Jiabs 29-2

JIABS
Journal of the International

Association of Buddhist Studies

Volume 29 Number 2 2006 (2008)
JIABS
Journal of the International


Oskar VON HINBER
Hoary past and hazy memory. On the history of early Bud-
dhist texts (Presidential address at the XVth Conference of
the International Association of Buddhist Studies, Emory
University, Atlanta, Georgia, June 2328, 2008) . . . . . . . . . . . . . . 193

Junjie CHU
On Digngas theory of the object of cognition as presented
in PS(V) 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
David HIGGINS
On the development of the non-mentation (amanasikra)
doctrine in IndoTibetan Buddhism . . . . . . . . . . . . . . . . . . . . . . . . . 255
Richard D. MCBRIDE, II
The mysteries of body, speech, and mind: The three eso-
terica (sanmi) in medieval Sinitic Buddhism . . . . . . . . . . . . . . . . . . 305
Hidenori SAKUMA
On doctrinal similarities between Sthiramati and Xuanzang . . . . 357
Jonathan STOLTZ
Concepts, intension, and identity in Tibetan philosophy of
language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383

Report on the XV
th
Congress of the International Associa-
tion of Buddhist Studies by Tom J.F. Tillemans, General
Secretary IABS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401

IABS Treasurer financial report by Jrme Ducor . . . . . . . . . . . . 407

Notes on the contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 411

watermark
The Journal of the International
(ISSN 0193-600XX) is the organ
of the International Association of
Buddhist Studies, Inc. As a peer-
reviewed journal, it welcomes
scholarly contributions pertaining to
all facets of Buddhist Studies.
nABS is published twice yearly.
Manuscripts should preferably be sub-
mitted as e-mail attachments to:
editors@iabsinfo.net as one single file,
complete with footnotes and references,
in two different formats: in PDF-format,
and in Rich-Text-Format (RTF) or Open-
Document-Format (created e.g. by Open
Office).
Address books for review to:
nABS Editors, Institut fur Kultur - und
Geistesgeschichte Asiens, Prinz-Eugen-
Strasse 8-10, A-I040 Wien, AUSTRIA
Address subscription orders and dues,
changes of address, and
business correspondence
(including advertising orders) to:
Dr Jerome Ducor, lABS Treasurer
Dept of Oriental Languages and Cultures
Anthropole
University of Lausanne
CH-I015 Lausanne, Switzerland
email: iabs.treasurer@uni1.ch
Web: http://www.iabsinfo.net
Fax: +41 21 6923045
Subscriptions to nABS are USD 40
per year for individuals and USD 70
per year for libraries and other
institutions. For informations on
membership in IABS, see back cover.
Cover: Cristina Scherrer-Schaub
Font: "Gandhari Unicode" designed by
Andrew Glass (http://andrewglass.org/
fonts.php)
Copyright 2008 by the International
Association of Buddhist Studies, Inc.
Print: Ferdinand Berger & Sohne GesrnbH,
A-3580 Hom
EDITORlAL BOARD
KELLNER Birgit
KRASSER Helmut
Joint Editors
BUSWELL Robert
CHEN Jinhua
COLLINS Steven
Cox Collet
G6MEZ Luis O.
HARRISON Paul
VON HINOBER Oskar
JACKSON Roger
JAINI Padmanabh S.
KATsURA ShOryil
KuoLi-ying
LOPEZ, Jf. Donald S.
MAcDONALD Alexander
SCHERRER-SCHAUB Cristina
SEYFORT RUEGG David
SHARF Robert
STEINKELLNER Ernst
nLLEMANS Tom
. HOARY PAST AND HAZY MEMORY
ON TIlE HISTORY OF EARLY BUDDHIST TEXTS'
OSKAR v. HlNOBER
On the occasion of the 215
th
meeting of the American Oriental Soci-
ety in Philadelphia in 2005, the Hittitist Gary Beckman from the
University of Michigan read his presidential address, "The limits of
credulity," in which he sketched modem approaches to the art of
writing history, presented a most useful overview or rather an extract
of the flood of theoretical literature on this topic and, above all, dis-
cussed how far it is advisable and possible to trust sources and how
to evaluate them.
1
All this is exemplified by ancient Middle Eastern,
first of all of course Hittite material. Although based on a culture
with a strong written tradition, much can be learned from this article
also for the thoroughly oral tradition of ancient India and early Bud-
dhism in spite of some marked differences.
In contrast to Beckman's after-dinner speech, the following
deliberations do not concentrate on historiography, neither on mod-
em historiography of the early Buddhist period; nor, and much less
so, on an early Buddhist historiography, which is deplorably absent
despite a remark made by Georg Btihler (1837-1898) to his friend,
the renowned Arabist at the University of Strasbourg, Theodor
Noldeke (1836-1930), as early as 1877:
Mit Deiner Idee, dass die Inder keine historische Literatur haben, stehst Du
auf einem veralteten Standpunkte.
2
This text was read as the Presidential Address on 23 June 2008 during the XV
th
IABS conference held at Atlanta (Georgia) from 23 to 28 June 2008. The oral form of the
presentation has been largely preserved. An enlarged and more detailed version dealing
with the early history of Theravada texts is under preparation.
1 JAOS 125. 2005, 343-352.
2 "Your idea that Indians do not possess literature on history is an outdated point of
Journal of the International Association of Buddhist Studies
Volume 29 Number 22006 (2008) pp. 193-210
194 OSKAR v. HINOBER
This is certainly true, if we remember KalhaI.1a's RajalarmigbJ-l and
the Nepalese Vaf!1savalfs, or almost a millennium earlier in the area
of Buddhism, the Dfpavaf!1$a and the Mahavaf!1sa. Important as both
vaf!1sas are as sources, as underlined by the research of Wilhelm
Geiger (1856-1943) and Erich Frauwallner (1898-1974),3 their
value for the history of texts of the very early period of Buddhism is
quite limited.
However, even if historiography begins too late for the period on
which the following considerations are focussed, and if this Buddhist
historiography contains little information on texts, we cannot con-
clude that this is due to a total lack of interest in history in general or
in the history of texts in particular in ancient India. Early evidence
proving the contrary is found in inscriptions. Already Asoka
vaguely, and it is true in a very general way, refers to the kings of
yore at the beginning of his seventh pillar edict, and compares the
successful propagation of his dhamma to the failure of those ancient
kings to educate their peoples. In the well-known res gestae of his
own reign, Kharavela looks back not only upon his own time.
Kharavela also records a King Nanda, ruling either three hundred or,
more likely, one hundred and three years before him, as having taken
away a Jina image, which he, Kharavela, brought back to his capi-
tal.
4
A similar memory is found in the Rudradaman inscription,
where the K ~ a t r a p a Rudradaman commemorates in the year AD 150
that he repaired and embellished the Sudarsana tank after it was
badly damaged by floods. This Sudarsana tank was originally built,
as Rudradaman reminds the readers of his inscription, by Candra-
gupta Maurya and was subsequently enlarged by Asoka.
5
This is
view," quoted from Julius Jolly: Georg Buhler. Grundriss der Indo-Arischen Philologie
und Altertumskunde. 1. Band, 1. Heft A. Strassburg 1899, p. 13.
3 The relevant articles are quoted in O. v. Hinuber: A Handbook of Ptili Literature.
Berlin 1996 (HPL), 182, 183.
4 Shashi Kant: The Htithfgumphti Inscription of Khtiravela and the Bhabru Edict of
Asoka. Delhi 22000, p. 11, line 6 and p. 17, line 12 of the inscription. For tivasasata
"103(7)" cf. terasavasasata "113",line 11.
5 Idal?l tat;/tikaJ?1 SudarsanaJ?1 ... mauryasya rtijfiai} Candraguptasya rti!jtriyeIJa Vaisye-
na PU!jyaguptena kiirita1]l Asokasya Mauryasya [krlte Yavanartijena TU!jtisphentidhi!jtiiya
HOARY PAST AND HAZY MEMORY 195
indeed a rather long historical memory stretching over almost half a
millennium. Interestingly, no traces of this memory are found in the
text of the inscription of the Hindu ruler Skandagupta, where he
records his repairs of the same tank three centuries later.
6
With the notable exception of the Bhairat edict by Asoka, refer-
ences to Buddhist texts are almost totally absent from early inscrip-
tions.?
However, historical memory is not only preserved in inscriptions,
but in Buddhist texts as well, and, of course, this historical memory
can be used with all due caution to date the texts that preserve them.
This can be done only if the historical memory refers to a datable
event in the political history, and this way of dating texts leads to
approximations at best. Hardly ever was a text composed at the very
time of the event being remembered, and never with the purpose to
simply give a straightforward record of a certain event in ancient
India. What we read is always an interpretation and a purposeful
message of the authors to their audience. The information handed
down by tradition thus depends on the intention and the will of the
authors to select and to convey certain facts. This intention or will to
shape the tradition being handed down is expressed both in the con-
tent and in the literary form of the texts, and both changed consid-
erably during the transition from Vedic to early Buddhist literature.
The intention why the collection later called Tipitaka was brought
together is very clearly stated in the report on the first council held at
Rajagaha. For we are explicitly told why the texts were assembled
and formalized:
pra/;tlilihhir alalTl/q'talTl, F. Kielhorn: Junagadh rock inscription of Rudradiiman; the year
72.EI8.1905-06,p.36-49,esp.p.43,8.
6 This inscription is published in Corpus Inscriptionum Indicarum III: Inscriptions of
the early Gupta kings revised by Devadatta Ramakrishna Bhandarkar. Delhi 1981, no. 28,
296-305, cf. also O. v. Hiniiber: Les documents epigraphiques indiens: Difficultes de leur
interpretation - Examples concernant l'irrigation. Academie des Inscriptions et Belles-
Lettres. Comptes Rendus des Seances de l'Annee 2004. Avril-Juin. Paris 2004 [2006], p.
989-1011, esp. p. 989foll.
7 Only very general references such as trepi!aka or vinayadhara are found occasion-
ally in Mathurii or Amariivan.
196 . OSKAR v. HINUBER
dhammaii ca vinayaii ca sal'J'lgiiyiima pure adhammo dippati dhammo pa!i-
biihfyati avinayo dippati, vinayo pa!ibiihfyati pure adhammaviidino bala-
vanta hanti, dhammaviidino dubbalii hanti ... , Vin II 285,4---8
"Come, let us chant dhamma and discipline before what is not dhamma
shines out and dhamma is withheld, before what is not discipline shines out
and discipline is withheld, before those who speak what is not-dhamma
become strong, and those who speakdhamma become feeble ... " (1. B.
Homer).
The purpose is obviously to preserve and to defend an orthodox tra-
dition. This must have been something quite new in ancient India at
that time, a new and considerable literary challenge to be confronted
not only by early Buddhists, but also by the followers of other new
systems created at that time in eastern India such as Jainas or A.jlvi-
kas. This change in paradigm, the preservation of orthodoxy and no
longer the continuation of the orthopraxy of the Veda, also called for
new literary forms. For this purpose veyyiikaralJas and dhammapari-
yiiyas, or suttantas as they were called later, were developed, per-
haps after some experiments with the prose of the Brahmlll,las, but
certainly based on this model. The model of Vedic prose is easily
detected in the Sagiithavagga of the Sal1lyuttanikaya, whose form
closely corresponds to the stories in the BrahmaI.1as and which even
continues Vedic topics such as the fight between gods and asuras.
The many short episodes telling the reasons for the rules in the
Mahiivagga and the Cullavagga of the Vinaya recall the structure of
Brahmal).as, as was observed by Erich Frauwallner long ago.
8
More-
over the story of the Buddha sneezing reads like an answer to a story
in the Jaiminfya-BriihmalJa, as noticed by William Dwight Whitney
(1827-1894) more than a century ago.
9
If the individual stories neatly connect Brahmal).a prose and early
Buddhist literature, the Buddhists went far beyond their model and
composed the first really long texts in ancient India, as shown by the
8 HPL 32.
9 Henry C. Warren: On superstitious customs connected with sneezing, JAOS 13.
1889, p. XVII-XX, esp. p. XX, where W. D. Whitney refers to Jaiminfya-Bri'ihmal}a II
155.
overall structure of the Khandhaka in "the Vinaya or the Mahiipari-
as an individual text. Moreover, the Mahiipari-
nibbiinasuttanta is the fIrst text ever composed in ancient India, as
far as we can see, with the explicit purpose of commemorating a
historical event, the death of the Buddha and thus, at the same time,
the first attempt to compose a long and coherent story.
The many difficulties encountered by those who shaped or made
use of this new literary form, perhaps for the first time, can be traced
easily in many details in the Mahiiparinibbiinasuttanta. It was, e. g.,
very obviously a matter of great effort to keep the story on track and
not to get lost in numerous narrative side alleys: Once eight reasons
for an earthquake are enumerated, which fits well into the story, a
cascade of totally unrelated groups of "eights" follows suit, as if in
the oral period of the early Buddhist tradition hearing or mentioning
the fIgure "eight" immediately and almost unavoidably triggered the
memory of the respective paragraphs from what we now call the
AIiguttaranikaya.
1O
In spite of evident diffIculties like these, which do not seem to
have found much attention in research, those monks who created the
Buddhist siUras had a very clear idea about the formalization of the
new texts. The idea of remembering the places where the Buddha
was supposed to have delivered a certain satra at the beginning of
each individual text was certainly an innovation. This happy decision
to provide the texts with a geographical frame, quite in contrast to
the earlier Vedic literature, where very little is found on topogra-
phy,l1 not only preserved a large of place names, both vil-
lages and towns, in the Buddhist literature. In addition, the particular
wording introducing these place names can tell us much about the
development of the literary form of early Buddhist texts and about
the historical memory of the early authors.
The opening formula of a satra is almost too well known to be
repeated here in the standard wording:
10 HPL 60.
11 On this point, see K. Hoffmann, as note 13 below, p.122.
198 OSKAR v. HINUBER
eva/I! me sutal'!! ekm'[!samaym?! bhagavti Stivatthiyal'!! viharati Jetavane Anti-
thapilJrj.ikassa tirtime
Thus I have heard. At onetime the Lord stayed at Savatthi in the Jetavana,
the park of AnathapiI:tt;lika.
This very wording continues after arame either by ... tatra kho bha-
gava bhikkhu amantesi, MN I 6,27foll. (No.2., Sabbasavasuttanta)
with the local adverb tatra, by the developed wording ... atha kho
bhagava pubba1J.hasamayaf!1. .. " MN I 160,27foll. (No. 26., Ariya-
pariyesanasuttanta) or, finally, by '" tena kho pana samayena
ayasmaAnando ... , MN III 189,27 (No. 132., Ananda-Bhaddekaratta
suttanta 2). Whereas the very beginning evaf!1. me sutaf!1. ekaf!1.
samayaf!1. has been discussed perhaps much more often than it really
deserves ever since John Brough's (1917-1984) article published
almost sixty years ago,12 little if any attention has been paid to the
much more interesting place names and to the way in which they are
introduced.
Besides this well-known introduction there are others, used much
more rarely and phrased in a slightly different way, such as:
eval'!! me sutal'!! ekal'!! samayal'!! bhagava kurttsu viharati - kammasa-
dhammal'!! nama kurttnal'!! nigamo - fatra kho bhagava bhikkha amantesi,
MN I 55,28foll. (No. 10., Satipatthanasuttanta)
... the Lord stayed in the land of the Kurus - there is a market place in the
land of the Kurus named Kammasadhamma - there the Lord addressed the
monks ...
The phrase "there is a market place in the land of the Kurus named
Kammasadhamma" syntactically forms a parenthesis, which does
not seem to be an exciting observation. However, almost half a cen-
tury ago Karl Hoffmann (1915-1996) demonstrated that this particu-
lar way of introducing place names can be traced back to Jndo-
12 Thus have I heard ... , BSOAS 13. 1950, p. 416-426, reprinted in J. Brough: Col-
lected Papers. London 1996, p. 63-73; for further references see HPL 53 and add: M.
Tatz: Thus have I heard: At one time, IIJ 40. 1997, p. 117fo11.; B. Galloway: A reply to
Professor Mark Tatz, IIJ 40. 1997, p. 367-371; F. Tala, C. Dragonetti: Ekaql samayam, ill
42. 1999, p. 53-55.
Iranian syntax.
13
For the Pali sentence just quoted is exactly parallel
to the Old phrasing found in the Achaemenian inscriptions of
Dareios (521-486) at Behistun: "when he arrived in Media - a town
Miiru by name is in Media - there he joined battle with the Medes."14
The use of phrases like this one continues through rare examples
from the Vedic language only into the earliest layers of Pali. Com-
parative syntax here allows the detection of a wording that is obvi-
ously a very early predecessor to the later common formula
introducing suttantas by the names of towns like Savatthi, Rajagaha
or others.
The place names mentioned in the older place name parenthesis
are quite different from these towns. Hardly any of the famous Bud-
dhist nagaras is mentioned, but only fifteen different market places
nigamas occur such as kammiisadhamma and once a Brahmin vil-
lage, a briihma1J.agiima in:
eva1!l me suta1!l eka1!l samaya1!l bhagavii Magadhesu viharati - piicfnato
Riijagahassa Ambasafltfii niima briihmaflagiimo - tass'uttarato Vediyake
pabbate tena kho pana samayena Sakkassa ... , D II 263,
2ff. (No. 21., Sakkapaiihasuttanta)
Thus I have heard. At one time the Lord stayed in Magadha - to the east of
Riijagaha there is a Brahmin village named AmbasaI.1c;lii - north 'of it ... in
the Indasiila cave '"
This rather exceptional formulation is due to a seemingly exact
description of the location of that particular cave.
These nigamas also occur in a slightly developed wording:
eva1!l me suta1!l eka1!l samaya1!l bhagavii Kurusu ciirika1!l caramlino mahatii
bhikkhusa1!lghena saddhi1!l yena thullakofthita1!l niima kuruna1!l nigamo tad
13 K. Hoffmann: Die Ortsnamen-Parenthese im Altpersischen und Vedischen (ZDMG
110. 1960), in: AuJsatze zur Indoiranistik ed. by J. Narten. Volume 1. Wiesbaden 1975, p.
120-129, on Piill S. 128f., cf. also G. E. Dunkel: Naming parenthesis in Indo-Iranian and
Indo-European, MSS 41.1982, p. 11-21.
14 R. Kent: Old Persian. Grammar, Text, Lexicon. New Haven 1953, p. 121: ya8ii
Madam pararasa - MaruS nama vardanam Madaiy - avada hamaranam akunalls, DB II,
line 22fo11. 22.
200 OSKAR v. HINDBER
avasari. assosU/!l kho Thullakot?hitii briihmaflagahapatikii ... , MN II
54,25foll. (No. 82, Ranhappalasuttanta)
... at one time the Lord walking on tour among the Kurus ... where there
was the market place of the Kurus called Thullakonhita, there he went ...
In such phrases, describing the Buddha and his followers travelling,
"the Lord walked in the land of ... , where there was a place called so
and so there he went", not only nigamas are mentioned but, in addi-
tion to a very few nagaras, also more frequently again brahnwlJaga-
mas, "Brahmin villages," which occur almost only in the following
formula:
eva'!! me suta,!! eka,!! samaya1J2 bhagavii Kosalesu ciirika'!! caramiino mahatii
bhikkhusaT!lghena saddhi,!! yena nagaravindaT?2 niima kosaliinaT!l briihmafla-
giimo tad avasari. assosu'!! kho Nagaravindeyyakii brahmaflagahapatikii ... ,
M III 290,26foll. (No. 150., Nagaravindeyyasuttanta)
Thus I have heard. At one time, the Lord, walking on tour in Kosala together
with a large group of monks, where there is the Brahmin village of Kosala
named Nagaravinda, there he went.
Interestingly, nine of the altogether fourteen Brahmin villages men-
tioned in the Theravada-Tipi!aka are situated in Kosala, four in
Magadha, and only one in the MalIa country. IS This compares well
with the evidence gathered from Vedic literature on the history and
geographical distribution of the Vedic schools. As research by M.
Witzel has shown, Kosala was at the eastern fringe of later Vedic
literature, and the Brahmins there used to study the Kal)va Sakha of
the Satapathabrahmal)a.
16
These then could well be the very Brah-
mins traced in ancient Buddhist literature.
In the immediate predecessor of the later formula, which men-
tions a place name such as Savatthi together with a monastery such
15 This is, at the same time, the only reference to the word used in the
Vinaya in the definition of majjhimadesa at Vin 1197,27, but not in the usual formula as in
Ud 78,5, the second of the only two references to the ThuI:1a in the Thera-
vada-Tipitaka, cf. also IlJ 45.2002, p. 79.
16 M. Witzel: The Development of the Vedic Canon and its Schools: The Social and
Political Milieu, in: Inside the Texts, Beyond the Texts ed. by M. Witzel. Harvard Oriental
Series. Opera Minora Vol. 2. Cambridge!Mass. 1997, p. 257-345, esp. p. 313foll. 5.2
as the Jetavana of a wording without parenthesis and
without naming a monastery occurs when a Brahmin village is
referred to:
eva1J1 me suta1J1 eka1J1 samaymJ1 bhagavil Kosalesu viharati sillilyal'rl" brilhma-
lJagilme. tatra kho bhagavil bhikkha ilmantesi, S V 144,12fo11. == 227, 12fo11.
Thus I have heard. At one time the Lord stayed in Kosala in the Brahmin vil-
lage SaIa. There the Lord addressed the monks ...
Interestingly, there were no vihiiras in Brahmin villages, but, much
more importantly, also not in the nigamas.
Taking together the very old place name parenthesis with the
Brahmin villages and market places (nigamas), the missing nagaras
and, above all, the missing vihiiras, it is more than evident that these
formulas belong to a very ancient layer of the formulation of Bud-
dhist texts as preserved within the Theravada-Tipi!aka. Moreover,
going back to the old parenthesis of place names, it is possible to
trace the reason for the word order - the town preceding and the
monastery following the verb viharati: SiivatthiYaJ!l viharati Jetavane
... - in the sentence opening satras, which is clearly conditioned by
the stylistic prehistory of this formula.
Furthermore, the preponderance of Kosala as a location of Brah-
min villages matches Vedic evidence. Consequently, we can be
fairly confident of finding here really ancient village names pre-
served in the memory of the early Buddhists. This is confirmed by
the simple fact that these early locations of the beginnings of Bud-
dhism very soon faded into the background and were superseded by
the five prominent cities enumerated at the beginning of the Mahasu-
dassanasuttanta, No. 17. in the Dighanikaya: Campa, Rajagaha, Sa-
vatthi, Saketa, Kosarnbi, D II 169,11. First of all, Savatthi
emerged as the prominent town, figuring at the beginning of 5 of 34
suttantas in the Dighanikaya, but already in 67 of 152 in the Majjhi-
manikayaand in innumerable texts of the SaqIyutta- and AIiguttara-
nikayas. Still later texts were almost flooded by references to Sa-
vatthi, to such a degree that the Millasarvastivinaya, as G. Schopen
202 OSKAR v. HlNOBER
has shown, suggested inserting Sravastl whenever a menk forgot the
place name when reciting a satraY
In the list of the five tdwns prominent in the Buddhist tradition,
one place name is conspicuous by its absence, namely Pataliputta,
the later Maurya capital. The reason is obvious. Pataliputta did not
exist during the time of early Buddhism. Its foundation is described
in a well-known paragraph at the beginning of the Mahiiparinibbii-
nasuttanta, where the Buddha makes the following prediction during
the reign of Ajatasattu: "As far as there are settlements of the Aryas,
as far as there are trading routes, this will be the first city (aggana-
gararrz) Pataliputta, a place where customs are collected (putabhe-
dana)," D II 87,33-88,1. At the same time, the Buddha changes the
name of the place from Pataligiima to Pataliputta and calls the new
city a puta-bhedana, most likely in a word play with patali-putta. No
mention is made of Pataliputta as the capital (riijadhiinf, e.g., D II
7,29) of the Maurya empire, in spite of the fact that Mauryas are
indeed referred to at the very end of the same text, when the relics
are distributed.
The Moriyas of Pipphalivana
18
learn very late of the death of the
Buddha, and when their envoy arrived at the site of the nibbbiina, all
relics had been distributed and only charcoal was left, over which the
Moriyas of PipphaIivana erected a stapa.
19
This rather meagre result
of the efforts made by the Moriyas to secure a share of the relics also
points to a time long before the ascent of the Maurya-dynasty. Later
this episode was obviously considered embarrassing and conse-
quently cancelled by the redactors of the Sanskrit version of the
M ahiiparinirvii1}asatra. 20
17 G. Schopen: If You Can't Remember, How to Make It Up: Some Monastic Rules for
Redacting Canonical Texts (1997), in: Buddhist Monks and Business Matter. Still More
Papers on Monastic Buddhism in India. Honolulu 2004, p. 395-407. In later texts, the
place names did not really matter, because there was no longer any immediate reference to
an old tradition anyway.
18 Although it is tempting to compare Pipphalivana to modem Piprahva, the difficulties
involved are considerable.
19 D II 166,2lfoll.; 167,17fo11.
20 A survey of the relevant material can be found in A. Bareau (1921-1993): Re-
Of course the prophecy of the Buddha about the future of Patali-
putta caught the attention of scholars at an early date, at least since
1879, when H.' Oldenberg (1854-1920) published the introduction to
his edition of the Mahiivagga of the Vinaya, which contains a
parallel to this part of the Mahiiparinibbiinasuttanta. One of the last
in this line of scholars dealing with this reference seems to be K. R.
Norman in his history of Pali literature a century later in 1983.
21
Strangely, all modem scholars seem to follow the conclusion
drawn by H. Oldenberg that the Mahaparinibbanasuttanta must have
been composed during the time of the Maurya dynasty because the
Buddha is assumed to be referring to the capital of the Maurya
empire, which he is certainly not. On the contrary, the Buddha men-
tions very clearly a place where merchandise is exchanged, an
important city certainly, but not a capital. This, however, was per-
haps not too evident before B. Kolver (1938-2001) finally clarified
the meaning of putabhedana,22 badly understood previously and still
most strangely misunderstood as "scattering its seeds far and wide"
(!?) instead of "market place" in a recent Dighanikaya-translation
published two years after B. Kolver's article.
23
Now, if the Mahiiparinibbiinasuttanta, or more cautiously, this
paragraph, was composed by the end of the fourth century, as K. R.
cherches SUT la biographie du Buddha dans les Siitrapitaka et les Vinayapitaka anciens II,
2: Les demiers mois, Ie et les fimerailles. Publications de l'Ecole
d'Extreme-Orient 77. Paris 1971, p. 303. However, Bareau's assumption that the Moriyas
of Pipphalivana were only introduced at a late date and only by the Theravadins obviously
turns the development of the text upside down. For, in addition to the reason given above,
it is easy to see that the obscure pipphalio was replaced by the much more common pip-
pala in: pippaliiyanai} mii(lQvai} pippalavatyiim migiirastr1pa1]l pratiNhiipayati, Mahiipari-
nirviilJasiitra 51.21. On the other hand, going back from pippala to pipphali does not make
any sense at all.
21 The Vinaya Pitaka1]l ed. by H. Oldenberg. London 1879, p. XXXVII; K. R. Norman:
Piili Literature including the canonical literature in Prakrit and Sanskrit of all the Hfna-
yiina Schools of Buddhism. A History of Indian Literature VII,2. Wiesbaden 1983 p. 38.
22 B. Kolver: Kautalyas Stadt als Handelszentrum: der Terminus putabhedana-.
ZDMG 135.1985, p. 299-311.
23 The Long Discourses of the Buddha. A Translation of the Dfgha Nikiiya by M. Wal-
she. Boston 1987, p. 238.
204 OSKAR v. HINtrBER
Norman conjectures, it would be rather strange tnat neither the
Mauryas nor Pataliputta, as the capital of a state or at least as a city
of any political importance, are referred to. This omission, on the
other hand, makes a lot of sense if the text is pre-Mauryan, that is
earlier than the accession of Candragupta Maurya in about 320 BC,
and most likely quite a while earlier, because it can and has been
demonstrated that the parallel to this part of the Mahaparinibbana-
suttanta preserved in the Vinayapitaka is linguistically a slightly
modernized version?4
Leaving aside the question of the date of this linguistic moderni-
zation of the story of the foundation of Pataliputta, a more interesting
question is whether the omission of the Mauryas can be used at all to
determine a date. An answer to this question can be found only by
investigating whether ancient Indian authors, given their presumed
lack of historical interest, did pay enough attention to changing
political situations to adapt their texts accordingly.
For this purpose it is necessary to look, at least very briefly, for
references to historical events in older Buddhist literature.
Even a superficial inquiry limited only to those Buddhist texts
preserved in Sanskrit yields some examples. A parallel story also
concerning Pataliputta and found in the Divyavadana, which is
derived from the MUlasarvastivadavinaya, is particularly illuminat-
ing. When the Buddha makes a prediction referring to Pataliputra in
the Asoka legend he says:25 tathagatasya
Pataliputre nagare Asoko namna raja caturbhagacakra-
vartf dharmiko dharmaraja yo me sarfradhiitfin vaistarikan
Divy 368,25foI1.:
24 This can be seen by comparing the place name Nadika, which is preserved in its
original eastern form only in tbe DIghanikaya (D II 91,15), but changed to Natika in the
Vinayapi!aka (Vin I 232,31); the old historical gen. pI. Ta/lfial!! (D II 87,2 etc.) is regularly
modernized as rajt7IW1!l (Vin I 228,30 etc.), cf. O. v. Hiniiber: Der Begilln der Schrift !lnd
friihe Schriftlichkeit ill Indien. Akademie der Wissenschaften und der Literatur, Mainz.
Abhandlungen der geistes- und sozialwissenschaftlichen Klasse, Jg 1989, Nr. 11, chapter
X, p. 46fo11.
25 Cf. J. Strong: The Legend of King AsiJka. A Study and Translation of the Asokava-
dana. Princeton 1983, p. 61.
A century after the Tathagata entered the Nirv1iI).a there will be a king named
Asoka in the city of Pataliputra, a conqueror of the four quarters of the
world, a righteous king, who will spread my relics.
Here, long after the Maurya period, this dynasty is to
Pataliputra. A similar reference comes from the of
the same Vinaya:
mama ... Kani:jko nama raja
bhavi:jyati. so 'smin pradde (i.e. Kharjarika) sUipa1Jl prati:jthiipayati, tasya
iti sa1?ljiia bhavi'fyati, GM ill 1,2,3 foIl .
... there will be a king named in the family. He will
establish a sUipa in this country (Kharjiirika) which will be called
Stiipa.
Now a third reference in an unclear fragmentary context found only
recently by R. Salomon can be added. Most likely it is some Ava-
dana text, where it is said: [maha]yiinasamprasthito nii[ma
riijii] , "a king named who has set forth on the Great
Vehicle.,,26
These three references demonstrate that Buddhist authors did
indeed pay some attention to historical events, if only to honour rul-
ers whose patronage was appreciated by the Buddhist saf!1.gha. The
Miilasarvastivadavinaya, which gained its shape much later than that
of the Theravadins, was evidently adapted to the politics of its time.
And, as we can be fairly certain about the year 127 as the date of
since the research done by H. Falk,27 the redaction can be
dated at the earliest to the first half of the second century.
Although it is true that the examples are few and far between,
they show nevertheless that there was much more awareness of his-
26 R. Salomon: A fragment of a Collection of Buddhist Legends, with Reference to
King as a Follower of the Mahayana, in: Jens Braarvig [Ed.]: Buddhist Manu-
scripts Volume II. Manuscripts in the Schl'lyen Collection III. Oslo 2002, p. 255-267. -
There seems to be even a portrait of recently discovered on a relief from Gan-
dhara, cf. F. Grenet: Note additionnelle on B. Marshak: Vne peinture kouchane sur toile,
Academie des Inscriptions et Belles Lettres. Comptes rendus des seances de l'annee 2006.
2006 [2008], p. 947-954 + 955-963, esp. p. 957.
27 H. Falk: The yuga of Spujiddhvaja and the era of the Silk Road Art and
Archaeology 7. 2001, p. 121-136.
206 OSKAR v. HJ:NlJBER
tory than mostly assumed, not only in Buddhist literature. As M.
Witzel has pointed out, the text of the Mahiibhiirata responds to
political developments lJP to the times, whereas peoples
intruding into India as late as during the Gupta period are included.
28
And as it seems, particularly the Buddhists always had an eye on his-
tory, as the tradition on the date of the Buddha and other indications
also demonstrate.
Consequently, it may be meaningful that the Buddha said to
have predicted a brilliant future for Pataliputra as a city of commerce
and not of politics. Moreover, given the great affection and admira-
tion for Asoka found everywhere in Buddhist texts, it is indeed hard
to conceive a date contemporaneous with, and still less likely after
Asoka.
29
Given the importance of the rise of the Maurya empire even
under Candragupta, who is better known for his inclination towards
Jainism, one might conjecture that the latest date for the composition
of the Mahiiparinibbiinasuttanta, at least for this part of it, is around
350 to 320 BC.
If this is not altogether too far off the mark, and if it is remem-
bered that the date of the nirviilJa can be assumed to be about 380
BC, this dating of the text certainly has also some consequences for
the assessment of the content. For a distance in time of roughly thirty
to sixty years from the event recorded to the text conceived allows
for a fair chance to trace true historical memory.
Of course it is not intended to tum the Mahiiparinibbiinasuttanta
as a whole into a full historical record now and to read it as a histori-
cal account instead of hagiography, which it is. However, while
28 M. Witzel: The Vedas and the Epics: Some Comparative Notes on Persons, Line-
ages, Geography, and Grammar, in: Epics, Khilas, and Purtil}as: Continuities and Rup-
tures. Proceedings of the Third Dllbrovnik International Conference on the Sanskrit Epics
and Purtil}as. September 2002 ed. by P. Koskikallio. Zagreb 2005, p. 21-80, esp. p. 63.
29 An extremely late date for the composition of the Mahtiparinibbtinasuttanta during
"the 2
nd
or 3'" century of the Common Era" (!) was recently assumed without any reason
given by C. Woodford Schmidt: Aristocratic Devotees in Early Buddhist Art from Greater
Gandhara, SASt 21. 2005, p. 25-45, esp. p. 25. This date, which is perhaps due to a
misreading of "AD" for "BC" (??), can be safely and confidently ruled out.
many references to Indian history found in the Tipi!aka remain
one .point should be beyond any reasonable doubt: the
death of the Buddha occurred at some point in history and at a cer-
tain place. A second point is no less important. The death of the Bud-
dha as the founder of the Buddhist sal'!lgha was an event of huge
consequences for the then contemporary Buddhists, and an event
witnessed by many monks and deeply penetrating into the collective
memory of all Buddhists of the time. In contrast, the bodhi, certainly
of prime importance for Buddhism and Buddhists, was not witnessed
by any future monk and no collective memory could spring up from
this event.
Although there were witnesses present at the nirvii1!a, mythologi-
cal features abound in the description of the death of the Buddha
because at that time no religious person could possibly die without
accompanying miracles, and at the time after the Buddha's death, no
text describing the career of the founder of any religion could have
possibly found acceptance without miraculous features.
Between fact and fiction are the earthquakes at the moment when
the Buddha gives up his vital force and again at the moment of the
nirvii1!a.
30
Although geophysics does not necessarily rule out that
they occurred, in all likelihood they did not; nobody would take them
to be historical events. And the divine flowers showering on the
deceased Buddha in such a quantity that they filled the whole small
town of Kusinara knee deep with heavenly mandiirava flowers
31
is
just as evidently mythology as it is indispensable in any record of an
event such as the Buddha's nirvii"(ta.
So, even if there had been the will or at least the intention of the
early Buddhist authors of the Mahiiparinibbiinasuttanta to portray a
more or less exact historical record of the nirvii1!a, miraculous and
supernatural events were impossible to avoid in an environment in
which the practice of all sorts of yogic achievements was common-
30 iiyusaJ?lkharaJ?l ossaji, D II 106,22 and D II 156,36.
31 D II 160,31.
208 OSKAR v. HINOBER
place. Miracles simply were part of the world-view of the authors of
the Mahiiparinibbiinasuttanta.
Consequently, if there are details that might be called historical,
they are necessarily buried in what we call mythology. However, we
can try to unearth a bit of history, as Gary Beckman has suggested in
his lecture, by replacing the rather credulous question "Why might it
be false?" by the more sceptical one: "Why should it be true?" if we
venture to attempt to separate fact from fiction.
32
For instance the route taken by the Buddha to Kusinara might be
such a detail based on remembered history, if the names of the many
otherwise unknown villages are recalled. The same might be said
perhaps in the case of the earliest record of what was possibly an
epidemic at Nadika, where twelve persons recently deceased are
enumerated by name, among them strange names such as
Nika!a or Ka!issabha not mentioned in any other source.
33
The name
of the very last monle ordained by the Buddha, Subhadda,34 may be
historical memory and, of course, the absence of nuns during the
35
In contrast, another famous paragraph cannot be historical as it
stands. These are the last words of the Buddha:
handa diini bhikkhave iimantayiimi va: vayadhammii saJTlkhiirii appamiidena
sampiidethii ti. ayafJ1 tathiigatassa pacchimii viicii, D II 156,lfo11.
Now, monks, I address you: Decay is inherent in all component things.
Work out your salvation without indolence.
32 JAOS 125,2005, p. 349.
33D IT 91,26-92,1l.
34 so bhagavato pacchimo sakkhisiivako ahosi, D IT 153,1l.
35 O. v. Hinliber: The Foundation of the BhikkhunTsaqJgha. A contribution to the earli-
est history of Buddhism. Annual Report of the Intemational Research Institute for Ad-
vanced Buddhology at Soka University for the Academic Year 2007 (ARIRIAB 11). Tokyo
2008, p. 2-29, esp. p. 22. - One of the most likely candidates for historical memory is per-
haps the famous name of the last meal of the Buddha. In all likeliness the obscure suka-
ramaddava (D IT 127,5) is the name of a local dish, which was piously preserved, while
the true meaning was soon forgotten: O. v. Hinliber: The Cause of the Buddha's Death:
The last Meal of the Buddha. A Note on siikaramaddava. JPTS 26. 2000, p. 105-117.
We can be sure that this is not exact historical memory, correct in
spirit and content at best, but certainly not in wording, because the
Buddha did not speak PilIi.
However this may have been in detail remains, to a considerable
extent, a matter of conjecture, because our sources never allow us to
go beyond more or less likely or probable conclusions about the
roots of the texts that reach far back into the period of early
Buddhism. On the other hand, determining the date of the closure of
a text remains a still thornier problem. However, we can be sure to
find quite a lot of very old material in the Theravada tradition and
we can, therefore, hope with a little confidence that the wonderfully
solemn and unique conclusion of the Mahiiparinibbiinasuttanta may
not be altogether wrong:
36
Abbreviations
ARIRIAB
BSOAS
D
Divy
EI
GM
HPL
IIJ
JAOS
36 D II 167,20.
evam etaf!L bhiltapubbaf!L
"Thus it was in the days of yore."
Annual Report of The International Research Institute for
Advanced Buddhology at Soka University
Bulletin of the School of Oriental and Mrican Studies
DIghanikaya
Divyavadana
Epigraphia Indica
Gilgit Manuscripts. Ed. by N. Dutt. Srinagar, 1939-1959.
O. v. Hintiber: A Handbook of Pilli Literature. Berlin
1996.
Indo-Iranian Journal
Journal of the American Oriental Society
210
JPTS
MSS
S
SASt
Ud
Yin
ZDMG
OSKAR v. HINDBER
Journal of the Pali Text Society
Mtinchener Studien zur Sprachwissenschaft
Sal1lyuttahikaya
South Asian Studies
Udana
Vinayapitaka
Zeitschrift der Deutschen Morgenliindischen Gesellschaft
ON DIGNA.GA'S THEORY OF THE OBJECT OF COGNlTION
AS PRESENTED IN PS(V) I'
JUNJIECHU
I. Introduction
1-1. In the first chapter of his magnum opus, the PramiilJasamuccaya
(PS) with its Vrtti (PSV),l Dignaga (ca. 480-540) asserts that percep-
tion, being free of conceptual construction, takes as its
object-field whereas objects other than this are
objects of conceptual construction, i.e., they are associated with
name, genus, etc., and thus are imagined, conventionally existent, or
the conceptualization of something that has been perceived earlier;
he also provides a list of objects that belong to the so-called "pseudo-
perception" Thus, for Dignaga the -
usually translated as "particular" - is the only real object of cogni-
* I am indebted to Prof. Dr. Eli Franco for his valuable comments on the first draft of
this paper, which allowed for great improvements in both content and language. I am also
indebted to Ven. Prof. Dammajoti who has made many valuable suggestions on a part of
this paper which I read as one of my lectures during my visit to the Centre of Buddhist
Studies of the University of Hong Kong in Feb. 2006.
I A new Sanskrit reconstruction based on linguistic materials collected from Jinendra-
buddhi's commentary has been offered by Steinkellner; see PS(V) 1. All quotations
appearing in this paper refer to this edition.
2 The Sanskrit words alambana, and artha refer to the object of cognition;
however they are used with some differences in meaning, for example, regarding to the
difference between and alambana Vasubandhu says: "Furthermore, what is the
difference between the object-field and object-support? If one thing has activity (karitra)
in respect to another thing, the former has the latter as its object-field; and that which is
grasped by the mind and mind-associates is the object-support." (AKBh 19,16-17: kab
punar yasmin yasya ktiritram, sa tasya yac cittacaittair
g,hyate, tad alambanam.) In this paper I try to render these Sanskrit words with different
English words:" "object-field" for "object-support" for alambana, and "object-re-
ference" for artha.
3 CfPS(V) 1.2-3, 7cd-8ab; Hattori 1968: 24-25; 28. Cf. also Chu 2004.
Volume 29 Number 2 2006 (2008) pp. 211-253
212 JUNJIECHU
tion. However, other than mentioning that is "inexpressi-
ble" (avyapadeya),4 he does not offer a clear explanation of the
term.
