps evolue avec Ie developpement
de l'etude de la chronologie du bouddhisme chinois. C'est ainsi que Hun-
tington se fourvoie en recourant au Catalogue of the Chinese Translation
of the Buddhist Tripitaka publie en 1883 par Nanjo Bun'yU, qui est "tout
a fait perime"64. C'est pourtant sur Huntington que se fonde encore Haes-
ner, apres tant d'autres
65

v' 60 Fus. 1999, p. 548, n. 57 .. Neanmoins, notre auteur indique plus loin (p. 577) que la
Pentade Buddhananda est "d'interpretation plus douteuse" que l'inscription de Govindnagar
et la triade B uddhamitra.
61 Le "Petit Siitra" n'entre pas en ligne de compte, puisqu'il n'a ete traduit pour la pre-
miere fois qu'en 402, par Kumarajiva (T. 366; cf. Ducor 1998). A signaler que sa seconde
traduction, attribuee a Xuanzang (T. 367), pourrait en fait etre imputable a un scribe de ce
demier, Dasheng Xun (Fujita 1998, p. 3-4, 21-23).
62 Pour les td. de Zhi Qian et S3Ip.ghavarman, Mallmann (p. 21 et n. 1)
donne respectivement les dates de 147-186, 223-279 et 252 ap. J.-C. Demieville ales
memes dates, sauf pour la traduction de Zhi Qian: 223-253 (Pensee unique, p. 233-235).
63 Cf. Lamotte, Concentration, p. 43-44; Demieville, YogiiciirabhUmi, p. 358; Harri-
son 1998b, p. 556, n. 15.
64 T. Rep., p. 4-5. Cf. Huntington 1980, p. 654-655; id. 1996, p. 45, n. 6.
65 Haesner, p. 427. Cf. notamment Fus. 1987, p. 74a, n. 34; id., 1999, p. 532, n. 17 et
p. 547(-548), n. 57.
LES SOURCES DE LA SUKHAVATI 369
Cependant, si·les donnees chinoises sont complexes
66
, il est acquis
depuis un certain· temps deja qu'An Shigao ("Shigao-l'Arsacide") n'a
jarnais traduit Ie "Grand Siitra"67. ·En effet, la mention de cette paternite
n'apparait pour la premiere fois que dans Ie catalogue de Fei Changfang,
datant de 597, dont Demieville a vigoureusement denonce les defauts
68
;
en outre, aucune des reuvres valablement identifiees d' An Shigao ne traite
de relevant du Grand Vehicule
69
.
D'autre part, il est generalement adrnis que Zhi Loujiachen ("*Lokak.-
Yuezhi") n'est pas Ie traducteur de T. 361: cette traduction n'appa-
rut ni dans ses biographies, ni dans les catalogues les plus anciens, et Ie voca-
bulaire ne correspond pas a celui de ses ouvrages reconnus
70
. Par contre, des
sources chinoises anciennes identifient T. 361 cornrne un texte de Zhu Fahu
datant de 308, que 1'on croyaitperdu
71
• D'autres
sources, enfin, 1'ont reconnu plutot cornrne l'reuvre de Bo Yan72, en 258:
c'est cette derniere attribution qui semblait prevaloir jusqu'il y a peu
73
.
66 Sur les attributions et les datations des td. conservees ou pretenduement perdues, v.
essentiellement Fujita, p. 23-96; id. 1985, p. 59-61. Cf. Blum, p. 148-149; Adachi, p. 21-
22; Kao Kuan-ju, EOB, 1-3, s.v. Amitabha-vyUha, p. 464a-465b; Mochizuki, p. 241-245;
Erac1e 1984, p. 12-17; id.1988, p. 80b-85a; Nattier 2003. Concordance entre les cinq td.
conservees in Inagaki 1978, p. viii-ix.
67 Fujita, p. 24-25. Premier traducteur arrive en Chine, An Shigao passe pour avoir ete
un prince d'origine parthe, qui arriva a Luoyang en 148 (T. Rep., p. 236a). Cf. Bagchi,
p. 8-37; Dantinne, p. 12; Demievilie, YogacarabhUmi, p. 343-347, et LCI. § 2070; Shih,
p. 4-12; Tsukamoto, p. 78 ss; Ziircher 1959, p. 32-34. V. Forte 1968.
68 V. Ducor 1998, p. 184, en note. Cf. T. 49, 2034, k. 4, p. 50b.
69 Dantinne, p. 12; Demieville, YogacarabhUmi, p. 353, n. 1; I.CI. § 2070, infine; Ziir-
cher 1959, p. 33.
70 Fujita, p. 35-37; Harrison 1998b, p. 556, n. 16. Zhi Loujiachen (abr. Zhi Chen) arriva
a Luoyang en 167 (T. Rep., p. 277b); i1 est souvent presente comme l'introducteur du Mahayana
en Chine. Cf. Bagchi, p. 37-44; Dantinne, p. 11-16; LCI. § 2071; Lamotte, Concentration,
p. 66-72; Tsukamoto, p. 98 ss; Shih, p. 13-17; Zurcher 1959, p. 35-36. Sur les Yuezhi,
v. Lamotte, Histoire, p. 495-497; id., Traitt! 3, p. IX et 1584, n. 2; Dantinne, p. 11, n. 40.
71 Cf. Fujita, p. 37-41; Blum, p. 148, n. 4 (3). Cetteattribution est soutenue par Mochi-
zuki (p.245; cf. LCI. §2073), Tsukamoto (p. 193 ss) et Hirakawa (p. 32-33). Zhu Fahu
etait d'une famille d'origine Yuezhi etablie a Dunhuang (T. Rep., p. 259a). Cf. Bagchi,
p. 83-114; Demieville, YogacarabhUmi, p. 348; LCI. § 2074; Lamotte, Concentration,
p. 81-85; Shih, p. 33-37; Ziircher 1959, p. 65-70.
72 "Bo" designe la dynastie royale de Kuca, a laquelle Bo Yan aurait ete affilie. Mais
i1 y a deux Bo Yan: Ie premier, qui passe pour avoir vecu sous les Wei (220-265) (T. Rep.,
p. 253a; Bagchi, p. 79-81; Shih, p. 19),n'a peut-etre jamais exist6; Ie second vecut SOllS les Liang
anterieurs (320-376); cf. Lamotte, Concentration, p. 79-81, 98-99; Ziircher 1959, p. 76-77.
73 Cette attribution etait admise dans la tradition bouddhique japonaise ancienne, notam-
370 rnR6ME DUCOR
A defaut des traductions d' An Shigao et de Zhi Loujiachen, c'est donc
celIe du disciple de ce del)1ier, Zhi Qian ("Qian-Ie-Yuezhi")14, qui s'est
imposee comme la plus imcienne (T. 362): les catalogues la datent de
rere Huangwu de Da Di, soit 222-229 ap. l_C.7S. Mais depuis quelques
decennies, certains specialistes, notamment japonais, considerent que
T. 362 pourrait etre la traduction originelle de Zhi Loujiachen, tandis que
T. 361 serait sa revision par Zhi Qian
76
. Cette hypothes.e a ete reprise
recemment par Paul Harrison 77. De fait, les deux traductions sont tres
proches et renvoient sans doute a une recension archa'ique distincte de la
tradition constituee par les autres traductions chinoises, ainsi que par la
version sanskrite nepalaise et la traduction tibetaine
78
.
ment par Shinran, Gyonen et Zonkaku (Blum, p. 148, n. 4 (1)). Elle l'etait encore par des
specialistes n:cents, dont Fujita qui penchait pour Ie Bo Yan des Wei en 258 (Fujita, p. 41-
51; Hob6girin 7, p. 997b). Cependant, celui-ci n'ayant peut-etre jamais existe, il faudrait
examiner la possibilite d'accorder l'attribution a son homonyme des Liang anterieurs (De
Jong, p. 354-356). .
74 Zhi Qian etait un laIc descendant a la 2· generation d'une famille immigree (T. Rep.,
p. 275b-276a); cf. Dantinne, p. 17; I.Cl. §2073; Lamotte, Concentration, p. 74-79; Shih,
p. 21-23; Tsukamoto, p. 193 ss; Ziircher 1959, p. 47-51.
75 T. 55, 2146, k. 1, p. 119b, etc. Cette anteriorite est soutenue par Ziircher (1959,
p.50; 1984, p. 190b), Lamotte (Traite 3, p. xxxiv-xxxv), Mochizuki (p. 243), Hubert Durt
(Hobogirin 7, p. 997b), et G6mez, p. 126, n. 3, et p. 130.
76 Cf. Fujita p. 55-61; id., 1985, p. 60; id., 1994, p. 14. La parente de T. 361 et 362 avait
deja ete signalee par Ie Pere Wieger (p. 12), qui counaissait son chinois; v. aussi Demie-
ville, "Pensee unique", p. 235.
77 Harrison 1998b, p. 556-557; et HHM, p. 179-181, 183, qui mentioune une etude
complete en preparation par Harrison. Cf. Fus. 1999, p. 532, n. 17; Nattier, p. 76, n. 13.
La question n'est pas abordee in Harrison 1993.
78 T. 361 et 362 sont ainsi reunis sous l'appellation de "Grand Siitra ant6rieur" (shoki-
Muryojukyo "Early L.Sukh" ou "Early Recension"), les autres td. chi-
noises ainsi que la version sanskrite nepalaise et la td. tibetaine etant c1assees comme "Gd
Siitra post6rieur" (koki-Muryojukyo "Later L.Sukh" ou "Later Recension")
(Fujita p. 167; id. 1980, p. 118; HHM, p. 179). n y a parente non seulement entre la
version sanskrite nepalaise (in! p. 365, n. 83-85) et la traduction tibetaine du IX" s. par Jina-
mitra, DiinaSila et Ye-ses-de (Tib. Trip. XXlI, 760-5), mais aussi avec la td. chinoise r6alisee
par Ie second Bodhiruci (572-727) et inc1use par lui dans Ie Ratnakiita aux environs de 706-
713, Ie Wuliangshou Rulai hui (T. 310-5), qui appartient donc a l'ere des
td. nouvelles (xinyi cf. Adachi, p. 22; HHM, p. 182-183. V. aussi la table de con-
cOI:dance entre Ie skr. et Ie tib. in Inagaki 1978, p. viii-ix.
LES SOURCES DE LA SUKHAVATI 371
Pour autant, ce ne sont pas ces traductionsarchaYques (guyi du
"Grand Siitra,,79 qui fournissent Ie plus ancien repere chronologique chi-
nois de l' existence du .culte d' Amitabha, comme nous Ie verrons plus loin.
2. Sources sanskrit
Les principales sources de la Sukhavafi dans la litterature bouddhique
indienne sont ces deux monographies que constituent Ie "Grand" et Ie
"Petit" Sukhiivatfvyuha-sutra; s'y ajoutent diverses mentions dans vingt-
neuf autres textes sanskrits
80
, sans parler des traductions conservees en
tib6tain ou en chinoiS
81
. Le "Petit Siitra" n'est accessible qu'a travers ses
editions japonaises en siddham du xvrn
e
S82., tandis que trente-huit
manuscrits nepalais du "Grand Siitra" sont conserves de par Ie monde
83
.
La plupart de ces derniers sont sur papier et datent du XIXe S84., mais deux
d'entre eux sont sur olles et remontent au XIIe s. Parmi ceux-ci, Ie meilleur
est Ie ms. Sakaki, aussi dit "Ashikaga" ou "Ryiikoku", qui date des envi-
rons de 1147-1167 ap. J._C.
85
.
79 On n'y inclura pas T. 360: son attribution a Kang Senkai ("*Saip.ghavannan-le-Sog-
dien"), avec la date de 252 (Bagchi, p. 77), est aujourd'hui abandonnee au profit de Bud-
dhabhadra (359-429) et son disciple Baoyun en 421, ce qui Ie rejette dans les
td. anciennes (jiuyi Cf. Fujita, p. 69-97 (id. 1994, p. 13; 1996b, p. 7; 1996c, p. 37),
repris par Durt (H8bfJgirin 7, p. 999ab), Inagaki (1994, p. 56, n. 71) et Nattier (2000, p. 76,
n, 13; 2003, p. 189); v. aussi G6mez (p. 126), Harrison (1998b, p. 557, n. 22) et Blum
(p. 147, n. 2).
80 Ceux-ci sont repertories in Fujita, p. 161-164; une douzaine d'entre eux sont utili-
ses in Schop. 1977.
81 Soit 290 textes chinois dans Ie Canon de TaishO (T.), repertories avec leurs equiva-
lents tiMtains in Fujita, p. 141-161. Cf. Yamada Isshi, I, p. 175-196; Blum, p. 162-164.
L'Amida-butsu setsurin acheve en 1770 par Ie Japonais ZenshO Keijo n'est pas une liste
de titres sanskrits (Fus. 1999, p. 537), mais une compilation de tous les extraits pertinents
tires de plus de 200 td. chinoises (Shinshu zensho, 7, p. 59-178; cf. Mbdj. 1, p. 73b-75a);
c'est la restitution de leurs titres en sanskrit qui est fournie par Kanaoka ShiiyU in EOB 1-3,
p. 425b-429a.
82 Ducor 1998, p. 99-105.
83 V. Fujita 1992, p. vii-xi; id. 1996a, p. v (cf. Fus. 1999, p. 531, n. 14). TIs ont ete edi-
tes par Fujita (1992, 1993, 1996a); v. aussi l'etude synoptique de Kagawa 1984. Aucun
ms. du "Grand Siitra" n'a ete trouve au Japon, contrairement II. ce qu'indique Huntington
(1980, p. 655; 1996, p. 44).
84 L'edition de Muller (1883, p. 1-78), aussi dite "edition d'Oxford", et sa traduction
(1894) se fondent sur cinq de ces ms. sur papier. .
85 Amene au Japon en 1923 par Sakaki Ryozaburo (1872-1946), il fut offert en 1958
372 rER6ME DUCOR
Cependant, l'bistoire de la tradition manuscrite du "Grand Sutra" a ete
revisitee tout recemment par Paul Harrison, Jens-Uwe Hartmann et Mat-
suda Kazunobu, qui ont ieuni et publie vingt-six fragments de trois folios
d'un manuscrit sanskrit de la Collection Martin Sch!Zlyen (SC 2382). Cette
nouvelle source apparalt d'une importance considerable: non seulement
elle remonterait aux environs des VIe_ VIIe siec1es, mais, en outre, son
contenu differe suffisamment de la tradition nepalo-tiMtaine et de la
recension chinoise archaique pour temoigner d'une troisieme tradition; de
plus, comme ces fragments proviendraient d' Afghanistan, peut-etre meme
de la region de Bamiyan, ils constitueraient aussi un jalon geograpbique
significatif, alors que Ie berceau meme de la tradition de la Sukhavatl
demeure inconnu
86
.
A defaut de pouvoir se lancer ici dans l'analyse du delicat probleme
de la datation du "Grand Sutra" , I' on se contentera de mentionner encore
les points suivants. Selon Lamotte, "l'original indien circulait deja ala
fin du Ier siecle de notre ere, et au plus tard au IIe", Schopen evoquant
aussi une periode anterieure au IIe siec1e, tandis que Harrison parle du
milieu du IIe siec1e ou plus t6t
87
. Quant a Fussman, qui la compare avec
Ie "Sutra du Lotus", l'actuelle version du "Grand Sutra" daterait "proba-
blement du IVe siec1e et des interpolations posterieures sont probables",
mais son noyau ancien, en sanskrit hybride, remonterait "aux environs du
Ier siec1e de n.e."88. Enfin, au terme d'une analyse serree des deux sutra
de la Sukhavatl, Fujita conc1ut sur une fourchette allant du Ier au TIe s. pour
leur version la plus developpee, tout en remontant a une periode situee
entre Ie Ier s. avo J.-C. et Ie Ier s. ap. J.-C. pour leur noyau ancien
89
. Quant
par la veuve de ce dernier au patriarche du Hompa-Honganji, qui Ie fit deposer 11 l'Univer-
site Ryiikoku (KyOto); son edition par Ashikaga Atsuuji (Ash.) a ere revisee dans la synopse
de Fujita. Le second ms. sur olles, conserve aux Archives Nationales de Katmlindu, est date
de 1152-1153 mais il est incomplet. Reproduction: Inokuchi 1986.
86 HHM, p. 181; Yamada 2002, p. 111-114. Cf. De Iong, p. 363-364.
87 Lamotte, Concentration, p. 57; Schop. 1977, p. 204; Harrison 1988b, p. 557. Cf. Dan-
tinne, p. 42.
88 Fus. 1999, respectivement p. 538-539, et p. 535. L'influence du "Grand Siitra" sur
Ie "Lotus" a ete relevee par Fujita (1980) et Huntington (1996, p. 95-99); elle parait "au
moins" douteuse 11 Fussman (1999, p. 527, n. 5).
89 Fujita, p. 231-235 (cf. id. 1996b, p. 9-10). D'autres font remonter les noyaux scrip-
turaires les plus archaYques jusqu'au milieu du n
e
s. avo I.-C., ou au ill
e
s. av., voire jusqu'aux
IYe_ille s .. av. (respectivement Huntington 1988, p. 87b; id. 1996, p. 96-99; Inagaki 1998a,
p ~ 37-47).
LES SOURCES DE LA sUKHA VATl
373
a l'eventuelle anteriorite de l'un de nos deux sutra sur l'autre, il ne sem-
ble pas possible de trancher
9o

ll. Interpretation
Les indianistes occidentaux a s'etre pencMs sur Ie contenu de ces sour-
ces litteraires canoniques sont encore relativement rares. C'est ce qui donne
un interet tout particulier a l'artic1e publie par Fussman en 1999, dont on
relevera ici quelques elements.
En guise d'introduction, ce dernier denonce, notamment, un contre-sens
qu'il juge enorme et caricatural, soit "la traduction anglaise generalement
acceptee de Sukhavatl par 'Pure Land' ou 'Land of Bliss»" (p. 528). En
fait, ces deux expressions renvoient a deux notions differentes
91
. Dans la
terminologie elaboree par Ie traducteur Kumarajlva (344-413), l'expres-
sion chinoise "terre pure" (jingtu ±) designe un purifie,
par opposition a un qui ne l'est pas, i.e. une "terre souillee"
(huitu En outre, l'expression "Terre Pure" en vint rapidement a
designer Ie champ de buddha purifie par excellence, i.e. celui d'Ami-
tabha
93
• D'abord reproduit en transcription phonetique
94
, son nom de
"SukhavatI" fut traduit en chinois par "Beatitude" (Anyang 3'i:l€) et
"Bonheur-Paisible" (Anle debut du IIIe S.; puis, Kumarajlva Ie
rendit par "Bonheur-Supreme" (lile usage enterine par Xuanzang
(602-664). On ne voit pas ce que Ies expressions "Pure Land" ou "Land
of Bliss" auraient de «sectaire» ou d' «ethere», pour reprendre les termes
90 V. Fujita, 1980, p. 118. Cf. Nattier 2000, p. 76, n. 13, et p. 92, n. 64; Fus. 1999, p. 535,
n.23.
91 Le glossaire de Gomez porte aussi it confusion puisqu'il donne sukhavatl comme equi-
valent de jingtu ("pure land", p. 318). L'entree "Land of Bliss" (p. 306) est plus claire; de
meme l'introduction (p. 9, n. 15).
92 Lamotte, Vimalaklrti, p. 397 ss.; H6b6girin 3, p. 199b-201b. L'expression "terre
pure" n'apparait pas encore dans la td. du "Petit Siitra" achevee par Kumarajlva en 402
(T. 366), mais el1e figure dans sa td. du "Siitra du Lotus" datant de 406 CT. 9, 262, k. 6,
ch. 16, p. 43c12; Robert, p. 288).
93 Fujita 1996b, p. 20-21; id. 1996c, p. 33-37.
94 Soit: "Xuhemati" @ilfiiJ lou qui renverrait it la gandhiirl "Suhamadi" (Fujita,
p. 432-433; De Jong, p. 361; cf. Harrison 1990, p. 31, n. 2). Plusieurs specialistes estiment
ainsi que ce prakrit du Nord-Ouest de I'Inde serait la source des traductions chinoises
archa'iques du "Grand Siitra" (Fujita, p. 239-247; Dantinne, p. 1-2; De Jong, p, 363; Naka-
mura, p. 205). Fussman est d'un avis oppose (Fus. 1999, p. 540, n. 41).
374 JEROME DUCOR
de Fussman (loc.cit.), qui traduit lui-meme "SukhavatI" par "terre de
bonheur" (p. 525). On relevera par ailleurs que, selon les sources sans-
krites, Ie champ de buddha d'Amitabha n'est pas tant une "terre"qu'un
univers (Sukhtivati lokadhtitu)95.