1-2. In another part of the PS(V} 1, as we will see in the following
(IT-I), in talking about the of cognition sphere (ayatana),
Dignaga also states that in respect to its cognition sphere, perception
is said to take the samanya, a term used here referring to many sub-
stances or atoms that constitute visible matter as a who'le, as its
object. This statement is criticized by his opponent as being contra-
dictory to his refutation of an object-support (alambana) that con-
sists of many atoms. In my opinion, this problem has arisen owing to
his being misunderstood as referring to the external things. However,
in following I shall argue: (1) Dignaga regards the samanya, no mat-
ter how it is interpreted differently by different schools, as an inter-
nal part of a cognition, the so-called self-cognizable (svasarrzvedya).
(2) Based on the Sautrantika's idea that the cognition sphere (aya-
tana) is not real, Dignaga denies the reality of an object having an
external existence; and furthermore, he maintains that the self-
cognizable or the internally cognizable is the source of cognition.
(3) Granting the Sautrantika's theory of cognition having aspects
(akiira) , he interprets the object's aspect as the self-cognizable.
Based on these points, I shall conclude that Dignaga's theory of
object of cognition presented in PS(V) 1 is consistent, and that he
can be characterized as "advocator of the theory of internal object"
(antarjfieyavadin), a term used by his commentator Iinendrabuddhi.
5
II. siimiinya
II-I. In PS(V) 1.4cd-5ab, in reply to an objection to his statement that
perception is free of conceptual construction, Dignaga gives some
additional information about his idea of The opponent
4 Cf PSV 1.1,24-25 (ad PS 1.2'cd'): svasiimiinyalak
e
a/;ziibhyiil!l hy avyapadesyavan:za-
tviibhyiil!l van:ziidi g,hftviinityatayii ciinitya'!l van:ziidfti manasii sandhatte, Cf. Hattori
1968: 24.
5 See below n. 46.
ON DIGNA.GA'S THEORY OF THE OBJECT OF COGNITION 213
argues that although it is said in the Abhidharma treatise that a sense
awareness is a direct awareness of something blue, etc., and not a
determination in the form "this is blue," or a cognition of a property
of the object,6 and thus that perception is free of conceptual construc-
tion; it is also stated in the Abhidharma treatise that the five groups
of awareness take an agglomerate (saficita)7 of atoms as their object-
support.
8
How can it be possible, the opponent asks, that these
awarenesses take an agglomerate as object-support, if they do
not image it to be a unity (ekataft)? And how is it to be understood
that such a sensory awareness takes the particular (svalak$Wza) as its
object-field in respect to the particular characteristic of the cognition
sphere (iiyatanasvalak$a1Ja) , and not to the particular characteristic
of the substantial element?9 Dignaga's answer is as follows:
[Perception] is said to take the siimiinya as its object-field in respect to its
[external] cognition sphere ([biihya-jiiyatana), because [perception] is pro-
duced by many substances [i.e. atoms], but not because of the conceptual
construction of non-difference with respect to difference. 10
11-2. Based on the original meaning of the word, Jinendrabuddhi, the
only Indian commentator of Dignaga's PS(V), interprets siimiinya in
6 Cf. PSV 1.2,20-21: abhidharme 'py uktam vijtinati no
tu nflam iti, arthe 'rthasaiijfifna tu dharmasaiijfifiti.
7 Dignaga uses the word saTtcita in the sense of samudtiya, as Iinendrabuddhi says.
See below n. 25 and n. 40.
8 Cf. AKBh 34,1-2: na caika indriyaparamtil}ur vti vijiitilWl?ljana-
yati, saficitasraytilambanatvtit pmlctintif!l vijntinaktiytintim. "No single atom of the sense
faculty or a single atom of the object-field can produce awareness, because the five groups
of awareness have the agglomerate of [atoms] as their basis [of cognition, i.e. the sense
faculty] (tiSraya) and object-support."
9 Cf. PSV 1.2,22-23: kathaf!l tarhi saiicittilambantib panca vijntinaktiytib, yadi tad
ekato na vikalpayanti. yac praty ete na dravyasva-
iti.
10 PSV 1.2,25-26 (ad PS 1.4cd): anekadravyotptidyatvtit tat sviiyatane
yam uktam, na tu abhedakalpantit. Cf. also PV 3.196: sa ca
ntintarel}tipartilwl}iin I tadektiniyamtij jfitinam uktaT!l stimtinyagocara,!l II "And the special
quality of atoms [of producing cognition] cannot [occur] without other atoms. Thus, since
the cognition is not restricted to a single [atom], it is said to have the stimtinya as its
cognition sphere." (Cf. also the translations in Dreyfus 1997: 87-88 and in Dunne 2004:
396-7.)
214 JUNJIECHU
the sense of the "common property" of all the atomscconstituting an
agglomerate and their thus becoming the cognition sphere, i.e., the
object of awareness. He writes:
These many atomic things, expressed [collectively] by the word "agglomer-
ate" (sancita) and the word "particular characteristic of cognition sphere,"
are common by being similar in the capacity of producing corresponding
awareness in dependence on each other. A common property is indeed
common [to many things] ... [followed by a grammatical explanation] 11
That is to say, cognition is produced by atoms, all of which have the
common property in their being capable of producing a cognition;
thus they are collectively the object-cause (alambanapratyaya) of
cognition and the single word "agglomerate" can be used for them.
Jinendrabuddhi emphasizes that Dignaga's reason for using the term
siimiinya is to eliminate the restriction (niyama) of the object of a
cognition to a single atom-substance. Nevertheless, this does not
imply, through the usage of the word siimiinya, that the perception
takes the universal characteristic of many atoms as its object-field:
thus the contradiction mentioned by the opponent does not exist. At
another point, in explaining why it is said in the Abhidharma treatise
that the five groups of awareness take an agglomerate as their object-
support, Jinendrabuddhi expresses also this idea:
Those atoms, generated through their own causes and conditions, inasmuch
as each being capable of producing awareness when they are situated in
close proximity to each other, are called "agglomerate" (sancita) ... For by
virtue of the conditions of getting close to each other they are agglomerated,
gathered together. Through these many atoms in such a state, [the five
groups of awareness] that bear their own appearance are produced. There-
fore, it is said: "[the five groups of awareness] take the agglomerate as
object-support" in consideration [of the fact] that [these awarenesses] sup-
11 PST 45,11-12: yat tat coktam anekal!! pa-
ramiil}uvastu tad eva pratiniyatavijl1iinajananasiimarthyena siidharmye1}a paraspariipek/fa-
yii samiinam. samiinam eva siimiinyam, ...
ON DIGNAGA'S THEORY OF THE OBJECT OF COGNITION 215
port themselves, without distinction, on all these [atoms], called "agglomer-
ate,,,12 but definitely not on a single substance.13
This samanya is also interpreted by Manorathanandin in his com-
mentary on Dharmaklrti's PV. In comparison to that of Iinendra-
buddhi, his interpretation is quite simple. According to him, taking
Samanya as the operation-sphere means taking the agglomerate,
namely a kind of collection (sanghata) of atoms, as the operation-
sphere and not the universal (samanya), which is different from the
individual atoms. Thus the undesired consequence (prasanga),.i.e.,
the self-contradiction, would not follow, namely, such a perception
would be accompanied by conceptual conception on account of hav-
ing taken the universal as its object.
14
II-3. Dignaga's above statement, which is in fact based on Vasuban-
dhu's idea as stated in the AKBh,15 is interpreted by Iinendrabuddhi
12 I do not follow the Tibetan translation which corresponds to: sarviirrzs tiin saFicitii-
Cf. Psrr D 21b5-6: de rnams thams cad la 'ang bsags pa zhes
bya ba'i khyad par gyis dmigs pa ste ...
13 PST 44,12-45,5: svahetllpratyayebhyo ye paramii7Javojiiyante, te 'nyonyasannidhii-
niivasthiipriiptii eva santaJ:z pratyekarrz vijfiiinopajananasamarthii/:z saficitasabdenoktii/:z ...
parasparopasarpa7Japratyayair hi te saiicitii!1 sarrzhatfkrtiiJ:z. tais tathiividhair anekaiJ:z sva-
pratibhiisii janyanta iti saiicitiilambanii ity uktiiJ:z, sarviirrzs tiin saiicitiikhyiin
bante, naikam eva dravyam iti krtvii.
14 Cf. PVV ad PV 3.196: siimiinyagocararrz jiiiinam
uktarrz tattvaviidinii, na tll tat kathal?z
tviit savikalpatvaprasmigaJ:z. (For the analysis of the compound saiicitaparamii7Jusalighiita-
cf. AKBh 23,24: paiicendriyadhiitavaJ:z paiica saFicitii/:z, paramii7Jll-
salighiitatviit. )
15 Cf. AKBh 7,22-24: nanu caivarrz samastiilambanatviit parka vijiiii-
nakiiyii!z priipnuvanti na praty ete
na ity "[Objection:] Since in this circumstance
the five groups of [sensory] awareness take an aggregate (samasta) [of atoms] as their
object-support, they must take the universal (siimiinya) as their object, not the particular
[Reply:] In respect to the particular characteristic of the cognition spheres,
these [five groups of sensory awareness] are accepted as having the particular characteris-
tics [of the cognition spheres] as their objects, not in respect to the particular characteristic
of the substances [viz. an atoms]. Therefore the fault does not exist." Cf. Hattori 1968: 26
Dab, and 89, n. 39. Cf. also MVBh 65a12-16: El i'i\"9:o
El i'iCflo -* El i'o El i'o l'f** El 1'1'1
El fo "Question: 'Why does tactile awareness take the
216 , JUNJIECHU
as an attempt to solve the problem that the perception of an object
consisting of many atoms would be not free of conceptual construc-
tion. However, another serious problem arises. As pointed out by
Hattori,16 Dignaga's idea as stated here is vehemently attacked by
Mallavadin in NC 86,6ff. The troublesome point is as follows: to say
that perception is produced by many atoms and therefore takes the
siimiinya of these atoms as its object seems to be similar to the thesis
Dignaga himself criticizes in the AP(V) and in PS(V) 1.14ff, regard-
less of whether it is referring to all things possessing the' common
property of being capable of producing a cognition, as interpreted by
Jinendrabuddhi, or simply to the entire collection of all atoms, as
interpreted by Manorathanandin. Thus it seems to contain a self-
contradiction in another sense. I shall argue in following discussion '
that this is actually a misunderstanding of Dignaga's idea on the part
of his opponents because of their realistic viewpoint.
11-4. Here, before we begin to discuss the problem of the siimiinya, it
is necessary to outline the thesis Dignaga criticizes, first in the
AP(V) and then again in PS(V) 1.14ff., for this is the starting point
of the discussion. It is well known that in the AP, Dignaga presents
the theory that expresses the two conditions a thing must fulfill to be
an object of cognition, i.e., (1) it must be a real existence so that it
can produce the cognition with its own form and thus be a real cause
of cognition; (2) it must have a gross form so that the cognition can
bear its appearance.
l
? There are three different alternatives to inter-
universal as its object-support, since the five kinds of awareness take the particular as their
object-support?' Answer: There are two kinds of the particular, one is the particular
characteristic of substance the other is the particular characteristic of
the cognition sphere In respect to the particular characteristic of
substance, the five kinds of awareness take also the universal as their object-support, in
respect to the particular characteristic of the cognition sphere, the five kinds of awareness
take only the particular characteristic [of the cognition sphere] as their object-support.
Therefore, there is no contradiction,"
16 Cf, Hattori 1968: 90, n.1.41.
17 Cf. APV 176,12-13: yuI zhes bya ba ni shes pas rang gi ngo bo nges par 'dzinpa yin
te de'i roam par skye ba'i phyir ro II "[We call something] object-field because
ONDIGNAGA'S THEORY OF THE OBJECT OF COGNITION 217
pret the object-support (iilambana), as stated both in A.P(V) and then
again in PS(V) 1.14ff., but none of them can fulfill both of these
conditions. The first alternative can fulfill only the first condition; it
holds that the object-support is the things as they are, i.e., the atoms
of color, etc., which become the cause of the cognition, even though
the cognition bears a different appearance than what they really
are. IS And the other two alternatives can fulfill only the second
condition; both of them maintain that the object-support is that with
the appearance of which the cognition arises with respect to the
color, etc.
19
According to the materials available to us, each of these
alternatives has been advocated by certain Indian philosophical
schools at various points in time. The first one is mentioned in the
VS as being held by the V This alternative does not need
to be discussed here, because the school advocating it is not the
school under examination. The second and third alternatives agree
on one point - a cognition bears the appearance of the object-sup-
port, and this appearance is produced by many atoms; they disagree
only on the nature of the appearance, i.e., whether it is of an
agglomerate of atoms as a whole, or of many individual atoms that
are gathered together. These two alternatives are attributed by the
commentaries to the Sautrantika
21
and the
vada respectively,22 and thus I will refer to them in this paper as the
"Sautrantika-theory" and the "Sarvastivada-theory." These two theo-
ries are described in the PSV as follows:
the cognition, grasping it as its proper form (svanlpa), arises with the aspect (akara) of
this [object-field]."
18 Cf. PSV 1.6,8-9: yathavidyamana anyabhasasyapi vijiianasya kara1}a1]l bhavanti.
19 Cf. PSV 1.6,7-8: yadabhasa1!! teeujiianam utpadyate, tatha ta alambanam ...
20 Cf. VS 6,27-28: yat tad nlpadikam ayatana1!! nlptidivijl1aptfnal!! pratyeka1]l vieaya(1
syat tad eka1]l va sytid yathavayavintpa1!! kalpyate vaisesikaih. However, according to Kui
ji, the Chinese commentator of Xuan zang's Vijiiaptimatratasiddhi, this is advocated by
SfupmitTya. Cf. CWSLSJ 269c16: Jlt<:piil-liltiE:A:ll1l.
21 Cf. TrBh 16,20-21: sa11citalambanas ca paiicavijiianakayas tadakaratvtit. "And the
five group of awareness have the agglomerate [of atoms] as their object-support, for
[awareness] has the appearance of that [agglomerate]." Cf. Kui ji's CWSLSJ 270a14: J1UJ1ii
*tfBgijiH ... Cf. also de la Vallee Poussin 1928: 43.
22 See n. 29, 31.
218 JUNJIECHU
[Sautrantika-theory:] If cognition arises from those [color, etc.,] in the form
as [they] appear, then the object-support exists only conventionally, since in
this case the five groups of awareness take an agglomerate (sancita) [of
atoms] as their object-support.
23
[Sarvastivada-theory:] To be sure (kiimam), in the case of the awarenesses
with appearances of something blue, etc., the awareness produced from that
object [i.e. the atoms of something .blue] must be the perception. That is to
say, in these [awarenesses with appearances of something blue, etc.], even
though the gathering (samudiiya) of those [atoms] is nominally 'existent, the
aspect (iikiira) of what substantially exists [i.e., each atom] is found.
24
II-5. Jinendrabuddhi does not offer a detailed explanation concerning
the Sautrantika-theory, but just paraphrases the word sancita used by
Dignaga with samuha, and adds that if we have in mind that the term
"things agglomerated" (sancita) has the same meaning as the term
"agglomerate of the things" (sancaya); and thus sancaya means
samuha.
25
However, interestingly, he attributes the Sarvastivada-
theory to "the advocators of the theory of objects having the aspect
of many [substances/atoms]" (anekiikariirthaviidin),26 and says that
they hold the following opinion:
Surely, the atoms which are definitely substantially existent have such an
aspect (iikiira). Precisely those atoms, inasmuch as they support one another
23 PSV 1.6,10-11: yadi yathabhasafT! teeu jiiiinam utpadyate, tatha saFicitiilambanatviit
panciiniil]'l vijniinakiiyiiniiIJ1 SafT!Vftisad eviilambanam iti.
24 PSV 1.6,12-13: kiimafT! nfladyiibhaseeu viji'iiineeu tato 'rthad utpannafT! vijliiilwfT!
pratyakeafT! syiit. tatha hi teeu tatsamudiiye prajFiaptisaty api dravyasadiikiiro labhyate.
25 PST 91,13-14: saiicitiilambanatviid iti sanu7hiilambanatvat salicital]'l sancaya iti
krtvii. salicayas ca samiihalJ.
26 The Tibetan translation of this term, mam pa du ma'i don du smra ba (PSTT D
42a3), "the advocator of the theory of object with multiple aspects," seems incorrect to
me. The relation between aneka and iikiira should be a not a Karmadharya, and
thus aneka should refer to "many atoms." Cf. dravyasatiim eva sa paramiilJiiniim iikiiralJ
quoted in n. 28.
(paraspartinugrhfta),27 appear [to cognition] in this way [i.e., with this
aspect]?8
Kui ji, the commentator of Xuan zang, identifies the advocator of
this opinion as Sanghabhadra who is called "Neo-Sarvasti-
vadin."29 However, as Kat6 has already pointed out,3D VinItadeva
attributes this opinion to "Vagbhata and others" (phakhol la sogs
pa)Y
27 But in PST 44,13, where Dignaga's own position is explained, anyonyasannidhiinii-
_ vasthii is used, cf. n. 13.
28 PST 92,2-5: nanu dravyasatiim eva sa paramiilJiiniim iikiiraf:!. ta eva hi paraspariinu-
grhftiis tathii pratibhiisanta ity anekiikiiriirthaviidinah ... (cf. VinItadeva's description of
this opinion quoted in n. 31.) This idea is reported and criticized also in Sthiramati's TrBh
and Xuan zang's CWSL. Cf. TrBh 16,26-27: ekaikaparamiilJur atfndriya/:z,
bahavas tu paraspariipeqii indriyagriihyii/:z. The same idea is also found in CWSL 4b: ;ff
fJfffl:i:'
1lo The word paraspariipeqii (cf. phan tshun nye bar 'gro ba in PVP: 221a4,
ad PV 3.196) should be understood as interchangeable with paraspariinugrhfta (=phan
tshlln phan btags pa) in the PST.
29 Cf. CWSLSJ 271al0: Cf. Cox's remark: "Though
Sailghabhadra's interpretations are generally consistent with the
position, especically as presented in the he is influenced by Vasubandhu's
criticisms and attempts to reformulate many Sarviistiviida-V positions in re-
sponse to them ... Sailghabhadra's works mark a turning point in the development of
thought; as a result, Sailghabhadra is acknowledged as the
inaugurator of the so-called period." (Cox 1995: 58) Cf. de la Vallee
Poussin 1928: 45, n.1.
30 Cf. KatO 1973: 134, Kata 1989: 179.
31 Cf. APT .189a7-b2): phyi rol gyi don du smra ba'i phyogs 'di la yang / pha kholla
sogs pa kha Gig / rdul phra rab mams 'dus pa 'i mam pa dag dbang po 'i mam par shes pa 'i
rgyllr 'dod do II de dag 'di skad du rdul phra rab mams la ni 'dus pa'i mam pa yang yod
do II rdul phra ba mams la gang cung zad cung zad du ci yod pa thams cad ni rdzas su yod
pa yin no II rdzas su yod pa'i phyir 'dus pa'i mam pa mam par shes pa'i rgyu nyid du
'gyur TO II rags pa'i phyir na mam par shes pa la rang gi ngo bo yang 'jog par 'gyur te / de
bas na rdul phra rab mams tshlll gzhan gyi yul nyid yin no zhes zer TO II Also among the
adherents of the theory of [the existence ot] the external object-reference (biihyiinhaviida),
some [masters], Viigbhata and others, hold that the gathered aspects of atoms are the cause
of the sensory awareness. They say: "Atoms have also the gathered aspect. Everything,
what ever exists among atoms as single particle, is substantially existent. Because of being
substantially existent the aspect of gathered [atoms] is the cause of the awareness, because
of the gross form (sthiila) it makes the awareness obtain its proper form (svariipa'. There-
fore, the atoms are the object-field [of the cognition] in a different way [than what they
exist]."
220 JUNJIECHU
II-6. From the report given in Sanghabhadra's Nyayanusara (NA) ,
we know that there was a sharp controversy on the topic of the
object of cognition between the (hereafter
only "Sautrantika") and the (hereafter only
"Sarvastivada"). According to Sanghabhadra, the Sautrantika's opin-
ion, advocated by "the venerable" (sthavira) - a title used to refer to
SriJata
32
- is as follows:
In this regard the venerable (sthavira, i.e. SrIlata) holds the foll.owing opin-
ion: Neither the basis [asraya, i.e. the sense faculty] nor the object-support
(alambana) of the five groups of [sensory] awarenesses are real existence,
because each individual atom cannot be the basis and object-support [of
those awarenesses], and because only an agglomerate CfOil) of many atoms
can constitute a basis and object-support. ... They, the master and the
followers, also .used worldly examples incorrectly; they use the simile of
blind men to prove their doctrine. It is said: Like blind men, who if separate
and alone, have no action of perceiving a visible form; when they gather
together they still have no action of perceiving. Atoms are the same:
when they are separate and alone, they do not perform the function of basis
and object-support; when many [atoms] are gathered together, they still do
not perform such a function. Therefore, the cognition spheres (ayatana, i.e.,
the sense faculties and their objects) are unreal; only the constituent ele-
ments (dhatu) are real.
33
This is to say, the Sautrantika holds that neither sense faculties nor
object-support really exist, because as an agglomerate they can be
reduced to atoms. Actually in AKBh, in reporting a debate on what
is the agent and what is the object of cognition, Vasubandhu ascribes
to the Sautrantikas the opinion that this kind of debate is useless, for
the visual awareness arises in dependence on the visual sense and the
visual matters, there is nothing that sees or is seen, there is merely
32 The sthavira mentioned in the NA is identified in the later sources as SrIlata, a
Sautrantika master (cf. Cox 1988: 71, n. 9.). Kui ji, for instance, idendifies him as SrIlata,
the second of "the three Sautrantika masters," who composed the Sautrantika- Cf.
CWSLSJ 358a,1l-12:
33 Cf. NA 350c5-17: XlA1%9:fQifo
j]J:i.X;FJTi&FM.w*il"5co ...... iJJi;o
{to llW.R;Jjio JtJtliJl'llWo
iJl'llWo Cf. Dhammajoti 2004: 89, also Dhammajoti 2004a: 20.
the dharma that is devoid of any activity, and there is merely cause
and effect. 34
Sanghabhadra himself, supporting Sarvastivada, holds the follow-
ing opinion:
The five groups of awareness do not take non-reality as their object-support,
because they take the gathered atoms as their object-support.
And because the five groups of awareness (*paiicavijiiiinakiiya) are free of
conceptual construction, they do not take an agglomerate of many atoms (m
1)\'x11J-6') as their object. There is no other kind of dhamJa that is called
"agglomerate" and that can be an object seen by, [heard by, smelled by,
tasted by,] up to touched by an [awareness] that is free of conceptual
construction (*nirvikalpaka), because that agglomerate is nothing other than
that which is grasped by the discriminating imagination (*abhinin7pa1}avi-
kalpa). But the five groups of awareness do not have the function of
discriminating, [and] therefore they do not take an agglomerate as their
object. That is to say, the atoms, inasmuch as they are gathered together and
structured [in a certain way] (*sannivi.y!a), are always the basis and object-
supports for the arising of the five groups of awareness, because there is no
atom that is not gathered; even if atoms were not gathered, they would still
be the basis and the object-support because they have the same nature.
However, because the five groups of awareness exclusively (*eva) take
gathered [atoms] as their object-support, they do not arise in taking that
[agglomerate] as an object-support.
35
Here, the Sautrantika-theory is criticized: an agglomerate of atoms
cannot be an object-support because it is grasped by conceptual con-
struction, but the five groups of sensory awareness are free of con-
ceptual construction. We shall see below that Dignaga's opponents
also criticize him in this way. For Sanghabadra, atoms are always a
real existent. The difference between the state of being gathered and
34 AKBh 31,12-14: atra sautrantika ahu!:z -kim idam akasa!!l khiJdyate. hi pra-
tftya riipa(zi cotpadyate latra katz pasyati, ko va drsyate. nirvyaparaJ!!
hfdaJ!l dharmamatra1J! hetuphaiamatra1J! ca. Cf. Dhammajoti 2004: 51-52.
35 NA 350c19-27: l'
oJ-F1p-;,4',jJlj I
nttlHt!t:5j-jJljmJ&o t:
f't;]i
mgyo
222 JUNJIECHU
not gathered is that the former is invisible, but the latter is visible;
the nature of being a real existent of the two states is the same.
36
II-7. The passages of NA quoted above suggest that the two words,
he-he (fOif) and he-ji used in this context separately for the
Sautrantika-theory and Sarvastivada-theory, were two different
words in the original Sanskrit text, and therefore, some modern
scholars hold that he-he and he-ji refer to two different theories
being advocated by the Sautrantika and the Sarvastivada, and also
suggest that their Sanskrit origins were different.
37
Of course, he-he and he-ji used in Sanghabhadra's context clearly
refer to two different things. According to my understanding, he-he
refers to a state in which many atoms are integrated into a single
unity with no space between them, and he-ji, a state in which many
atoms are merely gathered, but with spaces in-between. This kind of
difference, as implied by the two words used for the two theories, is
also mentioned by Vinltadeva in his commentary on the VS:
[The gross fonn such as a color, etc.,] as many atoms standing with inter-
space (*siintara), cannot be an object-field; it cannot be an object-field also
as atoms standing without inter-space (*nirantara).38
Kui ji interprets he-ji as follows:
36 NA533a10-11: ;ttR&t'ffo *!UiIt>f'i1o 1iI!?FMiSo
37 Cf. the discussion of these two words in Kato 1973: 137, Kato 1989: 180. According
,; to him, the original forms of he-ji used by Sarvastiviida and he-he used by Sautriintika
could be respectively *saficita and *sal'[lhata, the latter is used in VS for the Sautriintika-
theory; and it is also possible, KatO suggests again, based on the statemente in Abhidhar-
madlpa, that the venerable Vasumitra (sthaviravasurnitra), one of the four great masters of
the Sarvastivada, refutes the twenty-five true principles (tattva) of SiiiJkhya and destroys
the Sautriintika-theory of the agglomerate of atoms, mentioned as paramiinusaficayaviida
(AD 260,14-15: tad ebhyas caturbhyaT:z sarviistiviidebhyas trtfyaT:z sthaviravasurnitraT:z
paficavil'[lsatitattvaniriisf pararnlilJusaficayaviidonrniirthf ca ... ), that Sautriintika's he-he is
the renderings for saficaya. Cf. also Cox 1988: 74, n. 23. However, as we have seen above
(n. 23 and n. 24), in Dignaga's PSV, saficita is used for the Sautriintika-theory, and
sarnudiiya for Sarvastivada-theory; they are translated in Tibetan respectively as bsag pa
and tshogsltshogs pa (see Hattori 1968: 188,16b4-5, 189,97a6-7.) .
38 VST 219al-2 : rdul phra rab phrag can du gnas du rna yang yul rna yin no II du phra
rab dag bar rned par gnas pa yang yul rna yin te I
In the proximity of the same place is called he [*sal!!-], not integrated into a
single entity is calledji [*-udaya]; this is because they [i.e. atoms] are put in
proximity, but they are different entities.
39
ll-8. Nevertheless, I do not think that these two words under discus-
sion were necessarily different words in the Sanskrit original. Or if
they were indeed different words, then they must have had the same
meaning. The reason is as follows. First of all, Jinendrabudhi regards
all of the words in this semantic group to be synonyms.
40
And if we
analyze the compounds used by Sailghabhadra, zhong-wei he-he em
and he-ji ji-wei we find that the syntactic
relationship between the two elements of each compound is differ-
ent. This difference is expressed through the different word orders in
which these two elements are arranged. In the case of zhong-wei he-
he, we arrive at something like paramiilJusaficita, -smighiita, or -safi-
caya, which all appear in the AKBh,41 or paramiilJLlniiJ?1 sanghiitaJ:t,
which appear in V.s 7,11; thus the syntactic relationship between the
two elements is that paramiilJu depends on saficita or its variants
which as a noun is the modified element, and the meaning therefore
is "an agglomerate of many atoms." In the case of he-ji ji-wei, we
arrive at something like saJ?1hatiiJ:t paramiilJavaJ:t, which appears in
V.s 6,29, or paramiilJuJ:t sanghiitaJ:t in V.s 8,6, or the plural form
paramiilJava eva saiicitiiJ:t, which appears in TrBh 16,23-24; in this
case the syntactic relationship is that saJ?1hata or sanghiita as an
attribute modifies paramiilJu, and thus the meaning becomes "gath-
ered atoms." The difference between these two theories would thus
not necessarily be demonstrated through using different words, but
rather through the different syntactical relationships between the two
elements, i.e., through the different emphasis, whether on the
agglomerate, as in the case of the Sautrantika-theory or on the atoms
themselves, as in the case ofthe Sarvastivada-theory.
39 Cf. SWSLSJ 217a18-19: llP;&tJlIo
40 Cf. PST 44,1-2: saiicitasabdena samudiiya ucyate, bhiive saiiciti/:l
saiicitaf!! sal1caya(l samudiiya iti hi paryiiyii/:l.
41 See Hirakawa 1973: 225. Cf. also n. 14.
224 JUNJIECHU
Xuan zang also uses the two words, he-he !md he-ji, in his
translations of other texts in which only one word is used in the
original. For example,. in the Tibetan translation of the M(V), the
one term 'dus pa is used for both theories; however, Xuan zang
translates this single word with either he-he or he-ji. Also in the San-
skrit text of V.s 11 with its Vrtti, only one word, saf(lhatti, is used,
which is translated in Tibetan as 'dus pa;42 however, again Xuan
zang translates it with different words: "he-he, etc." in the
verse, and "he-he and/or he-ji" (flJil
in the Vrtti.43 Of course, the short form "he-he, etc." is the same as
"he-he and/or he-ji," it being abbreviated for metrical reasons. In this
case, Xuan zang regards saJ!lhata as referring to a kind of collection
that comprises the meanings of both he-he and he-ji. In short, Xuan
zang's translations of these two terms are of an explanatory nature.
II-9. Indeed, this kind of difference, i.e., an agglomerate of atoms or
gathered atoms, is significant only for the Sautrantika and the
Sarvastivada, as Bahyarthavada; both hold that the object-support is
the external existent. This difference is not significant for the Yoga-
cara, the Antarjfieyavada. For the latter, the object-support, regard-
less of whether it is "an agglomerate of atoms" or "gathered atoms,"
does not exist separately from an awareness; its external form is
merely an image or aspect (pratibhiisaltiktira) produced by the
awareness. The main purpose of PS(V) 1. 14ff. is to refute the
Bahyarthavada, the theory that the object of cognition exists
externally, and therefore the words used there, according to
. Jinendrabuddhi, are synonymous. Jinendrabuddhi tells us:
In this regard, the disputants are divided into two groups: Those who advo-
cate [that the object-support is] an internal cognizable (antarjiieyavZidin) and
those who advocate [that the object-support is] an external object (bahyZir-
thavZidin). Of these [two groups], for those who advocate [that the object-
support is] the internal cognizable, in the state that the reality is not per-
ceived, the [distinction between] the valid cognition and the cognizable
object, is only an imperfect [view] according to reality. Only people who are
42 Cf. VS
T
7b3-4.
43 Cf. vs
c
75c17-25. Cf. Kala 1973: 137, n. 28, Kala 1989: 179-180.
confused have this view of distinction between a valid cognition and a
cognizable object, based on their [false] vision. This has already been
explained.
44
Those who advocate [that an object is] an external object assert
that there are external cognizable objects absolutely in the sense of ultimate
reality and the valid cognition which takes that [object] as its In
order to explain how this [view] is incorrect, now he [Dignaga] says the
following, in order to undertake an examination with reference to the object-
support: "However, with reference to color-form, etc. the meaning of the
object-support should be explained." (PSV ad PS 1.14cd
45
).46
At the beginning of this section Dignaga states his hesitation in
accepting the popular saying that the Va.davidhi is a work of Vasu-
bandhu, holding that this is not to be accepted, or, if it is accepted, an
affirmation must has been made by Vasubandhu himself that this
work lacks quintessence.
47
On the last point of this statement Jinen-
drabuddhi offers a remark, saying that even if it is accepted that this
work is composed by him, it must be composed in his earlier life
when his insight is not perfect; later, being purified in understanding,
he made the affirmation of the lack of the quintessence in this
work.
48
That is to say, if the Va.davidhi is composed by Vasubandhu,
it must be composed by him when he was still a Bahyarthavadin;
later, when he became an Antarjiieyavadin, he changed his opinion.
Thus, the central topic in the section beginning with PS 1.14 is a
refutation of the statement found in the Va.davidhi, which is formu-
lated according to Bahyarthavada doctrine, that perception is an
awareness arising directly from a certain object-reference (tato
'rthiid vijiia.naJ!l Jinendrabuddhi emphasizes that the
44 Cf. PST 73, 10-74, l.
45 Cf. Hattori 33 D.
46 PST 90,6-11: iha dvaye viidino 'ntarjiievaviidino biihviirthaviidinas ca. tatriintarjiie-
vaviidiniil1l adr{atattviivasthiiyii1"(l pra111eya1!1 caparinipanna111 eva tattvatalJ.
kevalaJ?l bhriintiiniil!l yathiidarsana111 idaJ?l 111iinal1leyavyavasthitidarsana111. etae ea prii,?
eva pratipiidita111. biihviirthaviidinas tu para111iirthata eva biihYaJ!l pra11leya1!1 tadviayaJ?l
ca praI1liiIJaJlI icehanti. tae ea yatha na yujyate, tathii pratipiidayitul1l idiinfin iila11lbaniidhi-
viciiram iirabdhu11l iiha - rupiidieu tv iilambaniirtho vaktavya iti.
47PS I.13ab: na viidavidhir iieiiryasyiisiiro veti nikayalJ /
48 PST 86,10-12: yady api ca tena e.Q viracita(J, tathiipi prathamal1l anupajiitaprajiiiiti-
sayena satii. pasciid vyavadiitabuddher asva tatriisiiraniseayo jiita iti.
226 JUNJIECHU
purpose of this section is not merely to refute the t!leory of the Wida-
vidhi, but to demonstrate in general that the division of cognition and
its object, based on the assumption that the object exists externally,
is incorrect. He says again that in this regard, among those who
advocate that objects exist externally, the Buddhists Csvayuthyiif:t, i.e.
the Hlnayanists) are very strong, and when they are defeated, the
others will also be defeated, just as when the chief wrestler is
defeated, the other wrestlers are also defeated; therefore the only
opponent in his discussion is the Buddhist who advocates that
objects of cognition exist externally.49 Here svayuthyiif:t apparently
refers to the Sarvastivada and the Sautrantika. Actually, the
refutation of the object-support was carried out in different steps:
First, the theory of the object-support as advocated by the non-
Buddhists, the who hold that the externally existent
atoms, even if not in a gross form, are the object-support, is refuted
from the standpoint of the Buddhist, the Sarvastivada, who accepts
the existence of atoms, but asserts that the gathered atoms in their
gross form is the object-support. Then, the theory of the object-
support as advocated by the Sarvastivada is refuted from the
standpoint of the Sautrantika, as SrTIata had done, who accepts the
existence of atoms, but asserts that the agglomerate of atoms in its
gross form is the object-support. And finally, all the theories of the
object-support advocated by the Sarvastivada and the Sautrantika are
refuted from the standpoint of the Yogacara, who does not accept
any external thing as being the object-support, as Vasubandhu and
Digana had done, for instance, in the VS and the AP.50
II-lO. According to Jinendrabuddhi in his conclusion to this section,
Dignaga's answer to the question about the meaning of the object-
support Ciilambaniirtha) asked at the beginning of the section was
already given in PS 1.5cd, namely: "The visible matter that is self-
49 PST 91,2-6: tasmtlt samanyenaiva bahyarthtisrayil}l pramapadivyavastha na ghatata
iti pratipadanaparo 'yam lirambha iti veditavyam. tatrapi bahyarthavadisu balinal;z sva-
y,tthyah. itare nihata eva bhavanti, iva nihate tadanye maUa iti
tair eva saha viCara/!I karoti.
50 Cf. also the analysis in Mimaki 1972: 88.
:Cognizable and inexpressible is the operation-sphere of the
Jaculty.,,51 This is also the same as that which is stated in AP 6,
where, employing the Yogaciira's idea of the "image-part" of cogni-
tion,52 he states that only the internal cognizable form (antarjiieya-
rupa) that appears as if externally is the object of cognition, for the
feason that it has the form of this cognition and it is its cause.
53
Thus, in following Jinendrabuddhi, it is quite clear that Dignaga
is an Antarjfieyaviidin, advocating the doctrine that accepts only the
mternal object, without, however, committing himself to the exis-.
tence of external things, holding only that external things are not the
object of cognition. As we shall see below Jinendrabuddhi mentions
dgain this distinction between the Antarjfieyavada and the Biihyiir-
thavada in another context. 54 This special term, antarjiieyaviida, is
thus used by Jinendrabuddhi only in the epistemological sense, and
therefore is a proper designation of the epistemological theories
advocated by the special group of the Y ogiiciiras which I shall dis-
cuss in more detail in the following pages. 55
11-11. To sum up, Dignaga holds perception to take siimiinya as its
object-field, because perception is produced by many atoms. This
51 PST 90,12-91,4: hi manyante - eviiyam iirambha iti, tesiim
yad uktam - svasll1!'vedyll1!' tv anirdeyll1!' riipam indriyagocara ity atra gacariirtha vaktav-
ya/:t. kiTp yadiibhiisaTp tatra jiiiinam utpadyate tatoo tad gacara ityiidi tulyal;t paryanuyaga
. iti pratividheyam.
52 Cf. MS 2.11 (p. 29,8-12): de Itar mam par rig pa 'di dag thams cad ni don med pas
de tsam yin la I mig la sags pa'i mam par rig pa dag ni gzugs la sags pa rgyu mtshan yin
pa dang I de dag gi mam par shes pa Ita ba yin pa nas Ius kyi mam par shes pa Ita ba yin
pa'i bar gyis rgyu mtshan dang Ita ba dang bcas pa'a II "Thus, all these representations are
merely themselves (tanmiitra), for they lack objects. The visual representation, etc.
(calquriidivijiiapti), has visible matter, etc., as its image[-part] (nimitta[-bhagaJ), and the
awareness of these [visible matter, etc.] as the perceiving[-part] (dar.fana[-bhagaD, up to
having [the tangible object as its image[ -part] and the awareness of the tactile repre-
sentation as its perceiving[-part]. Thus it has the image[-part] and the perceiving[-part]."
53 AP 6 (quoted in TSP 582,11-12): yad antarjiieyarupaTp tu bahirvad avabhasate I
so 'rtha jiiiinarupatviit tatpratyayatayiipi ca II
54 Cf. n. 104.
55 Cf. ill-3 and V-3.
228 JUNJIECHU
siimiinya, I repeat,56 is interpreted by Iinendrabuddhi as "common
property," i.e. all atoms have the capacity of the producing cogni-
tion, that is to say, all atoms are active in generating the cognition,
they are the cause of the cognition. This seems to be similar to the
Sarvastivada-theory. Manorathanandin on the other hand interprets
this siimiinya as "an agglomerate, or a kind of collection of atoms,"
i.e., all atoms agglomerated as awho1e. This is probably the Sautran-
tika-theory. To be sure, if this siimiinya is understood in. the sense of
an external object, it is similar to the theory rejected by Dignaga
himself, and thus, of course, Dingaga is involved in a contradiction.
However, if we understand Dignaga as making this statement from
the viewpoint of Antarjfieyavada, the siimiinya made up of many
atoms no matter what it may be, is thus a reflection of awareness as
external many individual atoms or many atoms as a whole, which
becomes the cognition sphere (iiyatana) that is unreal. In this case he
is not involved in a contradiction and his theory is consistent.
III. iiyatana
111-1. In the AKBh, ayatana is "etymologically" explained as having
the meaning of the iiyadviira, "the gate of approaching [the mind and
mind-associates]," and dhiitu as having the meaning of gatra, "ge-
nus," which is further glossed as iikara, "origin" or "sources.,,57
According'to Sanghabhadra, the Sautrantika holds iiyanata to not
really exist; only the dhiitu is real existence. Sanghabhadra himself
on the other hand, as an advocator of Sarvastivada, has the opposite
opinion, and in his NA criticizes the former at length. The Sautran-
tika's opinion that the iiyatana is not real is derived from its position
56 Cf. n. 11 and n. 14.
57 Cf. AKBh 13,19-21: yathaikasmin parvate bahiiny
dhatava ucyante. evam ekasminn asraye santane va dhatava uc-
yante. akaras tatra ucyante. "Just like in a mountain there are many genera [of
elements] such as iron, copper, silver and gold, which are called elements, in the same
way, in a single basis [of personal existence, i.e. the body, = atmabhava] or a [single]
continuum, there are eighteen genera [of elements], which are called the eighteen [psycho-
physical] constituent elements (dhatu), here the origins (akara = 'byung gnas = 1=:;$:) are
called genera." Cf. also Dhammajoti 2004a: 20.
.. that the basis and the object-support of the five groups of awareness
not real. As we have seen above, the argument that only the
agglomerate (he-he) of many atoms can constitute the basis and ob-
ject-support is concluded by stating that "the iiyatana is unreal, only
(the dhiitu is Here the iiyatana, the cognition sphere, includes
the five sense faculties that are the basis (iisraya) of cognition and
,the five object-fields that are the object-support (iilambana) of cogni-
tion. Actually, the already held the opinion that cogni-
. tion can take the non-existent as its object-support,59 just like cogni-
tions that take magic creations, Gandarva-city, fire-brand, mirages,
etc. as their object-
. support. The in contrast, holds that all cognitions take
only existent object-references as their object-support.
6o
Following
the tradition, the Sautrantika maintains that even when a
cognition arises from the non-existent, the two [requisite conditions,
i.e., iilambana and iisraya] are still determined (*niscita) by the
reasoning through which the causality [between cognition and its
conditions] is mutually inferred.
61
The cognition sphere is unreal
because the agglomerate of atoms that constitutes this cognition
sphere is only conventional reality, since when the agglomerate is
analyzed into its pieces, according to the Sautrantika's theory of two
realities, there is no cognition of this agglomerate.
62
111-2. The idea that the cognition sphere is not real is repeatedly criti-
cized by Sanghabhadra in his NA. He says that if the Sthavira main-
tains that the iiyatana is unreal and the dhiitu is real, his statement
violates the utterances in the Siitras. He quotes some passages from
the Siitras, one of which is also quoted in the AKBh and AKV, to the
58 Cf. n. 33.
59 Cf. Cox 1988: 43ff.
60 MVBh 228b21-24:

61 NA 628c4-5: dF=J9:t:lEa
62 Cf. AICBh 6.4: yatra bhinne na tadbuddhir anyilpohe dhiyil ca tat / ghatilmbuvat
sQ/p.vrtisat paramiirthasad anyathii II Cf. also AKBh (ad lac.) 334,1-7. This theory is
reported by SaiJ.ghabhadra in his ASP 666a9-27.
230 JUNJIECHU
effect that the phrase "all things" mentioned by -the Blessed One
(bhagavat) refers to the twelve ayatanas.
63
Moreover, he says, the
Blessed One would not say something is ultimately real if it were
unreal, and also, one would not accomplish perfect Buddhahood
(*samyaksaY(lbuddha) if one were only aware of unreal things. Only
the nihilist would make this kind of statement, not
those who respect the Buddha as their teacher.
64
His conclusion is as
follows:
Therefore, all twelve iiyatanas are real existents. One cannot talk about the
ultimate reality with respect to unreal things. Thus, in the statements of
Sthavira many contradictions are found between that which was said earlier
and that which was said later (*pLlrvottaravacanavyiighiita). Those who have
faith without insight respect him; however, those who have both faith and
insight would certainly not follow him. And the simile of the many blind
persons
65
is also contradictory to his doctrine. The theory that each atom
cannot possess the nature of being the basis and object-support, and the
agglomerate of atoms can be the basis and object-support does
not conform to the simile of blind persons at all. The theory that the gathered
atoms (fQ:*t!lli*J) are the basis and object-support is not contradictory to the
simile of blind persons, because it holds that each atom is the basis and
object-support. If it is held [by you Sautrantika] that each atom is invisible,
then the agglomerate of atoms would also be invisible, because it is just like
what is said in the simile of the blind persons - it is the same as in the state
of not being agglomerated (fQ1f: e1f?). Therefore, the five groups of sen-
sory awareness definitely do not take the agglomerate as their object. How-
ever, there must be an object, and thus, it is established that they take real
things as their object. 66
63 Cf. AKBh 301,8: sarvarrz astfti briihmaIJa yiivad eva dviidasiiyataniini. AKV 5,28-
29: sarvarrz sarvam iti briihmaIJa yiivad eva dviidasiiyataniinfti satre vacaniit.
64 NA 352a5-12:
'ft:!'l:-tDo
tD11fo
7FH!lilUto
65 Cf. n. 33.
66 NA 352a12-21: M
Wmffilo 0 FiiJmW)c*o Ao i!tl1Hl *0
Jl:t
The key-point of the difference between the Sautrantika and Sarvas-
tivada is the fo.llowing: For the Sautrantika the agglomerate consist-
ing of many atoms is the basis and object-support, i.e., the iiyatana,
but it is not the real existent, because it can be analyzed into the
individual atoms, whereas for the Sarvastivada, each atom is the
basis and object-support, i.e., the iiyatana, and they are the real exis-
tent, and furthermore, when many atoms are gathered together they
are visible.
1II-3. The Sautrantika's idea that the cognition spheres are unreal
conforms to Dignaga's consistent thought quite well. When discuss-
ing the sources of our knowledge, he prefers to use negative
formulations. Formulas such as "P is not/is free of Q," and "Qj, Q2>
etc., are not P," are used in his definition of perception in PS 1.3cd
kalpaniiporjharrz niimajiityiidiyojanii /I). Also in his ear-
lier work, the *HastaviilaprakaralJa,67 basing himself on the famous
snake-fOpe-hemp simile in MS 3.8,68 he states that all phenomena
which are examined according to their own nature (*svan7pa) are the
operation-spheres of conventional cognition (kun rdzob shes pa'i
spyod yul, *sarrzvrtijiiiinagocara); even the partless infinitesimal
particle, the atom, does not really exist, for such particles are not
able to be the object-support owing to their ungraspable nature, just
like sky-flowers, hare-horns, etc.
69
That which cannot be perceived
cannot be held as an existent object of cognition; since the object
does not really exist, a cognition of this object must also be errone-
fY'ilfIlj,{f!!N!l)g;o Yf.1lf 0 Ali:il'&:o.,g.iJ'J'J!lIFF Yf.o !liJliill.
;lij:o
67 Cf. the critical edition of the Tibetan translation in Frauwallner 1959: 153-156; a
Japanese translation and Sanskrit reconstruction are found in Harada 1993. There is also
an early edition and English translation, see Thomas & Ui 1918.
68 Cf. MS 3.8. (p. 52,12-16): mun khung na sbrul du snang ba'i thag pa bzhin du 'jug
ste I dper na med pa'i phyir thag pa la sbrul ni nor ba ste I de'i don rtogs pa rnams ni med
pa la sbrul gyi blo ldog eing I thag pa'i blor gnas so II de yang rnam par phra mar bya na
nor ba ste I kha dog dang I dri dang ro dang reg bya'i mtsan nyid yin pa'i phyir ro II Cf.
Nagao 1991: 67ff.
69 Cf. HVP karika 3ab (154,13-14): eha med brtag par bya min phyir I tha rna yang ni
med par mtshung I and the prose that follows.
232 JUNJIECHU
ous. Also in another earlier work, the *Upadayaprajfiaptiprakara-
lJa,70 he denies the existence of an object of cognition in the same
way. There, in opposition to the extremist ideas of sameness (-'fi,
*ekatva), separateness *prthaktva) and non-existence (?-I=ff,
*asattva) , he explains that the Buddha designates things merely
according to conditioned labels *upadayaprajfiapti).
Thus, phenomena exist nominally: they appear either as a conglo-
merate *smighata, or similar expression), i.e., as th,e conceptu-
ally constructed unity of spatial manifoldness; or as a continuum (.t
t', *santana), i.e., as the conceptually constructed unity of temporal
manifoldness; or again as a special condition *avastha-
either as a particular component in the conglomerate or a
particular stage of the continuum. The conclusion is that phenomena
in all three forms exist merely nominally and thus do not really exist.
Such concepts are found in the Bodhisattvabhitmi of the Yogiiciira-
bhiimi. As has been demonstrated by Schmithausen, in this chapter
we can find a nominalist philosophy according to which all things
ultimately are merely a "Setzung der Sprache," or, as is said at
another point in the same text, that all objective appearances have
their root in a conceptual construction (vikalpa). According to
Schmithausen, this kind of theory is a special form of Mahayanistic
illusionism and is a preliminary stage of Yogacara idealism.
71
I
believe that this kind of theory, as a special branch of the Y ogacara
system, is the basis of Dignaga's thought, and that he fully developed
it in his last work, the PS(V).72
1II-4. The link between Dignaga's concept of ayatana and that of
Vasubandhu is very clear. As mentioned above, the background of
70 Cf. Frauwallner 1959: 83-164 (on this treatise: 121); Ui 1958, 167-232; Kitagawa
1957: 126-137.
71 Cf. Schmithausen 1973: 166f.
72 Also with regard to another part of the iiyanata, "the basis," i.e. the sense faculties,
Dignaga accepts the basic idea of the Sautrantika's position. In the ApV he maintains:
"The sense faculty is inferred from its effect [i.e., from cognition] as something the nature
of which is power, not as [something material] that is constituted by elements (bhautika)"
(MV 179,5-6: dbang po ni rang gi 'bras bu las nuspa'i lio bo Hid du rjes su dpag gi
, byung ba las gyur pa Hid du ni ma yin no /!)
Dignaga's discussion on this problem is a passage found in the
AKBh.
73
According to Y asomitra' s commentary, "the particular of
the cognition sphere" refers to "the condition of being cognizable by
visual awareness, etc., or[, put in other words,] the condition of
being the cognition sphere such as visible matter, etc.;"74 Jinendra-
buddhi and Vibhiiti also say that this is "the condition of being
graspable by visual awareness" This
cognizable or graspable object is simply that which we perceive.
Thus, it is said that cognition arises in respect to the particular of this
cognizable or graspable object, not in respect to the particular of
substance, which, according to Yasomitra, refers to homogenous
substances, such as the color blue?6 In this case, a mental generaliza-
tion takes place and consequently the object of the cognition is no
longer the particular, but the universal. The conclusion is that a
cognition which arises with respect to the cognition's appearance,
namely, with respect to self-awareness, is perception; a cognition
that arises with respect to the thing itself, namely, with respect to the
atoms or substances, is a conceptual construction. Actually, Dignaga
also says that even a mental construction (kalpana) is perception
insofar as it perceives itself with respect to self-awareness, but it is
not perception with respect to the thing itself (artha), because in that
case it conceptualizes its object.77
III-5. Thus, for the the ayatana is the atoms, which are
ultimately real, but for both the Sautrantika and the Y ogacara this
ayatana, i.e. that which we perceive, is not real. The question arises:
if it is not real, where does it come from? As far as I can see, this
question is not answered by the Sautrantika, and this is the weak
73 Cf. n. 15.
74 Cf. AKV 30,15-16: iiyataniinii1J1 cakeurvijiiiina-
viji1eyatviidi ntpiiyatanatviidi vii.
75 Cf. PST 44,6 and PVV 176, nA (Vibhii).
76 Cf. AKV 30,16-17: dravyiilJiil?J nfliidikiiniil?1 svalakeww1J1
niidiviji1eyatvam, nfliikiiriidi vii.
77 Cf. PS(V) 1.3,12-13: yadi riigiidisvasa1J1vitti{l pratyakeam, kalpaniijl1iinam api nama.
satyam etat. kaZpaniipi svasa1J1vittiiv niirthe vikaZpaniit / (l,7ab)
234 JUNJIECHU
point of their theory, a point open to attack by -the or
other opponents. In this regard, the Yogacara goes further, offering a
clear answer, namely; that it is a mental product. Vasubandhu, the
Kosakara,78 says in the VS(V) that a cognition bearing the appear-
ance of visible matter (rapa) comes forth on account of its own seed,
and these two, i.e., the seed, or the visual sense, and the visible mat-
ter, are said to be the cognition's two kinds of cognition sphere.
79
In
other words, only those things whose appearance the cognition bears
while arising are the cognition spheres, i.e., the visible matter, etc.
This is also as Dignaga sees it. Based on the Yogacara's common
ground with the Sautrantika, the object of sensory cognition is
reduced to the "cognizable" or "graspable" of the visual awareness.
Thus, the object of cognition is dependent on the cognition itself, and
only an object of this nature is its cognition sphere. That is to say,
what a cognition cognizes is sense data, not an independent sub-
stance. This idea is clearly stated in PS I.Scd. There, Dignaga con-
cludes that visible matter that is self-cognizable (svasaY!lvedya) and
inexpressible is the operation-sphere (gocara) of the senses.
80
This
half-verse is repeated in PSV ad PS 1.4Iab, where Dignaga states:
The object-reference of the senses that, although possessing manifold
properties,Sl appears to the senses in its particular character, is the cause of
the arising of the cognition bearing its appearance. As a part of the cogni-
78 I follow Schmithausen 1967: 136 in holding Vasubandhu, the author of the AK(Bh),
is also the author of the VS and TriI)1s; this is accepted in Frauwallner 1994: 425. The
latter earlier proposed the theory that there were two persons bearing the name Vasuban-
dhu (Frauwallner 1951), Vasubandhu the younger who was the author of the AKCBh), and
Vasubandhu the elder, the brother of AsaiIga, the Mahayanist; Frauwallner did not, how-
ever make a decision about which of the two authored the VS and TriI)1s (cf. ibid. 56). Cf.
Franco 1997: 77, n. 27; Kritzer 2005: xxiv-xxvi.
79 VS(V) 9 (5,25-6,1): yataJ:! svab"ijiid vijiiaptir yadiibhiisii pravartate I dvividhiiyata-
natvena te tasyii munir abrav"it II (9) kim uktal)1 bhavati. n7papratibhasa vijiiaptir yatal;
svabfjat utpadyate, tac ca bfjal)1 yatpratibhasii ca sii te tasyii
vijiiapte yathakramal)1 bhagaviin abravft.
80 PS 1.5cd: svasal)1vedyam hy anirdeyal)1 n7pam indriyagocaral; II = NMukh 16cd: III
Cf. n. 51.
81 Cf. PS 1.5ab: dharmiIJo 'nekarrtpasya nendriyiit sarvatha gatil; I
tion,s2 it is exclusively individually self-cognizable; and owing to its nature
[as individually self-cognizable] it is inexpressible, since an expressible
thing is the object that consists in the universal. 83
The "self-cognizable," according to Jinendrabuddhi, refers to some-
thing that is not cognized through communication, i.e., is of an inter-
nal nature.
84
Here it is quite clear that, based upon Y ogacara' s com-
mon grounds with the Sautrantika, Dignaga says that the "particular"
refers to the cognition sphere that depends on sensory awareness;
further, as a Yogacara, he says that this "particular" is the self-
cognizable, a part of the cognition, and thus, what the cognition cog-
nizes is the cognition itself. However, I do not suggest that Dignaga
went as far as those Y ogacaras who belong to the fully-developed
idealist branch of Y ogaciira, according to whom the sense faculty
and the object-support such as visible matter are the transformation
(pari1:ujma) of storehouse-awareness.
85
IV. aMra
IV-l. Another interesting point that demonstrates Dignaga's position
on the nature of the object of cognition is his treatment of the so-
called "aspect" (iikiira) , a term used in the Sautrantika system to
82 This translation, following Steinkellner's reconstruction, is based on Vasudhara-
translation: shes pa'i rang gi eha shas bzhin du. Kanakavarman's translation
deviates from this: shes pa'i rang gi bdag nyid bzhin du. This possibly can be recon-
structed as jfilinasvlitmavat, "the cognition itself."
83 PSV ad PS 1.41ab: anekadharmo 'pfndriylirtho yo 'slidhliral}enlitmanendriye 'va-
bhasamlina(1 svtibhtisajfitinotpattihetu(1, sa pratytitmavedya eva jiilinasvlil?lSavat. sa tad-
titmantiSakyanirdesa(1, nirdeiyasya stimtinyavieayatvtit (reconstructed by Steinkellner, ava-
bhasamtinas tadtibhasajiitinotpattihetu(1 is changed to mtinab svtibhasao, according to the
suggestion kindly given by Eli Franco, which is also accepted by Steinkellner per e-mail).
84 PST 47, 6-7: svasaTflvedyam antigamikam.
85 For example, cf. TrBh 16,2-4: tatra litmlidivikalpavtisantiparipoetid rLlptidivikalpa-
vtisantiparipoetie ca tilayavijJ1lintid titmlidinirbhtiso vikalpo rLlptidinirbhlisas ea utpadyate.
"With regard to this [pari!llima or transformation], owing to the complete development of
the latent psychological impression (vtisanti) of the conceptual construction of 'self,' etc.,
owing to the complete development of the psychological impression of the conceptual
construction of 'the visible matter,' etc., the conceptual construction with its manifestation
as 'self,' etc. and as visible matter, etc., arises from the storehouse-awareness."
236 JUNTIECHU
refer to something that connects a cognizing subJect with the cog-
nized object. Dignaga does not reject the Sautrantika's presuppo-
sition, but he develops upon it, coming to the conclusion that an
object of cognition must be internal, since the Sautrantika also main-
tains that the object of a cognition can only be cognized by means of
the fact that the cognitioq has an aspect (akara) as its medium to
grasp the object. The basis of this kind of development may be the
fact, I would like to suggest, that for this special branch of Yogacara
system (vide supra III-3) the state of being existent or non-existent is
not an ontological assertion, but rather a phenomenological descrip-
tion of what has been experienced by awareness; thus, the essential
point of the discussion is whether an experience is obtained inter-
nally or externally, not whether a perceived object really exists
externally or not.
IV -2. In holding this, Dignaga would not be involved in a doctrinal
conflict with the Sautrantika, because such a viewpoint is, to some
extent, actually also shared by the Sautrantika. For the Sautrantika,
an object is not directly perceivable, there is no direct connection be-
tween an object and its subject, and cognition is produced only
through this (the object's) "aspect." According to
this is the Sautrantika's tenet:
The Sautrantika maintains: everything that appears in the form of something
blue, etc., is nothing but the cognition [itself], not the external object,
because an insentient thing is not able to illuminate [itself to the cognition].
[This is] just as it is said that the visible [objects] that produce the cognition
in their own aspects are not the operation-spheres of sense faculties (indriya-
gocara).86
This intermediate thing, the "aspect" (aMra), also called "appear-
ance" (or abhiisa, pratibhiisa, or expressed in the verb form prati-
..fbhas, "to appear," etc.), must be the productive cause of internal
cognition, and this is cognition itself, i.e., self-awareness. In contrast,
86 TBh 63,17-20: sautrtintiktinti'!l matam - jiltinam evedQlJ! sarva,!! nfltidytiktirel}a
pratibhtisate, na btihyo 'rtlla!!, jat;lasya praktiStiyogtit. yathoktam - svtiktirajl1tinajanakti
drsyti nendriyagocarti!:z iii.
the external object, insofar as being independent of cognition since it
is insentient, cannot be perceived directly.
IV-3. The Sautrantika defines this akiira as "a specific way (prakara)
in which all minds and mind-associates grasp the object-support."87
It is also mentioned that the mind and mind-associates are accompa-
nied by an object-support (stilambana) , because they grasp their
object-field; and they have an aspect (sakara) of the object, because
precisely this object-support gives its aspect to the mind and mind-
associates in this specific way (prakara).88 Put in other words, the
mind and mind-associates never come forth without an object-sup-
port, and the object-support appears in the mind and mind-associates
only in a special way, i.e., through an aspect that presents the
particular character of the object-support. This idea is explained in
the AKV as follows:
The reason why these [mind and mind-associates] are accompanied by an
object-support is that precisely this object-support is grasped in a specific
way. How? For an awareness is aware of a blue or yellow thing, it perceives
it [in this way] - this is the meaning; a sensation experiences precisely this
thing as its object-support as such [i.e., as blue or yellow]; the ideation deter-
mines [this thing in that way], and the intention (cetana) forms [a volition
about this thing in that way], and other [cognitions] work the same way.89
That is to say, without this "aspect" the object-support cannot be
perceived in its specific way; the distinct visible form of an object is
made known only by virtue of this "aspect," because an object itself
in its real form, i.e., in the form of atoms, as the real cause of a
cognition, does not present a distinct form. Thus, this "aspect,"
though reflecting a thing that is not really existent, not the real cause
87 Cf. AKBh 401,20-21: eva'!l tu yukta'!l syiit -sarve
e
iilJ1 cittacaittiiniim iilal1lbanagra-
ha(wprakiira iikiira itL Yasomitra idendifies this as Sautrantika opinion (AKV 830, 32:
eva1J1 tu yuktaJ!l syiid iti sautriintikal1latam. Cf. Kritzer 2005: 374.
88 Cf. AKBh 62,5-6: ta eva hi cittacaittiii2... siilambanii vieayagrahaTJiit. siikiiriis
tasyaiviilal1lbanasya prakiirasa iikiira(liit.
89 AKV 165,17-21: yena te salambaniilJ tasyaiviilambanasya praktireTJa grahaTJiit.
kathal1l. vijiiiina,!l hi nfia,!l pfta1J1 vii vastu vijiiniiti, upalabhata ity arthab. tad eva tathii-
lal1lbana,!l vastu vedaniinubhavati. safijiiii paricchinatti. cetaniibhisaJ?lSkarotfty evam iidi.
238 JUNITECHU
of cognition, is the only indispensable link the object and
subject.
IV-4. When discussing, in PS(V) 1.8cd ff., the identity of the means
of cognition with its result, Dignaga asserts that self-awareness is the
result. In doing so, he very clearly presents his Y ogacara position
with regard to the problem of the object of cognition. Although he
bases his idea evidently on Y ogacara concepts, he tries to make the
Sautrantika's position fit his Y ogacara understanding of the object of
cognition.
90
A complete translation of this section is found in Hattori
1968: 28ff. In the following, I will briefly paraphrase those points of
this section that concern our topic, and introduce some relevant
interpretations of Jinendrabuddhi. In some places I deviate from Hat-
tori.
IV-5. In PS(V) 1.8cd-9a Dignaga states that his position is different
from that of those who maintain that an object of cognition exists
externally, since for him the means of cognition is the result. And
this result, i.e., the cognition, is apprehended as accompanied by an
act of cognizing, for it arises with the aspect of an object-field
yiikiira) , and thus, although without action, it is metaphorically
called a means of cognition.
91
Or more precisely, Dignaga rephrases,
90 Cf. Iwata 1991 I: 1-4, where the different doctrinal backgrounds of Sautrantika and
Y ogacara of each statement in PS 1.8ab-9cd are analysed.
91 As Tosaki pointed out (1979: 44), this Sautrantika idea has its origin in Vasuban-
dhu's AKBh 473,25-474.2: yat tarhi vijlliina1J1 vijiiniitfti siitra uktam, kilT! tatra vijHiinaIJ!
karoti. na kiHeit karoti. yathii tu kiiryaIJ! kiiral}am anuvidhfyata ity ucyate siidrsyeniitma-
liibhiid akurvad api kiileit, eva1J1 vijHiinam api vijiiniitfty ucyate siidrsyeniitmaliibhiid akur-
vad api kil1cit. ki/?l punar asya siidrsyam. tadiikiiratii. ata eva tad indriyiid apy utpannaIJl
vieaya/?l vijiiniitfty ucyate nendriyam. "Then, as to what is said in the Sutra that awareness
is aware [of an object], what does awareness do in that case? It does nothing. However,
just as it is said that an effect always conforms to the cause, on account of it acquiring its
existence by being similar [to the cause], although it does nothing whatsoever, in the same
way, it is also said that an awareness is aware [of an object] owing to the fact that it ac-
quires its existence [from the object] by being similar [to it], although [the awareness]
does nothing whatsoever. Now, what is its similarity? It is the fact that it has the aspect of
that [object]. Precisely for this reason, although it is also produced by the sense faculties, it
is said that [the awareness] is aware of the object and not of the sense faculties." (Cf. the
translations in Cox 1988: 39 and in Dhammajoti 2004: 90f.)
self-awareness is the result. Cognition arises with two appearances:
the appearance of the cognition itself and the appearance of the
object-field. A self-awareness of these two appearances is the
result.
92
Jinendrabuddhi adds here: "Regardless of whether there is
an external object or not, the cognition is cognized as bearing both
,,93
appearances.
IV -6. In the next passage Dignaga offers two interpretations that are
based on the Yogacara system and the Sautrantika system, respec-
tively. He explains in PS 1.9b that the reason for an awareness being
the result is that a determination of an object-reference, i.e., the
cognition of an object-reference, has the nature of that awareness
(tadriipa).94 This can be interpreted in two different ways. First,
according to the Yogacara-system, which does not accept external
objects, a cognition that is accompanied by [the aspect] of its object-
field is the object-reference, and thus the object-
reference is apprehended as being pleasant or unpleasant according
to the form of self-awareness (svasaf!1.vedaniinuriipam artham).95
And secondly, according to the Sautrantika-system, a system origi-
nating in the Sarvastivada, which accepts external objects, the exter-
nal object-reference is a cognizable object, and thus the cognition's
simply possessing the appearance of the object-field
sataiva) is the means of cognition.
96
Dignaga emphasizes that
although an object-reference as the proper form of cognition (sva-
92 PS(V) 1.Se-9a: atra ca savyiipiirapratltatviit pramii1}aTfl phalam eva sat. (Sed) na hy
atra Mhyakiiniim iva pramiipiid arthiintaraTfl phalam. tasyaiva tu phalabhatasya jiiiinasya
vi'fayiikiiratayii utpattyii savyiipiirapratlti/:l. tam upiidiiya pramiipatvam upacaryate nir-
vyiipiiram api sat. svasaTflvitti1;t phalaTfl viitra (9a) dvyiibhiisaJ?l hi jiiiinam utpadyate svii-
bhiisaTfl vi'fayiibhiisaTfl ca. tasyobhayiibhiisasya yat svasa'!lvedana'!l tat phalam.
93 PST 69,13-14: saty asati vii Milye 'rtha ubhayiibhiisaTfljiiiina'll saTflvedyate.
94 PS 1.9b: tadrapo hy arthaniicaya/:l /
95 Cf. PSV ad PS 1.9b: yadii hi savi'fayaTfljfiiinam artha/:l, tadii svasaTflvedaniinurapam.
artha'll pratipadyata i'ftam ani'fta'll vii.
96 Cf. PScY) 1.ge-d1: yadii tu Mhya eViirtha/:l prameya/:l, tadii vi'fayiibhiisataiviisya
pramii1}Qm.
240 JUNJIECHU
n7pam)97 is the self-cognizable within the cognition, nevertheless
under the Sautrantika's ontological presupposition mentioned above,
that the object-referem::e exists externally, this point is disregarded in
order to avoid a doctrinal conflict with the Sautrantika-system; thus,
by seeking common ground by putting aside differences, both
systems may agree on the point that the object-reference is cognized
simply through the cognition's possessing the appearance of the
object.
98
With regard to this point, linendrabuddhi states:
Even though at any time [the object-reference as] the proper form [of
cognition] is the self-cognizable [of the cognition], nevertheless, irrespec-
tive of this [nature, we say that] only the cognition's bearing the appear-
ance of its object-field in respect to the external cognizable is the means of
cognition, not its bearing the appearance of [the cognition] itself, because
this [bearing the appearance of itself] is unacceptable as producing [the
cognition] (sc1dhanatva) with respect to an external object, and it is unaccept-
able because it takes another thing [i.e. the internal aspect] as its object. For
insofar as the aspect of the grasping subject takes itself to be the object-field,
how could it be the means of cognition with respect to the external thing?
Indeed it is not possible to be means of cognition (prc1/nc1lJya) for one thing
with the object-field in another thing.
99
Thus, it is quite clear that under the Sautrantika's ontological pre-
supposition it must be said that a cognition bears the aspect of an
external object, not the internal aspect of the cognition. Dignaga's
purpose of stating this, I believe, is to show that even under the
Sautrantika's presupposition, it is still valid to say that a cognition
cognizes its object by means of an aspect and not directly.
97 This "proper form of cognition" is regarded as "the grasped part of cognition," for
instance, Vinltadeva paraphrases the phrace vijifiinasvariipe in TrBh 16,8 as mam pa shes
pa'i gzung ba'i cha (TrT 7bl). Cf. n. 17 andjfiiinasviil!lsa in n. 83.
98 Cf. PSV ad PS 1.9c-d
1
: tadii hi jfiiinasvasal!lvedyam api svanlpam
bhiisataiviisya pramiir.zam.