1. Le titre du "Grand SukhavatIvyiiha"
Abordant la question du titre du "Grand Sutra", Fussman (p. 575-577)
attire notamment I' attention sur les appellations figurant dans la conclu-
sion du texte:
"Fin du chapitre Agencement (vyuha)96 de la SukhavatI d'Amitabha, [con-
stituant] la celebration des qualites du Bienheureux Amitabha [et presentant]
l'entree des bodhisattva dans Ie stade irreversible"97.
Fussman en conclut que ce sutra "est donc un chapitre (parivartaM
extrait d'un ouvrage plus long aujourd'hui perdu, dont nous ignorons
Ie titre" (p. 576). Cependant, l'expression "chapitre" ne doit pas etre
95 "Grand Siitra", Ash. 30:9, § 53, passim; cf. "Petit Siitra", § 2. L'adjectif substan-
tive Sukhiivati est feminin, tandis que lokadhiitu est ferninin ou masculin: on peut donc
traduire par "la Bienheureuse" ou par "l'univers Bienheureux". Burnouf a "la terre for-
tunee" (Intro., p. 89, n. 1); LVP: "Ie bienheureux univers"(Dogme, p. 70; cf. Boud-
dhisme, p. 415, ad p. 215) ou "the happy world" (HERE, s.v. Cosmogony, p. 137, n. 6);
Filliozat: "1' element du monde (lokadhiitu) appele Sukhavatl, I"Heureuse'" (lCI. § 2331).
Lamotte ne traduit presque jamais, mais lorsqu'ille fait c'est par "univers Fortune" (<<Maii-
jusri», p. 18), ce qui l'amene a confondre sukha et bhadra (cf. Vimalakirti, p. 385 et 399).
Enfin Dantinne, citant Ie "Petit Siitra" (§ 2), a "cet univers est appele 'Felicite'" (p. 120,
n. 192), ce qui traduit sukha mais non pas avatf (cf. Fujita 1996c, p. 37-38). BHSD, p. 464a,
597a.
CO 96 Filliozat traduit ce terme complexe par Burnouf traduit ce terme complexe par "cons-
titution" (Intro., p. 88) ou "enumeration" (Lotus, p. 133), Filliozat par "developpement"
ou "manifestation" (ICI. § 2003, 2011), et Dernieville par "omementation" (id. §2113).
Lamotte Ie rend par "deploiement", "splendeur", "manifestation", voire "merveilles"
(Vimalakirti, p. 100, 120; Somme, p. 320; Concentration, p. 225); Dantinne a "splendeur"
ou "omementation" (p. 124). Fussman traduit par "merveilles" (p. 576). Cf. Schop. 1977,
p. 196; Gomez, p. 3, n. 1; BHSD, s.v. 2, p. 520a. V. aussi l'hypothese de Nattier 2000,
p. 73(-74), n. 6.
97 Skr. bhagavato 'mitiibhasya gU1Japarikfrtana/'fl bodhisattviiniim avaivarttikabhiImi-
pravdalJ. Amitiibhasya Sukhiivatf-vyuha-parivartalJ. samiiptalJ. (Ash. 67 :2-4). Supprirner Ie
point apres pravesaQ. dans Ash. 67:3, la negligence du sa/'fldhi etant l'une des caracteris-
tiques des mss nepalais (cf. Ash. v:25-vi:3). Le titreproprement dit est donc Amitiibhasya
Sukhiivati-vyuhao, precede des deux "sous-titres" bhagavato 'mitiibhasya gU1Japarikfrta-
na/'fl et bodhisattviiniim avaivarttikabhitmipravdalJ..
LES SOURCES DE LA SUKHAVATI 375
prise ici au sens strict du tenne - soit une subdivision d'un ouvrage plus
grand - mais au sens large de "texte". Lamotte a montre que des mots
comme sutra, nirdesa, dhannaparyiiya, dhannamukha ou parivarta. peu-
vent constituer autant d'equivalents
98
• Le VimalakirtinirdeSa, comme bien
d'autres siltra du Mahayana, se voit ainsi attribuer plusieurs titres dans sa
conclusion, dont celui de "Chapitre de la liberation inconcevable" (Acin-
qui ne pretend pas renvoyer a un autre texte:
l'expression constitue simplement un sous-titre paraphrasant de maniere
condensee la doctrine essentielle du siltra. Que si l' on tenait malgre tout
a conserver la signification "chapitre" stricto sensu pour Ie "Grand Siltra",
on pourrait bien etre tente d'y voir une allusion au Ratnakuta; comme Ie
fait Hirakawa Akira
100
. Cet ensemble canonique, conserve en chinois et
en tibetain, est en effet constitue de quarante-neuf siltra, dont Ie cinquieme
n'est autre que notre "Grand Siltra". Cependant, si la version tibetaine du
Ratnakuta classe bien ces differents siltra en "chapitres"(le'u), elle donne
de notre texte Ie titre sanskrit Arya-Amitiibha-vyuha-niima-mahiiyiina-
sutra; tandis que sa traduction chinoise les intitule "assemblees" (hui it,

2. La nature de la Sukhiivatf
Fussman (p. 574-575) signale aussi 1'importance du deuxieme de ces
sous-titres du "Grand Siltra", soit: "entree des bodhisattva dans Ie stade
irreversible" (bodhisattviiniim avaivarttika-bhami-pravesa); et il affirme
que ce sous-titre correspond "expressement" au 21 e vceu du futur Amitii-
bha. Or, ce vceu ne mentionne pas Ie stade irreversible, mais bien 1'ultime
etape ou Ie bodhisattva n'est plus separe de la manifestation du parfait
eveil que par une seule naissance (ekajiitibaddha)102: il s'agit donc d'un
98 Lamotte, Virnalakirti, p. 64-65, qui renvoie a Filliozat, LCI. § 2003, ou parivarta est
rendu par "tour d'horizon, tour autour d'une question"; cf. BHSD, s.v. 2, p. 329ab.
99 Selon Ie tibetain, ou parivarta est traduit par le'u; tandis que Xuanzang afarnen
soit "exposition de la Loi" (dharrnaparyaya). V. Lamotte, Virnalakirti, p. 392-393 et n. 42.
100 Hirakawa, p. 33-34.
101 Cf. sup. p. 370, n. 78. L'histoire du RatnakU{a, compile par Bodhiruci de 706 a 713,
est mal connue: v. ICI. § 2011,2037,2113; Lamotte, Traite 4, p. 1884-1887; id., Virnala-
kirti, p. 64; Eracle 1988, p. 77-79; Dantinne, p. 28-34; Pagel, p. 53-78.
102 Ash. 14:14; (Schop. 1977, p. 201); cf. Ash. 46:25 (§96-18) et 48:23-24 (§100;
Dantinne, p. 171). Sur ekajatifpratilbaddha, v. Ducor 1998, p. 74, n. 230. Contrairement
376 JEROME DUCOR
stade distinct et posterieur 11 l'irreversible
103
• C'est Ie 46
e
vreu qui assure
l'irreversible (avaivarttika, Ash. 21 :5), tandis que Ie 11 e vreu y fait
allusion, en promettant que les etres nes en Sukhavati ne sont pas sans
etre "fixes" (niyata )104 dans la suite de leur parcours jusqu' 11 l' atteinte du
nirvfu.1a, raison pour laquelle Ie stade irreversible est aussi nomme niyata
bhiimilO5. En outre, les vreux 511 9 garantissent l'obtention des cinq scien-
ces sublimes (abhijfia, Ash. 11: 14-12: 11), dont la plenitude est precise-
ment propre 11 ce stade
106
. Le 47
e
vreu precise egalement que l'atteinte de
l'irreversible en Sukhavat! est due 11 l'acquisition, en celle-ci, des trois
degres de l'endurance envers la non-naissance des choses (anutpattikad-
Enfin, Ie 16
e
vreu assure que les etres en Sukhavati
ignorent jusqu ' au nom du mal (akusalasya namadheya, Ash. 13: 15): cette
particularite nous confirme que ce champ de buddha est parfaitement
purifie et que ses habitants demeurent au
moins au 8
e
stade, parce qu'ils ont coupe toutes les passions (sarvakle-
sasamucchinna)I09, ainsi que l'atteste la doctrine c1assique de l'''Exegese
sur la Perfection de sagesse" (Prajfiaparamitopadesa)llO.
a ce que laisse entendre Fusman (p. 599, n. 80), Ie 21
e
vceu ne traite pas non plus de la
mesure de la vie des etres nes en SukhavatI, ce qui incombe au 14
e
vreu
(Dantinne, p. 153-154).
103 Ducor 1998, p. 34-36.
104 Ash. 12:17 (et §85; Ash. 40:19). Surla notion d'etre "fixe", v. Ducor 1998, p. 73
et n. 226.
105 Lamotte, Concentration, p. 157, alinea 4c. Sur Ie parcours des bodhisattva, v. Ducor
1998, p. 30, n. 39.
106 Cf. Ash. 46:14-17; §96. Lamotte, Traitt! 4, p. 1855-1860; Schop. 1977. p. 190-191,
n; 20.
107 Au cours de la carriere du bodhisattva. cette endurance passe par trois stades: adhe-
sion purement verbale, conviction preparatoire intense et acquisition defmitive; c'est cette
derniere qui donne acces a l'irreversible. V. Lamotte, Vimalakfrti, p. 411-412; Traitt! 4,
p. 1788, n. 1; et Concentration, p. 160-162, en note.
108 Lamotte, Traitt! 5, p. 2344.
109 Cf. Schop. 1977, p. 189-191, qui cite Ie Samiidhiriija-sutra.
110 Lamotte, Traitt! 5, p. 2343-2345. Premier traite a evoquer Amitabha et la SukhavatI,
I' Upadda attribue a Nagarjuna fut edite en chinois par Kumarajlva de 404 a 406 (Dazhi-
dulun, T. 1509); son auteur se situerait au debut du IVe siec1e, dans Ie nord-ouest de l'Inde
(Lamotte, id. 3, p. viii ss.; 4, p. xv). Fussman (1999, p. 539) n'y trouve que trois mentions
de la doctrine de la SukhavatI, soit: Lamotte, Traitt!, p. 2228, 2230 et 2308. On y ajou-
tera: p. 300,465,556,601, 1361, 1927,2335 et 2432. Encore Lamotte ne traduit-il "que"
Ie premier tiers des cent volumes de l' ouvrage, et d' autres references se trouvent dans Ie
texte chinois: T. 25, p. 342ab, 343a, 529c, 708c (td. in Lamotte, Vimalakfrti, p. 401), et
LES SOURCES DE LA SUKHAVATI 377
Bref, au vu du2l e vreu, d'une part, et des neuf autres vreux mention-
nes ici, d'autre part, on constate que la Sukh1ivatl englobe deux des uIti-
mes etapes de la carriere des bodhisattva, que ce soit les deux demieres
des quatre etapes selon les Prajfiaparamita, ou la 8
e
et la loe bhumi du
Dasabhumikasutra
lll
. C'est ce que Ie "Petit Sutra" (§ 10) resume en affir-
mant que les etres nes en ce champ de buddha sont de "purs bodhisattva
(Suddhii bodhisattvalJ), irreversibles (avinivartanfya), lies par une seule
naissance (ekajatipratibaddha)"l12.
Mais il y a plus. A partir du stade irreversible, la carriere du bodhisattva
se deroule sans que celui-ci n'ai plus a fournir aucun effort (anabhoga),
de maniere quasi automatique
1l3
. Tant et si bien que la g
e
bhumi est, elle
aussi, comprise dans la Sukh1ivatl. Les vreux ne la citent pas nomme-
ment, puisqu'elle est necessairement incluse entre Ie 8
e
et Ie lO
e
stade.
Mais Ie 2g
e
vreu mentionne bien que les etres en Sukh1ivatl seront pour-
vus des quatre pleins savoirs (pratisaT[lvid), qui font Ie bon predicateur.
Or ceux-ci sont precisement acquis dans la g
e
bhumi, stade ou Ie bodhi-
sattva se cons acre specialement a enseigner aux etres
114
, cette activite
etant d'ailleurs assuree par Ie 24
e
vreu. Tel que l'evoquent les deux Sukhii-
vatlvyuha, Ie champ de buddha d'Amit1ibha constitue donc, pour ses habi-
tants, un univers qui n'est autre que la sphere d'activite dans laquelle se
deroule automatiquement Ie parcours irreversible des bodhisattva (avini-
vartanacarya)ll5, ultirne tranche de leur carriere, entre Ie moment ou ils
atteignent l'irreversible et celui ou ils deviendront effectivement buddha.
Cette interpretation de la Sukh1ivatl est d'ailleurs bien connue, puisqu'elle
n'est autre que celIe qui allait se transmettre en Chine, notamment dans
712a; cf. id, Traite, 3, p. xxxv; Fujita, p. 143. Concernant les omissions mentionnees par
Fussman (loc.cit.), l'absence d'Amitayus en Traite 2336 n'a pas de quoi etonner puisque
ce passage ne concerne que les buddha de l'univers Saha; tandis que Traite 403-406, 442-
456 et 2352-2364 constituent un traitement general, ou aucun champ de buddha n'est cite
nomement. En revanche, Fussman aurait ete justifie de mentionner l' omission flagrante en
Traite 594-595 (cf. plus bas, p. 391).
III Lamotte, Traite 5, p. 2374, 2376.
112 Cf. Dantinne, p. 172.
113 References in Ducor 1998, p. 30, n. 42.
111 Lamotte, Traite 3, p. 1614 ss; id., Concentration, p. 157"158; id., Somme, p. 202,
293-294; Rahula, p. 159, n. 11 et p. 166; Dayal, p. 259-267, 291.
115 Lamotte, "Maiijusn", p. 12; Dantinne, p. 125, n. f.
378 mROME DUCOR
la tradition dite de Shandao
Jl6
. Ainsi, panni ses premiers representants,
Tanluan (476-542) expliquera comment les etres nes en Sukhavat! n'ont
pas a y passer par chaque bhiimi successivement, conformement au prin-
cipe canonique du parcours par saut (vyutknintaka)ll7 rendu possible a par-
tir de la 8
e
bhiimi
Jl8
.
Comme on peut Ie cons tater, la naissance en SukhavatI constitue plus
qu'un simple "enjoyment of its sensual and spiritual pleasures, so lavishly
described throughout the sutra"Jl9. Certes, la description des ornements
(alal'{lkiira) de ce champ de buddha a de quoi fasciner par sa richesse, mais
elle reste relativement banale si on la compare a d'autres descriptions,
comme celle, notarnment, de l'Uttarakuru
l20
. Et, de fait, ce n'est pas Ie
"monde-receptacle" (bhiijanaloka) de la Sukhavat! avec son decor qui se
trouve au creur des vreux. Car ce sont bien ces derniers qui constituent
la part originale, la substantifique moelle, du "Grand Sutra". Or, seuls
trois VlEUX sont cons acres aces ornements 121, tandis que leur tres grande
majorite, soit trente-neuf VlEUX sur quarante-sept, concerne les etres a
delivrer: vingt-sept vreux sont ainsi lies aux etres nes en Sukhavat!, tan-
dis. que les douze autres s'appliquent aux etres vivants dans les autres
champs de buddha. En outre, il n'est pas besoin de recourir a une inter-
116 Sur la tradition de Shandao, v. Ducor 1999, p. 145.
Il7 Chin. chaoyue (Tanluan: T.40, 1819, k. 2, p. 840c2; Inagaki 1998a, p. 253-257).
118 Cf. Demieville, Yogiiciirabhumi, p. 429-430; et Jacques May, art. chOjo, chOgyo
et chOotsu-shO du HobOgirin 4-5, p. 353a-360a, 346b, 360b-371a. n faut attendre Shinran
(1173-1263), fondateur du IOdo-Shinshii, pour voir Ie stade irreversible considere comme
atteignable en cette vie-ci par la foi seule, et la naissance en Sukhavat! interpretee comme
la realisation immediate du parfait eveil au moment de la mort (cf. Fujita 1996c, p. 50;
Ducor 1998, p. 36-37). Incidemment, on se demande pourquoi Fussman rend ShinshU Ii.*
par "Nouvelle Secte <de la Terre Pure>" (p. 527, n. 6); idem chez Magnin, qui a "Nou-
veau courant de la Terre pure" (p. 398).
119 Harrison 1978, p. 52. Cet article est deja ancien, et l'interpretation de Paul Harrison
a evolue: "I now believe that Sukhavat! is the forest hermitage or monastery writ large,
a perfect environment for strenuous religious practice, not for pleasurable enjoyment"
(communication personnelle).
120 Tucci, p. 217-218. Idem pour l'Abhirati Kwan, p. 109-113; Nattier,
p. 75. Cf. Fujita, p. 491-494.
121 Vceux 30-32: (Iuminosite de la SukhavatI), 31" vceu (vases de parfums pour les offran-
des aux buddha et bodhisattva), et 32· vceu (ondees de fleurs avec nuages de musique); Ie
28" vceu mentionne indirectement l'arbre de l'eveil (bodhivrkea). On nuancera donc Ie
propos de Nattier declarant (2000, p. 96-97): "the majority of his [Dharmakara'sl vows are
devoted to a description of the precise qualities of the Buddha-field he intends to create".
LES SOURCES DE LASUKHAVATI
379
pretation symbolique pour comprendre la fonction de ces ornements 122,
puisque les sutra parlent d'eux-memes. Ainsi des fameux oiseaux crees
artificiellement par Amitabha, qui sont mentionnes dans Ie "Grand Sutra"
(§ 67), et dont Ie "Petit Sutra" (§ 6) precise qu'ils chantent les trente-sept
auxiliaires de l'eveil, de concert avec la musique celeste, pour amener
les etres a commemorer les Trois Joyaux. De meme des diverses rivieres
qui repercutent Ie son de nombreuses rubriques de la Loi, dont les moin-
dres ne sont pas piiramitii, bhumi, silnyatii, et anutpattikadharma-k$iinti
123
.
Plus significatif encore, les etres, conformement au 28
e
vreu, peuvent voir
l'arbre de 1'eveil (bodhivrk$a) d'Arnitabha, meme s'ils n'ont que peu de
racines de bien (parlttakusalamilla)124; et des qu'ils Ie voient, ils obtien-
nent les trois degres d'endurance tout en devenant irreversibles
125
. Bref,
Ie commentaire de l' Upadesa a propos de l'arbre de l'eveil peut s'appli-
quer a tous les autres ornements de la SukhavatI: ils ne sont 13. que "pour
faciliter aux etres 1'audition de la Loi"!26.
Mais si ces ornements font ainsi "reuvre de buddha" (buddhakiirya)127,
cette fonction revient evidemment en premier lieu a Amitabha lui-
meme
128
, lui qui est "en pleine possession des perfections de son champ
de buddha"129. Conformement ala formule consacree, "il y demeure, s'y
tient, y subsiste et preche la Loi"!3O. C'est ce dernier point qui donne la
vraie dimension de la Sukhavati, comme l'a bien vu Ie regrette Bernard
Frank en la qualifiant de "Terre de predication" d'Amitabha!3!. Le sutra
precise encore que les etres "ne sont plus jamais separes de la vision du
buddha ni de l'audition de la Loi, mais sont exempts de rechute jusqu'a
leur accession a l'eveil"132. Comme l'assure Ie 45
e
vreu, ils peuvent ainsi
122 Comme Ie fera l'Upadda; cf. L ~ m o t t e , Traite 4, p. 1971.
123 Ash. 36:8, 10 et 13; v. §65 et 68. Cf. Dantinne, p. 216-217.
124 Ash. 16: 19. Cf. Dantinne, p. 216.
125 Ash. 48:15, §99 (v. plus haut, p. 376, n. 107). Cf.I'Upadda cite in Lamotte, Vimala-
kfrti, p. 401.
126 T. 25, 1509, k. 93, p. 712a15.
127 Cf. Lamotte, Vimalakfrti, p. 341, n. 7; et p. 364, n. 10.
128 Cf. Ducor 1998, p. 84, n. 278.
129 Skr. buddhak:jetrasal'{lpada samanvagataJ:z (Ash. 26: 18-19).
130 Ash. 26: 11, §40. V. "Petit Sfitra", § 2. Cf. Lamotte, Vimalakfrti, p. 247, n. 8; Dan-
tinne, p. 78 et 80.
131 Frank, p. 38.
132 Skr. sarva te 'virahita buddhadharsanena dharmasrava!:zenavinipatadharmalJ.o
380 JER6ME DUCOR
entendre la Loi a volonte, du Buddha en personne. Des lors, leurs pen-
sees sont, notamment, pour la pratique de la perfection de sagesse (praj-
iiaparamita, Ash. 51: 17, § 111), et ils obtiennent les cinq y com-
pris celui de buddha
133
• Tant et si bien qu'ils peuvent a leur tour precher
, la Loi avec omniscience (sarvajiiata, Ash. 51 :5, § 109), conformement
au 24
e
vreu.
En resume, l'on peut considerer que la Sukhiivatl fonctionne comme
une caisse de resonance du Dharma, et1'on retiendra ce juge'ment perti-
nent du Pere de Lubac: "Cette «Terre Heureuse» n'offre pas de plaisirs
oil I' on s' enlise. Elle est avant tout comme une vaste ecole, OU I' on se
rejouit essentiellement des progres qu'on y accomplit dans l'intelligence
de la Loi"134. Cependant, cet equipement en connaissance (jnanasGJ?7.bhara)
serait insuffisant sans son equivalent en merites (pu1JyasaTflbhara). Les
habitants de la Sukhiivatl obtiennent ce dernier principalement en allant
rendre hommage aux myriades de buddha de toutes les directions 135,
conformement au 22
e
vreu, tant il est vrai que les buddha constituent Ie
supreme champ de merites Et, comme l'affinne
Ie sutra, cela est rendu possible parce que les etres ont ete "embrasses
(parigrahe1Ja) par Ie secours des vreux primordiaux (pilrva-
pra1Jidhana) du Tathiigata AmitiiYUS"137.
Aucune de nos sources epigraphiques ou litteraires - qui seules nous
occupent ici - ne presente la Sukhiivatl comme un "paradis" (svarga),
yavadbodhiparyantalfl (Ash. 49: 18-19, § 104; cf. Lamotte, Concentration, p. 211, en note;
cf. p. 164). Sur Ie bienfait de ne plus etre separe du Buddha, v. id., Traite 4, p. 1924-
;:-1927. Sur Ie Buddha enseignant en son champ, v. Rowell, E. B. VI-4, p. 381-385.
133 Ash. 51:22, §112. Sur les cinq yeux, v. Ducor 1998, p. 83, n. 277.
134 De Lubac, p. 52; cf. p. 262. Selon La Vallee Poussin aussi, la Sukhavatl "is not an
everlasting paradise" mais "a blissful purgatory and a school, not only a place of retribu-
tion" (HERE, s.v. "Blest, Abode of the", p. 688b).
135 Ash. 47:2-5, 50:2-51:3; §96:20-21, §105-108. V. "Petit Siitra", §5 (Dantinne,
p. 114, n. 165). Cf. Lamotte, Traite 5, p. 2431-2433; Rowell, E.B. VI-4, p. 402.
136 Cf. Lamotte, Concentration, p. 231, n. 266; Traite 1, p. 524; ibid.'4, p. 1970; H6b6-
girin 3, p. 200ab; Rowell, op.cit., p. 402-405.
137 Ash. 51: 1, §108 (cf. id., 48: 19, §99). Fussman traduit: "parce qu'ils ont ere englo-
bes autrefois dans Ie vreu solennel et I'hero'ique resolution du tathligata Amitayus" (p. 569,
574). Burnouf rend adhi-rthiina par "benediction" "Lotus" p. 363; cf. L VP, Kosa, ch. 3, p. 31,
n. 2), comme Lamotte (Concentration, p. 269), qui a aussi "intervention 8urnaturelle"
(Vimalakirti, p. 321). Demi6ville Ie rend par "gr§ce efficiente" (YogacarabhUmi, p. 386), et Girard
par "force de soutien" (p. 209). Cf. Rowell, E.B. Vll-2, p. 154 S8.; BHSD, S.V. 3, p. 16a.
LES SOURCES DE LA SUKHAVATI 381
quoi qu'en dise Fussman (p. 558-559), car ce terme ne s'applique qu'aux
divers sejours des dieux (deva) au sein du saf{lsara. Certes, Ie 38
e
vreu
garantit que les etres nes dans Ie champ de buddha d'Amitabha gofitent,
pour dire Ie moins, Ie comble du bonheur concevable dans Ie cycle des
naissances et des morts: ce sommet est celui du moine arhat entre dans
Ie 3
e
dhyana, qui fait naltre dans Ie paradis des dieux SUbha1qtsna138.
Cependant, il y a mieux que ce bonheur divin (divyasukha) , pourtant
superieur au bonheur humain (manu$yasukha): Ie 33
e
vreu assure Ie bon-
heur qui depasse celui des dieux et des hommes (devamanu$yasamati-
kranta-sukha, Ash. 17: 23). De fait, les sources canoniques anciennes s' ac-
cordent a dire que Ie bonheur des dieux est lui-meme depasse par Ie
bonheur du nirvaI.1a (nirvalJ-asukha) 139. Et celui-ci est bien Ie bonheur
constitutif du champ de buddha d' Amitabha tel que Ie definissent les
antiques stances du "Grand Sutra", soit "Ie bonheur du plan du nirvaI.1a"
(nirvalJ-adhiitusaukhya, § 19_8)140. En definitive, la raison d'etre de la
Bienheureuse, comme l'affirme une autre de ces stances, n'est autre que
de procurer ce bonheur (sukha) aux etres confrontes ala souffrance (dulJkha)
qui constitue la premiere des quatre Nobles Verites
141
• Et l'on ne s'eton-
nera donc pas que Ie mot dulJkha n'existe meme pas en Sukhavati
142
, confor-
mement au 16
e
vreu.
En outre, la naissance en Sukhavati n'est pas une fin en soi. Certes, Ie
21 e vreu promet que les etres qui y sont nes ne sont plus lies qu' a une seule
naissance (ekajatibaddha) avant de devenir buddha (plus haut, p. 375);
mais il assure aussi que ses habitants, fixes dans la pratique d'Excellence-
Universelle (samantabhadra-carya-niyata)143, peuvent continuer Ie par-
cours des bodhisattva (bodhisattvacarya) pour se consacrer au nirvaI.1a
complet (parinirvalJ-a) de tous les autres etres, comme Ie confimie Ie
138 Cf. Dantinne, p. 196, n. 67. References in Ducor 1998, p. 62, n. 165; ajouter Inagaki
1998a, p. 279.
139 Lamotte, Traite 3, p. 1710; pour 1es sources palies: HobOgirin 7, p. 935ab.
140 Seion la lecture de Fujita 1996c, p. 47 et n. 32. Ash. 8: 10 a nirviilJalokad1uitusaukhya.
141 Ash. 21:22, §29-2; cf. "Petit Siitra" , § 2. On s'etonne que Fussman traduise ici dUQkha
par "malheur" (1999, p. 533, n. 20), comme Gomez qui a "misfortune" (p. 76). La td. Miil-
1er est viciee par des mss defectueux (1894a, p. 23).
142 Ash. 36:20, §69; Dantinne, p. 139-140. Cf. Fus. 1999, p. 558, n. 78.
143 Cf. Schop. 1977, p. 201. Sur Ies textes lies au Bodhisattva Samantabhadra, v. ibid.,
p. 183-184; Girard, p. 23-24.
382 JEROME DUCOR
sutra
144
. Ce faisant, iis accomplis sent l'ideal du Grand Vehicule et peu-
vent achever leur propre champ de buddha, en prenant exemple sur celui
d'Amitabha (Ash. 44:24,46:27; § 96-6, 96_19)145.
En conclusion, la doctrine de la Sukhavati montre que l'eveil d'Ami-
tabha regnant dans son champ de buddha est solidaire du progres spiri-
tuel de ses habitants: celui-Ia se doit d'y poursuivre son ceuvre de bud-
dha (buddhakarya) en prechant a ceux-ci, tandis que ces derniers peuvent
y achever Ie parcours irreversible (avinivartanacarya) des trois derniers
stades des bodhisattva
146
.
3. Les habitants de la Sukhavatf
Puisque Ie champ de buddha d'Amitabha constitue Ie cadre privilegie
ou accomplir Ie parcours irreversible des bodhisattva apres la mort, on ne
sera pas etonne que Ia doctrine de la Sukhavati fut, des ses origines, recu-
peree par les adeptes du Mahayana en general. En effet, elle offrait pro-
prement un raccourci a ceux d'entre eux qui en auraient deja franchi les
premieres bhrtmi en ce monde-ci, leur permettant d'economiser ainsi une
centaine de periodes cosmiques, au moins, sur Ie parcours vers l'eveil
147

Cette recuperation de la Sukhavati a ete abordee par Gregory Scho-
pen148. Mais, outrepassant les conclusions de ce dernier, Fussman estime
que l'acces meme a ce champ de buddha serait exclusivement reserve
aux plus hauts bodhisattva. C'est ainsi qu'il declare que "la renaissance
dans la Sukhavati est reservee a des bodhisattva dont c'est la derniere
naissance" (p. 550, n. 62). Ce faisant, Fussman confond la naissance en
Sukhavati avec la derniere naissance mentionnee par Ie 21 e vceu (ekajati-
144 Ash. 48:22-49:2, §100; Dantinne, p. 171.
145 A noter que c'est dans la 8
e
bhilmi que Ie bodhisattva obtient la souverainete sur la
purete de son futur champ de buddha. Cf. Lamotte, Vimalakfrti, p. 403-404; id., Concen-
tration, p. 157, et 164, n. 122; Rowell, ibid., p. 387-389.
146 La Vallee Poussin notait deja: "La doctrine de la carriere du futur Bouddha se conci-
lie cependant avec la doctrine de la Terre [Bienheureuse]" (L VP, Morale, p. 248, n. 1).
147 Ce ne sont donc pas les siitra de la Sukhiivatl qui constituent un "gauchissement"
du systeme des bhilmi, comme Ie croit Fussman (Annuaire du College de France 1998-1999,
p. 707-708; cf. Fus. 1999, p. 577-578). Sur la longue duree de la carriere des bodhisattva,
v. Lamotte, Traite 1, p. 246-249; id., Histoire, p. 693-695.
148 Schop. 1977, notamment p. 189-193, qui se rerere au Samiidhirajii-satra et au Bha-
dracariprw:zidhiina-siltra.
LESSOURCESDELASUKHAvKrt 383
baddha)I49, alors 'que notre auteur affmne plus loin - non sans raison-
que "la Sukhavau'sera l'etape precMant immediatement ['existence ou ils
obtiendront [la bodhz]" (p. 575, n. 109). .
Et Fussman de s'interroger sur deux passages du "Petit Sutra" mention-
nant la presence, en Sukhavau, d'hommes (manu:fya), d'une part, et d'au-
diteurs du Petit Vehicule (sriivaka), d'autre part
I50
• Au sujet des premiers,
il cite (p. 565-566) l'extrait du sutra concernant l'infinite de la vie
d'Amitayus et de ces "hommes" (§ 8), tout en prenant Ie parti des tra-
ducteurs tibetains, qui ont omis cette derniere expression lSI. Mais les
"hommes" sont deja mentionnes dans un passage anterieur du meme
sutra, y compris dans sa traduction tibetaine: il y est question des hom-
mes (manu:fyiilJiif!l, tib. mi de dag) qui entendent les sons delicieux reson-
nant dans la Sukhavau (§ 7; Bzwe. 346: 11)152. Quant au "Grand Sfitra" ,
il suffit de rappeler ici que les hommes et les dieux sont expressement
mentionnes par Ie 4
e
vreu.
Concernant les sriivaka, Fussman (p. 566-567) se rerere a un autre pas- .
sage du "Petit Sutra" (§ 9), lequel indique qu' Amitabha est pourvu d'une
communaute d'auditeurs (sriivakasangha). Et Fussman d'estimer que cette
indication constitue une "inadvertance", parce qu' elle serait contredite
par la suite du texte (§ 10). Comme nous l'avons vu plus haut (p. 377),
ce dernier affmne, en effet, que les etres nes en Sukhavau sont de "purs
bodhisattva, irreversibles et lies par une seule naissance"; de sriivaka, i1
n' est donc pas question. Le sutra poursuit en declarant que I' on doit faire
149 Meme confusion chez Gomez: selon lui, la derniere naissance a laquelle sont lies
les etres "is clearly birth in Sukhavati": "only this one birth remains for them", de sorte
qu'ils y realiseraient Ie parfait eveil "right there" (§ 96-18, § 100, p. 97, n. 79; cf. Ash.
46:25 et48:23-24); idem chez Miiller 1894, p. 51, n. 1; cf. Ash. 48:23-24). Or, Ie 21· vreu
mentionne sans equivoque les "etres qui sont nes" (sattvii upapanniil;) dans ce champ de
buddha (v. aussi Ash. 46:22-25, § 96-18); idem pour Ie "Petit Siitra" (§ 10; cf. plus haut,
p. 377). Leur realisation du parfait eveil en Sukhavati serait d'autant plus impossible que
plusieurs buddha ne peuvent coexister dans un meme champ de buddha (cf. Ducor, p. 1998,
p. 35, n. 63; ajouter Nattier 2000, p. 89).
150 La presence des sriivaka en Sukhavati etait deja signalee par La Vallee Poussin
(HERE, s.v. "Avalokitesvara", p. 258b, note Selon Huntington, il s'agit Ia de l'une des
"ambiguities" du "Petit Siitra" (1996, p. 71-72).
151 Cf. Ducor 1998, p. 145, n. 532. Dans Ie 14· vreu, Ie passage correspondant dit sim-
plement "etres" (sattviiniil]'!, Ash. 13:7); cf. Dantinne, p. 153-154.
152 Les "hommes" apparaissent aussi au § 6, mais Ie tibetain a "les etres
qui y sont nes" (Bzwe. 346:2).
384 JER6ME DUCOR
un vceu vis-a-vis de ce champ de buddha, c'est-a-dire, y souhaiter naitre.
n en explique la raison par Ie fait que l'on s'y trouvera reuni avec des
"heros excellents de cette espece"153, ceux-ci n'etant autres que les bod-
hisattva residant deja en SukhavatI, comme indique plus haut par Ie sutra;
ici non plus, il n'est pas question de sriivaka. Pourtant, la presence des
sriivaka en SukhavatI est bien attestee par nos deux sutra Car tous les bud-
dha obtiennent un entourage (pariviira), qu'il comprenne exclusivement
des sriivaka ou des bodhisattva, ou qu'il soit melange
154
. La definition
de l'entourage d'un buddha fait donc partie des listes classiques des qua-
lites propres a son champ155. Ainsi, tandis que l'entourage du Buddha
SugandhakUta ne compte que des bodhisattva
156
, celui de Sakyamuni est
melange!57, car sa predication du Mahayana profite aussi aux adeptes du
Petit Vehicule
158
. n en va de meme pour Ie Buddha en son
propre champ de buddha, Abhirati
159
. Et Arnitabha ne fait pas exception:
il dispose, lui aussi, d'un entourage melange, dont son 12
e
vceu se fait
l'ech0160. A son entourage appartiennent egalement les Bodhisattva Avalo-
153 La traduction de Fussman (p. 566) fait probleme: "Parce qu'on s'y trouve avec des
etres purs semblables <3. soi?>." (tathiirilpai/:t saha samavadhii-
naf!! bhavati; tib. 'di ltar skyes bu darn pa de Eta bu dag dang phrad par 'gyur ba'i phyir
TO, Bzwe. 348:2-3); idem chez Gomez (p. 19): "Because they will meet persons like them-
selves, who practice the good". A noter que Ie terme designe notamment des
bodhisattva laIcs (Harrison 1978, p. 57; id. 1990, p. 6, n. 7; Hobi5girin 7, 793a); Kumara-
jIva Ie rend par "hommes de bien superieurs" (shangshan-ren et Xuanzang par
"grands heros" (dashi *±): l'expression s'applique particulierement aux bodhisattva irre-
versibles (HobOgirin 6, 675ab).
154 Lamotte, TraUe 5, p. 2237-2240; id., "Vajrapfu.li", p. 145 ss .
•• }55 References in Ducor 1998, p. 31, n. 48. Ajouter: Kagawa 1985, p. 162-165; Inagaki
1998a, p. 56-58.
156 Lamotte, Vimalakfrti, p. 320.
157 Selon Ie prologue du "Grand Sutra", Siikyamuni etait entoure d'une "communaute
de 32.000 moines" (bhikijusangha) comprenant de grands auditeurs (mahiisriivaka) (Ash.
1: 13-14, § 2), ainsi que de nombreux bodhisattva-mahiisattva, parmi lesquels seul Mait-
reya est nomme (Ash. 2: 12-13, § 4), bien qu' AvalokiteSvara intervienne aussi dans Ie sutra
(cf. plus haut, p. 366). Le "Petit Sutra" (§ 1) mentionne egalement une "communaute de
1.250 moines" (bhikijusangha) comprenant des rnahiisriivaka, ainsi que des bodhisattva-rnahii-
sattva: quatre d'entre eux sont nommes - Ie premier etant MafijusrI suivi d'Ajita/
Maitreya - mais non pas Avalokitesvara.
158 Lamotte, Traite 1, p. 238-240; id., Vimalakfrti, p. 124, n. 87.
159 Dantinne, p. 231, et .p. 254-255, qui se rMere aux deux Sukhavatfvyilha; Nattier
2000, p. 83-84, 92-93.
160 Selon ce dernier, les sriivaka de l'entourage d'Arnita:bha seront innombrables (v. aussi
LES SOURCES DE LA SUKHAVATI 385
kitesvara et Mahasth1i.maprapta: si Ie "Grand Siltra" est Ie seul ales nom-
mer, il explique aussi que leur eclat depasse ce1ui des autres bodhisattva
de la Sukhavatl et, afortiori, de ses sriivaka
161
• Bref, les auditeurs d'Ami-
tabha n'apparaissent pas comme etant nes en tant que tels en SukhavatI,
mais plutat comme faisant partie des retributions de ce buddha, ces "qua-
lites de champ de buddha dont s'ome ce champ de buddha", comme Ie
rappelle d'ailleurs Ie "Petit Siltra", aussitat apres sa mention des sriivaka
(§ 9, in fine). Enfin, la suite de ce texte (§ 10) indique aussi que la com-
munaute des auditeurs (sriivakasangha) fait partie du cortege d' Amitabha
apparaissant aux mourants
162
. Au passage, on notera que la presence
simultanee en SukhavatI des sriivaka et des bodhisattva n'est pas sans evo-
quer Ie topique des stades communs aux deux Vehicules (siidhiiralJa-
bhumi) 163 .
4. La methode du "Petit Sukhavatlvyilha"
Pour demontrer que seuls des bodhisattva deja tres avances peuvent
entrer en SukhavatI, Fussman (p. 567-568) renvoie a un passage ante-
rieur du meme § 10 du "Petit Siltra". Comme nous l' avons vu, Ie siltra
y exhorte les etres a produire Ie vceu de naitre en SukhavatI parce qu'ils
s'y trouveront reunis avec ces heros excellents que sont les bodhisattva.
Cependant, il ne peut s'agir d'un simple vceu pieux, et Ie siltra de donner
cet avertissement avant de livrer sa methode:
"Ce n'est pas par une maigre racine de bien seulement, Sanputra, que les
etres naissent dans Ie champ de buddha du Tathagata ArnitaYUS"164.
Ash. 28: 11, § 46). Pour autant, l' Upadda indique que "dans Ie champ du buddha Amita,
Ie sangha des bodhisattva est nombreux, et celui des sriivaka peu nombreux" (cf. Lamotte,
Traite 5, p. 2335).
161 Ash. 49:3-4, 9-10; §101-l02. Cf. Lamotte, Traite 5, p. 2308, n. 2.
162 Passage d'ailleurs mentionne par Fussman: p. 550, n. 62. Huntington (1996, p. 71,
n. 24) releve que Ie texte chinois ne mentionne que les "Holy ones" (i.e. les saintes fou-
les, zhu sheng zhong T. 12,366, p. 347b13). C'est vrai de la traduction de Kuma-
rajlva, mais celIe attribuee a Xuanzang mentionne bien "la foule des disciples auditeurs
et des bodhisattva" (T. 12,367, p. 350aI2); idem pour Ie tibetain (Bzwe. 348:9).
163 Sur ce sujet complexe, v. Lamotte, Traite 5, p. 2377 ss; id., Concentration, p. 246-
251; id.,Histoire, p. 696-697.
164 Skr. niivaramiitrakera Siiriputra tathiigatasya
sattvii upapadyante; cf. Thomas, p. 84; Dantinne, p. 101, n. 115. Miiller traduisait: "Les etres
386 JEROME DUCOR
De cette litote, Fussman que, retrospectivement, Ies candidats a
Ia naissance en SukhavatI. "doivent d'abord avoir fait Ie vreu soiennel
d'atteindre Ia bodhi et avoir accumule d'innombrables racines de bien"
(p. 567). Pour toute justification de cette interpretation litigieuse du texte
sanskrit, Fussman (p. 568) se contente de deux renvois a des sources chi-
noises. Le premier se re:fere au passage correspondant du sutra dans Ie texte
de Kumarajlva mais selan Ia traduction libre de Gomez (1996, p. 148),
qui ne nous retiendra pas. Quant au second renvoi, Fussman tie fait que
mentionner l'UpadeSa dans Ia traduction de Lamotte (Traite 5, p. 2363,
§5.1), qui merite cependant d'etre cite ici plus litteralement:
"il y a des etres, aux merites mms et aux entraves tenues, qui doivent obte-
nir l'eveil dans Ie futur (ying dang de dao mais s'ils entendent
Ie nom des buddha, ils obtiennent l'eveil a l'instant meme (ji shi de dao

Cela etant, on ne voit pas en quoi I' UpadeSa infirmerait ici Ia methode
offerte par Ie "Petit Sutra", puisque celui-ci ne parle pas d'obtenir l'eveil
a I'instant meme de l'audition du nom
166
. Par contre, l'interpretation de
Fussman est contredite par Ie "Grand Sutra", 28
e
vreu et §93 (Ash. 16: 19,
42:23).