99 PST 72,6--9: yady api sarvakiilal!l svasar[lvedyam asti svariipam, tathiipi tad ana-
jiiiinasya bahye prameye pramiir.zam, na sviibhasatii, biihye 'rthe
tasyall siidhanatviiyo giit. ayo gas tv apariirthatviit. griihakiikiiro hy atmavi"ayaJ:t kathal!l
biihye 'rthe pramiir.zal'!! syiit. na hy priimaflyal!l yuktam.
IV-7. Dignaga continues: "the object-field is cognized as this or that
form, exactly according to the way in which the aspect of the object
(arthiikara) appears in the cognition, as being pleasant or unpleasant,
etc."lOO Jinendrabuddhi says straightforwardly that this sentence sim-
ply means that "an external object is determined by means of the
aspect of the cognizable of the cognition."lOl This "aspect" is, in the
ultimate sense, self-awareness. Thus, according to Dignaga's inter-
pretation, the Sautrantika's theory agrees with that of the Yogacara
in any case. To this point Jinendrabuddhi states:
And in this case, just as it is said: "The fire is inferred from the smoke," but
this [fire] is not directly inferred from that [smoke], [but] rather from the
cognition of the smoke that is caused by that [smoke], in the same way, even
though it is said: "this object is cognized through that [i.e., through the
cognition's possessing the appearance of the object]," yet it should be
understood as being through self-awareness, which is brought about by that
[appearance]. This is to say, exactly according to the way in which the
aspect of object-reference (arthakara) takes its form (sannivisate)102 in the
cognition in the form of pleasant or unpleasant, etc., the self-awareness
comes forth (prathate); and exactly according to the way in which this [self-
awareness] becomes known (khyiiti), the object is cognized as a pleasant or
unpleasant thing, etc. For, if [the cognition possessing] this aspect be pro-
duced, it must be the awareness of [the cognition] itself with such an
[aspect], and thus on account of this [self-awareness] the object-field should
be established, not otherwise. For this reason, possessing the appearance of
the object-field is the means of cognition.
103
100 PSV ad PS 1.9d
z
: yathti yathii hy arthtikiiro jiiiine pratibhtiti subhtisubhtiditvena, tat-
tadntpab sa pramfyate.
101 PST n,11:jl1iinasyajfieyiikiiravasena biihyo 'rtho nisefyata ity arthab.
102 The word sannivisate is used to paraphrase the word pratibhtiti that is used in the
PSV.
103 PST 72,11-73,2: atra ea yathii dhiimeniignir anumfyata ity ueyate, na ciisau
teniinumfyate, ki/?l tarhi taddhetukena dlulmajfiiinena, tathii yady api - so 'rthas tena mf-
yata ity ucyate, tathiipi tatsiidhanayii svasa1JLVideti veditavyam. tathti hi yatha yathartha-
karo jiiane sannivisate subhasubhadinlpe(la, tathii tatha. svasa1Jlvittily prathate. yathti
yathti ca sii khyiiti, tathii tathtirtho nise/yate subhtisubhtidiriipadily. yadi hi tadiikiiram
utpanna1Jl syiit, tadii tiidrsasyiitmanab sa1!lvittib syiit. talas ea tadvasiid
bhavet, niinyathti. tasmiid prama1}am.
242 JUNJIECHU
That is to say, the object-field of this or that form,(rilpa) is cognized
according to the aspect (aMra) that the cognition possess when it
arises, and this results in a self-awareness; it is precisely through this
self-awareness that the object is cognized, not through the object-
field itself, Thus, this position of the Sautrantika does not contradict
that of the Y ogacara.
IV-S. On the above-mentioned twofold interpretation of the cogniza-
ble, Iinendrabuddhi comments: .
In this regard, in the tenet of the internal cognizable
accompanied with its object-field [means accompanied] by an object-field
that is characterized as the grasped part [of the cognition], because only in
respect to this [grasped part] is the object-field established; in the tenet of
external object-reference however, [it means accompa-
nied] by the external [object-field]. Since in [our system of] representation-
only, [i.e., in the system of Antarjfieyavada,] nothing exists that is separate
from awareness, it is exclusively the cognition that determines the object-
reference to be pleasant when it experiences its own aspect as pleasant. From
the opposite [experience, i.e. one that is unpleasant, the object is determined
as being] the converse [i.e., unpleasant].104
Thus, the cognitive aspect is a necessary medium for cognizing an
object, and only through this aspect can the cognition be produced.
Iinenedrabudhi continues:
In this case, only the aspect of something blue, etc., alone is experienced.
This [aspect] must be necessarily accepted as the cognition itself (vijfia-
nasyatmabhata), otherwise this [cognition] would have no connection with
the object-reference. And a thing (vastu) external to or separated from this
[cognition], no matter whether with the aspect of this [thing] or without the
aspect of this [thing], is never perceived, nor is it possible to be the object-
support. How is this impossible? In [the section] "examination of the Vada-
vidhi" [i.e., PS(V) 1.13-16] it is explained in which sense it is impossible.10
5
104 PST 70,6-10: tatrantarjFieyapakse grahyarrzsalak:ralJena vi:rayeIJa tatrai-
va vi:rayavyavasthanat. bahyarthapakse tll bahyena. tatra vijiiaptimatratayam vijFianavyati-
riktasya vastuno ' bhiivad buddhir eva yadeerarrz svam aka ram anllbhavati, tade:rram artharrz
niscinoti, viparyayad viparftam.
105 PST 68,4-7: iha nfladyakara eka evanubhiiyate. sa vijFianasyatmabhiito 'vasyam
anyathii tasyarthena sambandho na syat. na ca tasmat tadakaram atadakararrz
Therefore, according to Jinendrabuddhi, Dignaga's thought in this
regard is Only the internal cognizable, the so-called "as-
pect," is the object of cognition, not the external object, because this
is not perceived.
IV-9. Jinendrabuddhi's following remark gives the gist of Dignaga's
thought in this regard:
Therefore, no cognition (saY[Lvitti) of something other than awareness (vijiiii-
na) is possible. Rather, only a self-apprehended awareness arises. Thus, only
self-awareness is the result. There may be an external object-reference; even
so, the object-field is determined exclusively according to [self-] awarenes s.
Therefore, it is only possible for this [self-awareness] to be the result. For it
is not the case that the object-reference is experienced according to its inde-
pendent existence (svabhava), so that according to this [experience] the
distinct proper-form [of this object-reference] could be determined, because
otherwise the undesired consequence would follow that all [different] cogni-
tions would have the same form. However, the subjective representations
(viji'iapti) are of many [different] forms. For instance, it is observed that with
respect to a single thing alone, different cognizers acquire their awarenesses
with [different] aspects, according to [their mind] being sharp or slow. But a
single thing does not have many [different] aspects, because otherwise it
must follow that it would not be a single thing. 106
Therefore, in Dignaga's system, regardless of whether the external
object exists or not, that which is cognized is the aspect of self-
awareness which is the SUbjective product: either the aspect of the
cognition itself as the grasped part of cognition according to Y oga-
cara, or the aspect of the external object according to Sautrantika. It
is never the external object itself. To this point Jinendrabuddhi states
quite decisively:
va bahir vyatiriktal[! vastapalabhyate. na ciilambanafJ! ghatate. kathal[! ca na ghatate.
yatha ca na ghatate, tatM vadavidhiparfksayam
106 PST 68,11-69,3: tasmad na vijiianavyatiriktasya kasyacit sal[!vitt* sambhavati. vi-
jfianam eva tu svasaJ'[!viditam utpadyata iti svasal[!vittir eva phalam. bhavatu nama
bahyarthaJ:t, tatMpi yatMsafJ!vedanam eva niscfyata iti tad eva phalal[! yuktam. na
hi yathasvabhavam anubhavo 'rthasya, yato yathasau vyavasthitasvarfipas tatha sakyeta
niicetum, sarvajiiananam ekakaraprasGligat. anekakaras tu vijfiaptayaJ:t. tatha hy ekasminn
eva vastuni pratipattrbhedena patumandatadibhir akarair anugatani vijiianany upalabh-
yante. na caikal[! vastv anekakaram, anekatvaprasGligat.
244 JUNITECHU
First, when [the cognition] is free from conceptual construction, the aspect
of grasping is the means of cognition, [namely] perception which is free
from conceptual construction; the particular, as the aspect of the grasped
object with a clear appearance, is the object of cognition. I07
V. Conclusion
V-I. We have mentioned above that Dignaga's theory was falsely
interpreted by his opponents as coming from a realistic standpoint,
and thus that it contained an internal contradiction. However, as we
have discussed, if we understand Dignaga as maintaining that the
object is a mental product, an internal aspect of object that appears as
externally existent, this problem will disappear. For Dignaga, an
object in any combined form of many atoms is not real. Thus he
refutes both the Sautrantika-theory and the Sarvastivada-theory. For
him, the cognition sphere (iiyatana) is unreal in any case, as it also is
for the Sautrantika, as reported by Sangabhadra. Thus, although the
cognition sphere may be analyzed as an agglomerate of atoms as a
whole, or as the gathered individual atoms, this is not important for
him. From the Y ogacara viewpoint, it is not necessary to make a
distinction between the two. For in any case it is clear for him that
external existents cannot be the cause of our cognition, since as
atoms the cognition cannot bear their appearance, and as an agglom-
erate of atoms they cannot be real existent, i.e., the real cause. There-
fore, the only possible conclusion is that the object of cognition is
the internal cognizable, is the grasped part of cognition, is the self-
cognizable, and any cognition is self-awareness in nature. He also
says that cognition arising with respect to self-awareness is percep-
tion; with respect to an object itself, namely outside this self-aware-
ness, it is conceptual construction. !Os Thus, every true cognition can
only be self-awareness in its true sense. Therefore, for Dignaga, the
svalak:far:ta cannot be an independent external thing, for in that case
it would either be unreal or something of which a cognition cannot
107 PST 74,11-12: nirvikalpe liivat griihakiikiira(z kalpaniipoljhaqz pramii-
r:zam, griihyiikiira(z prameyam .. ,
108 Cf. n. 77.
have an appearance; only as a cognition sphere of cognition, as self-
cognizable, or. as internally cognizable, can it be a real cause of
cognition. In this way, Dignaga's theory of object of cognition pre-
sented in PS(V) 1 is consistent.
V-2. Dharmaklrti explains svalako!ana along other lines. He explic-
itly states that the object of a perceptioI:l is capable of fulfilling a pur-
pose (arthakriyii), but an object of inference is not. And the particu-
lar and the universal are different: one is capable of fulfilling a pur-
pose and is real in the ultimate sense, the other is real only in a
conventional sense. 109 This description of svalaksalJa is, at least
prima jacie, based on something external. Thus DharmakIrti is a
realist in this regard, although in other parts of PV 3 he presents
other views. As expressed by Dunne, Dharmaklrti presents different
"scales of analysis," which in some cases represent the viewpoint of
the "epistemic idealist."llo This is not the case in Dignaga's work,
since for him, as an Antarjiieyavadin, an ultimately real existence,
being an existent outside mind, is non-perceivable, since a cognition
can not bear its appearance. What can be perceived is only an
appearance or an image, it does not exist in its real form, i.e., in the
. form of atoms. Dignaga expresses his skepticism about external real-
ity being perceivable; the idea of fulfilling a purpose is unknown to
him. The difference between DharmakIrti and Dignaga is clear:
DharmakIrti believes that there is a cognition of certainty based upon
an external real existence that can independently fulfill a human
being's purpose, and searches for the rules of its cognition. Dignaga
is skeptical in this regard. Thus the words such as "non-erroneous"
lO9 Cf. PV 3.1a-c: miinalJ'l dvividhalJ'l meyadvaividhyat I arthakriyiiyii'11;
PV 3.3 arthakriyiisamarthaJ'(l yat tad atra paramiirthasat I anyat saJ?lvrtisat proktaJ?l te sva-
II. And in PVSV 84,3-11: sa piiramiirthiko bhiivo ya eva
I (l66ab) idam eva hi vastvavastunor yad arthakriyayogyata ayogyata ca iti
I sa ca arthakriyayogyo artha!l niinveti yo anveti na tasmiit II
(166cd) tasmat sarvaJ'(l samanyam anarthakriyiiyogyatvad avastu I vastu tu viseea eva tata eva
tanni
e
patte!l I (Cf. English trans!. in Dunne 2004: 80f. n.4l.)
IlO Cf. Dunne 2004: 53ff, 59.
246 JUNJIECHU
(abhriinta), "fulfilling a purpose," "trustworthiness" (avisaf!lviida),111
etc., can only be found in Dharmaldrti's vocabulary. Dignaga tells us
what is not a real object, whereas Dharmakirti tells us what real cog-
nition is.
ll2
Because of this difference, Dharmaldrti develops his own
complete theory of perception. In the PVin and later in the NB, he
adds the new element "non-erroneous" to the definition of percep-
tion.ll3
V-3. As for the problem of Dignaga's doctrinal affiliation, I would
not like to give decisive conclusion, because the distinction between
the Sautrantika and Yogacara as two different schools 114 is an
unsolved problem, at least for me.
m the case of Vasubandhu, by whom Dignaga was undoubtedly
strongly influenced, more and more evidence has come to light
showing that many "Sautrantika" theories found in Vasubandhu's
AKBh are similar to those advocated in the Y ogacarabhumi. Some
scholars maintain that the former have their origin in the latter, and
thus, Vasubandhu was already a Mahiiyanist Yogacara when he com-
posed the AKBh, not changing his doctrinal affiliation from Hina-
yana to Mahayana. liS However, in a recently published paper
111 The word avisafJ,viida was used by Dignaga, but in another context, cf. Chu 2004:
115, n. 12.
112 This characteristic of Dignaga' s thought is also presented in his theory of inference.
Hayes has correctly pointed out: "[L]ogic for Dinnaga is not intended to serve as a means
of adding to our knowledge but rather as a means of subtracting from our opinions. The
criteria of inferential certainty that Diimaga puts forward are very strict, so strict that very
few of our opinions can measure up to them. Whereas Dharmaldrti tended to see this
strictness as a weakness in Diimaga 's logic, since it was in Diimaga's system virtually
impossible to arrive at any sort of reasonable inductive certainty, I shall argue that the
strictness of Diimaga's criteria was not oversight on this part but was quite deliberate, for
it served his overall skepticism quite well." (Hayes 1988: 35)
ll3 Cf. PVin 1.4ab: pratyakeal!l kalpaniipoifham abhriintam; NB 1.4: tatra pratyakeal]'
kalpaniipodham abhriintam.
114 Cf. the discussion on the use ofthe word "school" in Franco 1997: 90-92.
115 Cf. Kritzer 1999: 20, 204; 2005: xxvii. Since the 1980s, a number of Japanese
scholars have been involved in studies on Sautrantika and have re-examined the
information on the history of Sautrantika contained in Chinese materials. The most
important results can be found in Kati5 1989; cf. also Kritzer 2003, especially, 218f.
Ohammajoti maintains that "the early and (Sarvasti-
vadin) Y ogacara-s all belong to the same Sarvastivada tradition
originally."Jl6 And based on the assumption of this historical back-
ground he says:
Of course, being within the same milieu, the and
Mahayanic Yogacara - particularly those who are praxis-oriented - must
have been mutually influenced doctrinally. Accordingly, it should not be
surprising to find doctrinal parallels between what Vasubandhu identifies as
Sautrantika doctrine in AKB on the one hand, and some of the doctrines in
the Yogacarabhami on the other hand. This does not necessarily imply that
Vasubandhu bases his Sautrantika doctrines on the Yogacarabhami. 117
Actually, half a century ago Yin shun has already pointed out that in
the sectarian period of Buddhism there was no theory of awareness-
only in the ontological sense, however the theory of awareness-only
without external object in the epistemological sense was already
well-developedYs Thus, it may be safe to say that some Yogacara
theories, especially those in the field of epistemology, are much ear-
lier than the Yogacara that is generally regarded as a Mahayanic
"school." The persons who hold special Yogacara theories may be
called "Yogacara-Sautrantika" based on the differentiation between
the Hmayana Sautrantika and the Mahayanist Sautrantika made in
the Uighur version of the Tattviirtha,119 or "Yogacara with Sautran-
tika presupposition" in following Schmithausen's influential paper
"Sautrantika-Voraussetzungen in ViIpsatika und TriIpsika. ,,120 Based
on my studies in this paper, I tend to assume that Dignaga's theory of
the object of cognition also belongs to this kind of Yogacara theo-
ries. However, since the nature of these special Yogacara theories -
whether they are Mahayanic or remain IDnayanic - has not yet been
116 Dhammajoti 2006: 195
1I7 Dhammajoti 2006: 195-196.
118 Yin shun 1970: 200.
119 Cf. Dhammajoti 2006: 199.
120 Cf. Franco 1997: 94; Schmithausen: 1967. I do not suggest that the difference be-
tween this special group of Yogiiciiras and the rest of Yogiiciiras is the basis of the division
of Yogiiciira system into lung gi rjes 'brang and rigs pa'i rjes 'brang in the later Tibetan
grub mtha' texts.
248 JUNJIECHU
clarified, it seems to me too early to make decision with regard to
Dignaga's doctrinal affiliation. Thus, in this paper, I would like to let
it remain open. Actually, for me to label him as Yogacara or a
Sautrantika is less important than to understand what his statements
about the object of cognition really mean. In any case, one point is
clear: Dignaga treats the object of cognition as something inside
cognition, and regards its appearance as external thing as unreal.
This point is clearly close to the Mahayanic Y ogacara system,
however it is apparently also shared by the Hlnayanic Sautrantika to
a certain extent. Thus, we are not in the position to label Dignaga
simply as Yogacara or Sautrantika, rather, it seems to me proper to
refer him as a "theoretician of an internal object of cognition"
(antarjiieyaviidin), as Iinendrabuddhi did. 121
Abbreviations and literature
1. Primary literature
AD Abhidharmadfpa: Abhidharmadfpa with Vibhii:;iiprabhiivrtti. Ed.
by P. S. Iaini, Patna 1959.
AKBh
AKV
ASP
AP
APv
CWSL
CWSLSI
121 Cf. n. 46.
Abhidharmakosabhii:;ya: Abhidharmakosabha:;yam of Vasu-
bandhu. Ed. by P. Pradhan, Patna 1975.
Abhidharmakosavyakhya: The Abhidharmakosa & Bha:;ya of
Acarya Vasubandhu with Sphutarthii Commentary of Acarya
Yasomittrii. Vol. I-IT. Ed. by S. D. SastrY, Varanasi 1998.
*Abhidharmasamayapradfpika = T 1563 (vol.
29T29).
Alambanaparfk:;a: in Frauwallner 1930: 176-179.
Alambanaparfk:;iivrtti: in Frauwallner 1930: 176-179.
Cheng wei shi lun = Vijiianamiitrasiddhi.astra]: T
1585 (vol. 31).
Cheng wei shi lun suji T 1830 (vol. 43).
HVP
MS
MVBh
NA
NC
NB
NMukh
PST
PS(V) 1
PV
PVP
PVin
PVV
ON DIGNA.GA'S THEORY OF THE OBJECT OF COGN1TION 249
Hastaviilaprakara1J.a: in Frauwallner 1959: 153-156.
Mahiiyiinasmigraha: in E. Lamotte: La Somme du Grand
Vehicule d'Asmiga, Tome I, Versions tibetaine et chinoise
(Hiuan-tsang), Louvain 1973.
= A pi da mo da pi po sha lun
T 1545 (vol. 27).
*Nyiiyiinusiira = A pi da mo shun zheng li lun
] T 1562 (vol.29).
Nayacakra: Dviidasiiral?l Nayacakral?l of A.ciiraya Srf Mallaviidi
with the Commentary Nyiiyiigamiinusiiri1J.f of Srf
Sif!1hasLlri Ga1J.i Viidi 3 vols. Ed. by Muni
JamMvijayajI, Bhavnagar 1966.
Nyiiyabindu: in Pa1J.r!ita Durveka Misra's Dharmottarapradfpa.
[Being a sub-commentary on Dharmottara's Nyayabindup1ca, a
commentary on Dharmakirti's Nyayabindu]. Ed. by D. Malvania.
Patna 21971.
Nyiiyamukha = Yin-ming zheng li men lun T
1629 (vol. 32), and in Katsura 1982.
linendrabuddhi's Visiiliimalavatf Pramii1J.asamuccayatfkii, Chap-
ter 1, Pari I: Critical Edition. Ed. E. Steinkellner et aI., Beijing -
Vienna 2005.
Pramii1J.asamuccayatfkii (Tibetan translation) = Yangs ba dang dri
ma med pa dang ldan pa zhes bya ba tshad ma kun las btus pa'i
'grel bshad, Derge ed., Tshad rna, vol. 20, no. 4268, Ye 1b1-
314a7.
Pramii1J.asamuccaya( vrtti): Digniiga ' s Pramii(wsamuccaya,
Chapter 1, A Hypothetical Reconstruction of the Sanskrit Text
with the Help of the Two Tibetan Translations on the Basis of the
Hitherto Known Sanskrit Fragments and the Linguistic Materials
Gainedfrom linendrabuddhi's Tlkii. Ed. Ernst Steinkellner, pub-
lished at: http://www.oeaw.ac.at/ias!Mat/dignaga_PS_1.pdf.
Pramii(wviirttika: in PVV.
Pramii1J.aviirttikapafijikii (Devendrabuddhi) = Tshad ma mam
'grel gyi dka' 'grel, TP vol. 130, 5717(b), Che, 1-390a8.
Pramii1J.aviniscaya: Dharmakfrti's Pramii1J.aviniscaya, Chapter 1
and 2. Ed. E. Steinkellner, Beijing - Vienna 2007.
Pramii1J.aviirttikavrtti: R. SaiJlqtyayana (ed.), Dharmakfrti's Pra-
250 JUNJIECHU
with a commentary by Manorathanandin. Patna
1938-1940.
T Taish8 Shinsha Daizokyo Ed. Takakusu Jun-
JITO (il1li iffl IX and Watanabe Kaikyoku Uli' jl
Tokyo: TaishO Issakikyo Kankokai 1924-1932.
TP The Tibetan Tripitaka, Peking Edition - Kept in the Library of
the Otani University, Kyoto. Ed. by Daisetz T. Suzuki, Tokyo-
Kyoto 1957.
TBh Tarkabhii-?ii: in H.R. Rangaswami Iyengar (ed.), Tarkabhii-?ii and
Viidasthiina of Mok-?iikaragupta and Jitiiripiida. Mysore 1952.
TrBh Tril]1sikiivijiiaptibhii-?ya: in S. Levi, Vijiiaptimiitratiisiddhi, deux
traits de Vasubandhu. Vil]1satikii (la vingtaine) accompagnee
d'une explication en prose et Tril]1sikii (la trentaine) avec Ie eom-
mentaire de Sthiramati, original sanscrit publie pour la premiere
fois d'apres des manuserits rapportes du Nepal. Fe Partie: Texte.
Paris 1925.
TrT
Vibhii
V.s
V.s
T
VST
Tril]1sikiitfkii = Sum eu pa'i 'grel bshad, TP 5571 (vol. 114), Ku,
lal-69a5.
Vibhilticandra's marginal notes: cf. PVV.
Vil]1satikii Vijiiaptimiitratiisiddhi: see TrBh.
Tibetan translation of V.s = Nyi shu pa'i 'grel pa: in TP 5558
(voL 113), Si, 4bl-11al.
Chinese translation of V.s = Wei shi er shi lun: in T 1590 (vol.
31).
= Rab tu byed pa nyi shu pa'i 'grel
bshad, TP 5566 (vol. 113), Si, 201b81-232a8.
II. Secondary literature
Chu 2004
Cox 1988
J. Chu, "A Study of Sataimira in Dignaga's Definition of
Pseudo-Perception (PS 1.7cd-8ab)," in Wiener Zeitsehrijt
fUr die Kunde Sudasiens 47 (2004) 113-149.
C. Cox, "On the Possibility of a Nonexistent Object of
Consciousness: Sarvastivadin and Theories,"
in Journal of the International Association of Buddhist
Studies 11/1 (1988) 31-87.
Cox 1995 C. Cox, Disputed Dharmas: Early Buddhist Theories on
Existence. Tokyo 1995.
Dhammajoti 2004 Bhikkhu KL. Dhammajoti, Abhidhamw Doctrine and
Controversy on Perception. Dehiwala 2004.
Dhammajoti 2004a Bhikkhu KL. Dhammajoti, Sarvastivada Abhidharma.
Centre for Buddhist Studies, Sri Lanka, 2004.
Dhammajoti 2006 Bhikkhu KL. Dhammajoti, "Sarvastivada,
Sautrantika and Yogacara - Some Reflections on Their
Interrelation," in Journal of the Centre for Buddhist Stud-
ies Sri Lanka 4 (2006) 184-214.
Dreyfus 1997 G. Dreyfus, Recognizing Reality: Dharmakfrti's Philoso-
phy and Its Tibetan Interpretations. Albany 1997.
Dunne 2004 J. D. Dunne, Foundations of Dharmaklrti's Philosophy.
Boston 2004.
Frauwallner 1930 E. Frauwallner, "Dignagas Text, Uber-
setzung und Erlauterungen," in Wiener ZeitschriJt flir die
Kunde des Morgendlandes 37 (1930) 174-179.
Frauwallner 1951 E. Frauwallner, On the Date of the Buddhist Master of the
Law Vasubandhu. Serie Orientale Roma 3. Roma 1951.
Frauwallner 1959 E. Frauwallner, "Dignaga, sein Werk und seine Entwick-
lung," in Wiener ZeitschriJt flir die Kunde Siid- und Ost-
asiens 3 (1959) 83-164.
Frauwallner 1994 E. Frauwallner, Die Philosophie des Buddhismus. 4th edi-
tion, Berlin 1994.
Franco 1997 E. Franco, Dharmakfrti on Compassion and Rebirth. Wien
1997.
Harada 1993 W. Harada (W,S3fr:J*), (J) HastavalaprakraIJa
& Vrtti" - 2:: Skt. -j in
, 6 (1993) 92-110.
Hattori 1968 M. Hattori, Dignaga, On Perception, being the Pratyak:ja-
pariccheda of Dignaga's PramaIJasamuccaya. Cam-
bridge, Mass. 1968.
Hayes 1988 R. Hayes, Dignaga on the Interpretation of Signs.
Dordrecht 1988.
Hirakawa 1973 A. Hirakawa, Index of the Abhidharmakosabha:jya, Part
One, Sanskrit-Tibetan-Chinese. Tokyo 1973.
252
Iwata 1991
Kato 1973
Kato 1989
Katsura 1982
Kitagawa 1957
Kritzer 2003
Kritzer 1999
Kritzer 2005
La Vallee Poussin
1928
La Vallee Poussin
1971
Levi 1932
Mimaki 1972
Nagao 1991
JUNJIECHU
T. Iwata, Sahoplambhaniyama, Struktur und Entwicklung
des Schlusses von der Tatsache, daj3 Erkenntnis und Ge-
genstand ausschliej3lich zusammen wahrgenommen wer-
den, auf deren Nichtverschiedenheit. Teil I-II. Stuttgart
1991.
J. Kato c fOE - C
tjWif{O):rJE::tj] -ll in 1 (1973)
129-137.
J. Kato (Etude sur les
Sautriintika). Tokyo 1989.
S. Katsura en.), in
42 (1982) 84--100.
H. Kitagawa (::it! 11*.&Ij): "A Study of a Short Philosophi-
cal Treatise Ascribed to Dignaga," in Sino-Indian Studies
5/3&4 (Liebenthal Festschrift), 1957, 126-137.
R. Kritzer, "General Introduction [to the issue of the
International Association of Buddhist Studies entitled The
Sautriintikas]," in Journal of the International Association
of Buddhist Studies, 26.2 (2003) 201-224.
R. Kritzer, Rebirth and Causation in the Yogiiciira Abhi-
dharma. Wien 1999.
R. Kritzer, Vasubandhu and the YogiiciirabMimi, Yogii-
ciira Elements in the Tokyo
2005.
L. de La Vallee Poussin, La Siddhi de Hiuan-Tsang.
Traduite et Annotee. Tome I-II. Paris 1928.
L. de La Vallee Poussin, L'Adhidharmakosa de Vasuban-
dhu. 6 vols. 2
nd
edition, Brussels, 1971.
S. Levi, MateriaL/x pour l' etude du systeme Vijiiaptimiitra.
Paris 1932.
K. Mimaki (iW!:9:R:c), Q
Sautriintika IDl.ll , in 43 (1972) 77-92.
G. M. Nagao, Miidhyamika and Yogiiciira, A Study of
MahiiyiinaPhilosophies, Selected Papers of C. M. Nagao.
Edited, collated, and translated by L. S. Kawamura in
collaboration with G. M. Nagao. New York 1991.
Schmithausen 1967 L. Schmithausen, "Sautrantika-Voraussetzungen in Virp.-
satikii und TriI1Isikii," in Wiener Zeitschrift flir die Kunde
Siid- und Ostasiens 11 (1967) 109-136.
Schrrrithausen 1973 L. Schrnithausen, "Spirituelle Praxis und Philosophische
Theorie im Buddhismus," in Zeitschrift fUr Missions-
wissenschaft und Religionswissenschaft 57 (1973) 161-
186.
Thomas & Vi 1918 E. W. Thomas and H. Vi, "'The Hand Treatise,' A Work
. of Aryadeva," in The Journal of the Royal Asiatic Society
of Great Britain and lrland, 1918,267-310.
Tosaki 1979 H. Tosaki -
'(7' 7 7--;- r'f 7 -}vT 1 tJ.!l O);m;'rnH - ..t. ' -r
!1fU . Tokyo 1979.
Vi 1958 K. Vi . Tokyo 1956.
Yin shun 1970 Yin shun (I=P)IWi), . Taipei 1970.

Volume 29 Number 2 2006 (2008) pp. 255303
ON THE DEVELOPMENT OF THE NON-MENTATION
(AMANASIKRA) DOCTRINE IN INDOTIBETAN BUDDHISM
*

DAVID HIGGINS
I. Introduction
The first widespread use of amanasikra
1
(non-mentation) as a spe-
cific description of Buddhahood occurs within the Indian Siddha
movement, although the term is not unknown in the Pali canon.
2
The

* I would like to take this opportunity to thank Dr. Tom Tillemans for his valuable
comments and suggestions on earlier drafts of this paper. I am also much indebted to the
late Herbert Guenther with whom I had the opportunity to study and discuss many of the
texts I have cited. The paper also benefited from dialogues with several participants of the
2006 Graduate Students Conference at Hsi Lai Temple, Los Angeles, where I presented a
much abridged version of this paper. In particular, I would like to thank Elon Goldstein
and Ching Keng, both of Harvard University. It would be remiss not to offer sincere
thanks to the editors of JIABS, Helmut Krasser and Birgit Kellner, and the two reviewers
for their close reading of the paper and many cogent suggestions and clarifications, espe-
cially Klaus-Dieter Mathes who kindly made available his forthcoming paper on
Maitrpas Amanasikrdhra during final revisions of this paper.
1
Non-mentation and non-egocentricity are two possible translations of amanasi-
kra (T. yid la mi byed pa), a technical term that is as rich in nuances as its counterpart
manasikra/manaskra (T. yid la byed pa). In Abhidharma exegesis, manaskra occurs as
the last in a sequence of five omnipresent (Skt. sarvaga, Tib. kun tu gro ba) mental events
(caitta) that are present, overtly or covertly, in all conscious processes. See, for example,
Abhidharmasamuccaya p. 6. As the natural culmination of the third omnipresent mental
event intentionality (cetan) which describes the general object-directedness of mind,
manaskra has the function of bringing to mind or setting ones mind upon (focusing
on) a particular object and remaining involved (conceptually and affectively) with it.
When used as verbs, I have translated the terms accordingly. While mentation and non-
mentation are generally adequate as translations of manasikra and amanasikra in their
deployment as abstract nouns (and have been adopted throughout this paper for the sake of
consistency), ego-centricity and non-egocentricity are more precise in contexts where
the Sanskrit or Tibetan terms have been interpreted in line with the nuanced Cittamtra
conception of manas as both an intentional (object-intending) and reflexive (I-intending)
operation that structures experience in terms of an I (subject) and mine (object). See
below sections VIII and XII and notes 37 and 41.
2
In Majjhimanikya (I 436) and Aguttaranikya (IV 425), the amanasikra of the

DAVID HIGGINS 256
term and its Apabhraa variant amaasira are associated in
particular with the mystical songs (doh or vajragti) of Saraha, the
most famous of the early Siddhas,
3
and a cycle of texts attributed to
his commentator Maitrpda (aka Maitrpa, b. 1007 or 1010)
4
re-
ferred to in Tibet as the Yid la mi byed pai chos skor or The Cycle
of Teachings on Non-mentation.
5
The term also appears in certain
songs of Tilopa (9881069),
6
the Siddha from whom the Tibetan
bKa brgyud tradition claims descent. We find in the relevant pas-
sages of these authors the seeds of an Indo-Tibetan hermeneutical
tradition according to which amanasikra is understood as a descrip-

concept of the multiple (nnattasa) is identified as the goal in the ksnacyatana.
On later Mahyna sources, see note 37 below.
3
The dates of Saraha remain subject to controversy. The problem is compounded not
only by textual references to more than one Saraha but by the numerous appellations used
for the different Sarahas, such as Rahlabhadra, abari, Sarojavajra, Saroruha, and
Saroruhavajra (Shahidullah 1928: 29). Moreover, Tibetans refer to the Siddha Saraha by
various names such as Sa ra ha, Sa ra ha chen po, Bram ze chen po, mDa bsnun pa (=
Sarahan, one who has shot the arrow"), Ri khrod pa chen po Saraha. This Saraha seems
to have lived sometime between the seventh and early eleventh centuries CE. See Kvaerne
(1977) and Ruegg (1981) for an examination of the available research on possible dates.
Rahl Sktyyan (1957: 139) establishes Sarahas lineal successors as contemporaries
of King Devapla (810850), and links him with the beginning of the renaissance of Bud-
dhism during the Pla Dynasty, particularly at Nland university, under the reign of King
Dharmapla (970810). Shahidullah places Saraha at around 1000, a date supported by the
tradition that Saraha conferred initiation on a king named Ratnapla. This could refer to
the Ratnapla who reigned in Assam from 1000 to 1030. Support for this date is also
found in a Tibetan tradition which makes Maitrpa (eleventh century) a direct disciple of
Saraha. See Guenther (1969: 13). In support of this thesis, Guenther mentioned in personal
correspondence the existence of a Tibetan Brug pa bKa brgyud work he had seen in
Ladakh that purports to be a record of a dialogue in the form of question and answer be-
tween Maitrpa and his master Saraha.
4
On Maitrpda (alias Advayavajra and Avadhtipda), see Deb ther sngon po, da,
fol. 2af. (BA 841f.). See also Mathes (2006).
5
This cycle is listed in Bu stons gSan yig, fol. 58b1 and Padma dkar pos gSan yig,
fol. 33b2. Its history and general significance are dealt with in Padma dkar pos Phyag
chen gan mdzod, fol. 16a3. For a comparative listing of Bu stons and Padma dkar pos
versions supplemented with listings from the Advayavajrasagraha and the Tohoku cata-
logue to the bsTan gyur, see Broido (1987), Appendix B, p. 55f.
6
On Tilopa (alias Tillopa, Tillipa, Tailopa, Telopa, and Taillikapada), see Hoffman
(1956: 14045) and Guenther (1963: xiv).
NON-MENTATION DOCTRINE IN INDOTIBETAN BUDDHISM

257
tion of, and often also defended as a theory of, the highest experience
of Buddhist Tantrism known as Mahmudr (phyag rgya chen po).
This paper will inquire into the shifting meanings of amanasikra
as it developed in various contexts of Buddhist discourse, Indian and
Tibetan. Attention has recently been drawn to its occurrence in cer-
tain polemical contexts surrounding the Sino-Indian controversy be-
tween Instantaneist (cig car ba) and Gradualist (rim gyis pa) ap-
proaches to the goal of awakening (byang chub) as they were repre-
sented at the so-called bSam yas debate by the Chinese Chan master
Hva shang Mahyna (Mo ho yen) and the Indian Buddhist monk
Kamalala (750795) respectively.
7
We will begin by sketching the
genealogy of the term in these contexts, touching only briefly on
those issues concerned with the debate which have been treated at
length by others. Attention will then shift to the main focus of this
paper: a preliminary survey of non-mentation teachings in the tradi-
tion extending from Saraha and Maitrpda down through a long line
of Tibetan bKa brgyud masters. While these teachings by their na-
ture elude intellectual comprehension and resist easy classification,
they have been too influential with Tibetan bKa brgyud traditions to
simply leave aside. It is therefore hoped that this cursory overview
will broaden our limited understanding of this important tradition
and shed light on the still poorly understood complex of influences
that inspired the non-gradual Mahmudr teachings of Tibet.
II. amanasikra and bSam yas polemics
In the sBa bzhed,
8
an early Tibetan chronicle (probably eighth cen-
tury) on the bSam yas debate, amanasikra is presented as repre-
sentative of an instantaneist doctrine attributed to Hva shang advo-
cating the suppression of all mental activity. Kamalala criticizes a

7
See Demiville (1952), Ruegg (1989: 192f.), David Jackson (1994), Roger Jackson
(1982), van der Kuijp (1984), Broido (1987), and Gomez (1987: 96f.).
8
A succinct account of the bSam yas debate according to the early sBa bzhed is given
by Faber (1986). Relevant materials on the debate from the later (Zhabs btags ma) version
are found in Houston (1980). On the probable history of the earlier and later versions, see
Ruegg (1989: 67f.).
DAVID HIGGINS 258
similar doctrine (although the concepts used, such as sajisam-
patti are unmistakably Indian) in his first Bhvankrama without
identifying its proponent (Hva shang is mentioned nowhere in his
three Bhvankramas). The gist of his critique is that amanasikra
does not imply the suppression or cessation of mental activity but
rather its progressive refinement through the gradual elimination of
subjective distortions. The relevant passage from the Bhvankrama
I
9
reads:
When the Nirvikalpapraveadhra (NPD) states that non-mentation aban-
dons characteristics (mtshan ma = nimitta) such as [material] form (gzugs =
rpa), what it means is that there is no mentation directed toward things
that are not perceived when analyzed though discerning insight but it does
not [mean] a simple absence of mentation.
It is not an abandonment as in the conceptless absorptions (du shes med pai
snyom par jug pa = sajisampatti) and so forth, [i.e. an abandonment]
due to simply giving up mentation which has been attached to form and so
forth since beginningless time.
Thus, Kamalala singles out mistaking amanasikra for the absolute
non-existence of manasikra
10
as the major misinterpretation of this
concept, a point he further clarifies in his Nirvikalpapraveadhra-
k.
11
But this non-existence insofar as it is no thing whatsoever
(dngos po med pa), cannot serve as a cause of anything and thus
leaves no possibility of genuine conceptual analysis (yang dag par so
sor rtog pa = bhtapratyavek). And without such analysis, Kama-
lala argues, it is impossible not to mentally engage (yid la mi byed

9
Tucci (1958) 261. In the following passage, I have standardized Tuccis translitera-
tion: rnam par mi rtog pa la jug pai gzungs las yid la mi byed pas gzugs la sogs pai
mtshan ma spong ngo zhes gsungs pa gang yin pa de yang shes rab kyis brtags na mi dmigs
pa gang yin pa de der la yid la mi byed par dgongs kyis / yid la byed pa med pa tsam ni ma
yin te / du shes med pai snyoms par jug pa la sogs pa ltar / thog ma med pai dus nas
gzugs la sogs pa la mngon par zhen pai yid la byed pa spangs pa tsam gyis spong ba ni ma
yin no //
10
In Sanskrit, the negative prefix a- may be taken either as an absolute, non-affirming
negation or as a relative, affirming negation whereas in Tibetan, these are generally,
though not always, distinguished by the use of med for the former and min or ma yin for
the latter.
11
P v. 105 118.5.7f.

259
pa) in the characteristics of material forms and the other groupings
(skandhas) present to cognition. Thus there is no amanasikra apart
from this manasikra (yid la byed pa de las gzhan yid la mi byed pa
yang ma yin) taken in the sense of genuine conceptual analysis.
Kamalala goes on to show that what is intended by amanasikra is
none other than this genuine analysis, the former being the result
(phala) of the latters operation as a counter-agent (pratipaka) to
ordinary mentation (manasikra). Conceptual meditation, in other
words, is a necessary condition for non-conceptual realization:
12

The characteristics of genuine conceptual analysis are what is intended in
[using the term] non-mentation. While it is of the essence of divisive
conceptualizing, it is nonetheless burned away by the fire of genuine gnosis
arising from itself, just as a fire kindled by rubbing two sticks burns these
very pieces.
As recently noted by Klaus-Dieter Mathes,
13
Kamalalas gradualist
view of amanasikra as the non-conceptual outcome (phala) of
conceptual analytical meditation reinterprets the NPDs understand-
ing of amanasikra as direct, non-conceptual realization to bring it
into line with the traditional progressivist Mahyna-Madhyamaka
paradigm. This revisionist interpretation was not always endorsed by
later Tibetan bKa brgyud scholars. Gos lo tsa ba gZhon nu dpal
(13921481), for example, contrasts this gradualist interpretation of
amanasikra with the interpretation advanced by Maitrpa advocat-
ing direct, non-conceptual realization and the abandonment of men-
tal constructs through unmediated recognition of their luminous na-
ture. As gZhon nu dpal states in his commentary to the Dharmadhar-
matvibhga-kriks (Mathes translation):
14

As to what has thus been taught in the DhDhV, it is the meaning of entering
the non-conceptual that has been established [here]. There are obviously two

12
The following passage is quoted in Mathes (2005) 32, n. 48. The translation is my
own. yang dag par so sor rtog pai mtshan ma ni dir yid la mi byed par dgongs so // de ni
rnam par rtog pai ngo bo nyid yin mod kyi / on kyang de nyid las byung ba yang dag pai
ye shes kyi mes de bsregs par gyur te / shing gnyis drud las byung bai mes shing de gnyis
sreg par byed pa bzhin no //
13
See Mathes (2005) 12f.
14
Mathes (2005) 1314.
DAVID HIGGINS 260
traditions [, however,] of how to comprehend the meaning of this stra.
Kamalala maintains that the [interpretative] imaginations that must be
given up can be only given up on the basis of insight resulting from thorough
investigation. It is maintained in the commentary on Maitrpas Tattvadaa-
ka, by contrast, that they are not given up as a result of thorough investiga-
tion, but of a meditative stabilization which [experiences] reality exactly as
it is (Skt. yathbhtasamdhi). The latter knows the own-being of [even]
that which must be given up as luminosity. Here it is reasonable to follow
Maitrpa, who [re]discovered this treatise.
In Tibet, the association of amanasikra with a doctrine of the Chi-
nese Hva shang purportedly encouraging the supression of all mental
activity was introduced in several works by Sa skya Paita Kun
dga rgyal mtshan (11821251) in order to attack, by way of anal-
ogy, a certain non-Tantric present-day Mahmudr system (da
ltai phyag rgya chen po) which he describes as being for the most
part (phal cher) a Chinese doctrine or even as Chinese-style rDzogs
chen (rgya nag lugs kyi rdzogs chen).
15
This criticism and the subse-
quent charges of philosophical incoherence by the dGe lugs pas and

15
Sa skya Paitas critique of the Present-day Mahmudr or Neo-Mahmudr is
found in his sDom gsum rab dbye, Thub pai dgongs pa rab tu gsal ba and sKye bu dam pa
rnams la spring bai yi ge. Sa skya Paitas source appears to have been the sBa bzhed
since he refers to a dPa bzhed, dBa bzhed, or Ba bzhed in his discussions of Hva
shangs doctrines. It is important to note that that Sa skya Paita never directly criticizes
the amanasikra tradition of Saraha and Maitrpda. This is not surprising given that the
Siddha Virpa, who is regarded as spiritual progenitor of the Sa skya tradition, similarly
extols non-mentation in his spiritual songs. See for example his Dohkoa, in Phags yul
grub dbang rnams kyi zab moi doha rnams las khol byung mi tig phreng ba, pp. 1589. Sa
pans critique was mainly directed at some of the non-gradual Mahmudr teachings
associated with sGam po pa emphasizing the direct introduction (ngo sprod) to ones natu-
ral mind. These were repudiated on the grounds that (1) they were being taught independ-
ently of the Tantric system of mudrs elaborated by Nropa and transmitted in Tibet by his
disciple Mar pa, that (2) they represented newly introduced doctrinal innovations of
questionable (i.e., non-Indian) provenance and that (3) they advocated an erroneous non-
conceptual, non-gradual approach to goal-realization. See David Jackson (1994: 72f.) In
this connection, it is worth noting that the Mahmudr teachings of the Indian Siddhas
were decidedly non-gradual, encouraging direct, non-conceptual, spontaneous forms of
meditation and instruction, and were frequently taught independently of the Tantric system
of four mudrs. Whatever the non-Indian influences on sGam po pas varied discourses on
Mahmudr, it is in the teachings of the Indian Siddhas and their Tibetan successors (such
as Mar pa and Mila ras pa) that we find the major source of inspiration for sGam po pas
own non-gradual Mahmudr teachings.

261
Sa skya pas (who, by the sixteenth century, had both become politi-
cal rivals of the bKa brgyud pas)
16
led many bKa brgyud masters
including gZhon nu dpal, Padma dkar po (152792), Situ bsTan pa
nyin byed (dates unknown), Mi bskyod rdo rje (150754), Dvags po
bKra shis rNam rgyal (151287) to respond by (i) refuting allega-
tions that their tradition had any connection with Hva shang (the
rNying ma pas were less reluctant to acknowledge his influence or
Chinese influence in general);
17
and (ii) defending amanasikra both

16
The sixteenth century was marked by power struggles between the Karma pas and
dGe lugs pas for territory in Khams and gTsang. At this time the Karma pas, Sa skya pas
and dGe lugs pas all vied for patronage and assistance at the Mongol court of Gengzhiz
Khan and his successors. The Brug pa sect of the bKa brgyud was also drawn into the
political turmoil. Ngag dbang rnam rgyal (15941691), the recognized incarnation of
Padma dkar po (15271592), was forced to flee to Bhutan (in 1616) under the enmity of
the House of gTsang and Rva lung monastery, the seat of the Brug pa bKa brgyud, and
its affiliates were seized by the gTsang authorities. See Smith 1970: 6. By the seventeenth
century, the dGe lugs pas had gained the upper hand in these political struggles and were
to maintain a theocracy up to the time of the Chinese invasions of the last century. See
Stein (1962: 80f.), and Snellgrove & Richardson (1968, chapters 79).
17
Klong chen rab byams pa makes two intriguing references to a Ha shang Mah-
yna in his writings. In discussing the historical genesis of Mantrayna (gsang sngags ji
ltar byung ba'i tshul), Klong chen pa (Grub mtha' mdzod, fol 139b4f.) argues for a much
larger number of authentic stras and tantras than the gSar ma redactors allowed for. He
notes the presence in Tibet of many stras and tantras that were to be found in India, as
well as many that were not. Concerning stras, he mentions many stras translated into
Chinese before Sryasiddhi destroyed the manuscripts in a fire. According to Tra-
nthas History of Buddhism in India, pp. 1413, Sryasiddha (Tibetan Nyi ma dngos
grub) was an anti-Buddhist king responsible for a fire at Nland university that destroyed
a large number of Sanskrit Buddhist texts. Klong chen pa goes on to say that many of the
Chinese translations including the Avatasakastra, [Mah]parinirva[stra], and Vina-
ygama ('dul ba lung) were translated by Vairocana and Ba sang shi based on originals
deriving from the mind of the Chinese Pandit Ha shang Mahyna (rgya nag gi pan di ta
ha shang ma h y na'i blo las). It is not possible, Klong chen pa concludes, for any but
the omniscient to assess the measure/scope of a teaching or an individual (chos dang gang
zag gi tshad). Vairocana (renowned rDzogs chen scholar, translator, and student of Pad-
masambhava) and Ba sang shi (ta) have been clearly identified with the early residents at
bSam yas and are counted among the seven original Sangha members (sad mi mi bdun)
who were ordained by ntarakita in the eighth century. (Tucci 1958: 12f.) Ba sang shi
served as an envoy sent by the Tibetan king Mes ag tshom to the Tang court of China on
account of the emperors interest in Buddhism. Among the supporters of Hva shang in the
ensuing rift between Indian and Chinese factions are mentioned mNya Bi ma or sNa Bye
ma (possibly Vimalamitra) and Myang/Nyang Ting nge dzin bzang po, both important
figures in the early history of rDzogs chen. rDzogs chen works from the early period such

DAVID HIGGINS 262
as a valid description of the experience termed Mahmudr (Su-
preme Seal) and as an authenticated tradition of Mahmudr
instructions inspired by Saraha and systematized by his commentator
Maitrpa. The corpus of twenty-five texts known in Tibet as the Yid
la mi byed pai chos skor nyi shu rtsa lnga, most of which are avail-
able in Sanskrit in the Advayavajrasagraha,
18
form the main textual
basis for this tradition.

as the Chos 'byung Me tog snying po of Nyang ral Nyi ma od zer (11241192) and the
bSam gtan mig sgron of gNubs Sangs rgyas ye shes (probably eighth century) give an
impartial account of Hva shangs teachings. The latter interestingly considers both
Kamalalas and Hva shangs positions to be deviations (gol sa) from the more inclusive
rDzogs chen perspective.
One clear instance of Hva shangs influence on classical rNying ma thought is a passage in
Klong chen pas gNas lugs mdzod (fol 6b1 and auto-commentary, fol 33b2 f) which be-
gins: The sun of self-originated pure awareness (rang byung rig pa), the ultimate reality
(don dam), / Is equally obscured by the white or black clouds [of] virtue or evil See
Faber (1986: 478) for Hva shangs virtually identical statement recorded in the sBa
bzhed. Klong chen pas commentary on this passage is most interesting (33b6): Although
at the time the Mahpaita Ha shang made this statement, narrow-minded people could
not comprehend it, it in fact holds true. It is kept secret from those on the lower spiritual
pursuits; were they to denigrate it because their minds could not comprehend it, they
would only plunge, on account of this karma, into the lower destinies. (slob dpon chen po
ha shang gis gsungs pas de dus blo dman pai blor ma shong yang don la de bzhin du gnas
so / theg pa og ma gsang ba blor mi shong bas skur pa btab dus kho las des ngan song du
ltung bai phyir ro //)
Later rNying ma pas did not necessarily share Klong chen pas high estimation of Hva
shang. Mi pam rgya mtsho (18461912), for example, distinguishes the correct under-
standing of yid la mi byed pa non-mentation [resulting] from not seeing any character-
istics of objective references whatsoever once all attachments to entities have been under-
mined (dngos zhen thams cad khegs nas dmigs gtad kyi mtshan ma ci yang ma mthong nas
yid la mi byed pa ) from Hva shangs annihilationist version which supports the
suppression of all thoughts. Concerning the latter, Mi pham cites a passage from the
Dharmadharmatvibhga (pp. 378) which presents amanasikra as one of five factors
that are to be eliminated. See dBu ma rgyan gyi rnam bshad jam dbyangs bla ma dgyes
pai zhal lung, p. 57.1f.
18
See note 5 above and Bibliography under Maitrpa.