Certes, Ia Iitote du "Pet,it Sutra" peut surprendre, et c'est la, sans doute
sa fonction: attirer I'attention sur Ia methode qui suit immediatement dans
Ie texte, et qui pourrait paraitre anodine sans cette figure de rhetorique,
Iaquelle n'a d'ailleurs rien d'exceptionneF67. Or, cette methode ne fait
ne naissent pas dans ce Pays de Buddha du Tathagata Amitiiyus en recompense et par suite
dis bonnes actions accomplies dans Ia vie actuelle" (cf. Ducor 1998, p. 145, n. 537); cette
lecture a ete suivie par Sakaki Ryozaburo (Kaisetsu bongogaku, p. 259), Filliozat (l.CI. §2331,
p. 570), Huntington (1996, p. 71), et, tout recemment, par Paul Magnin (p. 462, n. 3).
165 T. 25, k. 34, p. 313c27-28. Cette citation est d'ailleurs a completer par Ia suite du
meme passage, qui lui donne un sens tout particulier: "De plus, c'est par Ie pouvoir magni-
fique des buddha qu'ils l'entendent et obtiennent aussiti'it la delivrance (ji de du
(ibid., p. 313c28-29).
166 Lamotte porte a confusion lorsqu'il affmne que, selon les Sukhiivativyuha, "Ie seul
fait d'entendre Ie nom du buddha Amitabha ( ... ) assure ipso facto une renaissance en
Sukhiivatf" (Traite 5, p. viii).
167 Par ex. dans Ie "Sutra du Lotus" (ch. XXIV): "TIs ne possederont pas une maigre
racine de bien, 0 Bienheureux, Ies etres qui entendront cette histoire du Bodhisattva-Maha-
sattva Avalokitesvara" (na te Bhagavan sattvii avarakelJa kusalamUlena samanviigatii
bhavi$Yanti; SP 374:2-3; Burnouf, "Lotus", p. 268; Watters, p. 238a; cf. Robert, p. 371).
V. aussi dans Dantinne, p. 101.
LES SOURCES DE LA SUKHAVATI 387
absolument pas allusion aux merites anterieurs ni a la production de la pen-
see d'eveil (bodhicitta), comme Ie prouve Ie texte du sutra lui-meme:
"Sariputra, tout fils de famille ou fille de famille qui entendra Ie nom de ce
Bienheureux Tathagata AmiUiyus, qui, l'ayant entendu, y portera attention
(manasikarifYati) de une a sept nuits, y portera attention d'une pensee sans
distraction (avik#ptacitto manasikarifYati)168, lorsque ce fils de famille ou
cette fille de famille mourra, ce Tathagata Amitiiyus, entoure de sa comrnu-
naute d'auditeurs, escorte d'une foule de bodhisattva, se tiendra face a lui
(purataJ:t sthasyati) au moment de sa mort, et il mourra avec une pensee
sans meprise (aviparyastacittaJ:t); mort, il naitra dans l'univers Sukhavatl,
champ de buddha de ce Tathagata Amitayus."
Si ron s'en tient au texte du sutra, sa methode se decompose donc selon
Ie processus suivant:
audition du nom> attention> absence de distraction> presence du Buddha>
absence de meprises > mort> naissance en SukhavatI.
Ces elements meritent d'etre detailles. Tout d'abord, les etres doivent
entendre Ie nom du Buddha Amitayus169. C'est dire 1'irnportance de l'au-
dition du nom (niimadheyasravalJ-a): en fait, c'est elle qui constitue la clef
de notre litote. II apparait en effet que si la methode detaillee dans Ie
"Petit Sutra" ne constitue pas une "maigre racine de bien", c'est bien parce
qu' elle s' amorce dans les "racines de bien produites par l' audition du nom"
(niimadheyal'{l srutvii tac-chravalJ-asahagatena kusalamulena)170. Or, ace
stade du sutra (§ 10), ce nom a d'ores et deja ete "entendu", tant par
l'auditoire de Sakyamuni que par Ie lecteur ulterieur du texte: l'appellation
d'Amit:lyus a, en effet, ete devoilee des Ie debut du sutra (§ 2), tandis que
la signification du double nom Amitayus/Amitabha a suivi un peu plus loin
(§ 8-9). II n'y a donc plus a s'interroger sur la quantite des merites accu-
mules pour obtenir d'entendre ce nom; et pas d'avantage, d'ailleurs, sur les
merites qui, de toute evidence, ont permis l' obtention d 'une existence
168 Pour Ie tenne technique aviklfiptacitto, Fussman traduit: "sans penser 11 autre chose"
(p. 567); idem (p. 574) dans un passage du "Sutra du Lotus" qui a aviklfiptena manasikii-
relJa (ch. XXVI, SP 388:6; Burnouf, id., p. 279).
169 La diffusion du nom est assuree par Ies autres buddha, selon Ie "Grand Sutra": 17
e
vreu
et §29-3, 89 et 95 (Ash. 21 :24-22:2,41 :25-42:4,4314-17). V. Adachi, p. 23 ss. Cf. Har-
rison 1978, p. 54 et n. 21; Lamotte, Traite, 5, p. 2364.
170 Pour reprendre l'expression des vreux 34, 42 et 43 (Ash. 18:5-6,20:1-2,7-8).
388 JEROME DUCOR
humaine ainsi que la rencontre avec ce sutra. En soulignant l'importance
de l'audition du nom du Buddha, Ie "Petit Sutra" - tout comme Ie "Grand
Sutra"l7l - s'inscrit dans un courant typique du Grand VehicuIe, ainsi
que I'attestent, notamment, les commentaires que lui cons acre l'Upadesa:
"II y a des buddha dont la commemoration suffit pour que les etres obtien-
nent l'eveil. ( ... ) II Y a des buddha dont l'audition du nom suffit pour que les
etres obtiennent l'eveil"172.
Bref, a l'audition du nom, Ie pratiquant est done amene a y porter atten-
tion (manasiktira), ce qui consiste a tenir son esprit sur Ie nom du Bud-
dha pris pour objet173; et c'est bien ce qU'a compris Kumarajiva en tra-
duisant manasiktira par "garder Ie nom" (zhichi minghao Pour
autant, cette attention ne doit pas etre excessive au point qU'elle abouti-
rait, paradoxalement, a une distraction (manasikiira Iaquelle
serait une dispersion provenant du doute
174
• C'est dire aussi que Ia foi
(sraddhii) n'est pas absente de cet acte d'attention: elle d'ailleurs expli-
citee dans Ies exhortations de Sakyamuni, vers Ia fin du sutra (§ 17). En
resume, nous avons donc affaire, deprime abord, a une pratique purement
mentale
175
, certes encore bien eloignee du nembutsu jaculatoire qui s'im-
171 Cf. §147 (Ash. 62:18-20); ainsi que les vreux 18, 19,34,35,36,41,42,43,44,46
et 47. V. aussi l'etude d'Adachi.
172 T. 25,1509, k. 34, p. 313c 9-10,12-13; cf. Lamotte, Traite 5, p. 2359, §7 et 9; v. ibid.
p. 2355-2364.
173 Selon la definition d'Asanga (Rahula, p. 7, qui restitue en citta-dhiiraIJa). Cf. LVP,
Kosa 2, p. 154; Burnouf, "Lotus", p. 413; Ducor 1999, p. 108, n. 60; BHSD, p. 418a.
174 Selon la definition d'Asanga (Rahula, p. 13-14), qui distingue six sortes de dis-
tractions: c' est la sixieme qui est la distraction provenant de l' attention. Fussman (p. 567,
n. 93) releve que la distraction constitue 1'une des six pensees mauvaises que Ie bodhisattva
doit repousser. Pour etre plus precis, elle constitue l'un des six contraires (vipakifa) des six
perfections (paramita) qui doivent etre ecartes par la perfection de sagesse (prajfia): la dis-
traction est Ie contraire de la perfection de la meditation (dhyana) (Lamotte, Traite 5, p.
2244-2245, n. 2). On soulignera cependant que Ie "Petit Sutra" ne fait pas une seule fois
allusion a la culture des six paramita.
175 Ainsi, Demieville rend "garde du nom" par "Ie retenir en son esprit" ("Pensee
unique", p. 235). La traduction attribuee a Xuanzang dit que les etres qui ont entendu Ie
nom doivent "y reflechir" (siwei (T. 12,367, p. 350a9). Contrairement a ce qu'indique
Dantinne (p. 42), les Sukhiivatzvyuha ne parlent pas de "la seule invocation du nom d' A-
mitabha au moment du trepas", principe qui sera soutenu par Ie fameux Sutra des contem-
plations de Vie-Infinie (Wuliangshou-guan-jing, T. 12,365, p. 345c, 346a; Yamada 1984,
p. 101, 109; cf. Ducor 1999, p. 116-117, 120).
LESSOURCESDELASUKHAVATI 389
posera dans la tradition sino-japonaise. n n'empeche que manasikiira peut
aussi equivaloir a "commemoration" (anusmrti)176, une pratique dont De
Jong a montre qu'elle peut bien avoir ete une invocation en Inde deja
177
.
De meme, Fujita a signale des exemples tires de stitra ou I'expression
chinoise "garde du nom" equivaut bel et bien ala prononciation de ce
derni
er
178.
Quoi qu'il en soit, I'acte d'attention sur Ie nom du Buddha a pour
consequence qu'au moment de Ia mort, Ie Buddha se tient face au pratiquant;
et cette presence permet a ce dernier de trepasser sans entretenir l'une des
meprises (viparyiisa) qui Ie retiendraient dans Ie cycle des naissances et
des morts
179
• En vertu de quoi, il na1:t finalement en SukhavatI
180

Neanmoins, Fussman comprend curieusement que Ie mourant decede
non pas libre de meprises, mais "sans avoir change d'etat d'esprit", c'est-a-
dire "sans renoncer a son vreu d'atteindre Ia bodhi" (p. 567 et n. 93). Or,
non seulement une telle interpretation ne tient pas compte de Ia specificite
du vocabulaire bouddhique, mais, surtout, elle fait violence au texte,
puisque Fussman restitue aviparyastacittaJ;. par Ie neologisme "avipa-
176 Comme l'atteste notre sutra: cf. § 6 (buddhamanasikiira) et § 7 (buddhiinusmrti);
v. Fujita 2001, p. 131-134; Watters, p. 234b. Cf. aussi "Grand Sutra", Ash. 43:6 et 10
(§ 94). L'equivalence est aussi admise par Lamotte (Traite 5, p. 2350). Enfin, Filliozat rend
manasikiira par "mise en esprit" ou "memoration, mentalisation", et smrti par "presence
d'esprit" (I.Cl. § 2253).
177 De Jong, p. 365-366. Lamotte remarque aussi: "L'audition du nom des Buddha
entrame naturellement a sa suite une reflexion (manasikiira) plus ou moins prolongee sur
ces memes Buddha, et cette reflexion se traduit souvent par une invocation orale (akranda) " \
(Traite 5, p. 2350; cf. p. 2355, n. 2). Cf. Yamabe, p. 128, 143-144.
178 Fujita 2001, p. 133-134, qui renvoie aT. 9, 263, k. 10, p. 129b, et aT. 9,278, k. 49,
p. 708a. V. aussi plus bas, p. 304, n. 207. C'est cette equivalence entre "garde du nom"
et "prononciation du nom" qui sera reprise par Ie Chinois Shandao (613-681) pour justi-
fier la pratique du nembutsu jaculatoire (cf. Ducor 1999, p. 108).
179 Ce sont les quatre meprises: prendre 1 'impermanent pour etemel, la souffrance pour
bonheur, l'absence d'lime (aniitman) pour lime, et, enfin, l'impur pour pur. Cf. Ducor
1998, p. 82; ajouter: Scherrer-Schaub, p. 142, n. 116; Lamotte, Vimalakfrti, p. 289, n. 14;
id., Traite 3, p. 1150; Rahu1a, p. 120, n. 1; Dayal, p. 90-91; Vievard, p. 97-98.
180 La traduction de Kumiirajlva precise qu'il va y naitre "aussitot" (ji de wangsheng
Ce subitisme sera l'objet de refutations, dont on trouve les premices dans
l' UpadeSa (Lamotte, Traite 5, p. 2362-2363). Elles seront pleinement elaborees par l'ecole
Y ogacara, notamment dans la "Somme du Grand Vehicu1e" (Mahayiinaslllf1graha) d' Asailga,
avec son interpretation diachronique (kiiliintariibhipriiya), dont les effets se feront sentir
jusque dans la Chine du VITe siecle (references in Ducor 1999, p. 133, n. 147).
390 JEROME DUCOR
ryasta<bodhi>citta/:L"181. Sa conclusion n'est pas moins absurde puisqu'il
pretend que Ie mourant, "au moins a cet instant", est fixe dans l'irrever-
sible (avinivartanfya): il y a Ia un double contre-sens, car l'irreversible
est consecutif a la naissance en Sukhavati
182
, et, par definition, il ne peut
etre ni ponctuel ni provisoire.
La conclusion du §1O du sutra estdetoumee d'une maniere similaire
lorsque Fussman affrnne que les etres sont exhortes a naitre en Sukhavan
en faisant Ie "vreu solenne1 <dit bodhi>citta"183. Car Ie texte dit seule-
ment que les etres doivent aspirer a Sukh1lvati en faisant un "vreu en leur
pensee" (cittapralJidhiinam). De bodhicitta, il n'est toujours pas question.
Par contre, si l'on replace ce vreu dans Ie deroulement de la methode du
"Petit Sutra", on constate qu'il ne peut qu'etre concomitant a l'acte d'at-
tention (manasikiira) suivant l'audition du nom: en precisant que ce vreu
doit etre fait" en pensee"; Ie sutra indique donc clairement que cette aspi-
ration a la naissance en Sukhavati doit proceder d'un acte d'attention
fonde sur Ia reflexion (cintiimaya), et non pas seulement sur l'audition
(srutamaya), meme si cette demiere s'appuie sur un sermon du Buddha
184

Mais revenons ala pensee denuee de meprises (aviparyastacitta), conse-
quence directe de I'apparition du Buddha
185
. Le sutra montre qu'elle
constitue l'ultime experience mentale du pratiquant au moment meme de
mourir; elle n'est donc rien d'autre que la pensee du moment de Ia mort
(maralJacitta), dont l'importanceest capitale dans les textes les plus
anciens du bouddhisme: car la demiere pensee d'une vie qui s'acheve
est "la cause suffisante et immediate" induisant la naissance suivante
186

.. , Or, une telle absence de meprise as simile soudainement Ie mourant ni
181 L'altemative mentionnee par Fussman, selou qui aviparyastacittalJ pourrait etre un
synonyme de ne tient pas non plus.
182 Plus haut, p. 376; et Schop. 1977, 190-19l.
183 Fus. 1999, p. 567 et n. 94. Meme chose dans Ie traitement du §17 du sutra, passage
que Fussman qualifie curieusement de "sectaire" Cp. 569).
184 Cf. L VP, Kosa, ch. 2, p. 328; ch. 4, p. 223. Sur les regles de l'herrneneutique boud-
dhique, v. les references in Ducor 1999, p. 126, n. 126.
185 La traduction attribuee 11 Xuanzang se signale en precisant: "Le secours de la bien-
veillance et de la compassion [du Buddha] fait que la pensee [du mourant] n'a pas de
meprises" CT. 12,367, p. 350aI3). Sur Ie mot "secours" (jiayou :!JOt'/), v. plus
haut, p. 380, n. 137.
186 LVP, Morale, p. 180; references in Ducor 1998, p. 82, n. 268.
LES SOURCES DE LA sUKHA VATI 391
plus ni moins qu'au 7
e
des dix stades des bodhisattva
l87
,avant qu'il
n'accede au 8
e
stade une fois ne en SukhivatI. La chose est d'autant plus
remarquable que Ie "Petit slUra" s'adresse soit aux "etres" (sattva) en
general, soit aux "fIls ou [Illes de farnille" (kulaputra,kuladuhitr),
sion designant notamment des Ialcs. Sans doute est-ce Hi la doctrine la plus
originale de notre sutra: celui-ci temoigne d'une tendance du Grand Vehi-
cule visant a. ouvrir au commun des mortels la voie ardue des bodhisattva.
Ce principe est confrrme par Ie fait que cette pensee denuee de meprises
suit l'apparition du Buddha face au mourant; mais, au contraire de la pra-
tique ordinaire des visualisations, ce dernier n'a pas besoin d'acquerir
prealablement 1'reil divin (divyacab;us) necessitant des meditations (dhyiina)
reservees aux saints et aux bodhisattva liberes du desir
l88
. Et c'est vrai-
semblablement pourquoi Ie "Petit Sutra" ne parle jamais de vision
l89
.
Cette absence de 1'reil divin nous impose un rapprochement significatif,
qui nous aidera a mieux situer la place des deux Sukhavatfvyuha dans Ie boud-
dhisme indien, alors que l'article de Fussman nous laisse sur notre faim mal-
gre son titre. En effet, Ie "Sutra du recueillement du face-a.-face avec les bud-
dha du present" (Pratyutpanna-buddha-saf!lmukhiivasthita-samiidhi-sutra )190
preconise une pratique qui s' adresse aussi bien aux religieux qu' aux hucs
et qui, comme dans Ie "Petit Sutra", pennet l'apparition d'un buddha face a
eux en ce monde-ci, sans recourir a ['ceil divin
l91
; en outre, elle consiste aussi
a. porter attention, d'un jour a une semaine, sans distraction, sur un buddha
dupresent, Ie Pratyutpanna dormant specifiquementAmitiiyus en exemple
l92
;
[87 Cf. Dayal, p. 276-277, 290.
188 Sur l'ceil divin, l'une des sciences sublimes (abhijiia), v. Ducor 1999, p. 137, n. 161.
[89 Au contraire, Ie "Grand Siitra" offre, entre autres, une methode pour proprement
voir Amitabha (Ash. 42:9-21, § 91-92 Dantinne, p. 224; Conze 1954, p. 206). Eile neces-
site non seulement un acte d'attention, mais aussi la production de la pensee d'eveil, une
haute resolution, l'accumulation des merites et leur transfer!, toutes choses qui n'appa-
raissent pas dans Ie "Petit Siitra".
190 Tib. Trip. XXXII, 801; T. 13,418. Cf. Ducor 1999, p. 135-136, n. 155; ajouter I.C!.
§2071, et Harrison 1990.
191 Cf. Ducor 1999, p. 138 et n. 164; ajouter: Lamotte, TraUt! 5, p. 2274, n. 2.
192 Cf. la td. du tib. par Harrison, qui a: "with undistracted thought
he concentrates [manasi-kr-] 'on the Tathagata <Amitiiyus>" (1978, p. 42-43; cf. 1990,
p. 31-32). Dans les traductions chinoises, Ie nom du Buddha est donne comme "Amituo"
tandis que la pratique consiste a Ie "commemorer" (nian Cf.la version en 3 vols
(T. 13,418, k. 1, p. 905ab): Lamotte, Traitt! 5, p. 2274, n. 3; Harrison 1998a, p. 17-19;
392 ffiR6ME DUCOR
enfm, cette methode resulte egalement dans l'acquisition de l'irreversi-
ble
193
.
Neanmoins, Ie Pratyutpanna se distingue du "Petit Sutra" sur
points importants. Tout d'abord, i1 requiert non seulement Ie respect com-
plet des preceptes (SUa), mais aussi seize conditions prealables, dont les
moindres ne sont pas de se priver de sommeil, et de pratiquer en marchant
durant trois mois
194
• En outre, la pratique de ce samiidhi implique une
visualisation complete du buddha avec toutes ses marques. De plus, il doit
etre accompli sans aucun attachement: en fait, il doit proprement s'allier
ala sagesse (prajfiii) percevant la vacuite (siinyatii)195. Enfin, I'apparition
du buddha procuree par ce samiidhi survient hic et nunc, sans qu'il soit
necessaire d' attendre Ie moment de Ia mort
196

Les deux sutra se retrouvent cependant dans leur finalite. En effet,
ce samiidhi n'est pas non plus une fin en soi: comme la naissance en
Sukhiivatl, Ie but de l'apparition qu'il procure est de pouvoir entendre
directement l'enseignement d'un buddha en personne
197
• Mais queUe est
donc la question que Ie pratiquant va enfin pouvoir poser a cet eveille?