263
III. Maitrpa on amanasikra
It is of interest to note that Maitrpas views regarding amanasikra
as set forth in his Amanasikrdhra
19
are characterized by a
conspicuous concern to defend the concept against possible misinter-
pretation. This concern, also evident in the Siddha literature we will
examine, again indicates that the Gradualist opposition to amanasi-
kra was prevalent in Indian, as well as Tibetan, polemical con-
texts.
20
Maitrpa begins his short text by noting that this term ama-
nasikra has been widely misunderstood. Maitrpa takes up first (i)
the objection by some people that the term is ungrammatical (apa-
abda) because in compound [the correct form] should be amanas-
kra.
21
Maitrpa responds by quoting Pinis Adhyy (6,3,14)
where it is stated that in the case of tatpurua and kt [the com-
pound] is irregular.
22
Maitrpa goes on to show that the term is both
found in the stras (here he quotes the NPD) and attested in the
tantras

in order to refute two further objections: (ii) that the term

19
Sanskrit passages are based on the Japanese edition of the Amanasikrdhra (here-
after AM; see Bibliography). I have incorporated corrections made by Mathes in his
forthcoming paper. Tibetan passages are from D v. 28, p. 276.4f.; P v. 68 p. 286.56f.
20
One must nonetheless be wary of the temptation to derive historical connections
from conceptual correlations. (See Gomez 1987: 139 n. 14) Gomez argues convincingly
that the sudden-gradual opposition only reflects a very general, sometimes vague, intui-
tion of a tension or polarity between two approaches to knowledge and action (p. 131).
21
AM 136,34: tatra kacid ha / apaabdo yam iti / samse manaskra iti bhavi-
tum arhati / Tib: D v. 28, p. 276.5: yid la mi byed pa zhes bya ba di la phal cher log par
rtogs pa ste / de la kha cig di skyon can gyi tshig tu smra ste / bsdus pas yid [la] mi byed
pa zhes rtogs par gyur ro / In the last sentence, the Tibetan yid la mi byed pa could be cor-
rected to yid mi byed pa to avoid confusion of amanaskra, which it translates, with
amanasikra, which is invoked in the beginning of the passage.
22
Constructions such as manasi + k are common in Sanskrit. Because a close
association developed between manasi and k, a syntactic compound came into being be-
tween the locative manasi and the verbal noun kra derived from k. When, against the
general rule, the case suffix of the first member is not dropped, the compound is called
aluk (non-deletion). I thank Dr. Aok Aklujkar for sharing these grammatical observa-
tions (in personal correspondence) and refer the reader to his Sanskrit: An Easy Introduc-
tion to an Enchanting Language (Richmond: Svadhyaya Publications 2003), sections 29.11
fn. 7, 33.16, 34.22.
DAVID HIGGINS 264
does not belong to Buddhism; (iii) that it belongs only to the stra
corpus but not to the tantras.
23

The remainder of Maitrpas short text addresses a fourth objec-
tion which acknowledges the presence of the terms meaning in the
tantras but absurdly contends (iv) that amanasikra there constitutes
an absolute, non-affirming, negation (prasajyapratiedha = med par
dgag pa) given that manasikra is the subject of the negative particle
(nao viaya = med pai yul) in a non-affirming negation. To this
Maitrpa replies:
24

That is not the case. [A non-affirming negation] is a negation of a different
type; it is a prasajyapratiedha (prasajya negation) in the sense that it does
not negate what is not applicable (aprasajya). An illustrative example: the
wives of the king do not see the sun. The meaning of this is as follows: the
wives of the king are kept hidden so that they do not see even the sun. This
does not mean that the sun is non-existent. So what does it mean? What we
negate is [only] what is applicable (prasajya), viz., that the kings wives see
the sun. In the case of amanasikra (non-mentation) too, what we negate

23
Maitrpa responds to the latter objection by saying (Mathes translation): That is not
so, for it is stated in the Hevajra[tantra], in the chapter on reality [I.5.1]: Neither mind
nor mental factors exist in terms of an own-being. Moreover, according to the sense, [HT
I.8.42ab may be] understood in terms of amanasikra [as well]: Therefore the whole
world is meditated upon [in such a way], because it should not be produced by the intel-
lect. AM 136,10138, 3: tan na / uktam hi hevajre tattvapaale / svarpea na citta
npi cetasam / tath / bhvyate hi jagat sarva manas yasmn na bhvyate / arthd ama-
nasikreeti gamyate / Tib. D v. 28, p. 277.3f.: de ni ma yin te / he ba dzra las gsungs pa /
ngo bo nyid kyis sems med cing / sems byung baang med pao / de bzhin / gang phyir yid
kyis ma bsgoms na / gro ba thams cad bsgom par bya / zhes pai don gyis yid la mi byed
pao zhes rtogs par byao //
24
AM 138,815: tan na / prakrntarasya pratiedha / nprasajya pratiidhyata
iti prasajyapratiedha / yathsryapay rjadhr / ayam artha / eva nma t gupt
rjadhr yat sryam api na payantti / atra na srybhva kta / kin nma rja-
dhr yat sryadarana prasajya tan niiddham / amanasikre pi na manasika-
raa yad grhyagrhakdi prasakta tan niiddham / na mana / ato na doa / Tib. D
v. 28, p. 277.5: de yi ma yin na / yod pai dngos poi dgag pa ni med par dgag pao / yang
na thal ba med pai thal bar gyur ba gog pao zhes pa ni / med pa dgag pao / ji ltar
rgyal poi btsun mo nyi ma mthong ba ni dii don to / di lta bu mi min zer zhes pa yang
rgyal poi chung ma de shin tu sbas pas nyi ma yang mi mthong bao / dir bkag pai nyi
ma med par ma byas bao / gang zhe na / rgyal poi chung ma rnams gang gis nyi ma
mthong ba yod na de gog pa ste / yid la mi byed pa la yang yid la byed pa gang gzung ba
dang dzin pa la sogs par yod pa de gog pa ste / yang ni ma yin te / di la skyon med do //

265
by means of the negative affix [a] is only that which is applicable, viz.
manasikaraa (mentation) consisting in [dichotomies] such as subject and
object. The mind itself is not [negated]. Therefore there is no mistake.
The non-affirming negation here applies to the dichotomizing acti-
vity of mind, not to mind per se. This will enable Maitrpa to argue
that the absence of dichotomic mentation, with its clinging to ex-
tremes of existence and non-existence, can allow minds true nature,
its radiant clarity to show itself. Maitrpa goes on to consider reifica-
tions associated with mentation, particularly constructs concerning
eternalism and nihilism and concludes that amanasikra refers to a
state in which all such superimpositions or denigrations have been
completely left behind (sarvavikalpanimittasamatikrmat).
25

Maitrpa is now in a position to indicate how amanasikra can
also be understood in terms of a relative, affirming negation (pary-
udsa-pratiedha = ma yin par dgag pa). As he states (Mathes
translation):
26

Even [when amanasikra is taken] in the sense of an affirming negation,
there is no fault. When [someone] says Bring a non-Brahmin, the bringing
of somebody similar to a Brahmin, a Katriya or the like [is intended], but
not a low-caste person of base origin, such as a wagon maker. Here, too,
[where amanasikra is taken as an affirming negation,] an awareness of
essencelessness is maintained. Hence the tenet of Myopamdvaya is estab-
lished. From what, then, does the [undesired] consequence of the view of
annihilation follow?

25
For a full translation of the relevant passage, see Mathes (forthcoming).
26
AM 140,811: paryudsapake pi na doa / abrhmaam nayety ukte brhma-
asadasya katriyder nayana bhavati / na tu vijtyasya kade / atrpi nisvabh-
vavedanasya sasthiti kt / etena myopamdvayavdasthito bhavet / kuta ucchedav-
daprasaga iti / Tib. D v. 28, p. 278.3f.: ma yin par brtags pai phyogs kyang skyon med
de / bram ze ma yin par khrid la shog ces pas bram ze dang dra bai rgyal po la sogs pa
khrid la shog ces par gsal gyi / rigs mi mthun pa shing shing rta mkhan la sogs pa ni ma
yin no / di la yang rang bzhin med pai rig pa la gnas par byas pa ste / de dag gis ni sgyu
ma lta bur gnyis su med par smra bar gnas par gyur ro / gang
a
las chad par lta bar gyur /

a
Text reads grang but P, v. 68, 287.2.7 correctly has gang.
DAVID HIGGINS 266
Understood in terms of an affirming negation, amanasikra negates
those base dualistic thought-processes that obscure reality while pre-
serving the latters essenceless, non-dual, illusion-like nature.
Maitrpa finally turns his attention to understanding the specific
meaning of the negative affix a/mi in amanasikra/yid la mi byed pa.
He notes that the negative particle is here taken only in a convention-
al (tha snyad pa) sense, viz. that everything is (i) illusory (sgyu ma)
and (ii) not truly established (yang dag par grub pa med pa). There-
fore the a neither negates existence nor non-existence and cannot, on
the basis of this reasoning, be taken in a nihilistic sense. He further
explains that amanasikra
is manasikra with the primary letter a. It is a compound in which the mid-
dle word is omitted, as in the case of king [fond of] greens. In this sense,
all and any mental engagement is of the nature of non-origination, i.e., a.
27

Padma dkar po will later (seventeenth century) help to clarify this
passage (along with its grammatical example; see below section V),
contending that non-mentation is the same as proper mentation
(tshul bzhin yid la byed pa = yoniomanasikra) or prajpramit, a
positive application of the mind that reveals its natural condition of
non-origination, unimpededness and non-duality.
Maitrpa concludes his short text with a response to objections
that his interpretation of the negative prefix a in terms of non-
origination, emptiness and the like do not derive from the Buddha.
We can briefly paraphrase the authors closing arguments in terms of
the principal meanings he finds ascribed to this privative prefix in
certain tantras.
28
(a) Firstly, in defence of understanding a in terms
of non-origination, Maitrpa quotes the Hevajratantra [I.2.1] which
states that the letter a is at the beginning because all phenomena

27
AM 142,13 akrapradhno manasikra / kaprthivavat madhyapadalop sam-
sa / etena yvn manasikra sarvam anutpdtmaka< / a> ity artha / Tib. D v. 28,
p. 278.5: a yig gtso bor gyur pai yid la byed pa ni yid la mi byed pa ste / lo mai rgyal po
bzhin tshig dbus ma phyis pai bsdus pao / de gang gis ni yid la byed pa gang thams cad ni
a ste skye ba med pai don do /. The grammatical example is found in Jayakas Subho-
din commentary on the Siddhntakaumud. See Mathes (forthcoming) n. 60.
28
For a full translation of this section, see Mathes (forthcoming).

267
have been unoriginated since the very beginning
29
and the Maju-
rnmasagti [V.1c-2b] which similarly mentions non-origina-
tion as one of the senses of this ultimate letter (paramkara = yi
ge dam pa). (b) Secondly, in defence of understanding a in terms of
essencelessness, Maitrpa notes that a is the bija or seed syllable of
Nairtmy, a feminine archetype of selflessness (antman). When
the Hevajratantra [II.4.22a] states The first vowel symbolizes
Nairtmya, the a here indicates that all acts of mentation, being de-
void of self, have no abiding essence.
30
(iii) Thirdly, a can refer to
radiant clarity:
31

Moreover, a is a term for radiant clarity (prabhsvarapada = od gsal bai
tshig), and manasikra is a term for self-inspiration (svdhihna = bdag la
byin gyis rlabs). Being both a and manasikra [i.e., amanasikra is a karma-
dhraya compound], it is called amanasikra.
IV. Maitrpas legacy
Before considering some key doctrinal developments of amanasi-
kra by the Siddhas and their Tibetan interpreters, it is worth briefly

29
AM 142,45: akro mukha sarvadharmm dyanutpannatvd itydi / The Ti-
betan here simply transliterates the Sanskrit D v. 28, p. 278.6f.
30
AM 142,1213: sarvamanasikro ntmako svabhva ity ukta ca bhavati /
Tib. D v. 28, p. 279.1f.: yid la byed pa thams cad bdag med pas rang bzhin med pa /
31
AM 142,1718: yadi v / a iti prabhsvarapada / manasikra iti svdhihnapa-
dam a csau manasikra cety amanasikra / Tib. D v. 28, p. 279.2: yang na a zhes pa
ni od gsal bai tshig la / yid la byed pa ni bdag la byin gyis brlab pai tshig ste / di yang a
yang yin la yid la byed pa yang yin pas yid la mi byed pao // The term svdhihna is
important in Tantric systems such as the Pacakrama where it constitutes one of the five
stages and involves intitation by and self-identification with the deity. In the Siddha
tradition, however, this ritual initiation and consecration is internalized such that the
individual becomes directly attuned to reality without ritual mediation. Thus Maitrpas
student Sahajavajra states in his Tattvadaakak: The expression fully adorned by self-
inspiration means being inspired in oneself (bdag nyid) in terms of the self-nature (bdag
nyid) of the continuum of ones mind stream becoming attuned to the self-nature of
genuine reality itself. What emanates from the nature of suchness naturally adorns one
rang byin brlabs pas rnam brgyan pao zhes bya ba ni rang nyid gnyug mai de kho na nyid
kyi bdag nyid du byor pai sems kyi rgyun dei bdag yid du byin gyis brlabs pao // de
bzhin nyid kyi rang bzhin las phro ba rang bzhin gyis rgyan pa Cited in Mathes
(2006) n. 59. Translation is my own.
DAVID HIGGINS 268
assessing the impact that the interpretations of Kamalala and espe-
cially Maitrpa exerted on subsequent bKa brgyud authors who
continued defending the term against possible misrepresentation.
Situ bsTan pai nyin byed (seventeenth century), for example, up-
holds Maitrpas elucidations of amanasikra against those who had
attempted to realign it with the instantaneist doctrine of Hva shang:
32

In the context of this statement Freedom from mentation (yid byed bral ba)
is Mahmudr [from stanza 19 of Rang byung rdo rjes Phyag chen smon
lam], some people supposed to be analytical (dpyod ldan) think that explain-
ing absence of mentation as being Mahmudr is absurd (ha cang thal ches).
But [we reply] there is no fault. [The reason is as follows:] The letter a
occurring in the Sanskrit term amanasikra conveys the meaning of
selflessness (bdag med) or non-origination (skye med) and the like, i.e.,
the emptiness (stong pa nyid) that transcends all discursive elaborations. The
remaining letters convey the idea of mentation divested of mentation (yid la
byed dang bral bai yid la byed pa) where there is no attachment (zhen) even
to emptiness itself, thus [amanasikra] is established as the Mahmudr of
integration free from the extremes [of eternalism and nihilism] (mtha bral
zung jug gi phyag rgya chen po).
Stated otherwise, yid la mi byed pa is a non-reifying understanding
of emptiness in which even the tendency to identify with emptiness
is abandoned. bsTan pai nyin byeds seemingly contradictory defi-
nition of the component manasikra as mentation divested of men-
tation (yid la byed dang bral bai yid la byed pa) points to a charac-
teristic ambivalence among the defenders of Maitrpas doctrine. On
the one hand, manasikra refers to a positive application of the mind
(akin to such terms as praj, dharmapravicaya and bhtapratyave-
k). On the other hand, it refers to a negative, egocentric operation

32
Nges don phyag rgya chen po smon lam gyi grel pa, fol. 34a1: yid la byed bral ba
di ni phyag rgya che / zhes pa dii skabs su dpyod ldan du zhal gyis ches pa kha cig yid
byed dang bral ba phyag rgya chen por chad pa ni ha cang thal ches so snyam du dgongs
mod kyi skyon med de / legs sbyar gyi skad du a ma na si ka ra zhes byung bai a yig gis
bdag med dang skye med la sogs pa spros pa thams cad las das pai stong pa nyid kyi don
ston la / yi ge lhag ma rnams kyis stong pa de nyid laang zhen pa med par yid byed dang
bral bai yid la byed pa ston pas mtha bral zung jug gi phyag rgya chen por grub pa yin
te //

269
of mind that hinders spiritual awakening (akin to the negative
application of terms such as vikalpa, mati, and smti).
V. Padma dkar pos defence
A more elaborate defence of the term is offered by Padma dkar po in
his Phyag chen gan mdzod.
33
There he delineates three conceptions
of amanasikra that represent the authentic interpretation of Maitr-
pas Yid la mi byed pai chos skor, a cycle of non-gradual Mah-
mudr teachings considered valid and important by bKa brgyud
traditions generally. Against this background, we may paraphrase his
summary:
(i) First, amanasikra is considered in terms of the locative case used with
reference to mind [indicated by the i of manasi/la of yid la]. Since the
locative case refers to a locus or founding basis, the privative a of amana-
sikra implies the absence of a locus or founding basis for any intentional
object (gang du dmigs pai gnas sam rten gzhi med pa). This is the view
expressed in the Savarodaya.
34
Since the sense of having such a locus is
refuted by the initial letter a, we speak of amanasikra. While this firm
mental fixation which holds tightly to its intentional object through a
mode of apprehension characteristic of the manasikra included in the
five omnipresent mental events is deemed necessary in the context of
establishing an ordinary calm abiding (thun mong gi zhi gnas), it is re-
futed here [in Padma dkar pos tradition] .
(ii) Second, when amanasikra is rendered as a tatpurua compound, this
manaskra/yid byed pa [in which -i/la is not present] seems to be con-
strued as a genitive form of the tatpurua and is glossed as yid kyi las =
yid kyi byed [i.e., activity of the mind], it is the activity of the mind which
is here claimed to stand in need of refutation. More precisely, it is the
arduous application belonging to the mode of apprehension of a grasping
intentionality (sems pa) among the five mental events that is refuted.
The mental event of intentionality refers to mental activity which builds
up conditioning factors in mind and has the function of setting the mind
on wholesome, unwholesome or indeterminate actions. Such conditioning

33
For the full passage from Phyag chen gan mdzod, v. 21, p. 38.5f. (paraphrased
above), see Appendix.
34
P v. 52, n. 2230.
DAVID HIGGINS 270
factors are refuted. While they may establish a calm abiding, Mahmudr
is beyond such activities and has nothing to do with karma that is built up.
Thus, he concludes, everything that functions as mental activity, having
been established in conjunction with an intentional object here it is
shown definitively that all [such] intentional objects are completely
stilled.
(iii) Third, the prefix (yar bcad) a in amanasikra may be understood in the
sense of non-origination (anutpda = skye ba med pa): Manasikra is ex-
plained as yid la byed pa; in this case the meaning of the letter a is that of
proper mentation (tshul bzhin du yid la byed pa) or taking things as
they are
35
and may be termed a yid la byed pa, bringing to mind a [non-
origination]. In that case, the intermediary mi does not figure just as [in
the case of] the king fond of greens (lo mai rgyal po) whose name is
[truncated to] king greens. In this case, any mental engagements (mana-
sikra) are of the nature of non-origination. Thus, Padma dkar po con-
cludes, a refers to the transcending function of discerning insight (shes
rab kyi pha rol tu phyin pa = prajpramit). By way of such expres-
sions as non-origination, unimpededness and so forth, such insight is
able to reveal all the teachings of non-duality. As the commentary on the
Majurnmasagti indicates, in the Mantrayna approach, non-
duality refers to the intrinsic essence of supreme bliss (bde ba chen poi
ngo bo nyid) in which there is no duality between discerning insight
(praj) and skilfull means (upya). In the Pramityna approach, all
dualities between subject and object, knowing and the known, I and
mine, which persist so long as there are the vacillations of ego-mind (ji
srid yid kyi rnam par g.yo ba), become resolved into non-duality. The lack
of individuating principle (bdag) in all phenomena devoid of discursive
elaborations is the intrinsic nature of non-duality. Since amanasikra is
revealed by meanings such as those [summarized above], we speak of the
Cycle of Teachings on Non-Mentation, teachings that were composed
by the great teacher Maitrpa, alias Advayavajra.
To summarize, amanasikra can be understood in any of three ways,
depending on how its grammar is construed: (i) no engagement in
mind [locative], (ii) no engagement of mind [genetive], or (iii)

35
In Phyag chen gan mdzod (p. 280.1), Padma dkar po quotes a stra entitled Sangs
rgyas yang dag par sdud pai mdo (not included in catalogues to the Peking or Derge edi-
tions) in which appropriate is given as the meaning of non-origination (skye ba med pa
ni tshul bzhin no).

271
proper mental engagement in the sense of prajpramit. Of parti-
cular interest in Padma dkar pos account is his equation of amana-
sikra with proper mentation (tshul bzhin yid la byed pa = yonioma-
nasikra), an equation he returns to frequently in his Phyag chen gan
mdzod and other works.
36
In a later passage of the Phyag chen gan
mdzod (101.6 f) which glosses two quotations from the Hevajratan-
tra, Padma dkar po describes proper mentation as the ever-present
primordial gnosis (elsewhere specified as pure awareness: rig pa or
even rig pa chen po, 265.2) that reveals itself through the purifying
(dag) of improper mentation (tshul bzhin ma yin pai yid la byed pa).
This improper mentation is the source of mental agitation or distract-
edness (yengs pa) that arises owing to the power of beginningless
latent tendencies (thog ma med pai bag chags kyi mthu las byung).
Mentation is a state of ignorance (ma rig pa) which, as mental agita-
tion, constitutes a deviation away from proper mentation (tshul bzhin
yid la byed pa las gzhan du phyogs pa). The retrieval of proper
mentation involves firmly deciding between pure awareness and
ignorance (rig ma rig gnyis su kha tshon gcod), the roots of nirvana
and samsara respectively.
Now the terms yoniomanasikra and its counterpart ayonioma-
nasikra have a long history in Buddhist thought, their earliest
deployment being found in the Pali canon.
37
Padma dkar pos spe-

36
For example, Phyag chen gan mdzod, pages 38.2, 102.2, 116.2, 265.3, 271.3, 280.1,
287.6 and Klan ka gzhom pai gtam bzhugs, v. 21, p. 559.3.
37
On yoniomanasikra (Pli yonisomanasikra), see Dghanikya III 227, Vibhaga
373. On ayoniomanasikra (Pli ayonisomanasikra), see Dighanikaya III 273, Vibha-
gtthakatha 148. The two terms are also found in certain works attributed to Asaga/Mai-
treya; yoniomanasikra: Mahynastrlakra I 16; ayoniomanasikra: Mahyna-
strlakra XI 4; Uttaratantra I 556, 589. Padma dkar pos understanding of improper
mentation as an adventitious distortion of appropriate mentation or pure awareness closely
resembles Uttaratantra I 56 where improper mentation, the basis of adventitious karma
and kleas, is said to be itself based entirely on the clarity of Mind (sems kyi dag pa la rab
gnas).
Where the term amanasikra occurs in Asaga/Maitreyas works, it is generally given a
negative valuation. In Mahynastrlakra XI 3, amanasikra and ayoniomanasikra
are both included in a list of sixteen obstacles to the accumulation of merit and knowledge.
In Dharmadharmatvibhga (p. 37 f), amanasikra is included in a list of five impedi-
ments (pratipaka) to non-conceptual primordial gnosis (avikalpajna) whereas manasi-

DAVID HIGGINS 272
cific (and decidedly Tantric) usage of a standard Buddhist term as a
synonym for amanasikra invites comparison with the synonyms in-
voked by both Kamalala genuine conceptual analysis (yang
dag par so sor rtog pa = bhtapratyavek)
38
and Maitrpa genu-
inely valid cognition (yang dag par rig pa). What these terms com-
monly imply is a mode of cognizing things that is not channelled
through the dualistic categories of representational thinking, the cru-
cial difference being whether this cognition is arrived at as the
fruition (phala) of a long drawn-out process of moral refinement and
elimination of hypostases through analytical investigation (Kamala-
la) or disclosed in its originary condition through a more radical
clearing of dualistic tendencies (Maitrpa and Padma dkar po).
VI. sGam po pa on manasikra in Madhyamaka and Mantra-
yna
The attempt made by various Indian and bKa brgyud authors to
equate amanasikra with a positive appraisal of manasikra may be
viewed in the light of their more general concern with legitimizing
controversial ideas by showing their continuity with Mahyna and
Mantrayna modes of discourse. A case in point is an analysis of
manasikra by sGam po pa bSod nams rin chen (10791153) in
which he compares Madhyamaka and Mantrayna conceptions in or-
der to show that both lead to goal-realization, albeit in different
ways.
39
The Madhyamaka involves bringing to mind the general idea
(don spyi yid la byed pa) of reality such that the negation (bkag) of

kra provides access to non-conceptual primordial gnosis. See, however, Madhyntavi-
bhga V 12 and Vtti where manasikra, understood as the belief in an I or mine
(ahakti = ngar dzin), is presented as one of six types of distractedness (vikepa) and is
said to be characteristic of the narrow-mindedness (chung ngui sems) resulting from the
application of manasikra in the Hnayna. In a similar vein, Abhisamaylakra V 289
(together with Haribhadras Sphurth) specifies not setting ones mind (amanasikra) on
the quintessence of awakening (bodhigarbha) as characteristic of the manasikra of the
rvaka family (gotra) amongst the Hnayna.
38
In Phyag chen gan mdzod, p. 116.2, Padma dkar po equates yid la byed pa with so
sor rtog pa.
39
Zhal gyi bdud brtsi thun mongs (sic!) ma yin pa, v. DZA, fol. 7b1f.

273
the four extremes (mtha bzhi) pertaining to the conventional (kun
rdzob) reality namely, existence, nonexistence, both or neither is
claimed to yield, on the ultimate level (don dam par), freedom from
discursive elaborations (spros bral) which eludes all positions (khas
len pa). Thus the Prajpramit [i.e., Prajpramithdaya] states
that although from the conventional standpoint, form is emptiness
and emptiness is form, these being non-dual, from the ultimate
perspective, no entities are perceived (mi dmigs par) at all. The
Mantrayna involves bringing to mind the actual reality (don dngos
yid la byed pa), as illustrated by Tilopa: Alas, self-awareness (rang
gi rig pa) is primordial gnosis of suchness (de kho na nyid kyi ye
shes); it is nothing whatsoever that could be shown by me! In this
case, states sGam po pa, view (lta ba) is devoid of any duality; felt
experience (nyams myong) flows unobstructedly (rgyun chad med);
cultivation (bsgom pa) does not succumb to bias (phyogs su ma lhung
pa); conduct (spyod pa) is free from acceptance or rejection (dgag
bsgrub dang bral ba); and the goal is free from hope and fear (re
dogs bral ba).
VII. amanasikra as description and explanation
The various interpretations of amanasikra formulated in response
to criticisms raised at the bSam yas debate or in Indian circles should
not allow us to overlook the fact that the idea first gained currency as
a description of goal-realization in contexts of a mystical, rather than
polemical, nature. Within the Siddha genre of mystical songs, ama-
nasikra is used to describe certain transformative experiences and
modes of being, characterized by ecstacy, radiant clarity and loss of
the customary sense of self, rather than to define or defend a particu-
lar line of Buddhist thought. This contrast can perhaps be best under-
stood in terms of the distinction between a description and an
explanation. A description is the immediate articulation, verbally
and conceptually, of an actual experience one is having. Where the
available words and concepts in their possible semantic combina-
tions are inadequate to convey the experience, new words or new nu-
ances of old words are coined. Explanation is any concept or theory
that attempts to go behind an experience or description of an experi-
DAVID HIGGINS 274
ence, to account for it or legitimize it, in terms of something other
than what is experienced, and often for purposes extraneous to
experience, be they doctrinal or tactical.
40
It would be wrong to make
this distinction a categorical one. Every description, by virtue of its
communicative thrust and its participation in the system of associ-
ated commonplaces (Max Black) surrounding every word, carries
with it the sedimentation of explanatory rules. Likewise, every
explanation, by virtue of its rootedness in the referential ground of
language, reveals something about the world.
While it is true that the descriptive and explanatory uses of
amanasikra have tended to overlap in the history of its usage so that
it could come to function simultaneously in soteriological and doc-
trinal contexts, the tradition of the Siddhas in which the term was
first widely used paid curiously little attention to sectarian issues of
doctrinal affiliation, even when adopting or reinterpreting standard
Buddhist concepts. A matter of greater importance to them was how
to actually convey an extraordinary experience by means of a me-
dium that almost invariably falsifies, distorts or conceals what it tries
to express. This factor accounts for the Siddhas frequent use of
unconventional and antinomian uses of standard terms, one obvious
example being the recurrent espousal of selfhood (tmya = bdag
nyid) in flagrant opposition to the central Buddhist doctrine of self-
lessness (antman). In this light, it is not hard to see how amanasi-
kra could constitute a semantic inversion of the notion of manasi-
kra, an idea which was generally favoured in the Buddhist tradition
and endorsed by the Buddha himself who frequently admonished his
disciples to apply their minds (Pali: manasi karotha) to his teachings.
It will become clear from the Siddha elucidations of amanasikra
examined below (section IX) that the later attempts to legitimize and
domesticate the concept by bringing it into line with such standard
notions as manasikra and ayoniomanasikra by no means reflect

40
Schmithausen (1981: 200) draws a similar distinction between an immediate
verbalization of an actual experience and the secondary transformation of such a pri-
mary verbalization effected for logical, doctrinal or even tactical reasons.

275
the intentions of those who first popularized the term as an apophatic
description of Buddhahood.
VIII. The Siddha conception of manas
Central to Tibetan developments of the amanasikra doctrine were
the varying conceptions of mind (manas) elaborated in the Indian
Buddhist philosophical systems. The Yogcra-Cittamtra concep-
tion of ego-mind
41
(manas) as a bipartite Janus-faced process of
constitution reflexively constituting itself as a self (I) by looking
back upon the horizon from which it emerges while intentionally
constituting its world by looking outward in the act of structuring its
sensory-ideational information into an intelligible unity and the
possibility of reversing (parvtti) this process, had particularly far-
reaching consequences for later Tibetan interpretations of amanasi-
kra. One such interpretation will be discussed below in section XII.
At this juncture, however, we will briefly consider another concep-
tion of manas which is of immediate relevance to the Siddha
interpretations we will be examining. Padma dkar po draws our
attention to a specifically Tantric and Siddha use of manas
42
which
diverges from both the Abhidharma conception of the mental faculty

41
This conception is traditionally distinguished from the Abhidharma-based rvaka
(Vaibhika and Sautrantika) interpretation of manas, as summarized, for example, by
mChims ston Blo bzang grags pa (probably fourteenth century) in his commentary on the
Abhidharmakoa (Chos mngon pa gsal byed legs par bshad pai rgya mtsho, fol. 27a2f.):
The two rvaka schools [Vaibhikas and Sautrntikas] maintain a six-fold pattern of
conscious operations (rnam shes tshogs drug). The two crya brothers [Vasubandhu
and Asaga], however, maintain an eight-fold pattern [which expands the six to in-
clude]: (1) a substratum consciousness (kun gzhii rnam par shes pa) which by objecti-
fying the totality of external and internal, phenomenal and individual, referents, re-
mains [itself] qualitatively non-distinctive (rnam pa mi gsal) and uninterrupted (ma
chad pa); and (2) an emotionally tainted ego-mind which, by objectifying this [sub-
stratum consciousness], has the aspect of believing in an I (ngar dzin pai rnam pa
can).
Whatever its philosophical limitations as an idealist construct, the Cittamtra eightfold
model of mind provided a fruitful and highly influential conceptual scheme for elucidating
the genesis, and possible transcendence, of dualistic experience.
42
Phyag chen gan mdzod, 270.3f.
DAVID HIGGINS 276
(yid dbang shes) that becomes apparent when the six modes of
perception cease and the Yogcra-Cittamtra conception of an I
and mine intending process (nga dang ngar sems pa). These latter,
Padma dkar po argues, fail to describe the mind of yogis (rnal byor
pai yid) who have gone to the heart of meditative composure
(mnyam bzhag gi ngo bor song ba). For such individuals, mental
activity (yid kyi las), by which we can understand thinking, feeling
and conceptualization in general, continues in various forms on the
level of subjective mind, but is now fed, so to say, from a higher
source. The Siddhas therefore use certain terms such as genuine
mind (gnyug mai yid) or mind devoid of objective reference (dmigs
pa med pai yid) to describe this transformed or ex-centric subjectiv-
ity. As Padma dkar po goes on to elaborate:
43

Next, if one has to purify the whole complex of mind involving active
dichotomic cognition (rnam shes bya ba dang bcas pai yid) at the time of
settling in the realization of genuine mind (gnyug mai yid), then it [is mind
(manas) which] serves to demarcate realization from non-realization. Here
in this context of calm abiding (zhi gnas), the type of mental engagement
that is to be abandoned (yid byed spang rgyu) consists in the improper forms
of mental engagement.
Mind thus emerges within the field of consciousness as the pivotal
factor from which the dual inclinations toward realization [direct
self-recognition] or non-realization [non-recognition] proceed.
IX. Siddha interpretations of amanasikra
The idea of non-mentation as it is developed in certain mystical
songs attributed to Saraha and Tilopa, serves as one among many
negative descriptors of an experience considered so rich and pro-
found as to defy expression in thought and language. It is akin to,
and frequently occurs alongside, a variety of other negative descrip-
tors such as transcending intellect (blo las das pa), devoid of

43
Phyag chen gan mdzod, 271.2f.: de las gnyug mai yid bsgrub pa la bzhag pai tshe
rnam shes bya ba dang bcas pai yid de thams cad dag pa dgos pa na rtogs ma rtogs kyi
mtshams byed pa la yin no / zhi gnas kyi skabs dir yid byed spang rgyu ni tshul bzhin ma
yin pai yid la byed pa de dag go //

277
representational thinking (bsam med, bsam du med pa), devoid of
subjective grasping (dzin med), free from discursive elaborations
(spros bral) devoid of meditation (sgom med, bsgom du med pa),
terms which commonly refer to a depth dimension of experience that
eludes the appropriations of dualistic mentation. It is of interest to
note that Saraha elaborates on the term in a relatively small number
of his many Dohs. The term scarcely occurs in his famous and
much commented upon Doh Trilogy (Doha skor gsum),
44
for exam-
ple, or in his Kakhasyadoh
45
(for which we have his Tippana), or in
the songs attributed to him that are included in Munidattas Caryg-
tivinicaya.
46
It is a central theme, however, in his Vajragti Quartet
comprising the Kyakomtavajragti, Vkkoarucirasvarajagti,
Cittakojavajragti, and Kyavccittmanasikranma,
47
as well as
in his Mahmudropadea.
48
It is unfortunate that the term is given
detailed attention in works for which we have no Indian or Tibetan
commentaries. In the case of Tilopa, the term occurs in several of his
mystical songs on Mahmudr including the Dohkoa, Acintyama-
hmudr, and Mahmudropadea.
49

At the outset, we must note a certain ambivalence in Sarahas use
of amanasikra and its variants. When regarded as a polemical posi-
tion, Saraha is as quick to criticize its (anti-intellectualist) supporters
as its (intellectualist) detractors. Early in the Kyakoa we find
Saraha critically assessing various Buddhist and non-Buddhist ap-

44
See Guenther (1993). A Hindi translation and transliteration of the trilogy has been
made by R. Sktyyan (1957) 38f., 83f. and 99f. respectively.
45
P v. 69, nos. 3113 and 3114 respectively. A Hindi translation and transliteration
from the Tibetan Kakhasya doh (but not ippaa) is given in Sktyyan (1957) 127f.
46
See Kvaerne (1977) songs 22, 32, 38, and 39.
47
P v. 69, nos. 31143118. In Sktyyan (1957) 141f., 185f., 203f. and 215f.
respectively.
48
P v. 69, no. 3119. In Sktyyan (1957) 249f.
49
P v. 69, no. 3128, vol 82, no. 4635, and v. 69, no. 3132 respectively. Even a cursory
comparison of existing Apabhraa and Tibetan versions of the Dohs reveals consider-
able differences in content and arrangement and indicates that various recensions of the
songs must have been in circulation. See R. Jackson (2004) 48f.
DAVID HIGGINS 278
proaches to realization, among which he includes the tendency to
find fault with either yid la mi bya ba or its opposite yid la bya ba:
50

Though true understanding varies with knowledge interests,
In non-reflection (dran med),
51
there has never been anything false.
Though the goal varies according to efforts on the path,
In reflection (dran pa), there has never been anything true.
Though each ones meditative fixation varies with his state of indifference,
In non-origination (skye med), there has never been any duality.
Though people ascribe fault to either mentation or non-mentation,
In transcending intellect (blo das), there has never been anything to search
for.
Sarahas critique of the opposing positions regarding amanasikra
provide further evidence that the type of contentious issues staged at
bSam yas during the Sino-Indian controversy were prevalent in India
(though dates are less certain). What is of particular interest, how-
ever, is Sarahas expos of the vain purposiveness common to both
intellectualist and anti-intellectualist positions, the former disparag-
ing non-mentation, the latter disparaging mentation. Both fail to go
beyond intellectual deliberation.
Tilopa begins a song entitled Dohkoa by admonishing his
listeners not to reduce amanasikra to popular prejudices or to
subjective demands:
52

50
P v. 69, p. 103.4.3; D v. 28, p. 196, 213.3.
mos pai shes pas rtogs pa tha dad kyang /
dran med di la brdzun pa yod re skan /
lam gyi rtsol bas bras bu so so yang /
dran pa di la bden pa yod re skan /
btang snyoms dbang gis re jog tha dad kyang //
51
On the four symbol (brda bzhi) terms dran pa, dran med, skye med, and blo das see
Guenther (1969) 11f. and 14. I have somewhat loosely rendered dran pa (smti) as reflec-
tion to cover the two basic cognitive operations it describes, namely, memory (mnemic
reflection) and attention (thematic reflection), both involving reflexivity, a reflecting on
experience. See Padma dkar pos cogent analysis of the concept as it relates to Mahmudr
teachings in his Phyag chen gan mdzod, 271.4f.
52
P v. 69, 131.2.3; D v. 28, p. 204, 271.6:
yid la ma byed gnyug mai rang bzhin la /
brdzun pa rnams kyis skur ba ma debs shig /
rang dbang yod pas rang nyid ching ma byed //

279
Non-mentation, [ones] genuine nature
Do not disparage it by way of lies!
Since it is present by its own power, do not hold it as ones own self!
A recurrent paradox in Siddha descriptions of goal-realization is that
what is most natural comes most naturally precisely when egocen-
tred identifications subside. Herein lies the basis for the Tantric
distinction between the sense of self (bdag) which, as a subjective
construct, is amenable to deconstruction, as in the Buddhist no self
doctrine (antman = bdag med), and authentic selfhood voiced in the
first person (bdag nyid) which is precisely what the Siddhas mean by
non-mentation. As Tilopa states it (Dohkoa):
53

The natural expression of stainless mind
Should be known by anyone [in] self-awareness (rang rig).
I myself am living being, I myself am Buddha.
I myself am non-mentation (bdag nyid yid la mi byed pa).
The term self-awareness (rang rig or rang gi[s] rig pa), as elaborated
by Saraha, describes the experience of recognizing spontaneous
ever-present gnosis as it is, stripped of all subjectivizing and
objectifying tendencies. When fully recognized, it is open awareness
(rig pa). When not recognized, it is ignorance or un-awareness (ma
rig pa), a process-product term which at once describes the non-
recognition of ones natural condition (not understanding it as it is)
and the ensuing mis-apprehension (taking it for something it is not,
viz. a self). Saraha begins his Cittakoa with the following pas-
sage:
54

53
P v. 69, 131.3.4; D v. 28, p. 204, 272.7:
dri med sems kyi rang bzhin la /
gang zhig rang rig shes par bya /
bdag nyid gro ba bdag nyid sangs rgyas te /
bdag nyid dri ma med cing bdag nyid yid la mi byed pa //
54
P v. 69, p. 107.3.3; D v. 28, p. 198, 230.5:
skye bo lhan cig skyes pai ye shes ni /
rang gi nyams su myong ba de kho na /
rig dang ma rig rang rig gsal ba de kho na /
mar me mun gsal rang gi rang gsal rang la sad /
dam gyi padma dam la ma zhen kha dog legs /
gzung dzin dri ma ma spangs snying po gsal /

DAVID HIGGINS 280
Spontaneous gnosis [in each] person
That alone is felt and experienced individually.
That alone is the radiance of self-awareness in awareness and un-awareness,
A darkness-illumining light, intrinsically self-luminous, aroused in oneself.
A lotus in a swamp resplendent in colours, unconditioned by the swamp,
Ones vital quintessence shines without removing the grime of subject and
object.
As a deer living in an alpine forest wanders alone,
That alone is goal-realization, unconditioned by causality.
Like the lumen naturale (natural light) of the medieval scholastics,
self-awareness comes to light naturally for the very reason that self-
illumination and self-disclosure are intrinsic to being human.
55
That
our vital quintessence (snying po) can shine forth without having to
remove the accretions of subject and object is only possible because
humans are always already pervaded by spontaneous gnosis (lhan
cig skyes pai ye shes).
56

The possibility of existential recovery would be of little relevance
if the tendency to glide away or go astray (khrul pa) from authentic
possibilities into a world of appropriated objects (yul) were not like-

nags khrod gnas pai ri dags gcig pur rgyu /
rgyu la ma zhen bras bu de kho na //
55
See Heideggers discussion of lumen naturale in Being and Time, pp. 133 and 170
and in several essays such as Moira, in Early Greek Thinking (1975: 97), and The End
of Philosophy and the Task of Thinking, in Basic Writings (1977: 286). In Moira (p.
97), Heidegger writes: The essence of aletheia [i.e., truth as unconcealedness] remains
veiled. The visibility it bestows allows the presencing of what is present to arise as outer
appearance [Aussehen] (eidos) and aspect [Gesicht] (idea). Consequently, the perceptual
relation to the presencing of what is present is defined as seeing (eidenai). Stamped with
this character of vision, knowledge and the evidence of knowledge cannot renounce their
essential derivation from luminous disclosure, even where truth has been transformed into
the certainty of self-consciousness. Lumen naturale, natural light, i.e., the illumination of
reason, already presupposes the disclosure of the duality [i.e., of the presencing of what is
present]. The same holds true of the Augustinian and medieval views of light not to
mention their Platonic origins which could only develop under the tutelage of an
Aletheia already reigning in the destiny of the duality."
56
According to sGam po pa bsod nams rin chen, spontaneous primordial gnosis
(lhan cig skyes pai ye shes) refers to ones natural awareness in the present moment (da
ltar gyi tha mal gyi shes pa) as it is primordially present (ye nas yod pa). In Zhal gyi bdud
rtsi thun mongs (sic!) ma yin pa, v. dza, fol. 7a2.

281
wise constitutive of the kinds of being (gro drug) we are. According
to a beautiful passage from Sarahas Kyakoa quoted below, going
astray is engendered by mnemic and thematic reflection (dran pa,
see n. 51) which mentally and subjectively schematizes its content
into particular appearances (snang ba). By this is meant not that
these reflected-on appearances (dran pai snang ba) conceal an origi-
nal something of which they are a mere copy or distortion but that
perception is normally reduced to certain aspects of what is per-
ceived as dictated by subjective demands (things appear as this or
that in accordance with this or that purpose). Just as the true enjoy-
ment of what life has to offer is only possible when one stops look-
ing for it, so authentic subjectivity, as a process of individuation,
57

can only blossom in the absence of subjective deliberations. The
Kyakoa states:
58

57
The term individuation was coined by Carl Jung to describe the process of becom-
ing a complete individual who is aware of his or her unique individuality. It refers specifi-
cally to the gradual opening of an individuals consciousness to the complete range of
possible elements that are already inherent in the individual at a preconscious level. The
Siddhas seemed well aware that the opening of ego-mind (manas) to non-egocentricity
(amanasikra), not unlike the conscious egos integration of unconscious elements in
Jungian psychology, does not lead to the annihilation of the subject (bdag med) but rather
to its fulfillment in authentic selfhood (bdag nyid).
58
P v. 69, p. 104.3.8; D v. 28, p. 196, 217.5:
gang la mi gnas chags pa med par spyod /
me tog sbrang rtsi sbrang mas thung dang dra /
so sor rtog pai ye shes thabs yin te /
ro dang phrad na ro la zhen pa med /
de ltar kun gyis shes par gyur ma yin /
snying poi don gyi gro drug khyab mod kyang /
gro ba dran pas bcings te pad trai srin /
sems las dran pa byung phyir khrul pai rgyu /
yid la mi byed shes na sangs rgyas nyid /
khrul pa de la thabs dang shes rab med /
kye ho dbyer med shes na thabs mchog de kho na /
sangs rgyas sems can chos rnams thams cad kun /
rang gi sems nyid dag dang lhan cig skyes /
yid la mi byed yid la skyes tsam na /
dran pai snang ba nub ste bden brdzun med /
de phyir de nyid kho nai yul ma yin /
dper na mig gi yul du sgra mi snang /
rnam par mi rtog rtog pai yul ma yin /

DAVID HIGGINS 282
Like a honey bee sipping honey from flowers,
Not staying anywhere [but] enjoying [each] without attachment,
So individually-conceiving gnosis is the skillful means
[To] encounter [aesthetic] flavours without attachment to [any] flavour.
It is not known by all in this way.
Though lifes vital quintessence pervades the six kinds of beings,
Beings become imprisoned by reflections [like] insects by a flowers petals.
Because mind-based reflections arise, they become the cause of errancy.
If one recognizes non-mentation, this is Buddhahood.
[But] in this errancy, there is neither skillful means nor insight.
Alas! If one knows their indivisibility, that alone is the highest means!
Buddhas, sentient beings, the whole phenomenal [world],
Arise together with ones own pure Mind itself.
At the time when non-mentation emerges in mind,
Reflections appearances vanish so nothing of true or false remains.
Therefore [non-mentation] is not an object for itself,
Just as sound does not appear as an object for seeing.
Non-conceptuality is not a conceptual object.
But when reflections become illumined by the condition of emptiness,
Reflections appearances vanish and there is no more looking.
This lengthy passage helps us to understand how amanasikra is able
to function in the Siddha mystical songs as a negative description of
a positive experience. The apophatic description serves as a counter-
tendency to any assertive claims about the experience, be they
epistemological (the attempt to reduce it to an object of knowledge),
ethical-axiological (to evaluate it as good as opposed to evil) or
ontological (to assert it as real as opposed to unreal). According to
the Kyakoa:
59

Where there is no egocentric deliberation, [that] is Mahmudr.

stong pai rkyen gyis dran pa gsal tsam na /
dran pai snang ba nub nas mthong ba med //
59
P v. 69, p. 105.2.6; D v. 28, p. 221.1:
gang la yid la byar med phyag rgya che /
mtshan mai dran rig sna tshogs ji snyed pa /
de nyid phyag rgya che la dbye ba med /
rtogs dang mi rtogs gnyi ga so so min //

283
However varied awarenesss reflections on its [self-imputed] characteristics,
These are indivisible with Mahmudr.
Conceptualization and non-conceptualization are not two separate things.
If apophatic descriptions point to what is ineffable in experience,
they are nonetheless frequently used alongside more positive (cata-
phatic) descriptions such as radiant clarity (od gsal) or ultimate
reality (chos nyid), expressions that emphasize the positive, fecund
character of mystical experience. A clear example of the mixing of
the apophatic and cataphatic orders of discourse is found in the open-
ing verse of Tilopas Acintyamahmudr:
60

Radiant clarity, ultimate reality unborn and unceasing is
The way of Mahmudr, inexpressible in thought and language.
It is non-mentation, beyond what can be identified.
Homage to the unconditioned, profound and calm!
The peculiarities of negative description have enabled us to see how
amanasikra could paradoxically serve as a primary description, and
later as a key hermeneutical definition, of the utterly positive experi-
ence known as Mahmudr. But given that amanasikra is pre-
cisely that which eludes positive assertions, how then is it to be culti-
vated (sgom)? Sarahas answer is, as we could expect, negative: The
best cultivation (sgom mchog) is that which comes of its own accord
in the absence of any willful cultivation. As Saraha states in his
Dohkoanmamahmudropadea:
61

In Mahmudr which is non-mentation,
Since there is not the slightest reason to meditate, theres no meditation.
Without meditating or being divorced from what its is about is the best
meditation!

60
P v. 82, n. 4635, p. 38.5.4 f; D v. 28, 490.4:
skye gag med pai chos nyid od gsal ni /
smra bsam brjod med phyag rgya chen poi lam /
ngos gzung dang bral yid la mi byed pao /
zab zhi dus ma byas la phyag tshal lo //
61
P 69, 110.5.3; D v. 28, p. 246.3:
yid la mi byed phyag rgya chen po la /
sgom rgyu rdul tsam med pas mi bsgom ste /
sgom med don pa bral med sgom pai mchog //
DAVID HIGGINS 284
The theme is elaborated in another passage from the authors Doh-
koanmamahmudropadea:
62

Mind cut off at its root is like the open sky.
There being nothing to meditate on, theres no mental engagement because
Ordinary awareness, perfectly natural in its own way of being,
Is not deceived by artificial thought objects.
There is no need to fake this naturally pure mind.
So, without holding or dismissing it, leave it where it is most happy!
As Ba ra ba rGyal mtshan dpal bzang (13101391) notes in con-
nection with these two preceding passages,
63
the term meditation
(sgom) is to be understood in the sense of preserving the Mah-
mudr experience (phyag rgya chen po skyong ba) or non-menta-
tion (yid la mi byed pa), terms which similarly indicate that there
has not for a moment been any cause to interfere with it (de nyid la
yengs rgyu skad cig kyang med pa). The term ordinary awareness
(tha mal [gyi] shes pa) also requires explication. According to
Padma dkar po (mDzub tshugs & Phyag chen gan mdzod),
64
it is syn-
onymous with natural awareness (rang bzhin gyi shes pa), tha mal
and rang bzhin both being translations of the Sanskrit prakta. In

62
P v. 69, p. 110.4.7; D v. 28, p. 245.7:
rtsa ba chod pai sems nyid nam mkha dra /
sgom du med pas yid la mi bya ste /
tha mal shes pa rang lugs gnyug ma la /
bcos mai dmigs pa dag gis mi bslad de /
rang bzhin dag pai sems la bcos mi dgos /
ma bzung ma btang rang dga nyid du zhog //
63
Ngo sprod bdun mai grel pa Man ngag rin po chei sgron me, v. 11, p. 234.1f.
Yang dgon pas Ngo sprod bdun ma is found in the Pha jo ldings edition of the authors
bKa bum (mGur bum collection, v. 3, p. 274.1). The song is not contained in the rTa
mgo edition which has a much shorter mGur bum but is otherwise identical in contents.
Ba ra ba was regarded as a re-embodiment of Yang dgon pa and was heir to his lineage,
the sTod Brug, through his own root teacher Zur phug pa Rin chen dpal bzang, a student
of Yang dgon pas leading disciple Spyan snga rin chen ldan (12021329). See Smith
(1970: 9f.) and Deb ther sngon po, ja, fol. 127af. (BA 692f.). As Smith (p. 7) notes, The
sTod Brug gave rise to a host of important schools: the Ne-rings bKa brgyud pa, the
Mdo bo che ba, and the Yang dgon bka brgyud pa among others. The Yang dgon school
produced ultimately the Ba ra bKa brgyud pa, a sect that had maintained its identity up
to 1959.
64
rNal byor bzhi mdzub tshugs, p. 484.3f. See also Phyag chen gan mdzod, p. 103.5f.

285
Padma dkar pos language of experience (myong bai skad), it re-
fers to that free-rising cognition (thol skyes kyi rig pa) which is aris-
ing continuously (shar shar ba) but only becomes fully manifest
once the net of concepts (rtog pa) that obscures it has cleared.
65

The unwilled cultivation of non-mentation is described in the Sid-
dha texts as a process of spiritual attunement (rnal byor = yoga).
This process, according to Sarahas own definition (Kyavccittama-
nasikra),
66
is a matter of remaining attuned to ones natural condi-
tion (rnal mai don la gnas pai rnal byor). Since this occurs only in
the absence of subjective interference either accepting or rejecting
of a grasping subject, it is known as the true concentration (bsam
gtan nyid) as distinct from ordinary fixation involving dualistic ob-
ject apprehension. When all representational and objectifying think-
ing has dissolved into the single flavour of non-mentation, gnosis is
present as ones vital quintessence. As stated in Sarahas Kyavc-
cittamanasikra:
67

Without accepting or rejecting, it is naturally free in every respect.
The attunement without grasping or egocentricity is the true concentration.
Since that which cannot be cultivated as anything or sought anywhere is
Inconceivable, Alas! it is the same flavour as non-mentation.
Gnosis, inconceivable and uncontrived like the sky, ones vital quintessence:
Alas! It is nothing that can be intellectually thought about or verbalized!
Mahmudr as an absolutely positive experience is characterized
negatively during the path of recovery where its self-disclosure is
made possible by a via negativa which gradually strips away the
egoic projections and appropriations that attempt to make of it some-
thing other than it is. From the perspective of self-disclosure, how-

65
rNal byor bzhi mdzub tshugs, p. 485.1f.
66
P 69, 108.4.7.
67
P v. 69, p. 108.5.6; D v. 28, p. 199, 237.3:
btang gzhag med cing rang bzhin rnam par grol /
dzin med yid la bya med rnal byor bsam gtan nyid /
gang la mi bsgom gang duang btsal ba med pa de /
bsam du med pas yid la mi byed ro snyoms kye /
ye shes mkha dra bsam bral ma bcos snying po don /
di la blos yis bsam zhing brjod du med do kye //
DAVID HIGGINS 286
ever, where the positive stands completely in the open and is no
longer cognized as other, propositions about mentation and non-
mentation no longer apply. The Kyakoa states:
68

Settled in what is neither mentation nor non-mentation,
Since self-awareness emerges as Mahmudr itself,
Mahmudr reveals itself to itself by itself.
X. The Siddha impact on bKa brgyud pa views of amanasi-
kra
What can we conclude in this final section of the paper about the
Siddha interpretation of amanasikra and its impact on later
developments? First, it must be recognized that the idea at this
formative stage in its development had not yet been codified into a
unified, systematic doctrine, even if contemporary opponents of the
term already identified it with the Siddha teachings. Perhaps the
polyvalent significations of the term in Siddha contexts and its resis-
tance to any univocal or unequivocal definition reflects the general
tenor of the Siddha movement: its spirited disavowal of intellectual
systems of any variety and of ideological identification in general.
This aspect of the movement should be borne in mind when examin-
ing later attempts by Tibetan authors to retrospectively identify the
concept with particular schools of Buddhist thought. For example,
Mi bskyod rdo rje (15071554), the eighth Karmapa, discusses
amanasikra in his commentary on Candrakrtis Madhyamakva-
tra and Bhya
69
as if it were a Madhyamaka doctrine (yid la mi
byed pai dbu ma) of Maitrpa in which he had synthesized the
Madhyamaka teachings of Saraha (younger and elder), Ngrjuna
and Candrakrti. Although the influence of Madhyamaka on Maitr-
pas doctrine is unmistakable, one is hard pressed to identify Sara-
has teachings with the Madhyamaka (apart from certain conceptual

68
P v. 69, 105.3.2; D v. 28, p. 197, 221.4:
yid la bya dang mi bya med par gzhag /
rang rig phyag rgya chen po nyid la byung /
phyag rgya chen po nyid la nyid kyis bstan //
69
dBu ma la jug pai rnam bshad, fol. 5a4f. See also Ruegg (1988) 125.

287
correlations) given his own frequent critiques of all the Buddhist phi-
losophical systems, Madhyamaka notwithstanding. In fact, Sarahas
Kyakoa, the work in which he gives the fullest elaboration of
amanasikra, begins with such a critique:
70

The Vaibhikas, the Sautrntikas,
The Yogcras, the Madhyamakas and the rest,
Find fault with one another and engage in [endless] disputes.
Oblivious to That itself, the sky[-like] sameness of appearance and empti-
ness,
They turn their backs on spontaneity.
Despite the multiplicity of connotations in the Siddha treatment of
amanasikra, we can extract from the passages we have examined
two overlapping deployments which strongly influenced later Ti-
betan interpretations: (i) Firstly, its use in didactic and rhetorical
contexts as a critique of subject-centered rationality in intellectual
and ethical pursuits. This is evident in the two passages of Saraha
and Tilopa that were cited at the beginning of the previous section.
(ii) Secondly, its use in descriptive contexts to provide a phenomeno-
logical-psychological account of the transcendence of subject-cen-
tered mind and the recovery of non-dual gnosis. This is apparent in
the apophatic descriptions of amanasikra examined throughout the
previous section. Some examples of these influences will now be
examined.
XI. amanasikra and the critique of purposive rationality:
sGam po pa and Rang byung rdo rje
What the use of amanasikra as a counter-measure to the subjecti-
vizing and objectifying tendencies of dualistic mind principally seeks
to undermine is the purposiveness or instrumentality that surrepti-

70
P v. 69, p. 103.3.5; D v. 28, p. 212.5:
bye brag pa dang mdo sde sngags pa dang /
rnal byor pa dang dbu ma la sogs te /
gcig la gcig skyon gel zhing rtsod par byed /
snang stong mkha mnyam de nyid mi shes pa /
lhan cig skyes la rgyab kyis phyog par gyur //
DAVID HIGGINS 288
tiously controls all rational-calculative thinking. We have seen that
non-mentation (yid la mi bya) becomes no less an impediment to the
free flow of experience than mentation (yid la bya) so long as willful
deliberation is involved. For several bKa brgyud pa authors includ-
ing sGam po pa bSod nams rin chen (10791153) and the third Kar-
mapa Rang byung rdo rje (12841339), Mahmudr is precisely
what comes to the fore in the absence of instrumental rational or
moral deliberation. sGam po pa, for example, defines Mahmudr as
follows:
71

Here, Mahmudr means not reducing anything to ego-mind (yid la ci
yang mi byed pa): one neither cultivates any qualities such as non-divided-
ness or emptiness nor eliminates any defects such as conceptualizing.
That which is naturally free (rang grol ba) from the intellect with its dualis-
tic beliefs is what is [meant by] Mahmudr.
The third Karmapa Rang byung rdo rje similarly disclaims the valid-
ity of a moralism which belies subjectivistic deliberations in his
commentary on Tilopas Mahmudropadea (stanza 10):
72

In the context of worldly appearances, it is commonly declared that awaken-
ing to Buddhahood is attained solely on the basis of accumulating stores of
merits. But when lifes vital quintessence is no longer reduced to ego-mind
(yid la ma byed pa), then good does not yield the slightest benefit and evil
does not bring the slightest harm. If one goes to the core of radiant clarity,
beyond all attachments and desires, and deeply understands it, then all
phenomena belonging to samsara and nirvana [in their] multiplicity assume
the single flavour of basic equality and all the masses of notions that arise in
ones ego-mind (rang gi yid), apart from becoming friends with the Dharma-
kya in its basic equality, do not become harmful.

71
Phyag rgya chen po rtsa ba la ngo sprod pa, v. YA, fol. 2b5f.
72
Phyag rgya chen po Gang ga mai grel pa, p. 42.1f. The arrangement of lines in this
version of the Dohkoa are quite different from the bsTan gyur version. jig rten pai
snang ngo la bsod nams kyi tshogs gsog pai rten ba zhig bsgrub cing di yis sangs rgya
bar byed zer yang / snying poi don yid la ma byed pa / dge bas phan spu tsam ma byas /
sdig pas gnod pa spu tsam yang ma bskyel zhing / zhen pa dang dod pa kun dang bral bai
od gsal ba de nyid khong du chud cing rtogs na ni / khor ba dang mya ngan las das pai
chos thams cad mnyam pa nyid du ma ro gcig par gyur zhing rang gi yid la byung bai
rtog pai tshogs thams cad ni / mnyam pa nyid chos kyi skui grogs su gyur pa ma gtogs pa
gnod par mi gyur ba yin te /

289
XII. amanasikra and transformed subjectivity: Yang dgon
pas Cittamtra-Mantrayna synthesis
The Siddha conception of amanasikra in terms of a goal-sustained
transformation of subject-centered consciousness was to be given its
most lucid expression in the Cittamtra-based interpretations of
amanasikra, such as we find, for example, in the writings of the
sTod brug mystic rGyal ba Yang dgon pa (12131258).
73
It is of
interest to note that Karmapa Mi bskyod rdo rje, who has traced
three distinct Madhyamaka lines of interpretation of the amanasikra
doctrine in Tibet, specifies the Cittamtra-based interpretation (more
specifically the sems tsam rnam rdzun gyi dbu ma) as the one which
follows the sense of the Dohs. This tradition, represented by Yang
dgon pa, who evidently received it from rGod tshang pa (1189
1258?),
74
is distinguished from the Mantra-Madhyamaka (sngags kyi
dbu ma) and Stra-Madhyamaka (mdoi dbu ma) interpretations
which are said to derive in their entirety from Marpa and Mila ras
pa. The Stra-Madhyamaka tradition was chiefly represented and
widely propagated by sGam po pa. The Cittamtra-Madhyamaka
interpretation, like the Dohs, emphasizes an awareness devoid of
the subject and object (gzung dzin gyis stong pai shes pa) character-
ized as self-radiant self-awareness (rang rig rang gsal).
The interweaving of Cittamtra and Siddha views regarding the
transformation of ego-mind is conspicuous in Yang dgon pas
interpretations of yid la mi byed pa as presented in certain of his
Mountain Teachings (Ri chos) texts. His most extensive account is

73
On Yang dgon pa, see above note 63.
74
dBu ma la jug pai rnam bshad, fol. 6a2f. This line is said to have been widely
represented in India and Tibet by Phyag na (Vajrapi) of India, a direct disciple of
Maitrpa. Yang dgon pa claims to have received a cycle of Mahmudr teachings from his
root teacher rGod tshang pa called the Phyag rgya chen po skor tsho that passed through
Maitrpa (Me tri pa) and Vajrapi (rgya gar Phyag na). See Ri chos kyi phyag len gsal
bai sgron me, v. 1, fol. 2b5f. On the importance of rGod tshang pa in the transmission of
Maitrpas Mahmudr teachings, as later noted by Gos lo tsa ba and Karmapa Mi bskyod
rdo rje, see David Jackson (1994) 82ff.
DAVID HIGGINS 290
given in first part of his Ri chos kyi rnal byor bzhi pa within the con-
text of a lengthy elucidation of Mahmudr:
75

The meaning of the term yid la mi byed pa [derives from] the term amanasi-
kra in the Sanskrit language. As derived from this term used in the locative
case, manas is rendered as yid, sikara is rendered as [la]byed pa and these
are negated by the a. One thus speaks of an absence of mentation (yid la
byed pa med pa) in the sense of not dwelling in ego-mind (yid la mi gnas pa),
being free from ego-mind (yid las grol ba) or transcending ego-mind (yid las
das pa). If we translate it as not dwelling in ego-mind, the meaning of the
term is easy to understand.
However [the translation] yid la mi byed pa (non-mentation; literally, the
ego-mind not focusing upon) [means that] divisive concepts (rnam rtog)
and hypostases (kun rtog) [i.e., the ego-mind] arisen from sources of error
are taken as the subject, and then [yid la mi byed pa] is these agents (de
byed pa) not doing. Given that understanding [of the term], then even when
there is non-mentation [in this sense], there will nonetheless be activity in
ones mind. The point is that however the all-pervasive substratum (kun
gzhi) and its five sensory operations arise, they are but the self-effulgence of
the conceptless and this is the intrinsic dynamics of Mahmudr. Thus,
when the emotionally-tainted ego-mind (nyon yid) gazes inwardly upon the
all-pervasive substratum, it holds it to be its self. And when the egoic con-

75
Ri chos kyi rnal byor bzhi pa Phyag rgya chen po snying poi don gyi gter mdzod, in
rGyal ba Yang dgon pa bKa bum (rTa mgo edition), v. 1, p. 247.5f.: yid la mi byed pa
zhes bya bai sgra don ni / sa tri [sic!] tai skad du / a ma na si ka ra la zhes bya ste / yi
ge bdun po dei sgra las drangs na / a ma na yid / si ka na [sic!] byed pa yin la de as bkag
pas / de ltar na yid la byed pa med pa zhes bya ste / yid la mi gnas paam / yid las grol
baam / yid las das pa zhes pa la jug la / de yid la mi gnas par gyur na sgra don go bde
ba la / yid la mi byed pa la khrul gzhi byung nas / rnam rtog dang kun rtog gis yul can
byas nas / de byed pa la mi byed par go nas yid la mi byed par rang yid la byed par yong
ba yang dug / dei don ni kun gzhi dang sgo lnga ji ltar shar yang rtog med du rang gsal
ba phyag rgya chen poi rang gshis yin / [248] de la nyon yid kyis kun gzhi la kha nang du
bltas nas bdag tu bzung / yid shed [sic!] kyis sgo snga la phyir bltas nas rigs su bcas
[sic!] / de ltar yid gnyis kyi bzung dzin ni khor bai chos dang / bzang ngan gyi dzin pa
thams cad yid yin la / de las das shing mi gnas pa phyag rgya chen po yin no / / Unfortu-
nately, both the available editions of Yang dgon pas bKa bum are rife with spelling er-
rors. Many of the works in this collection were committed to writing by Yang dgon pas
leading disciple, sPyan snga rin chen ldan. The passages pertinent to this essay are particu-
larly problematic because of the transliterations from Sanskrit, a language with which the
author, scribe and copyist were probably not too conversant.

291
sciousness (yid shes)
76
looks outward through the five sensory gates, it di-
vides it into the categories [of the life-world] (rigs su bcad).
77
Hence, all that
presents itself as samsara, as the subject object [duality] of the two-fold ego-
mind (yid gnyis), and all our beliefs in good and evil are what is meant by
ego-mind (yid). To go beyond this and not remain in it is Mahmudr.
Through a masterful synthesis of the Cittamtra model of consciousness
and the Siddha interpretation of amanasikra, Yang dgon pa is able to
account for both the genesis and transformation of subject-centered con-
sciousness. The emergence of ego-mind, the sense of self through
which experience is structured in terms of subject (I) and object
(mine) is elaborated according to the Cittamtra notion of an all-perva-
sive substratum (kun gzhi) and its concomitant intentional and reflexive
conscious operations. On this account, it would be incorrect to
characterize the transcendence of dualistic consciousness in terms of the
type of suppression or cessation of mental activity associated with Hva
shang. Yid la mi byed pa does not imply the suspension of all mental
activity but only of those subjectivizing and objectifying operations
which concurrently give rise to our hypostatized sense of self and ob-
jects.
How then is non-egocentric, selfless experience possible and how
is its possibility actualized? Yang dgon pa articulates the Cittamtra
eightfold ensemble of consciousness (rnam shes tshogs brgyad) as a
process of co-constitution, simultaneously constitutive of objects
(intentionality) and the subject to whom they belong (reflexivity).
We gather from his account that this process of co-constitution ac-
tually prescinds from what is non-constitutive, the pre-reflective
non-thematized flow of experience. In other words, when the inci-
pient phase of constitution is attended to closely, what is disclosed is
not a transcendental subject or foundational presence of any sort, but
rather a sheer absence, a non-constitutive, non-subjectivizing expe-
rience which makes possible the self-manifesting of dualistic expe-
rience. Thus Yang dgon pa can speak of the emergence of the perva-
sive substratum and its conscious operations as self-effulgence of the

76
The text incorrectly has yid shed.
77
The text incorrectly has rigs su bcas.
DAVID HIGGINS 292
conceptless and this latter as the intrinsic dynamics (rang gshis) of
Mahmudr. Going beyond subjectivizing experience really means
going back before it, but this going back is more properly not
dwelling (mi gnas) in it in the first place. What is actually meant by
this latter expression is clarified by the author when he returns to the
theme of amanasikra in his Ri chos Yon tan kun 'byung gi lhan
thabs chen mo:
78

Yid la byed pa is amanasikra in the Sanskrit language. Its meaning may be
rendered as not dwelling in ego-mind (yid la mi gnas pa) the ego-mind not
focusing upon [various objects] (yid la mi byed pa), or free from ego-mind
(yid las grol).
If it had been translated as not dwelling in ego-mind, this would have been
straightforward. But as it was translated as yid la mi byed pa, certain people
went somewhat astray. When they said the ego-mind does not focus upon
the past, the future or the present, the ego-mind served as the subject (yid
kyis yul can byas) and those [three times] served as the object (de dag gis yul
byas), and then they said that not focusing on them (de la mi byed pa) was
the ego-mind not focusing upon [objects]. But here, the past, the future, the
present, existence, non-existence, samsara and nirvana are all superimposi-
tions of ego-mind (yid kyi sgro btags pa), and the point of the above [render-
ings of amanasikra] is that Mahmudr [whether understood as] the ego-
mind not focusing upon or [ones] not dwelling in ego-mind (yid la mi
byed pa'am mi gnas pa), is, to put it concisely, not dwelling in existence,
non-existence, past, future, samsara or nirvana. Thus the terms "transcending
the intellect" (blo 'das), "free from discursive elaborations" (spros bral),
integration (zung 'jug), and Mahmudr (phyag rgya chen po) are all syn-
onymous.