Le Pratyutpanna est clair: la question du pratiquant conceme la methode
pour naitre dans Ie champ de ce buddha Et Ia reponse du buddha interroge
ne nous etonnera pas outre mesure: cette methode n'est autre que la com-
memoration du buddha (buddhiinu.smrtl), laqueUe fait obtenir Ie recueille-
ment de la vacuite (siinyatii-samiidhi)198.
En resume, si Ie Pratyutpanna fait bien Ie pont entre Ia doctrine des
Prajfiiipiiramitii et celIe des Sukhiivatzvyiiha
199
, il se distingue en ce que
la realisation de la vacuite y est concomitante a la pratique du samiidhi
Tsukamoto 2, p. 850-851. Et l'abrege en 1 voL (T. 13,417, p. 899ab): Inagaki 1989, p. 58-
60. V. aussi Fujita 2001, p. 135.
193 T. 13, 418, k. 1, p. 903b4-5; Harrison 1998a, p. 9; id. 1990, p. 18. Cf. Lamotte,
Traite 4, p. 1789.
194 Harrison 1990, p. 45-46; id. 1998a, p. 23-24.
195 Lamotte, Traite 4, p. 1927-1928; id. 5, p. 2268-2269; Harrison 1978, p. 45-46 et .
48-51; id. 1992, p. 221-223.
196 Harrison 1990, p. 33; id. 1998a, p. 18. Cf. Lamotte, Concentration, p. 163, n. 121;
id., Traite 4, p. 1930; ibid. 5, p. 2273, n. 2.
197 Cf. Harrison 1978, p. 52-54; 1990, p. 32-33; 1998a, p. 18.
198 T.B, 418, k. 1, p. 905b8-19; Harrison 1998a, p. 19; id. 1990, p. 36-38 (cf. 1978,
p. 52); id. 1992, p. 220-221.Tsukamoto 2, p. 857.
Harrison 1978, p. 40; Kajiyama, p. 16-17.
LES SOURCES DE LA SUKHAVATI 393
en cette vie, tandis qU'elie est une consequence de la naissance en Sukha-
van apres la mort. Cette nuance importante induit une derniere difference
entre Ie Pratyutpanna et Ie "Petit Siitra": celie du degre de difficulte de
leurs methodes respectives. De toute evidence, Ie pratyutpanna-samiidhi
implique un parcours plus difficile et sa pratique neces-
site plus d'effort (prayogika). Ce principe est confmne non seulement
par l' Upade.sa
200
, mais aussi par un autre commentaire, egalement attri-
bue a Nagiirjuna: Ie "Traite sur l'analyse des dix stades" (Dasabhilmi-
Or, si ce dernier texte accorde une place importante
au pratyutpcLnna-samiidhz"202, il offre aussi - non sans reticences - une
alternative au pratiquant manquant d'energie (vi"rya) sur ce parcours dif-
ficile. Ce succedane est celui du moyen habile de la foi (xin-fangbian
qui constitue un parcours facile (yixing JMt) pour atteindre l'irre-
versible
203
. n consiste a "garder et prononcer les noms" (zhichi cheng min-
ghao des buddha des differentes directions, d 'une maniere assez
similaire aux n!commandations apparaissant vers la fin du "Petit Siitra"
(§ 11-16)204, mais selon une liste apparentee a celie du "Grand Siitra,,205.
La recommandation de la garde des noms des buddha qui clot I'expose
de la methode du "Petit Siitra" montre les buddha des six directions
exhortant a la foi en ce texte par cette injonction:
200 T. 25, 1509, k. 21, p. 221b6-7; Lamotte, Traite 3, p. 1361.
201 Inagaki 1998b, p. 137. Ce traite est conserve seulement dans la td. chinoise de Kuma-
rajiva (Shizhu-piposha-lun; T. 1521), vers 408-409 ap. I.-C. L'identite de son auteui;,.li;St'
debattue; v. Kagawa 1985, p. 152-155; Fujita 1996b, p. 33; Lamotte, Traite 3, p. XLIV,'ii"qi
Harrison 1990, p. xxv, n. 25; Tanaka, p. 10, n. 47; Girard, p. 217; cf. Kuo Li-ying, p:!oS'2:.
Outre les dix bhibni, Ie titre chinois evoque leur equivalent anterieur que sont les dix stations
(vihiira): cf. Demieville, Yogiiciirabhumi, p. 430, et 434, n. 8; Girard, p. 208-209.
202 Cf. Inagaki 1998b, p. 10, 20, 66, 82-83.
203 T.26, 1521, k. 5, p. 41b. Inagaki 1998b, p. 139-140; Dantinne, p. 128; Hob6girin
1, p. 27b; Eracle 1981, p. 6.
204 La liste des noms des buddha dans Ie "Petit Siitra" a, depuis longtemps, ete mise
en parallele avec Ie "Siitra des noms des buddha" (Foming-jing, T. 14,440) (cf. Ducor 1998,
p. 84, n. 280; ajouter Fujita 2001, p. 71-79). Mais 1es fragments decouverts dans la Col-
lection Sch¢yen permettent aussi d' evoquer un lien entre ce deruier texte et Ie "Grand
Siitra"; cf. IffiM, p. 181 et 206.
205 Selon Inagaki (1998b, p. 11), la version du "Grand Siitra" utilisee par ce traite etait
proche de T. 360; cf. ibid., p. 147 ss., 163, n. 24; et les tableaux comparatifs des p. 201-208.
V. aussi Kagawa 1985, p. 169.
394 JEROME DUCOR
"Acceptez cette celebration des qualites inconcevables
206
, cet expose de Ia
Loi intitule «embrassement par tous Ies buddha» (sarvabuddhaparigraha)"
(§ 11-16).
Cette demiere fonnule est ensuite expliquee par Sakyarnuni de la maniere
suivante:
"Tous Ies fils ou fiUes de farnille qui entendront Ie nom de cet expose de Ia
Loi et qui retiendront Ie nom (nt1madheyaf!! de ces bien-
heureux buddha seront tous embrasses par Ies buddha (buddhaparigrhftt1) et
seront irreversibies quant a I'eveil correct, complet et insurpassabIe" (§ 17).
Fussman consacre de longs developpements (p. 568-574) aux deux
expressions sarvabuddhaparigraha et buddhaparigrhftii208. En particulier,
il considere que la premiere est Ie VIai titre du "Petit Sutra" (p. 568,571).
Or, ce tenne ne constitue pas un titre a proprement parler
209
• Ala maniere
des sous-titres apparaissant a la fin du "Grand Siitra" (plus haut, p. 374,
il fait partie d'une serie d'appellations louangeuses appliquees a bon
nombre de sutra
21O
• Le "Sutra du Lotus", pour ne prendre que cet exem-
pIe, l'applique au sennon que Sakyarnuni venait d'achever juste avant
de commencer son expose2l1; et Ie meme sutra se l' applique aussi a
lui-meme, sans y accorder non plus la valeur d'un titre stricto sensu,
ainsi que nous Ie montre d'ailleurs cette citation foumie par Fussman
(p. 571-572): "Ie texte de la loi dont Ie nom est SP [Saddharmapu"(uJarfka],
206 L' expression "celebration des qualites" (gul}aparikfrtanam) apparait aussi dans les
,'!!!lus-titres du "Grand Sutra" (cf. plus haut, p. 374).
;£ .2P7 Expression equivalente a "prononcer Ie nom". Cf. Fujita 2001, p. 272ab; Harrison
225, n. 45.
208 Cf. Gomez, p. 19, n. 8; Watters, p. 239b; BHSD, p. 321b.
209 Le titre de 1a traduction attribuee a Xuanzang, Chengzan jingtu fo sheshou jing
(Sutra des louanges de la Terre Pure et de l'embrassement par les
buddha, T. 367), renvoie bien aux expressions parikfrtana et sarvabuddhaparigraha du texte
du sutra, mais il s'agit d'une initiative du traducteur chinois (Fujita 2001, p. 42).
210 Cette expression apparait ainsi comme un equivalent de celles utilisees pour quali-
fier Ie "Grand Sutra" lors de sa transmission a Maitreya (Ash. 63:20-21, § 150): "loue
par tous les buddha" (sarvabuddhasa1]'!varl}ita), "celebre par tous les buddha" (sarvabud-
dhaprasasta) et "approuve par tous les buddha" (sarvabuddhiinujiiiita).
211 Burnouf ("Lotus", p. 3) rend sarvabuddhaparigraha par "qui a ete possede par
tous les Buddhas" (SP 4:4; Watters, p. 239b; cf. Robert, p. 49). D'une maniere generale,
les sutra precMs par Ie Buddha sont, entre autres, loues comme etant "adoptes par tous
les etres nobles" (sarviiryajanaparigrhftii); cf. Lamotte, Vimalakfrti, p. 379.
LES SOURCES DE LA suKHA VATI 395
un sutra developpe [mahavaipulya], instruction des bodhisattva [bodhi-
sattvavavada], possession de tous les buddha [sarvabuddhaparigraha]"212.
Quant a l'expression buddhaparigrhfta, elle s'explique, notamment, par
Ie 25" vreu.
Pour en revenir au Pratyutpanna, on relevera aussi que c' est bien ce
sutra qui foumit un rep ere historique capital dans la chronologie du cou-
rant de la Sukhavatl. En effet, sa traduction par Zhi Loujiachen (T. 418)
est particulierement bien datee et authentifiee
213
, et son anciennete en fait
Ie plus vieux document historique chinois lie au Buddha Amitabha/Ami-
tayus, puisqu' elle remonte a l' an 179 de notre ere. Enfin, tandis que la
methode preconisee par ce texte est presentee comme applicable a tOilS
les buddha du present dans les dix directions, celui de la SukhavatI est
Ie seul aetre nommement mentionne dans Ie texte; on peut donc en
conc1ure que I' epoque, forcement anterieure, OU la version indienne de ce
sutra fut composee connaissait deja un "well-developed and widespread·
cult of On l'admettra, cet argument est autrement plus
significatif que celui de la Pentade Buddhananda .
. 5. La place des Sukhavatlvyliha
Nous avons releve dans les deux sutra de la Sukhavatl un certain nom-
bre d'elements relevant en propre de la doctrine du Grand Vehicule. Cer-
tes, ceux-ci sont "mostly undevelopped" et necessiteraient une analyse
plus fouillee
215
, mais leur manque meme de systematisation atteste de
1'anciennete des origines de la doctrine du champ de buddha d' Amitabha.
Et c'est ainsi que nos deux sutra font bien partie de ces "textes qui repre-
sentent un Mahayana en fonnation" , au m.eme titre que les sutra de I'
sahasrika Prajfiaparamita, du Pratyutpanna, du SuraT{tgama et celui de
Vimalakirtz"216.
212 Ch. VII, SP 161:26-28; cf. Burnouf, id., p. 111.
213 Lamotte, Concentration, p. 66-69; ZUrcher 1959, p. 35, n. 93, 95; Harrison 1990,
p. 255-267; Mochizuki, p. 241-242.
214 Harrison 1978, p. 43-44; id. 1990, p. 31, n. 1. Cf. Fujita 1996b, p. 10; Kajiyama,
p. 13, 17.
215 Schopen 1977, p. 191.
216 Lamotte, Concentration, p. 44, 49.
396 JEROME DUCOR
Mis en perspective avec Ie Pratyutpanna, les deux Sukhiivatfvyiiha
temoignent pareillement d'un double souci: d'une part, permettre l'audition
du Dharma preche directement par un buddha en personne, malgre la dis-
parition de Sakyamuni; et, d'autre part, elargir a un plus grand nombre
l'acces a l'irreversible. Cependant, avec son accent sur la doctrine de la
siinyatii, Ie Pratyutpanna apparait aussi comme une tentative de recupe-
ration du courant de la Sukhavatl au profit de celui des Prajfi.iipiiramitii
217
:
celles-ci avaient en effet superbement ignore Amitabha et son champ de
buddha, alors meme que leurs redacteurs ne pouvaient pas ne pas en avoir
connaissance
218
, tout comme ils connaissaient bien et son Abhi-
ratF19. Neanmoins, cette tentative du Pratyutpanna sera elle-meme criti-
quee par les tenants de l'ecole Madhyamika, dans Ie cadre d'une contro-
verse entre "mystiques" et "rationalistes" qui se serait developpee au
Cachemire, au debut du Ne s., ainsi qu'en temoigne l'Upadda
220
. Pour
ce dernier, Ie pratyutpanna-samiidhi est non seulement plus difficile que
la vision du Buddha obtenue par I' ceil divin, mais, en outre, il "peut seu-
lement rassemblerla pensee distraite "221. Et I' Upadda d' in-
sister encore d'avantage sur l'importance d'unir la sagesse (prajfi.ii) a la
commemoration du buddha, a tel point que cette derniere y est elle-meme
presentee comme un "recueillement" (buddhiinusmrti-samiidhi)222 - alors
217 Kajiyama parle de "demythologisation" du culte d' Amitabha au moyen de la philo-
sophie de la vacuite (p. 18-19).
218 Lamotte, Concentration, p. 56. La possibilite pour les laYcs de voir les buddha sans
recourir 11 1'tEil divin apparaitra dans la Paficavif!1satisahasrikii PP; cf. Lamotte, Traite 5,
p. 2265-2266.
AP9 «Lamotte, Vimalakfrti, p. 361-362, en note; HobOgirin 1, p. 25b, 40a; Kwan, p. 133
S8.; Schop. 1987, p. 117, n. 50 (p. 131).
220 Cf. Lamotte, Traite 5, p. VI-VII. Cette controverse resurgit un siecle plus tard en
Chine, ou elle opposa KumarajIva, fondateur historique du Madhyamika chinois, et Huiyuan
du Mont Lu (334-417); cf. Traite 5, p. VI-VII, 2268-2272; Demieville, Yogacarabhumi,
p. 357(-358), n. 8; Zurcher 1959, p. 219 ss.; Tsukamoto 2, p. 851 ss; Mochizuki, p. 254-
256. Sur l'Upadda et Ie Pratyutpanna, v. Harrison 1990, p. xxiv, n. 24.
221 T. 25, k. 33, p. 306a; k. 34, p. 314a23-24 (Lamotte, Traite 5, p. 2274, 2369). On
relevera la similitude de cette fonction avec celle de l' attention portee au nom du Buddha
dans Ie "Petit Siitra" (sup. p. 388).
222 Cf. la paraphrase du Pratyutpanna (T. 13,418, k. 1, p. 905b8-14; Harrison 1998a,
p. 19) dans 1'Upadda (T. 25, k. 29, p. 276a18-22; Lamotte, Traite 4, p. 1927 - de prefe-
rence 11 id. 3, p. 1361, n. 2). L' Upadda fait la difference entre Ie buddhiinusmrti-samiidhi,
obtenu par "1'tEil de la pensee" et la buddhanusmrti, simple invocation orale
du buddha (Lamotte, Traite 1, p. 409, 414).
LES SOURCES DE LA SUKHAVATI 397
que cette qualification etait inconnue tant du Pratyutpanna que de la quasi-
totalite des sources' sanskrites conservees223. Pour autant, Ie but vise reste
des plus classiques, puis que cette methode permet au pratiquant de neplus
etre separe des buddha: ainsi, "il peut faire a volonte des offrandes aux
buddha de son vivant et les rencontrer derechef au moment de la mort,,224.
Enfin, Ie Madhyamika s' accorde avec Ie Pratyutpanna au su jet de la nature
de la vision du buddha obtenue: conformement a la doctrine des Prajiia-
paramita, il ne peut s'agir que d'irnagination subjective (sa1'[lkalpa) du
pratiquant
225
, car "c' est Ie mental qui fabrique Ie buddha, c' est Ie mental
qui est Ie buddha" (xin zuo fo, xin ski fo
Tout au contraire, selon la remarque judicieuse de Demieville, Ie champ
de buddha d' Amitabha considere comme but de la prochaine naissance
constitue un "«ailleurs» [tafeng {ill)]] transcendant [ ... ] hors du triple
monde": de fait, il se situe au-deIa de myriades de koti de champs de bud-
dha ("Petit Sutra", § 2; "Grand Sutra", § 40), tout en n'etant accessible
qu ' apres la mort
227
.
Cependant, cette distance inconcevabIe, expression de Ia dimension
supra-mondaniste (lokottara) d'Amitabha, l'exposait directement a la
concurrence du culte du Bodhisattva Maitreya. Ce dernier presentait en
effet un double avantage: d'une part, sa tradition etait anterieure au Grand
Vehicule, et, d' autre part, sa residence dans Ie paradis de indus
au sein de notre univers Saha, Ie rendait plus accessible, y compris en la
vie presente228. La transcendance de la SukhavatI, combinee avec les cri-
223 La Satasiihasrikii P P fournit une rarissime occurrence de l' expression sanskrite bud-
dhiinusmrti-samiidhi 6.1 Yamabe, p. 157-158). Cf. la Paiicavirrzsatisiihasrikii PP
(Dntt 5.3; Lamotte, Traite 1, p. 403), dont la td. par Kumarajlva (T. 8,223, k. 1, p. 217a26)
est citee par 1'Upadda (T. 25, 1509, k. 7, p. 108c21; Lamotte, id., p. 409).
224 T. 25, 1509, k. 30, p. 276b14-18; cf. Lamotte, Traite 4, p. 1930.
225 Cf. Lamotte, qui parle d'" autosuggestion": Traite 5, p. 2272, 2274 et n. 1; ibid. 4,
p. 1927(-1928), n. 1.
226 Cf. Upadesa: T. 25, k. 29, p. 276blO-12 (Lamotte, Traite 4, p. 1930); et Pratyut-
panna: T. 418, k. 1, p. 905c-906a (Harrison 1992, p. 225-226; id. 1998a, p. 21). Meme
chose dans Ie Sutra des contemplations de Vie-Infinie (T. 12,365, p. 343a19-22); a noter que
selon ce demier, Ie pratyutpanna-samiidhi peut etre obtenu apres la naissance en Sukhii-
vat! (ibid., p. 346b3); cf. Yamada 1984, p. 51 et 11l.
227 Demieville, Yogiiciirabhumi, p. 395, n. 3, et sa glose sur ta/eng fill:;!]. Fujita souli-
gne que cet "ailleurs" ("world of another dimension", j. tahO-sekai fill:;!]iltW) ne doit pas
etre confondu avec l'au-dela ("other world",j. takai {illW): 1996c, p. 44-47; 2001, p. 112.
228 Cf.le temoignage de Xuanzang (602-664): Demieville, Yogiiciirabhilmi, p. 389-390
398 ffiROME DUCOR
tiques du Madhyamika
229
, explique vraisemblablement pourquoi Ie culte
d'Amitabha n'a guere laisse de traces specifiques en Inde
23o
• Car l'image
de son champ de buddha y fut rapidement recuperee par un Mahayana
moins radical, qui la banalisa comme un archetype de voire
une simple comparaison litteraire c1assique: ce phenomene, qui se Mve-
loppa, approxirnativement, des rer_IIe s. jusqu'aux Ne_v
e
s. de notre ere
23
1,
cuIminera par la rehabilitation de Sakyamuni dans Ie KarU/JijpU/J-fjarfka
232

D'ailleurs, si l'Upadesa considere Ie Pratyutpanna comme digne de cri-
tique, il n'est que dedain pour l'iMal de la SukhavatI proprement dite,
puisqu'il ne la juge meme pas comparable avec l'univers PadmavatI du
Buddha Samantakusuma, ou Mafijusrl, parangon de la prajiiil, reside
notamment. Le pretexte invoque par Ie traite est pourtant des plus gratuits:
lorsque Ie futur Amitabha contemplait les univers de buddha pour preparer
sa propre SukhavatI, Ie pouvoir de ses merites aurait ete trop mediocre
pour lui permettre de voir les univers les plus purs, tant et si bien que son
propre champ de buddha ne pouvait que leur etre inferieur
233
. Mais l'au"
dace de cette these, etayee par rien, ne fait que confirmer l'insucces de
la SukhavatI dans Ie Madhyamika indien. Des lors, on ne sera pas etonne
que l' Upade.sa ne la cite meme pas en exemple lorsqu'il mentionne les
univers situes a l'ouest, lui preferant un champ de buddha bien moins
connu: l'Upasanta de Ratnarcis
234

5. Les origines de la SukhavatI
Quai qu'il en soit, plusieurs specialistes occidentaux allaient juger Ie
culte d' Amitabha suffisamment incongru avec l' enseignement Ie plus
(T. 53, 2122, k. 16, p. 406a3-7; T. 54, 2123, k. 1, p. 6c26-7a3). V. Lamotte, Histoire, p. 785-
787; de Lubac, p. 82-85.
229 Bient6t rejoint par l'eco1e Yogacara (cf. plus haut, p. 389, n. 180).
230 Le seul temoignage positif de pelerin chinois serait celui de Huiri (ou Cimin, 680-
748), qui voyagea en Inde et au Gandhara de 702 11 719; cependant sa relation dans la Bio-
graphie des religieux eminents des Song lui est posterieure de deux siec1es (T. 50, 2061,
k. 29, p. 890b9-20); cf. H6bi5girin 1, p. 25a; Fujita, p. 239; Tanaka, p. 3.
231 Schopen 1977, p. 194-199 et 201-204.
232 Yamada 1sshi, I, p. 164-166.
233 T. 25, 1509, k. 10, p. 134b5-1O; Lamotte, TraUe 1, p. 601. Cf. Hirakawa, p. 9.
234 T. ibid, p. 133b5-6; Lamotte id., p. 594-595.
LES SOURCES DE LA SUKHAVATI 399
ancien du bouddhisme pour se demander s'il ne resultait pas d'influen-
ces exterieures a celui-ci, ou meme a 1 , Inde. Des la fin du XIXe siecle fut
evoquee Ia possibilite que Ies concepts lies a Ia divinite VaruI:la et sa .cite
Sukha aient pu influencer l'apparition de l'enseignement sur Ia SukhavatI
d'Amitabha, d'autant que Ie nom de ce dernier apparait comme un nom
generique dans Ie Vi.glupUral}a
235
• Cherchant plus loin, on a imagine des
rapports avec 1 'Iran zoroastrien, ou meme avec Ie christianisme nestorien
ou chaldeen. Revenant a 1 , Inde, on a enfin evoque Ie courant indien de
devotion (bhakti) represente par Ia Bhagavadgfta
236

Voila une cinquantaine d'annees, Ie tableau de ces diverses hypotheses
a ete brosse par Ie Pere de Lubac
237
, qui arrivait deja a ceUe conclusion:
"Ie Mahayana, auquel appartient l'amidisme, est en continuite profonde
avec Ie bouddhisme primitif"238. Fondees sur des etudes recentes, Ies conclu-
sions de Fussman ne disent pas autre chose: "L'eclat infini d'Amita-
bha, son infmie duree de vie sont Ia consequence Iogique de la complete
transformation du buddha en «super-dieu». Point n'est besoin d'invoquer
l'influence de l'Iran pour cela" (p. 557). Bien plus, Ies sources etudiees
par Fussman lui permettent de conclure qu'aux deux premiers siecles de
l'ere chretienne, Ie culte d' Amitabha "n'etait en rien contradictoire avec
un mahayana «orthodoxe»" (p. 523)239. On ne s'explique donc pas cette
autre assertion du meme auteur: "ll n'y a que Ies sectateurs de la Terre
Pure pour penser que Ia croyance a Amitabha et l' aspiration a renaitre dans
Ia SukhavatI sont I'aboutissement logique et naturel d'idees-forces du
bouddhisme" (p. 552).
235 Cf. respectivement Muller 1894, p. xxii (mentjonne par Filliozat, I.Cl. § 2 ~ 3 1 ) , et
Renou: I.Cl. § 1018; etc.
236 Notarnment L VP, Dogme, p. 70-71; Tucci, p. 196; Conze 1971, p. 165 ss. Cf. Ams-
tutz (1998) et Beyer (1977); v. Dantinne, p. 42, n. 208; Fujita, p. 562-565; id. 1996b, p.
9; Harrison 1992, p. 223.
237 De Lubac, Amida, ch. X, qui donne 1es references. V. aussi Fujita, p. 262-278,464-
474; Machida 1988.
238 De Lubac, op.cit., p. 248. De Lubac utilise 1es quatre arguments suivants: 1. "la rigou-
reuse doctrine du karman, malgre 1es apparences, reste intacte" (p. 257-260); 2. Amida
"n'est pas dieu en un sens quelconque" (p. 260-261); 3. Ia Terre Pure "n'est pas encore
1a Delivrance" (p. 261-264); 4. Amida "ne conserve provisoirement l'aspect d'un sauveur
personnel que pour Ie vulgaire etn'est en realite ( ... ) qu'une manifestation relative de
l'Absolu" (p. 264-268).
239 Deja Fus. 1994, p. 38, infine. Cf. Nattier 2000, p. 90-91.
400 JER6IvIE DUCOR
Reglee la question des influences exterieures, il reste a preciser d'avanc
tage la jonction entre la doctrine des deux Sukhiivatlvyi1ha et les
elements anterieurs connus, notamment a travers la stratigraphie des
traductions chinoises archalques du "Grande Sutra" et de sa version
sanskrite, l'evolution de la cosmologie bouddhique
240
ainsi que l'inter-
pretation supra-mondaniste (lokottara) de la bouddhologie de Sakya-
muni, depuis la longevite infinie attribuee aux buddha dans Ie courant
Mahasaqtghika
241
jusqu'a la "quasi-infinitude" du Sakyamuni duo "SiUra
du Lotus "242, sans oublier Ie developpement des grandes figures sote-
riologiques, comme Maitreya, A valokitesvara ou Dne telle
analyse necessite la reunion des ressources pluridisciplinaires des india-
nistes et des sinologues, tant orientaux qU'occidentaux. Gageons que
l'ampleur de l'entreprise impliquera aussi un travail d'equipe, a l'exem-
pIe de la remarquable contribution de Harrison, Hartmann et Matsuda
(HHM 2002).
Bibliographie
Adachi, Toshihide: "«Hearing Amitabha's Name» in the Sukhiivatf-vyiiha".
Kagawa Takao Hakase koki kinen Tonshii (Bukkyo-gaku JOdo-gaku kenkyU,
2001), p. 21-38.
Amstutz, Galen: "The Politics of Pure Land Buddhist in fudia". Numen, vol. XLV,
No.1 (1998), p. 69-96.
Ash. = Ashikaga, Atsuuji: Dai-Muryojukyo bompon, Sukhavatlvyiiha. Kyoto,
Hoz6kan, 1965; reimp. Ashikaga Atsuuji chOshaku zenshii, 2 (T6kai Dai-
gaku, 1988), p. 105-182.
Bagchi, Prabodh Chandra: Le Canon bouddhique en Chine. Paris, Librairie Orien-
taliste Paul Guenther, 1926.
240 Comme la multiplication de buddha contemporains (Yamada 1sshi, I, p. 121-153),
ainsi que les descriptions de I'Uttarakuru (Tucci, p. 197-202; Kwan, p. 69-82), du para-
dis Trayastrirpsa (Tucci, p. 202-209), ou de la cite du Cakravartin (Tucci, p. 209-215). Cf.
de Lubac, p. 240-241.
241 Fujita, p. 333-335; Lamotte, Histoire, p. 690-692; Bareau 1955, p. 59.
242 Cf. Frank, p. 38; Fujita 1980, p. 124; id. 1996b, p. 14-15.
243 Sur AvalokiteSvara, v. Schop. 1977, p. 200; Tucci, p. 174-178; de Lubac, p. 104-
106. Selon Dantinne, aurait precede Ie "Grand Sutra" (p. 39-45), mais
Fujita les considere comme quasirnent contemporains (p. 232-233; id. 2001, p. 154-155);
cf. Harrison 1978, p. 40, n. 4; Schop. 1987, p. 112-113; EOB 1-3, p. 365a; Nattier 2000,
p. 99-102; Lamotte, Concentration, p. 59; Kwan Tai-wo (1985).
LES SOURCES DE LA SUKHAVATI 401
Bareau 1955 = Bareau, Andre: Les sectes bouddhiques du Petit Vehicule (Publi-
cations de l'EFEO, Vol. XXXVIII). Paris, 1955.
Bareau 1966 = id.: "Le bouddhisme indien". Les religions de l'Inde, III (Paris,
Payot, 1966), p. 7-246.
BEFEO = Bulletin de I'Ecole fran"aise d'Extreme-Orient. HanoI, Saigon, Paris,
1901-.
Beyer, Stephan: "Notes on the Vision Quest in Early Mahayana". Prajfiaparamita
and related Systems: Studies in honour of Edward Conze (ed. by Lewis Lan-
caster; Berkeley Buddhist Studies Series, 1; 1977), p. 329-340.
BHSD = Edgerton, Franklin: Buddhist Hybrid Sanskrit Grammar and Diction-
nary, Vol. IT: Dictionnary (1953); Delhi, Motilal Banarsidass, 1985.
Blum, Mark L.: The Origins and Development of Pure Land Buddhism, A Study
and Translation of Gyonen's Jodo Homon Genrusho. New York, Oxford
Univ. Press, 2002.
Brough, John: "Amitabha and Avalokitesvara in an Inscribed Gandharan Sculp-
ture". Proceedings of the Conference-seminar of indological studies, spon-
sored by the Royal Swedish Academy of Letters, History and Antiquities
(Stockholm, Indologica Taurinensia, X (Torino, 1982), p. 65-70.
(Reimp. Brough, John: Collected Works (London, SOAS, 1996), p. 469-473).
Buddhist Mahayana Texts, Two Parts. "The Sacred Books of the East" (ed. by
F. Max Miiller), vol. XLIX (Oxford, 1894); rpI. New York, Dover Publica-
tions, 1969.
Burnouf, Intra. = Burnouf, Eugene: Introduction a l'histoire du Bouddhisme
indien (1844). 2· ed., Paris, 1876.
Burnouf, Lotus = id.: Le Lotus de la Bonne Loi, traduit du sanscrit (1852); reimp.
Paris, Adrien Maisonneuve, 1989.
Bussagli 1984 = Bussagli, Mario: L'arte del Gandhiira (Storia univerale dell'arte,
sezione seconda: Le civilitiJ. dell'Oriente). Torino, UTET, 1984.
Bussagli 1996 = id.: L' art du Gandhiira (td. par Beatrice Amal). Encyclopedies
d'aujourd'hui, Histoire universelle de I'art; La Pochotheque, Le Livre de poche,
1996.
Bzwe. = Bon-zo-wa-ei gappeki liJdo-sambukyo (ed. by Ogiwara Unrai, Kawaguchi
Ekai, Takakusu Junjiro). JodoshU zensho, bekkan (1931); reed. Tokyo, Daito
Shuppansha, 1961.
Conze 1954 = Conze, Edward: Buddhist Texts Trough The Ages (1954); rpr.
Boston, Shambala Publications, 1990.
Conze 1971 = id.: Le Bouddhisme dans son essence et son developpement (td.
par Marie-Simone Renou). Petite BibliotMque Payot, 187; Paris, 1971.
Czuma, Stanislaw: Kushan Sculpture, Images from Early India. Cleveland, Cle-
veland Museum of Art, 1985.
Dantinne, Jean: La Splendeur de l'Inebranlable (Aksobhyavyuha), Tome I, Ch. 1-
III, les Auditeurs (Sriivaka). Publications de I Orientallste de Louvain,
29. Louvain-Ia-Neuve, 1983.
402 rnROME DUCOR
Dayal, Har: The Bodhisattva Doctrine in Buddhist Sanskrit Literature (1932); rpr;
Delhi, Motilal Banarsidas, 1975.
De Jong, Jan W.: c.r. Fujita Kotatsu, Genshijodo shishO no kenkyu. T'oung Pao,
Vol. LVIII (1972), p. 352-366.
Demieville, Paul: "Les sources chinoises [du bouddhisme]", I.Cl. § 2048-2169;
reimp. in Paul Demieville: Choix d'etudes bouddhiques (Leiden, Brill, 1973),
p. 157-222.
Demieville, CondIe = id.: Le CondIe de Lhassa, une controverse sur Ie quiitisme
entre bouddhistes de l'Inde et de la Chine au VIlle siecle, I (Bibliotheque de
l'Institut des Hautes Etudes Chinoises, vol. VII; 1952); reimp. Paris, Col-
lege de France, 1987. -
Demieville, "Pensee unique" = id.: "Surlapensee unique". BEFEO XXIV /1924
(1925), p. 231-246.
Demieville, YogacarabhUmi = id.: "La YogacarabhUmi de Sarigharllllka". BEFEO,
XLIV-2 (1954), p. 339:436.
Ducor 1998 = Ducor, Jerome: Le Sutra d'Amida precht par Ie Buddha [T. 366]
(Societe Suisse-Asie, Monographies, vol. 29). Bern, Peter Lang, 1998. Addenda
et corrigenda: www.pitaka.ch/corrig.htm.
Ducor 1999 = id.: "Shandao et Honen, It propos du livre de Julian F. Pas: «Visions
of SukhavatI»". JIABS 22-1 (1999), p. 93-163. Addenda et corrigenda:
www.pitaka.ch/corrig.htm#shandao.
Dutt, Nalinaksha: Paiicavirrzsatisahasrikii Prajiiaparamita (Calcutta Oriental
Series, 28); London 1934.
EOB = Encyclopaedia of Buddhism (ed. by G[unapal] P[iyasena] Malalasekera).
Ceylon, Government Press, 1961-.
Eracle 1981 = Eracle, Jean: Le Chapitre de Nagarjuna sur la pratique facile
[T. 1521, ch. 9], suivi du Sutra qui loue la terre de Purete [T. 367]. Bruxelles,
Institut BeIge des Hautes Etudes Bouddhiques,1981.
Eracle 1984 = id.: Trois Soutras et un Traite sur la Terre Pure [T. 360,365,366,
1524]. Geneve, Aquarius, 1984.
Eracle 1987 = id.: L'Inde aux mille visages. Geneve, Musee d'ethnographie, 1987.
Eracle 1988 = id.: "Le soutra en une feuille sur Ie Bouddha Amita, l'un des plus
anciens textes du bouddhisme MaMyiina?" [T. 373]. Bulletin du Musee
d'ethnographie de Geneve, 30 (1988), p. 69-95.
Filliozat, Jean: "Le bouddhisme".I.Cl. §1929-2386.
Fisher, Robert E.: L'art bouddhique. Paris, Thames & Hudson, 1995.
Forte, Antonio: "An Shih-kao: biografia et note critiche". Annali del Istituto
Universitario Orientale, Vol. 56, Nuova Serie xvm (Vol. 28), Fasc. 2
(Napoli, 1968), p. 151-194
244

244 Ie n'ai pas eu acces a Forte, Antonio: The Hostage An Shigao and his Offspring
(Kyoto, Istituto Italiano di Cultura, Scuola di Studi sull' Asia Orientale, 1995). .
LES SOURCES DE LA SUKHAVATI 403
Foucher 1917 = Foucher, Alfred: The Beginnings of Buddhist Art. Paris & London,
1917.
Frank, Bernard: Amour, colere, couleur, Essais sur Ie bouddhisme au Japon. Bib-
liotMque de l'Institut des Hautes Etudes Japonaises. Paris, Diffusion de Boc-
card, 2000.
Fujita = Fujita, Kotatsu: Genshi JOdo-shiso no kenkyu (1970). 7
e
ed. Tokyo,
Iwanarni shoten, 1997.
Fujita 1980 = id.: "Pure Land Buddhism and the Lotus Siltra". Indianisme et
Bouddhisme, Melanges offerts a Mgr. Lamotte (Publications de l'Institut
Orientaliste de Louvain, 23; 1980), p. 117-130.
Fujita 1985 = id.: "JOdo-kyoten no shujuso". Jodo-shiso, p. 55-80.
Fujita 1992, 1993, 1996a = id.: The Larger Sukiivafivyuha, romanized text of
the Sanskrit manuscripts from Nepal, Part I-III. Tokyo, Sankibo, 1992, 1993,
1996.
Fujita 1994 = id.: "Muryojukyo". Jodo-bukkyo no shiso, I (1994; 2
e
ed. Tokyo,
Kodansha, 1995), p. 1-241.
Fujita 1996a = v. Fujita 1992.
Fujita 1996b = id.: "Pure Land Buddhism in India" (trsL by Taitestu Unno). The
Pure Land Tradition, p. 1-42.
Fujita 1996c = id.: "The Origin of the Pure Land" (trsL by Rebecca Otowa). East-
ern Buddhist, N. S., 29-1 (1996), p. 33-52.
Fujita 1998 = id.: "Genjo yaku Shi5san-jodo-busshoju-kyo ko". Indo-tesugaku
bukkyo-gaku, 13 (1998), p. 1-35.
Fujita 2001 = id.: Amidakyo kokyu. Kyoto, Shinshii Otani-ha shUmusho shuppan-
bu,2001.
Fus. 1974 = Fussman, Gerard: "Documents epigraphiques kouchans, III: Bud-
dha de l'An 5". BEFEO, LXI (1974), p. 54-58, pI. XXXI, 27.
Fus. 1987 = id.: "Numismatic and Epigraphic Evidence for the Chronology of Early
Gandharan Art". Investigating Indian Art, Veroffentlichungen des Museums
fiir indische Kunst, VoL 8 (Berlin, 1987), p. 67-88.
Fus. 1994 = id.: "Upaya-kausalya, L'implantation du bouddhisme au Gandhara".
Bouddhisme et cultures locales, Quelques cas de reciproques adaptations
(ed. par Fukui Fumimasa et Gerard Fussman; EFEO, Etudes thematiques, 2;
Paris, 1994), p. 17-51.
Fus. 1999 = id.: "La place des Sukhiivatf-vyuha dans Ie bouddhisme indien".
Journal Asiatique 287-2 (1999), p. 523-586.
G h o ~ a , Pratapacandra: Satasiihasrikii Prajfiiipiiramitii (Bibliotheca Indica); Cal-
cutta, 1902.
Girard, Frederic: Un moine de la secte Kegon a l'epoque de Kamakura, Myoe
(1173-1232) et le "Journal de ses reves" (publications de l'Ecole Fran9aise
d'Extreme-Orient, vol. CLX). Paris, Adrien-Maisonneuve, 1990.
Gomez, Luis 0.: Land of Bliss, The Paradise of the Buddha of Measureless Light,
Sanskrit and Chinese Versions of the Sukhiivatf-vyuha Sutras . Kyoto, Higashi-
Honganji ShinshU Otani-ha; Honolulu, Uillv. of Hawaii Press, 1996.
404
. .JEROME DUCOR
Guenee, Pierre: Bibliographie analytique des ouvrages parus sur I 'art du Gand-
hiira entre 1950 et 1993 (Memoires de l' Academie des Inscriptions et Belles-
Lettres, N.S., T. XVI). Paris, Diffusion de Boccard, 1998.
GvM = Mitterwallner, Gritli von: "The Brussels Buddha from Gandhara of the
Year 5". Investigating Indian Art (Veroffentlichungen des Museums fUr
indische Kunst, Vol. 8; Berlin, 1987), p. 213-247.
Haesner, Chhaya: "Paradise Scenes in Central Asian Art: Autochthonous or
modelled after Gandhara Art?". Coins, Arts and Chronology: Essays on the
pre-Islamic History of the Indo-Iranian Borderlands (ed. by Michael A)ram &
Deborah E. Klirnburg-Salter; Wien, Verlag der Osterreichischen Akademie
der Wissenschaften, 1999), p. 423-452.
Harle, James C: "A hitherto Unknown Dated Sculpture from Gandhiira: A Pre-
liminary Report". South Asian Archaeology 1973 (ed. by J.E. van Lohuizen-
de-Leeuw & J.M.M. Ubaghs; Leiden, Brill, 1974), p. 128-135, pI. 71-75.
Harrison 1978 = Harrison, Paul M.: "Buddhiinusm[1i in the Pratyutpanna-bud-
dha-saT(lmukhiivasthita-samiidhi-sutra". Journal of fudian Philosophy, 6
(1978), p. 35-57.
Harrison 1990 = id.: The Samadhi of Direct Encounter with the Buddhas of the
Present, An Aunotated English Translation of the Tibetan Version of the
Pratyutpanna-Buddha-SaT(lmukhiivasthita-Samiidhi-Sutra (Studia Philolo-
gica Buddhica, Monograph Series, VI). Tokyo, The futemational fustitute
of Buddhist Studies, 1990.
Harrison 1992 = id.: "Commemoration and Identification in Buddhiinusmrti". The
Mirror of Memory (ed. by Janet Gyatso; Albany, State University of New
York, 1992), p. 215-238.
Harrison 1993 = id.: "The Earliest Chinese Translations of Mahayana Buddhist
Siitras: Some Notes on the Works Buddhist Studies Review,
Vol. 10, No.2 (London, 1993), p. 135-177.
Harrison 1998a = id.: The Pratyutpanna Samiidhi Sutra, translated by
[T. 418] (Bukkyo Dendo Kyokai English Tripitaka, 25-ll). Berkeley, Numata
.:.Center for Buddhist Translation and Research, 1998.
Harrison 1998b = id.: "Women in the Pure Land: Some Reflections on the Tex-
tual Sources". Journal of fudian Philosophy 26-6 (Dec. 1998), p. 555-572.
HERE = Hastings, James (ed.): Encyclopaedia of Religion and Ethics, 13 vol.
(1908-1927); reed. T. & T. Clark's Publications, 1980.
HHM = Harrison, Paul; Hartmann, Jens-Uwe; Matsuda Kazunobu: "Larger
SukhavatIvyiihasiitra". Manuscripts in the Sch¢yen Collection ill, "Buddhist.
Manuscripts", Volume IT (ed. by Jens Braarvig; Oslo, Hermes Publishing,
2002), pp. 179-214.
Hirakawa, Akira: "Jodo-shiso no seiritsu". lodo-shiso, p. 1-54.
H6biigirin, Dictionnaire encyclopedique du bouddhisme d'apresles sources chi-
noises et japonaises. Tokyo, Maison Franco-Japonaise; Paris, Maisonneuve,
1929-.
LES SOURCES DE LA SUKHAVATI 405
Huntington 1980 = Huntington, John c.: "A Gandharan hnage of Arnitayus' Sukha-
vati:". Annali dell'Istituto Orientale di Napoli, Vol. XL (N.S. XXX; 1980),
p.651-672.
Huntington 1988 = id.: "Mathura Evidence for the Early Teachings of Maha-
yana". Mathurii, The Cultural Heritage (ed. by Doris M. Srinivasan; New
Delhi, American Institute of Indian Studies, 1988), p. 85-92.
Huntington 1996 = id.: "Rebirth in Arnitayus' SukhavatI". The Pure Land Tra-
dition, p. 43-105.
I.CI. = Renou, Louis; Filliozat, Jean, et alii: L'Inde classique, manuel des etudes
indiennes. 2 vol.; T. I (1947), reimp. Librairie d'Amerique et d'Orient, 1985;
T. II (1953), reimp. Paris, EFEO, 1985.
Inagaki 1978 = Inagaki, Hisao: Index to the Larger Sukhiivafivyuha Sutra, A Tibetan
Glossary with Sanskrit and Chinese Equivalents. Kyoto, Nagata Bunshodo,
1978.
Inagaki 1985 = id.: Glossary of the Smaller SukhavatIvyilha Siltra, Sanskrit-
Tibetan-Chinese. A Trilingual Glossary of the SukhavatIvyilha Siltras, Part II
(Kyoto, Nagata Bunshodo, 1985), p. 275-323.
Inagaki 1989 = id.: Pan-chou-san-mei-ching [T. 417]. Indian Philosophy and
Buddhism: Essays in Honour of Professor Kotatsu Fujita on His Sixtieth
Birthday (ed. by Hokkaido Univ.; Kyoto, Heirakuji shoten, 1989), p. 49-88.
Inagaki 1994 = id.: The Three Pure Land Sutras, A Study and Translation from
Chinese (in collaboration with Harold Stewart) [T. 360, 365, 366]. Kyoto,
Nagata Bunshodo, 1994.
Inagaki 1998a = id.: T'an-luan's Commentary on Vasubandhu's Discourse on
the Pure Land, A Study and Translation [T. 1819]. Kyoto, Nagata Bunshodo,
1998.
Inagaki 1998b = id.: Niigiirjuna's Discourse on the Ten Stages (DaSabhilmika-
A Study and Translation from the Chinese of Verses and Chapter 9
[T. 1521] (Ryukoku Literature Series, V). Kyoto, 1998.
Ingholt, Harald: GandhiiranArt in Pakistan. New York, Pantheon Books, 1957.
Inokuchi, Taijun: Bombun Mury6juky6 shahon shUsei (Ryilkoku Daigaku zem-
pon sosho, 6); Kyoto, Hozokan, 1986.
lOdo-Sambukyo (ed. by Nakamura Hajime, Hayashima Kyosh6, Kino Kazuyoshi).
Iwanami bunko 6593-6594; Tokyo, Iwanami shoten, 1963.
lodo-shiso. Koza Daijo Bukkyo, 5 (1985); reed. Tokyo, Shunjilsha, 1996.
Kagawa 1984 = Kagawa, Takao: Muryojukyo no shohon taishi5 kenkyft. Kyoto,
Nagata Bunshodo, 1984.
Kagawa 1985 = id.: "Ryilju no Jodo-shiso". lodo-shiso, p. 151-178.
Kajiyama, Yuichi: "Transfer and Transformation of Merits in Relation to Empti-
ness". Y. Kajiyama, Studies in Buddhist Philosophy (ed. by Katsumi Mimaki
et aI.). Kyoto, Rinsen shoten, 1989), p. 1-20.
Kuo, Li-ying: Confession et contrition dans le bouddhisme chinois du ve au x
e
sie-
cle (publications de l'EFEO, Monographies, 170). Paris, 1994.
406
. ffiROME DUCOR
Kurita, Isao: GandlulranArt, 2 vols. Revised and enlarged edition. Tokyo, Nigen-
sha,2003.
Kwan, Tai-wo: A Study of the Teaching Regarding the Pure Land of Aksobhya Bud-
dha in Early Mahayana. Ph.D. University of California; Los Angeles, 1985
(UMI's order nr. 8513133).
Lamotte, Concentration = Lamotte, Etienne: La Concentration de la marche
heroique (SftraT(!gamasamadhisfttra) (MCB XllI). Bruxelles, Institut BeIge
des Hautes Etudes Chinoises, 1965.
Lamotte, Histoire = id. : Histoire du bouddhisme indien, Des origines ii l' ere
Saka (Publications de l'Institut Orientaliste de Louvain, 14; 1958);' reimp.
Louvain-Ia-Neuve, Universite de Louvain, 1976.
Lamotte, "MafijusrI" = id.: "MafijusrI". T'oung Pao, Vol. XLVIIl (1960), p. 1-
96.
Lamotte, Somme = id.: La Somme du Grand Vehicule d'Asahga (MahtiyanasaT(!-
graha), 2 vol. (Bibliotheque du Museon, 7; 1938); reimp.: Publications de
l'Institut Orientaliste de Louvain, 8; Louvain-la-Neuve, Universite de Lou-
vain, 1973.
Lamotte, Traite = id.: Le Traite de la grande vertu de sagesse de Nagarjuna
(Mahaprajfzaparamitasastra) [T. 1509, k. 1-34,49-50],5 vol. Vol. I-II (Bib-
liotheque du Museon, vol. 18; 1949), reimp. 1966, 1967; vol. III-V (Publica-
tions de l'Institut Orientaliste de Louvain, 2, 12 et 24), Universite de Lou-
vain, 1970, 1976 et 1980.
Lamotte, "Vajraprup." = id.: "Vajraprup. en Inde". Melanges de Sinologie offerts
ii M. Paul Demieville, I (Bibliotheque de l'Institut des Hautes Etudes Chinoi-
ses, XX; Paris, PUP, 1966), p. 113-159.
Lamotte, Vimalakfrti = id.: L' enseignement de Vimalakfrti (Vimalakfrtinirdda)
(Bibliotheque du Museon, vol. 51). Louvain, 1962.
Lee, Junghee: «The Origins and Developments of the Pensive Bodhisattva Ima-
ges of Asia». Artibus Asiae 53 (1993), p. 311-357.
LeRoy Davidson, Joseph: Art of the Indian Subcontinent from Los Angeles Col-
iections. Los Angeles, UCLA Art Galleries, 1968.
Lubac, Henri de, S.J.: Amida ("Aspects du Bouddhisme", II). Paris, Seuil, 1955
(reedition a paraitre aux Editions du Cerf, in: Cardinal de Lubac, "(Euvres
completes", t. XXI).
LVP, Bouddhisme = La Vallee Poussin, Louis de: Bouddhisme, Opinions sur
['Histoire de la Dogmatique. Paris, Beauchesne, 1909.
LVP, Dogme = id.: Le dogme et la philosophie du bouddhisme. 2
e
ed. Paris,
Beauchesne, 1930.
LVP, Kosa = id.: L'Abhidharmakosa de Vasubandhu. 6 tomes (1923-1931); n. ed.
anastasique presentee par Etienne Lamotte: MCB XVI (1971); reimp.
1980.
LVP, Morale = id.: La morale bouddhique (1927); reimp. Editions Dharma,
2001.
LES SOURCES DE LA suKHA VATl 407
Machida, Soho: "Life and Light, The Infinite: A Historical and Philological Analy-
sis of Amita Cult". Sino-Platopjc Papers, 9 (Univ. of Pennsylvania, Dec.
1988), p. 1-46.
Magnin, Paul: Bouddhisme, unite et diversite. (ColI. "Patrimoines, bouddhisme")
Paris, Cerf, 2003.
Mallmann, Marie-Therese de: Introduction it l'etude d'Avalokitefvara. Annales
du Musee Guimet,75. Paris, 1948.
Mbdj. = Mochizuki Shinko: Bukkyi5 daijiten, 10 vol.; 4
e
ed., Tokyo, Sekai-seiten
kanko kyokai, 1958-1963.
MCB = Melanges Chino is et Bouddhiques (Bruxelles, Institut BeIge des Hautes
Etudes Chinoises), 1932-.
Mochizuki, Shinko: Pure Land Buddhism in China, A Doctrinal History, ch. 2-4
(trl. by Leo M. Pruden). Pacific World, Third Series, Nr. 3 (Berkeley, Insti-
tute of Buddhist Studies, 2001), p. 241-275.
The Monuments of Siind (Marshall, John; Foucher, Alfred; Majumdar, N.G; 1940);
rpr. Delhi, Swati Publishings, 1982.
Muller 1881 = Muller, Friedrich Max: Textes sanskrits decouverts au Japon.
Annales du Musee Guimet, 2 (1881), p. 1-37.
Muller 1883 = id. & Nanjio Bunyiu [Nanjo Bun'yfi]: Sukhiivatf-vyuha, Descrip-
tion of Sukhiivatl the Land of Bliss. Anecdota Oxoniensia, Aryan Series,
Vol. I, Part. II (Oxford, 1883); rpr. Amsterdam, Oriental Press, 1972.
Muller 1894 = id.: "The Larger SukhavatI-vyfiha", "The Smaller SukhavatI-
vyfiha". Buddhist Mahiiyiina Texts, II.
Nakamura, Hajime: Indian Buddhism, A Survey with Bibliographical Notes
(1980); rpr. "Buddhist Traditions", vol. I; Delhi, Motilal Banarsidass, 1987.
Nanjio Bunyiu [Nanjo Bun'yfi]: A Catalogue of the Chinese Translation of the
Buddhist Tripitaka, the sacred canon of the Buddhists in China and Japan.
Oxford, Clarendon, 1883; rpr. Taiwan, 1975.
Nattier 2000 = Nattier, Jan: "The Realm of A Missing Piece in the
History of Pure Land Buddhism". JIABS 23-1 (2000), p. 71-102.
Nattier 2003 = id. : «The Indian Roots of Pure Land Buddhism: Insights from the
Oldest Chinese Versions of the Larger Sukhiivatl-vyuha». Pacific World,
Third Series, Nr. 5 (Berkeley, Institute of Buddhist Studies, 2003), p. 179-201.
Pagel, Ulrich: The Bodhisattvapitaka. Buddhica Britannica, Series Continua V;
Tring, The Institute of Buddhist Studies, 1999.
The Pure Land Tradition: History and Development (ed. by James Foard, Michael
Solomon and Richard K. Payne; Berkeley Buddhist Study Series, 3). Univ.
of California at Berkeley & Institute of Buddhist Studies, 1996.
Quagliotti 1977 = Quagliotti, Anna Maria: "Osservazioni suI Buddha di Bruxelles".
Rivista degli Studi Orientali, Vol. LI (Universita di Roma, 1977), p. 137-
140 + 3 ill.
Quagliotti 1996 = id.: "Another Look at the Mohammed Nari Stele with the
So-called «Miracle of SravastI»". Annali del Istituto Universitario Orientale,
Vol. 56 (Napoli, 1996), pp. 274-289.
408
. JEROME DUCOR
Rahula, Walpola: Le compendium de la super-doctrine (Philosophie) (Abhidhar-
masarp.uccaya) d'Asanga (Publications de l'EFEO, vol. LXXVill). Paris,
1971. .
Robert, Jean-Noel: Le Sutra du Lotus, Suivi du Livre des sens incomparables et
du Livre de la contemplation de Sage-Universel, traduit du chinois [T. 262,
276,277]. Collection "L'espace interieur". Paris, Artheme Fayard, 1997.
Rowell, Teresina: The Background and Early Use of the Buddha-kyetra Concept.
Eastern Buddhist, Vol. VI, Nr. 3 (1934), p. 199-246; Nr. 4 (1935), p. 379-
431; Vol. VIT, Nr. 2 (1937), p. 131-176.
Sakaki, Ryozaburo: Kaisetsu bongogaku. Kyoto, 1907.
Salomon & Schopen = Salomon, Richard; Schopen, Gregory: "On an Alleged
Reference to Amitabha in a Inscription on a Gandharian Relief".
nABS, Vol. 25, No. 1-2 (2002), p. 3-31.
Scherrer-Schaub, Cristina Anna: Commentaire ala soixantaine
sur Ie raisonnement ou Du vrai enseignement de la causalite par Ie MaItre
indien Candrakfrti. MCB XXV (1991).
Schop. 1977 = Schopen, Gregory: "Sukhavatl as a Generalized Religious Goal in
Sanskrit Mahayana Sfitra Litterature". Indo-Iranian Journal, vol. 19, no. 1-2
(1977), p. 177-210.
Schop. 1985 = id.: "Two Problems in the History of Indian Buddhism" (Studien
zur Indologie undlranistik, 10, 1985); reed. in Schop. 1997, p. 23-55 (reference
a la reed.).
Schop. 1987 = id.: "The Inscription on the Image of Amitabha and the
Character of the Early Mahayana in India". nABS, Vol. 10, No.2 (1987),
p.99-137.
Schop. 1997 = id.: Bones, Stones and Buddhist Monks. Honolulu, Univ. ofHawai'i
Press, 1997.
Serinde, Terre de Bouddha: Dix siecles d'art sur la Route de la Soie. Paris, Reu-
nion des Musees Nationaux, 1995.
Sharma 1984 = Sharma, Ramesh Chandra: Buddhist Art of Mathurii. Delhi, Agam
Prakashan, 1984.
1989 = id.: "New Inscriptions from Mathura". Mathurii, The Cultural
Heritage (ed. by Doris Meth Srinivasan; New Delhi, American Institute of
Buddhist Studies, 1989), p. 308-315.
Sharma 1995 = id.: Buddhist Art, The Mathurii School. New Delhi & London,
Wiley Eastern Ltd., New Age International, 1995.
Shih, Robert: Biographies des moines eminents de Houei-Kiao, premiere partie:
Biographie des premiers traducteurs [T. 2059] (Bibliotheque du Museon,
Vol. 54). Louvain, Institut Orientaliste, 1968.
Shinshu zensho. 75 vol. (1913-1916); reed., Tokyo, Kokusho kankokai, 1971.
Silburn, Lilian (ss la dir. de): Le bouddhisme (1977); reed. Aux sources du boud-
dhisme. Paris, Fayard, 1997.
LES SOURCES DE LA SUKHAVATI 409
SP = U. Wogihara [Ogiwara, Unrai] and C. Tsuchida [Tsuchida, Chikashi]: Sad-
dharma-pwyfarfka Sutram. Tokyo, Sankibo Buddhist Bookstore, 1958.
T. = TaishO shinshU Daizokyo, 100 vol. (ed. by Takakusu, Junjiro; Watanabe, Kaigyoku;
Ono, Gemmyo; Tokyo, 1924-1935). Reimp. Tokyo, 1960; Taipei, 1983.
T. Rep. = Paul Demieville, Hubert Durt, Anna Seidel (compil.): Repertoire du
canon bouddhique sino-japonais, edition de TaishO. Fasc. annexe du Hob6gi-
rin, ed. rev. et augm. Paris et Tokyo, 1978.
Tanabe, Katsumi: "Iconographical and typological investigations of the Gandha-
ran fake Bodhisattva image exhibited by the Cleveland Museum of Art and
Nara National Museum". Orient XXVI (Tokyo, 1988), p. 84-107.
Tanaka, Kenneth K.: The Dawn of Chinese Pure Land Buddhist Doctrine, Ching-
ing Hui-yuan's Commentary on the Visualization Sutra [T. 1749]. State Uni-
versity of New York Press, 1990.
Thomas, Edward Joseph: The Perfection of Wisdom, The Career of the Predes-
tined Buddhas, A Selection of Mahilyana Scriptures translated from the
Sanskrit. Wisdom of the East Series (1952); rpr. Rutland, Charles E. Tuttle,
1992.
Tib. Trip. = The Tibetan Tripitaka, Peking Edition; 168 vol.; Tokyo, Tibetan Tri-
pitaka Research Institute, 1955-1961.
Tissot 1985 = Tissot, Francine: Gandhara. Paris, Adrien Maisonneuve, 1985.
Tissot 1987 = id.: Les arts du Pakistan et de ['Afghanistan (Coll. "Ecole du Lou-
vre, Les grandes etapes de I'art"). Paris, Desc1ee de Brouwer, 1987.
Tsukamoto, Zenryu: A History of Early Chinese Buddhism, 2 vol. (trsl. by Leon
Hurvitz). Tokyo, Kodansha International, 1985.
Tucci, Giuseppe: "A propos Avalokitesvara" (Buddhist Notes, I). MCB IX (1948-
1951), p. 173-219.
Vievard, Ludovic: Vacuite (sfinyata) et compassion (karm;la) dans le bouddhisme
madhyamaka (College de France, Publications de I'Institut de civilisation
indienne, 70). Paris, Diffusion de Boccard, 2002.
Watters, Thomas: "The A-mi-t'e Ching". China Review: or, Notes and Queries
on the Far East, Vol. 10, no. 4 (Hong-kong, 1881-1882), p. 225-240.
Wieger, Leon, S.J.: Amidisme chino is et japonais. Imprimerie de Hien-hien,
1928.
Yamabe, Nobuyoshi: The Sutra on the Ocean-like Samadhi of Visualization of
the Buddha: The Interfusion of Chinese and Indian Culture in Central Asia as
Reflected in a Fifth Century Apocryphal Sutra [T. 643]. Ph. D. Yale Univer-
sity, 1999 (UMI's order nr. 9930977).
Yamada 1984 = Yamada, Meiji (under the dir. of): The Sutra of Contemplation
on the Buddha of Immeasurable Life as Expounded by Siikyamuni Buddha
[T. 365]. Kyoto, Ryukoku University, 1984.
Yamada 2002 = Yamada, Meiji: "Buddhism of Bamiyan". Pacific World, Third
Series, Nr. 4 (Berkeley, Institute of Buddhist Studies, 2002), p. 109-122.
410 JEROME DUCOR
Yamada, Isshi: Kan.uJapUl:!(jarika, The White Lotus o/Compassion; 2 vols. New
Delhi, Heritage Publishers, 1989.
Zurcher 1959 = Zurcher, Erile The Buddhist Conquest o/China, The Spread and
Adaptation 0/ Buddhism in Early Medieval China, 2 vol. (Sinica Leidensia,
vol. 11; 1959); reed. Leiden, E.J. Brill, 1972.
Ziircher 1984 = id.: "Par-deliiIa porte de Jade: Ie bouddhisme en Chine, au Viet-
nam et en Coree". Le monde du boudhisme (paris, Bordas, 1984), p. 185-203.
Abstract
During the past few decades the Western Indianists have published an
increasing number of studies on Amitabha/Amitayus and his Sukhavatl.
One of them is Gerard Fussman's long and interesting article in Journal
Asiatique 1999, p. 523-586. Unfortunately, however, some rather basic
mistakes seriously detract from this study, as when it confuses the 21 st and
46
th
Vows (p. 574-575) or when it asserts that the very birth in SukhavatI
is the ekajiitibaddha's birth itself (p. 550, n. 62). Due to some confusions
about Buddhist terms, Fussman seems to have misunderstood the origi-
nal method of the Smaller Sukhiivatf (§ 10) and concludes that birth in
this buddha-field necessitates bodhicitta and countless roots of merit as
prerequisites (p. 567). This unlikely interpretation was supposely justified
by a quite baffling transformation of aviparyastacittam into «aviparyasta
<bodhi>cittam», that is «without renouncing his [i.e. bodhisattva's] vow
to reach the bodhi».
In the present article I try to gather the scattered information concer-
ning Amitabha's epigraphy and representations: apart from the early
Govindnagar and the late SancI inscriptions, there is no conclusive evidence
in any sculpture, even in the Buddhamitra Triad or the Mohammed Nari
Stele. Neither is there any evidence in the Buddhananda (<<Bruxel1es»)
Group that Fussman defines as the terminus ante quem for Amitabha's
cult (p. 548, n. 57), despite the uncertainty over both its identification
and its dating.
The comparison between the various su.tras' s methods finally leads me
to the hypothesis that the Pratyutpanna-samiidhi-sutra could be an attempt
to incorporate the concept of SukhavatI into the Prajfiiipiiramitii tradi-
tion, this attempt being in tum criticized by the Madhyamika as it appears
in the Upade.sa.
HIMALAYAN CONUNDRUM? A PUZZLING ABSENCE
IN RONALD M. DAVIDSON'S INDIAN ESOTERIC BUDDHISM
DAVID N. GELLNER
Ronald M. Davidson's Indian Esoteric Buddhism: A Social History of
the Tantric Movement (2002, New York: Columbia University Press) is,
in many ways, a masterpiece. It synthesizes an enormous mass of mate-
rial, fearlessly crosses conventional disciplinary boundaries, and makes
a qualitative leap forward in the understanding and interpretation of early
Buddhist Tantrism. Davidson has marshalled internal textual evidence in
great quantity, and related it to epigraphical and other historical evidence,
in satisfying and original way. For once, the eulogistic citations on the
book's back cover (in this case by David F. Germano, Matthew Kapstein,
and Phyllis Granoff) do not exaggerate.
Early scholars of Buddhism found Tantricism to be so offensive that
they could not bring themselves to study it dispassionately - and as
Davidson emphasizes, it is offensive, no matter how apologists, whether
inside or outside the tradition, have tried to explain its antinomianism
away. This early attitude was to a considerable extent shared by David
Snellgrove; that and the intellectual rationalism of his approach (and his
concomitant lack of interest in the details of political arrangements) meant
that his mature synthesis Indo-Tibetan Buddhism: Indian Buddhists and
their Tibetan Successors does not represent anything like as important an
achievement from the point of view of historical sociology. Geoffrey
Samuel's Civilized Shamans: Buddhism in Tibetan Societies is a similar
highly laudable attempt at synthesis and overview, but covers a later
period and deals only briefly with the Indian background.
Davidson's thesis is that the new, politically more insecure conditions
in India after the Gupta period generated a new kind of feudalism, which,
following BDChattopadhyaya, he calls samanta feudalism (p. 137). Power
was more decentralized than it had been in the preceding period, long-
distance trade (which had been strongly supportive of Buddhism) was in
Journal of the International Association of Buddhist Studies
Volume 27 • Number 2 • 2004
412 DAVID N. GELLNER
decline, militant Shaivism was increasing in importance, regional art
traditions and languages were flourishing, and popUlations were disper-
sing from old to new centres. In this context, Buddhism had to adapt
and one way it did so was to take on the language and ritual of political
imperialism. Davidson demonstrates in fine detail how the terminology
of the initiation of Tantrics parallels exactly that of the feudal ruler.
He concludes:
Buddhists derived their mandala forms and functions, not so much from the
theoretical treatises of Indian polity as from their immediate observation at
the disposition and execution of realpolitik in their environment ... by obser-
ving the actual relationships of the overlords and their peripheral states ...
Indeed, the Buddhist mandala is a classic analysis of the system of samanta
feudalism in early medieval India, all sufficiently sanctified for the monas-
tic community. (p. l39)
It was not that the founders of Buddhist Tantrism were
sycophantic actors imitating the domain of mere politics ... [Rather t]hey
attempted to transform power and hierarchy into community and congregation.
Swimming in the sea of samanta feudalism, they tried to see it as an ocean
of gnosis and to engage it in the creation of merit for all beings. (p. 161)
In doing so they went far beyond any previous Mahayanist adaptions
in the name of 'skilful means'. As Davidson points out, esoteric Buddhism
spread through royal patronage outside India, and it would have been
strange if the same were not true inside India as well (p. 115). In spite of
the use of tribal imagery in Tantrism, everything points to its being the
creation of urban intellectual elite (p. 238).
Davidson emphasizes that the Tantric Buddhist corpus was produced
and systematized within a remarkably short time: "in a matter of a few
decades, not over the centuries sometimes proposed" (p. 338). One would
have expected him therefore to welcome Alexis Sanderson's demonstra-
tions that numerous passages of the Buddhist Y ogini Tantras were lifted
directly from Shaiva originals (Sanderson 1994,2001). But Davidson is
concerned rather to insist (p. 203ff) that Buddhist Tantras took their mate-
rials from other sources as well, and not just from the Shaiva Kapalikas.
Be that as it may, Davidson has certainly produced the most sophisti-
cated account yet of what he calls "the esoteric conundrum": how
it could be that a religion based on renunciation, moderation, and non-
HIMALAYAN CONUNDRUM? 413
violence could adopt rituals of sex, violence, and the infraction of all
social rules, and introduce symbolic forms fully reflective of this dra-
matic antinomianism.
The aim of this communication, however, is not just to praise David-
son, but also to point out a surprising and mysterious lacuna in his book,
namely Nepal, and more particularly the Buddhist culture of the Kath-
mandu Valley (which in the period he is dealing with, and indeed well
into the twentieth century, went by the narne 'Nepal').
It is not that Davidson is hostile to anthropological or ethnographic
evidence as such. In his introduction, he writes:
Since the time of Herodotus' description of the Skythians in his Histories,
however, participant-observation data have proved of extraordinary value in
assessing foreign cultures, whether accrued by the historian, by anthropolo-
gists, or both, as in the case of this book. We cannot underestimate [sic] the
value to the historian of learning the colloquial languages of these cultures,
living in villages, or (in our case) in Buddhist monasteries. Indeed, many
Indologists would affirm that they did not truly understand much of this
complex society until they had lived and worked there among the descendants
of those very people under investigation. (p. 21)
Surely, then, he must deal somewhere with the Newars who are the
last remaining Mahayana South Asian Buddhists, as a long line of schol-
ars have recognized - from Sylvain Levi, to David Snellgrove, Siegfried
Lienhard, Dhanavajra Vajracharya, Hemraj Sakya, John Locke, Mary
Slusser, Michael Allen, Niels Gutschow, Bernhard KOlver, Gautam Vajra
Vajracharya, Kashinath Tarnot, Ian Alsop, Karunakar Vaidya, Min Bahadur
Sakya, Todd T. Lewis, Kimiaki Tanaka, K. Yoshizaki, Bruce Owens,
Gregory Sharkey, Alexander von Rospatt, Will Tuladhar-Douglas, John
Huntington, Dina Bangdel, and myself!. Surely Davidson will at least
mention the one South Asian society where the Namasangiti is still recited
in Sanskrit as part of the daily liturgy and where each complex Buddhist
ritual begins with a samkalpa that locates the action as taking place in the
mandala of Sri Samvara in the northern Pancala country of Bharatavarsa?
But no, Davidson proceeds to claim, in the one statement of the book
I See references for main publications. One should also consult Levy (1990) and Tof-
fin (1993) on Hindu (Shaiva) Tantrism in the Kathmandu Valley.
414 DAVID N. GELLNER
which I feel wholly competent to say is a fundamental, and very unfortu-
nate, error of fact and interpretation, " ... there are no continuously sur-
viving Indian Buddhist institUtions ... " (p. 24).
This is very strange. Nepal does not figure in the index of Indian Eso-
teric Buddhism, but there are in fact a number of mentions, as well as
at least one to Newars, and two to Newari, and he implies that he has been
to Nepal
2
• Davidson is not averse to citing anthropological authors on
India. Thus Jonathan Parry and Lynn Teskey Denton are quoted oIl. asce-
tics in Banaras (p. 95), Gross is cited on modern Indian sadhus (p. 207,
238), Elvin, Gell, Vitebsky, and Boal are invoked on the Murias and their
tribal neighbours (p. 321). Davidson also mentions his own field visits to
some Indian sacred sites (e.g. p. 309 to Brahrnaur in Himachal Pradesh),
but not his visit to Kathmandu. He castigates other scholars for ignoring
tribal ethnography (p. 321) and for labelling phenomena as Tibetan when
they are actually Indian (p. 374 n. 97). So what can possibly explain the
fact that he ignores Nepal and the Nepali descendants of the Buddhists
he is discussing in this way?
Nepal was only recognized by the British as a separate kingdom, that
is to say, as different in statUs from the princely states of India, in 1923.
Had history been a little different, had Nehru not been such a gentleman,
Nepal could easily have ended up inside the Indian union. Even today the
border is open and the citizens of both countries are free to travel to and
work in the other country without the need for any documentation. At any
period before the nineteenth centUry what is now called Nepal was just
another region of the subcontinent. Surely Davidson cannot have fallen
for t ~ e fallacy, which he would (and does) castigate in other scholars, of
reading back into history the contingent political arrangements of the
present day?
2 Davidson notes, "My experience with Buddhist scholars in India, Nepal, and Tibet
has been exclusively textual, with little interest displayed towards epigraphy, archaeology,
or other sources" (p. 352 n. 52). I cannot speak for India or Tibet, but several Buddhist
scholars in Nepal have indeed paid considerable attention to epigraphy, notably Hemraj
Sakya in his numerous historical works (see Sakya 1977 for his magnum opus on the history
of the sacred complex of Svayambhu). For various reasons to do with Buddhist identity
and Nepali nationalism, several Nepali scholars have also been very concerned with the
epigraphy and archaeology of Lumbini.
HIMALAYAN CONUNDRUM? 415
Not being personally acquainted with Davidson, and not being able to
fmd one in the book itself, I have no answer to this question. As some-
one who has written at length on Newar Buddhism - precisely in the
framework of Sylvain Levi (whom Davidson would surely honour as a
great Sanskritist, Tibetologist, Sinologist, and Buddhologist) that Nepal
is "India in the making" (Levi 1905 I: 28) and that Newar Buddhists are
indeed the last South Asian Mahayana Buddhists - such neglect is both
galling and puzzling. Had Davidson deigned to look at the history of the
Medieval Malla kingdoms of the Kathmandu Valley, he would have found
ample evidence to illustrate how Tantric Buddhism there evolved a close
and symbiotic relationship with royal power, and with Shaivism, as indeed
he argues was the case for medieval India. Had he looked at the traditional
and still current practice of Buddhism among the Newars he would have
found as a clear case as he could have wished of Vajrayana' s efficacy at
"transform[ing] power and hierarchy into community and congregation".
I shall close by quoting the words of another prominent historian of
Buddhism in India, Greg Schopen:
It is a curious fact that scholars interested in Mahayana Buddhism in India
have paid so little attention to Nepal - indeed it may actually be perverse.
It was Nepal, after all, that first revealed, and continued to supply, most of
the Mahayana literature that we have in Sanskrit. .. Ironically, while some
students of the Mahayana have understood this [the importance of studying
ethnographic accounts of actually practised Buddhism], they have generally
not looked in the most obvious places ... where, one wonders, are the names
of those who are working in an actual South Asian Mahayana culture - John
Locke, David Gellner, and Todd Lewis? There is something rather strange
here and it is time - indeed long overdue - that Newar Buddhism assume
its rightful place in Buddhist studies. (Schopen 2000: ix-xi)
References
Allen, M. 2000. Ritual, Power, and Gender: Explorations in the Ethnography of
Vanuatu, Nepal, and Ireland. Delhi: Manohar.
Bangdel, D. 1999. 'Manifesting the Mandala: A Study of the Core Iconographic
Program of Newar.Buddhist Monasteries in Nepal'. PhD, Ohio State Uni-
versity. UMl AAT 9941281.
Douglas-Tuladhar, W. 2002. 'The Fifteenth-Century Reinvention of Nepalese
Buddhism'. DPhil, University of Oxford.
416 DAVID N. GELLNER
Gellner, D.N. 1992. Monk, Householder, and Tantric Priest: Newar Buddhism and
Its Hierarchy of Ritual. Cambridge: Cambridge University Press.
Gellner, D.N. 2001. The Anthropology of Buddhism and Hinduism: Weberian
Themes. Delhi: OUP.
Gellner, D.N. & D. Quigley (eds) 1995. Contested Hierarchies: A Collaborative
Ethnography of Caste among the Newars of the Kathmandu Valley, Nepal.
Oxford: Clarendon.
Gutschow, N. 1997. The Nepalese Caitya: 1500 Years of Buddhist Votive Archi-
tecture in the Kathmandu Valley. Stuttgart: Edition Axel Menges. .
Huntington, I.e. & D. Bangdel 2003. Circle of Bliss: Buddhist Meditational Art.
Serindia.
KOlver, B. & H. Sakya 1985. Documentsfrom the Rudravanya-Mahtivihiira. Sankt
Augustin: VGH Wissenschaftsverlag.
Levi, S. 1905. Le Nepal: etude historique d'un royaume hindou (3 vo1s). Paris:
Leroux. (Reissued 1991, Delhi: Asian Educational Services.)
Levy, R. 1990. Mesocosm: Hinduism and the Organization of a Traditional Hindu
City in Nepal. Berkeley: University of California Press.
Lewis, T.T. 1984. 'The Tuladhars of Kathmandu: A Study of Buddhist Tradition
in a Newar Mercant Community'. PhD, Columbia University (UMl8506008).
Lewis, T.T. 2000. Popular Buddhist Texts from Nepal: Narratives and Rituals of
Newar Buddhism. Albany: SUNY Press.
Lienhard, S. 1984. 'Nepal: The Survival of Indian Buddhism in a Himalayan
Kingdom' in H. Bechert & R. Gombrich (eds) The World of Buddhism. Lon-
don: Thames and Hudson.
Locke, I.K. 1980. Karunamaya: The Cult of Avalokitesvara-Matsyendranath in
the Valley of Nepal. Kathmandu: Sahayogi.
Locke, I.K. 1985. Buddhist Monasteries of Nepal: A Survey of the Biihiis and
BahlS of the Kathmandu Valley. Kathmandu: Sahayogi.
Owens, B .M. 1989. 'The Politics of Divinity in the Kathmandu Valley: The Fes-
tival of Bungadya/Rato Matsyendranath'. PhD, Columbia University.
Pal, P.:1974. The Art of Licchavi Nepal. Bombay: Marg Publications.
Pal, P. 1974. Arts of Nepal. Leiden: E.I. Brill.
Petech, L. 1984. Mediaeval History of Nepal (c. 750-1492) (Serie Orientale Roma
44, 2
nd
ed.). Rome: ISMEO.
Regmi, D.R. 1985. Inscriptions of Ancient Nepal. Delhi: Abhinav.
Riccardi, Theodore, Ir. 1980. 'Buddhism in Ancient and Early Medieval Nepal'
in A.K. Narain (ed.) Studies in the History of Buddhism, pp. 265-81. New
Delhi: BK Publishing.
Sakya, H. 1977 (1098 NS). SrI SvayambhU Mahiicaitya. Kathmandu: Svayambhu
Vikas MandaI.
Sakya, M.B. See <www.nagarjunainstitute.com>
Samuel, G. 1993. Civilized Shamans: Buddhism in Tibetan Societies. Washington:
Smithsonian Press.
HIMALAYAN CONUNDRUM? 417
Sanderson, A. 1994. 'Vajrayana: Origin and Function' in Dhammakaya ed. Bud-
dhism into the year 2000, pp. 87-102. Bangkok and LA: Dhammakaya Foun-
dation.
Sanderson, A. 2001. 'History through Textual Criticism in the study of Saivism,
the Paiicaratra, and the Buddhist YoginItantras' in F. Grimal (ed.) Les sources
et le temps, pp. 1-47. Pondicherry: Ecole d'Extreme Orient.
Schopen, G. 2000. 'Foreward' in Lewis (2000).
Sharkey, G. 2003. Buddhist Daily Ritual: The Nitya Puja in Kathmandu Valley
Shrines. Bangkok: Orchid.
Slusser, M. 1982. Nepal Mandala: A Cultural Study of the Kathmandu Valley
(2 vols). Princeton: Princeton University Press.
Snellgrove, D. 1957. Buddhist Himalaya. Oxford: Bruno Cassirer.
Snellgrove, D. 1987. Indo-Tibetan Buddhism. Boulder: Shambala Press.
Tanaka, K. & K. Yoshizaki 1998. Nepiiru Bukky6. Tokyo: Shunjiisha.
Toffm, G. 1993. Le palais et le temple: Lafonction royale dans la vallee du Nepal.
Paris: CNRS.
Vaidya, Karunakar 1986. Buddhist Traditions and Culture of the Kathmandu Val-
ley (Nepal). Kathmandu: Sajha.
Vajracharya, D.V. 1973. Licchavikiilkii Abhilekh. Kathmandu: INAS.
Vajracharya, G.V. 1987. 'An Interpretation of Two Similar Nepalese Paintings
in the Light of Nepalese Cultural History' in N. Gutschow & A. Michaels
(eds) Heritage of the Kathmandu Valley (Nepalica 4), pp. 29-42. Sankt
Augustin: VGH Wissenschaftsverlag.
von Rospatt, A. 2000. 'The Periodic Renovations of the Thrice Blessed Svayam-
bhiicaitya of Kathmandu'. Habilitation thesis, Hamburg University.
NOTES ON CONTRIBUTORS
Oliver FREIBERGER is Assistant Professor of Classical Indian Bud-
dhism at the University of Texas at Austin. He received his Ph.D. from
the University of Gtittingen/Germany in 1999. He is the author of Der
Orden in der Lehre: Zur religiosen Deutung des Sangha imfriihen Bud-
dhismus, a study of concepts of the sangha in Pali canonical texts, and
has published several articles on early Buddhism and on methodological
issues in the study of religion.
Klaus-Dieter MATHES works as a research fellow at the Institute for the
History and Culture of India and Tibet at the University of Hamburg. His
research in progress is dealing with the Indian origins of Tibetan mahamu-
drii traditions. He obtained a PhD from Marburg University with a study
of the Dharmadharmatiivibhaga and recently completed his Habilitation
at the University of Hamburg with a work on 'Gos Lo tsa ba gZhon nu
dpal's mahamudrii interpretation of the Ratnagotravibhaga. At the
moment he is a visiting professor at the University of Vienna.
Richard D. MCBRIDE, IT is presently a post-doctoral fellow in Korean
Studies and Buddhist Studies at Washington University in St. Louis.
Another related article, "Dharal).I and Spells in Medieval Sinitic Bud-
dhism," will appear in JIABS, vol. 28. He is currently revising his Ph.D.
dissertation "Buddhist Cults in Silla Korea in their Northeast Asian Con-
text"(UCLA, 2001) for publication.
David N. GELLNER is Lecturer in the Anthropology of South Asia, Uni-
versity of Oxford. He is the author of Monk, Householder, and Tantric
Priest: Newar Buddhism and its Hierarchy of Ritual (CUP, 1992) and The
Anthropology of Buddhism and Hinduism: Weberian Themes (OUP,
2001). He has edited several other books, including (with Declan Quigley)
Contested Hierarchies: A Collaborative Ethnography of Caste among
the Newars of the Kathmandu Valley, Nepal (OUP, 1995) and Resistance
and the State: Nepalese Experiences (Social Science Press, 2003).
Rebuilding Buddhism: The Theravada Movement in Twentieth-Century
Nepal (Harvard University Press), a monograph co-authored with Sarah
LeVine, is due out in 2005.
Journal of the International Association of Buddhist Studies
Volume 27 • Number 2 • 2004
The International Association of Buddhist Studies
Jikido Takasaki
President
Ernst Steinkellner
Vice-President
Tom Tillemans
General Secretary
Jerome Ducor
Treasurer
Regional Representatives:
Janet Gyatso (Cambridge, MA, USA)
Kazunobu Matsuda (Kyoto)
David Seyfort Ruegg (London)
Peter Skilling (Bangkok)
Directors at large:
Nalini Balbir (Paris), Colette Caillat (Paris),
Georges Dreyfus (Williamstown, MA), Robert M. Gimello (Cambridge, MA),
Oskar von Hiniiber (Freiburg), Sh5ryu Katsura (Hiroshima),
Liying Kuo (paris), Richard Salomon (Seattle), Cristina Scherrer-Schaub (Lausanne),
Lambert Schmithausen (Hamburg)
The International Association of Buddhist Studies, founded in 1976, is devoted to
promoting and supporting scholarship in Buddhist Studies in all its aspects, past and
present, around the world. Membership is open to scholars of all academic disciplines.
Membership dues are USD 40 for full members, USD 20 for student members,
USD 1000 for life members. Subscriptions to the JIABS for libraries and other institu-
tions are USD 70. Dues and subscriptions may be paid in the following ways:
1. by direct bank to bank transfers in US dollars or in the equivalent amount in
Euros or Swiss francs. The transfers should be to the lABS accounts at the Banque
Cantonale Vaudoise, PI. St-Fran90is, Ch-1001 Lausanne, Switzerland (SWIFT CODE:
BCVLCH2L, clearing no. 7677),· account numbers, 983.51.04 for US dollars,
5042.09.82 for EUR and 983.51.02 for Swiss francs.
2. by Visa or Mastercard
3. by cheque payable to the" Association Internationale d'Etudes Bouddhiques
(lABS)." There will be no supplementary charges for cheques drawn in Swiss
currency on Swiss banks. Otherwise, please add 10 dollars or 15 Swiss francs to cover
our processing charges.
4. by Eurocheque in Swiss francs.
Prospective members from developing countries may contact the Treasurer concerning
subsidized membership rates. Dues and subscriptions are payable per calendar year
by December 31 of tlie previous year. Payments other than direct bank transfers should
be sent to Dr. Jerome Ducor, Section de langues et civilisations orientales, Universite
de Lausanne, BFSH 2, CH-! 0 15 Lausanne, Switzerland.

Sign up to vote on this title
UsefulNot useful