78
rGyal ba Yang dgon pa bka bum (rTa mgo edition), v. 2, 76.4f.: yid la mi byed ces
pa ni / sang kri tai skad du na / a ma na sri [sic!] ka ra / dei don yid la mi gnas paam /
mi byed paam / yid las grol zhes bya ba la jug ste / yid la mi gnas par bsgyur na bde ba la
/ yid la mi byed pa zhes bsgyur nas / ga zhig cung zad nor nas / das pa yid la mi byed /
ma ongs yid la mi byed / da lta ba yid la mi byed ces / yid kyi[s?] yul can byas / de dag gis
yul byas / de la mi byed pa la yid la mi byed zer / dir das paam / ma ongs paam da lta
baam / yod paam / med paam / khor baam / das paam / de thams cad yid kyi sgro
btags pa yin cing / de la phyag rgya chen po yid la mi byed paam / mi gnas pa de / mdor
na yod med du mi gnas / das ma ongs la mi gnas / khor das la mi gnas te / gong ma
rnams kyi don no / blo das zhes bya / yid la mi byed pa zhes bya / spros bral zhes bya /
zung jug zhes bya / phyag rgya chen po zhes bya / de thams cad don gcig pa yin //

293
Not dwelling in ego-mind means not remaining stuck in constructs of
what is itself already a construct. It is not enough to go about
suppressing the subjectivizing or objectifying tendencies of thought
while leaving intact the insidious habit of labeling experience and
identifying with these labels. It is in and through language that hu-
mans constitute themselves as subjects. This is not to say that lan-
guage causes subjectivity, but only that it determines the sense the
experiencer has of being a psychic unity that transcends actual expe-
riences.
Yang dgon pas elucidation of amanasikra here brings the
deconstructive strategies of Madhyamaka philosophy to bear upon the
Cittamtra phenomenology of constitutive experience. This move is not
arbitrary, for any attempt to examine the constitutive activity of expe-
rience must eventually grapple with the role of language in this activ-
ity.
79
The emphasis here shifts from what we experience to how we
experience what we experience by means of the sedimented cultural-

79
Language plays a central role in self-representation, the construction of an I who
experiences which lies at the heart of subjectivizing-objectifying awareness. One contribu-
tion of recent neurophysiological research has been to elucidate some of the neurological
processes at work in the construction of a sense of self. In the words of one neurophysiol-
gist, Sam Harris:
The sense of self seems to be the product of the brains representing its own acts of
representation; its seeing of the world begets an image of a one who sees. It is impor-
tant to realize that this feeling the sense that each of us has of appropriating, rather
than merely being, a sphere of experience is not a necessary feature of conscious-
ness. It is, after all, conceivable that a creature could form a representation of the
world without forming a representation of itself in the world. And, indeed, many spiri-
tual practitioners claim to experience the world in just this way, perfectly shorn of
self.
A basic finding of neurophysiology lends credence to such claims. It is not so much
what they are but what they do that makes neurons see, hear, smell, taste, touch, think,
and feel. Like any other function that emerges from the activity of the brain, the feel-
ing of self is best thought of as a process. It is not very surprising, therefore, that we
can lose this feeling, because processes, by their very nature, can be interrupted.
While the experience of selflessness does not indicate anything about the relationship
between consciousness and the physical world (and is thus mute on the question of
what happens after death), it has broad implications for the sciences of mind, for our
approach to spirituality, and for our conception of human happiness.
See his The End of Faith: Religion, Terror and the Future of Reason (New York: Norton,
2005) p. 212.
DAVID HIGGINS 294
linguistic patterns at our disposal. The subject can never be present to
himself as a pure prelinguistic subjectivity but only as an emergent cen-
tre of operations who is inexorably caught up in the flux of experience
and linguistic self-implication. The point of not dwelling in ego-mind is
to catch oneself in the act, so to speak, of linguistically implicating one-
self as the proprietor of this or that situation as it unfolds. It is in this
moment of losing ones self that the dominative and instrumental
deliberations of dualistic thought lose their hold. An account of the
contemplative practices for experiencing non-mentation or loss of self
obviously falls outside the scope of this preliminary survey. A wide
range of Buddhist practices may be said to have amanasikra as their
aim. Thus Yang dgon pa can mention several synonyms of amanasi-
kra
80
(see above quotation) deriving from a variety of Buddhist teach-
ings: Dissociation from discursive elaborations (spros bral) is a term
used in Madhyamaka thought but also constitutes the second of sGam po
pas four yogas (rnal byor bzhi)
81
. Transcending the intellect (blo [las]
'das) constitutes the fourth of Sarahas four symbol terms (brda bzhi)
82

and is met with frequently in teachings on Mahmudr. Integration
(zung 'jug) is a term well known in Buddhist Tantrism which refers, in
particular, to the indivisibility of insight (praj) and action (upya).
Mahmudr is of course the sine qua non of Tantric and Siddha
praxis. Suffice it for the present to close this essay with a passage from
Yang dgon pa that indicates how the four yogas facilitate the
transformation of egoic mind through the clearing of the eightfold con-
sciousness as anticipated in his account of amanasikra:
83

80
A much longer list of synonyms is given in the Ri chos kyi rnal 'byor bzhi pa Phyag
rgya chen po snying po don gyi gter mdzod, 4b2 f.

81
Namely, rtse gcig, spros bral, ro gcig, and sgom med. Yang dgon pa diverges from
the majority of Tibetan authors on the rNal 'byor bzhi who take pains to establish Indian
sources for this doctrine, in the stras as well as the tantras. According to Yang dgon pa,
these teachings on the rNal 'byor bzhi are not actually expressed in the tantra corpus
because their content corresponds to things as they are (don ji lta ba). They are therefore
special teachings that cannot be compared to the tantra corpus but which represent the
doctrine of Dvags po rin po che (i.e., sGam po pa) (ibid., fol. 2b1 f.).
82
See above note 51.
83
rGyal ba Yang dgon pa bka bum (rTa mgo edition), fol. 8b3f.

295
The manner in which the presencing of the eight-fold pattern of conscious-
ness (tshogs brgyad) is cleared away ('dag) in the post-composure awareness
(rjes shes) through a deep understanding (rtogs) which is free from the five
constituents [that make up an individual] in the composure state (mnyam
bzhag) is as follows: During time of single-pointed focusing (rtse gcig), one
is free from the constituent of form (gzugs) and the five sensory operations
are cleared away. During the time of freedom from discursive elaborations
(spros bral), one becomes free from the constituents of feeling (tshor ba) and
conception ('du shes) and subjective consciousness (yid shes) is cleared
away. During the time of one-flavouredness (ro gcig), one is free from the
constituent of motivational tendencies ('du byed) and the emotionally-tainted
subjectivity (nyon mongs pa can gyi yid) is cleared away. During the time of
non-meditation (bsgom med), one is free from the constituent of conscious-
ness (rnam par shes pa) and the all-pervasive substratum awareness (kun
gzhi'i rnam par shes pa) is cleared away.

Appendix
Phyag chen gan mdzod, v. 21, p. 38.5f. (cf. n. 33 above)
de yang yid la mi byed pa zhes pai don la rnam pa gsum gsungs pai dang po /
(1) a ma na si k ra zhes pai sii yig ni / yid la zhes pai la yig bdun pai don
yin / bdun pa di la gnas gzhii rkyen zhes bya bar sgrai mdor /
gang kun nas dzin pa de gnas gzhio zhes byung //
de ltar gnas gzhi can gyi don de thog mai a [39] yig gis bkag pas / a ma na si ka
ra zhes pa / gang du dmigs pai gnas sam rten gzhi med pa la bya dgos par shes
te / sdom byung las
dngos med dngos po la rten nas / brten pa med pai bsgom pa bya /
yid med yid kyi byas nas su / cung zad tsam yang mi bsam mo //
zhes gsungs so / de bas sems byung yid la byed pai dzin stangs kyis / dmigs pa
la bsgrims nas sems dzin dam por byed pa thun mong gi zhi gnas bsgrub pai
skabs la dgos kyang dir de bkag pa yin no / de yang thogs med kyis / de la jog
par byed pa dang / yang dag par jog par byed pa la mi bsgrims te jug pai yid
la byed pa yod do zhes pao //
(2) gnyis pa a ma na si k ra zhes pai bdun pai dei skyes bui bsdu ba byas
nas yid mi byed pa zhes pa / la yig mi mngon par byas pai bshad pa gnyis pa
mdzod do / dei don ltar naang / yid mi byed pa zhes pa yid kyi las dir dgag
DAVID HIGGINS 296
byar bzhed pa ste / mngon par /
yid kyi las gang zhe na / sems pa yid kyi las yin no //
zhes sems byung sems pai dzin pai dzin stangs la nan tan du byed pa de [40]
dgag pao / sems byung sems paang sems mngon par du byed pai yid kyi las
te / de nyid du /
dge ba dang mi dge ba lung du ma bstan pa rnams la sems jug par byed pai
las can no zhes bao / don mngon par du byed pa dgag pao / nyes pa lnga
spong bai du byed brgyad lta bu zhi gnas bsgrub pa la yin gyi / phyag rgya
chen po la ni byas pa rnams dang bral zhing bsags pa las min
zhes dang /
nga ni gro ong mi len mi dor ro
dang / dgyes rdor las /
gang phyir yid kyis mi bsgom par /
zhes pas so // des na /
yis kyis de nyid dmigs pa bcas //
zhes yid kyi las su bya ba gang yin thams cad dir dmigs pa dang bcas par gzhag
nas dmigs pa thams cad nye bar zhi ba cig nges par bstan no / des bas na /
kun tu rtog pas ma brtags pa / rab tu mi gnas pa yi yid /
dran pa med cing yid byed min / dmigs pa med la phyag tshal dud //
ces ston pas gsungs pa de legs par bshad do / dran pa med pa sogs rgyas pa og
tu byung ngo /
(3) gsum pa / a ma na si k ra zhes pai a yar bcad nas / a yig skye ba med pai
don du [41] byas te / ma na si k ra yid la byed par bshad do / de ltar na a yig gi
don tshul bzhin du yid la byed pa ni / a yid la byed pa zhes byar te / de yang / bar
gyi tshig mi mngon par byas pa lo ma la dgai rgyal po la lo mai rgyal po zhes
pa bzhin no / dir a ni shes rab kyi pha rol tu phyin pao / a nu tpa nna / a ni ro
dha zhes pa lta bui sgo nas / skye med gag med sogs gnyis su med pai rnam
grangs thams cad mtshon nus so / mtshan brjod las /
a ni yig bru kun gyi mchog / don chen yi ge dam pa yin /
kong nas byung ba skye ba med //
sogs kyis so / mtshan brjod kyi grel chen las /
sngags kyi tshul gyis ni shes rab dang thabs ni gnyis so / de gcig tu gyur pa ni
gnyis su med pa ste / shes rab dang thabs gnyis su med pa bde ba chen poi
ngo bo nyid ni gnyis su med pa yin par dod de / de las byung bao / pha rol
tu phyin pai tshul gyis ni / gzung ba dang dzin paam / bdag dang bdag

297
giam / shes pa dang shes bya ste / ji srid yid kyi rnam par g.yo ba de srid du
gnyis so / g.yo ba de srid du gnyis so / g.yo ba thams cad dang bral zhing /
spros pa med pa chos thams cad bdag med pa ni gnyis su [42] med pai ngo
bo nyid chos nyid kyi bdag nyid can gyi sku byung ste / de bas na gnyis su
med par byung bao / gnyis su med par byung ba yang mi skye bai rnam
pas khyad par du dbye bai phyir / mi skyei chos can zhes bya ba smos te
zhes gsungs pao // de lta bui don gyis yid la mi byed pa ston pas yid la mi byed
pai chos skor zhes byao / de thams cad slob dpon nges par sbyangs pa gnyis su
med pai rdo rje zhes sam / grub pai slob dpon chen po mnga bdag Mai tri pas
mdzad pao //
Abbreviations and bibliography
D Derge Tibetan Tripiaka. The Tibetan Tripiaka, Taipei Edition. Taipei,
Taiwan: SMC Publishing 1991.
P Peking Tibetan Tripiaka. The Tibetan Tripiaka, Peking Edition. Tokyo-
Kyoto: Tibetan Tripiaka Research Institute 1957.
I. Primary sources
A. Indian sources (Sanskrit and Tibetan by author):
Kamalala
Bhvankrama I. Sanskrit and Tibetan texts edited by Tucci in Minor Bud-
dhist Texts (Part II). See Tucci.
Avikalpapraveadhrak. Tibetan translation in P vol. 105, no. 5501.
Maitreya/Asaga
Abhidharmasamuccaya. Edited by Pralhad Pradhan. Santiniketan: Visva
Bharati 1950.
Abhisamaylakra. Edited by Stcherbatsky. Bibliotheca Buddhica, vol. 23.
Leningrad 1929.
Dharmadharmatvibhga. Edited by J. Nozawa. In: Studies in Indology and
Buddhology Presented in Honor of Prof. Susumu Yamaguchi on the
Occasion of his 60th Birthday. Kyoto: Hozokan 1955.
Madhyntavibhga. Edited by G.N. Nagao. Tokyo 1964.
Mahynastrlakra. Edited by S. Levi. Paris: Champion 1925.
Uttaratantra. Edited by E.H. Johnson. Patna: Bihar Research Society 1950.
DAVID HIGGINS 298
Maitrpda
Amanasikradhara. (AM) In Advayavajrasagraha, Edited by M. H. Shas-
tri. GOS, vol. 60. Baroda: 1927, 6062. Also in Advayavajrasagraha.
Ed. by the Study Group on Sacred Tantric Texts. Annual of the Institute
for Comprehensive Studies of Buddhism (AICSB) 11 (March 1989) 136
143.
For reproduction of original Sanskrit manuscript, Nepal German Manu-
script Preservation Project B 22/24.
For Tibetan translations, Peking Tanjur (P), v. 68 p. 286.5.6 f and Taipei
Tanjur (D), v. 28, p. 276.4f.
Saraha
Cittakojavajragti. Tibetan translation in P vol. 69, no. 3117.
Kyakomtavajragti. Tibetan translation in P vol. 69, no. 31135
Kyavccittmanasikranma. Tibetan translation in P vol. 69, no. 3118.
Mahmudropadea. Tibetan translation in P vol. 69, no. 3119.
Vkkoarucirasvarajagti. Tibetan translation in P vol. 69, no. 3116.
Tilopa
Acintyamahmudr. Tibetan translation in P vol. 82, no. 4635.
Dohkoa. Tibetan translation in P vol. 69, no. 3128.
Mahmudropadea. Tibetan translation in P vol. 69, no. 3132.
Vasubhandu
Abhidharmakoa. With Bhya and Yasomitras Vyakhya. Edited by Swami
Dwarikadas Shastri. 4 vols. Varanasi: Bauddha Bharati 197073.
Trimikvijapti. With Bhya. Edited by S. Levi in Vijaptimtrasiddhi.
Paris: Champion 1932.
Various authors
Phags yul grub dbang rnams kyi zab moi doha rnams las khol byung mu tig
phreng ba (A Selected Anthology of Dohs from private collection of
Herbert Guenther; no publication information given)
B. Tibetan sources (by author)
Ba ra ba rGyal mtshan dpal bzang
Ngo sprod bdun mai grel pa Man ngag rin po chei sgron me. In: A Tibetan
Encyclopedia of Buddhist Scholasticism: The Collected Writings of Ba
ra ba rgyal mtshan dpal bzang, vol. XI. Dehra Dun: Ngawang Gyaltsen
and Ngawang Lungtok, 1970.

299
sBa gSal snang, sBa Sang shi, et al.
sBa bzhed ces bya ba las sBa gsal snang gi bzhed pa bzhugs. Edited by
mGon po rgyal mtshan. Beijing: Mi rigs dpe skrun khang 1982.
Bu ston Rin chen grub
bKa drin rjes su dran par byed pai gSan yig (= gSan yig). In The Collected
Works of Bu-ston, vol. LA (26). Edited by Lokesh Chandra. ata-Piaka
Series 66. Delhi: International Acamdemy of Indian Culture 1971.
Blo bzang grags pa, mChims ston
Chos mngon pa gsal byed legs par bshad pai rgya mtsho. Sarnath: Central
Institute of Higher Tibetan Studies 1986.
sGam po pa bSod nams rin chen
Collected Works of sGam po pa. Dvags lha sgam po edition. Darjeeling
1982.
Phyag rgya chen po rtsa ba la ngo sprod pa, vol. VA.
Zhal gyis bdud brtsi thun mongs (sic!) ma yin pa, vol. DZA.
Gos lo tsa ba gZhon nu dpal
Deb ther sngon po. Translated as The Blue Annals (= BA) by G. Roerich.
Delhi: Motilal Banarsidas 1949.
Klong chen rab byams pa
Grub mtha rin po chei mdzod. Gangtok: Dodrup Chen Rinpoche edition
1969.
gNas lugs rin po chei mdzod. Gangtok: Dodrup Chen Rinpoche edition
1969.
Mi pham rgya mtsho
dBu ma rgyan gyi rnam bshad jam dbyangs bla ma dgyes pai zhal lung. In:
Collected Writings of Jam-mngon Ju Mi-pham rgya-mtsho. Gangtok
1971. vol. XII.
Mi bskyod rdo rje
dBu ma la jug pai rnam bshad dPal ldan dus gsum mkhyen pai zhal lung
dvags brgyud grub pai shing rta. Rumtek: Dharma Chakra Centre 1969.
gNubs Sangs rgyas ye shes
bSam gtan mig sgron. vol. 74 of Smanrtsis Shesrig Spendzod series. Leh:
S.W. Tashigangpa 1974.
Nyang ral Nyi mai od zer
Chos byung me tog snying po sBrang rtsii bcud. In: Rin chen gter mdzod
chen poi rgyab chos, vols. 5 & 6. Paro: 1979.
DAVID HIGGINS 300
Padma dkar po
The Collected Works of Kun-mkhyen Padma-dkar-po. Gnam Brug edition.
Darjeeling: Kargyud Sungrab Nyamso Khang 1973.
bKa brgyud kyi bka bum gsil bu rnams kyi gSan yig. (= gSan yig). vol. 4,
no. XII.
Klan ka gzhom pai gtan bzhugs. vol. 21, no. XI.
rNal byor bzhii bshad pa don dam mdzub tshugs su bstan pa. (= rNal byor
bzhi mdzub tshugs). vol. 21, no. IX.
Phyag rgya chen po man ngag gi bshad sbyor rgyal bai gan mdzod. (=
Phyag chen gan mdzod). vol. 21, no. I.
Rang byung rdo rje (Third Karmapa)
Nges don phyag rgya chen poi smon lam. Blockprint (Lahul, n.d.)
Phyag rgya chen po Gang ga mai grel pa. In: gDams ngag mdzod, vol. 5.
Compliled by Kong sprul Blo gros mtha yas. Edited by Nawang Lung-
tok and Nawang Gyaltsen, Delhi 19712.
bsTan pai nyin byed, Situ
Nges don phyag rgya chen poi smon lam gyi grel pa Grub pa mchog gi zhal
lung. From microfilm of Rin chen ri bo edition (1517).
Yang dgon pa, rGyal ba
Collected Writings of rGyal ba Yang dgon pa rGyal mtshan dpal. rTa mgo
edition. Thimphu: Tango Monastic Community 1982.
Ri chos kyi rNal byor bzhi pa Phyag rgya chen po snying poi don gyi gter
mdzod. In vol. 1.
Ri chos kyi Phyag len gsal bai sgron me. In vol. 1.
Ri chos Yon tan kun byung gi lhan thabs chen mo. In vol 2.
II. Secondary sources
Broido, Michael M.
1987 Sa-skya Paita, the White Panacea and the Hva-Shang Doctrine,
Journal of the International Association of Buddhist Studies 10/2, 27
68.
Caputo, John.
1982 Hermeneutics and the Recovery of Man, Man and World 15, 343
67.
1987 Radical Hermeneutics: Repitition, Deconstruction and the Hermeneu-
tic Project. Bloomington: Indiana U. Press.

301
Demiville
1952 Le Concile de Lhasa. Bibliotque de lInstitut des Hautes Etudes
Chinoises VII. Imprimrie Nationale de France, Paris.
Faber, Flemming.
1986 The Council of Tibet According to the sBa bzhed, Acta Orientala
47, 3361.
Gomez, Louis.
1987 "Purifying Gold: The Metaphor of Effort and Intuition in Buddhist
Thought and Practice. In: Sudden and Gradual: Approaches to
Enlightenment in Chinese Thought. Hawaii: Kuroda Institute, 67165.
Guenther, Herbert V.
1963 The Life and Teachings of Naropa. Oxford: Clarendon Press.
1969 The Royal Song of Saraha. Seattle: U. of Washington Press.
1993 Ecstatic Spontaneity: Sarahas Three Cycles of Doh. Berkeley:
Asian Humanities press. Nazan Studies in Asian Religions 4.
Harris, Sam.
2005 The End of Faith: Religion, Terror and the Future of Reason. New
York: W.W. Norton & Co.
Heidegger, Martin.
1962 Being and Time. Translated by J. Mcquarrie & E. Robinson. New
York: Harper & Row.
1975 Early Greek Thinking. Edited by F. Capuzzi and D. Krell. New York:
Harper & Row.
1977 Basic Writings. Edited by D. Krell. New York: Harper & Row.
Hoffman, Helmut.
1961 The Religions of Tibet. New York: Macmillan.
Hofstadter, Albert.
1979 On the Interpretation of Works of Art. In: The Concept of Style.
Edited by Beryl Lang. U. of Pennsylvania Press 6791.
Houston, Gary.
1980 Sources for a History of the bSam yas Debate. Sankt Augustin:
Monumenta Tibetica Historica I, 2.
Jackson, David
1994 Enlightenment by a Single Means. Vienna: sterreichische Akademie
der Wissenschaften.
DAVID HIGGINS 302
Jackson, Roger
1982 Sa skya paitas Account of the bSam yas Debate: History as Po-
lemic, The Journal of the International Association of Buddhist Stud-
ies 5, 8999.
2004 Tantric Treasures: Three Collections of Mystical Verse from Buddhist
India. New York: Oxford University Press.
Kierkegaard, Soren.
1983 Fear and Trembling/Repitition. New Jersey: Princeton U. Press.
van der Kuijp, Leonard
1984 Miscellanea to a Recent Contribution on/to the Bsam-yas Debate,
Kailash, Vol. 11/34, 149184.
Kvaerne, Per.
1986 An Anthology of Buddhist Tantric Songs: A Study of the Carygti.
Second edition of 1977 version. Bangkok: White Orchid Press.
Mathes, Klaus-Dieter.
Forthcoming. Maitrpas Amanasikrdhra (A Justification of Not
Becoming Mentally Engaged), Journal of the Nepal Research Cen-
tre 13.
2006. Blending the Stras with the Tantras: The Influence of Maitrpa and
his Circle on the Formation of Stra Mahmudr in the Kagyu
Schools. In: Tibetan Buddhist Literature and Praxis: Studies in its
Formative Period 9001400. Ed. by Ronald M. Davidson and Chris-
tian K.Wedemeyer (Proceedings of the Tenth Seminar of the IATS,
2003, Oxford, vol. 10/4). Leiden: Brill 201227.
2005 Gos Lo ts ba gZhon nu dpals Commentary on the Dharmat
Chapter of the Dharmadharmatvibhgakriks. In: Studies in In-
dian Philosophy and Buddhism, University of Tokyo, Vol. 12, p. 13ff.
Ruegg, David S.
1981 The Literature of the Madhyamaka School of Philosophy in India.
Wiesbaden: Otto Harrassowitz.
1988 A Karma bKa brgyud Work on the Lineages and Traditions of the
Indo-Tibetan dBu ma (Madhyamaka) In: Orientalia Iosephi Tucci
Memoriae Dicata, vol. 3. Roma: Istituto Italiano per il Medio ed Es-
tremo Oriente, 1249-80.
1988 Buddha-nature, Mind and the Problem of Gradualism in a Compara-
tive Perspective. London: School of Oriental and African Studies.

303
Sktyyan, Rhul.
1957 Doh-ko. Patna: Bihr Rr-bh Pariad, Pan. In Hindi.
Schmithausen, Lambert.
1981 On Some Aspects of Descriptions of Theories of Liberating In-
sight and Enlightenment in Early Buddhism, In: Klaus Bruhn and
Albrecht Wezler (eds.), Studien zum Jainismus und Buddhismus:
Gedenkschrift fr Ludwig Alsdorf. Wiesbaden: Franz Steiner, 199
250.
1987 layavijna: On the Origin and the Early Development of a Central
Concept of Yogcra Philosophy, 2 vols. Tokyo: International Insti-
tute for Buddhist Studies.
Shahidullah, Mohammed.
1966 Buddhist Mystic Songs. (2
nd
revised Edition). Bengal Academy,
Dacca.
Smith, Gene.
1970 Introduction to Dkar brgyud gser phreng: A Golden Rosary of Emi-
nent Gurus. Compiled by Mon rtse pa Kun dga dpal ldan. vol. 3 of
Smanrtsis Shesrig Spendzod series. Leh: S. W. Tashigang.
Snellgrove, David & Richardson, Hugh.
1968 A Cultural History of Tibet. Boston: Shambhala.
Stein, R.A.
1971 Tibetan Civilization. Stanford: Stanford U. Press.
Trantha
1970 Chattopadhyaya, Debiprasad (eds.), Taranathas History of Buddhism
in India. Translated from the Tibetan by Lama Chimpa Alaka Chatto-
padhyaya. Simla: Indian Institute of Advanced Study. 1
st
edition.
Tatz, M.
1987 The Life of the Siddha-Philosopher Maitrgupta, Journal of the
American Oriental Society 107, 695711.
Tucci, Giuseppe
1958 Minor Buddhist Texts (Part II). 1981 reprint. Delhi: Motilal Banarsi-
das.
THE'MYSTERIES OF BODY, SPEECH, AND MIND:
THE THREE ESOTERICA (SANMI) IN MEDIEVAL SINITIC BUDDHISM
RlCHARD D, MCBRlDE, II
When observed through a Japanese Shingon lens the "three eso-
terica" or "three mysteries" (Ch. sanmi, Jpn. sanmitsu .::::. W, Skt.
*tri-guhya) is generally considered a seminal component of the radi-
cal, ritual-oriented approach of "Tantric" or "Esoteric" Buddhism!
by which a practitioner attains buddhahood in this very body (soku-
shin jobutsu )JX;1J1l). The three esoterica, as a ritualized replica-
tion of the body, speech, and mind of the Buddha, is seen as a key
practice described in the Da Piluzhena chengfo jing A JJX: 1J1l
fr..& (Satra on Vairocana's attaining buddhahood, T 848), which Ku-
kai 2m (774-835) mastered during his visit to China.
2
Because of
its centrality in "esoteric" ritual, the meaning of the three esoterica
was a source of debate in the medieval Tendai Esoteric tradition
An earlier draft of this paper was presented at the panel "Tantric Buddhism through the
Chinese Looking Glass," at the Annual Meeting of the American Academy of Religion,
Washington, D.C., November 21, 2006, The author of this article would like to thank
George A. Keyworth, III, Charles D. Orzech, and the peer reviewer, whose comments and
suggestions greatly improved the quality of this essay.
1 Although the category Tantrarrantric is a modem scholarly invention, I still prefer it
to Esoteric Buddhism, which is just as problematic. On problems with the term Tantra see
Hugh B, Urban, Tantra: Sex, Politics, and Power in the Study of Religion (Berkeley: Uni-
versity of California Press 2003) 271-281; see also McBride, "Is there really 'Esoteric'
Buddhism?" Journal of the International Association of Buddhist Studies 27/2 (2004) 329-
356.
2 See Ryiiichi Abe, The Weaving of Mantra (New York: Columbia University Press
1999) 129-132. Abe also notes that Kiikai wrote three short essays on the body, speech,
and mind - Sokushin ji5blltsugi HP (Attaining buddhahood in this very body, or
Transforming one's body into the realm of enlightenment) T 2428, 77; Shi5ji jissi5gi *
(Voice, Letter, Reality) T 2429, 77; and Unjigi (On the Sanskrit Letter
Hiiq:t) T 2430, 77 - which represent the "core of Kiikai's philosophy." See Ryiiichi Abe,
"Buddhahood in this Lifetime," in Buddhist Scriptures, ed. Donald S. Lopez, JT. (London:
Penguin 2004) 497.
Volume 29 Number 22006 (2008) pp. 305-355
306 RlCHARD D. MCBRlDE, II
(Taimitsu :lW): Rival interpretations deriving from Tiantai Zhiyi's
(538-597) commentaries on the Lotus Sutra and the views
of Tendai thinkers influenced by Shingon j (Tomitsu texts
vied for supremacy.3If so important in Japan, what of this concept
when viewed through a Chinese looking glass? This paper will ad-
dress three questions - how did medieval Chinese Buddhist
intellectuals understand the term, how did early Tantric practitioners
understand it, and did its meaning change?
In this paper I address these interrelated questions by analyzing
several of the most instructive instances of the use of the term "three
esoterica" and related terms by Buddhist scholiasts in medieval
Sinitic Buddhism.
4
After first discussing the origins of the term in
Chinese Buddhist literature during the Northern and Southern
Dynasties period (Nanbeichao, ca. 317-589), I will analyze the term
as it was used by eminent translators and scholars of the Sui-Tang
transitional period (ca. 589-712), the High Tang period (ca. 712-
756), and the mid and late Tang period (ca. 756-907). The concept
of the three esoterica has not been examined in detail by Western
scholars of either mainstream Chinese Buddhism or Chinese Tantric/
Esoteric Buddhism.
s
In earlier research I have sought to place the
3 See, for instance, Kubota Tetsumasa EB 1!f iE, "Nihon Tendai no okeru yuso san-
mitsu hoben setsu" B ;$:7i:in;:: tnt (On arguments about the meaning
of sanmitsu [Three Mysteries] in the Japanese Tendai Sect), Nihon Bukkyo gakkai nenpo
B 57 (1992) 145-162; Okubo Ryoshun "Taimitsu no
sanmitsu ron" (The Three Secrets Theory of Taimitsu), Tindai gakuho 7i:it
34 (1992) 109-113; Otsuka Nobuo "Sanmitsu shiso ni tsuite"
":J,' l (On the concept of the tri-guhya), Indogaku Bukkyogaku kenkyii 34/1 (1995) 174-
176.
4 What I call China's medieval period refers to the years between 317 and 907 C.E.
5 Tajima Ryujun treats the three mysteries from a Japanese Shingon perspective but
not the earlier exegetical tradition in his Etude sur Ie Mahtivairocana-sL7tra (Dainichikyo)
(paris: Adrien Maisonneuve 1936) 44-54, 91-98; see also Tajima, "A Study of the
Mahavairocana-sntra," in The Enlightenment of Vairocana, trans. and ed. Alex Wayman
(Dehli: Motilal Banarsidass 1992; rpt. 1998) 245-247, 293-298. Much more detailed
assessments are available, of course, in Japanese. A detailed treatment may be found in
Mochizuki Shinko ed., Bukkyo dai jiten (Encyclopedia of Bud-
dhism), rev. ed. 10 vols. (Kyoto: Seikai Seiten Kanko Kyokai 1954-1963) 2:1682c-1684b,
s.v. "sanmitsu" and "sanmitsu kaji." Nakamura Hajime rpH:7I; only treats the mikkyo
interpretation in his Bukkyogo dai jiten (Dictionary of Buddhist technical
THE MYSTERIES OF BODY, SPEECH, AND MIND 307
rhetoric and practices of the nascent Tailtric tradition within the con-
text of mainstn;:am Sinitic exegetical Buddhism.
6
This, I believe, is a
fruitful approach to understand what is shared and what is new and
provides important insight into the dynamic nature of medieval
Sinitic Buddhism. By understanding the exegetical and rhetorical
usage of the term over time we can better understand how the Tan-
tras were seen as a continuation of the Mahayana and we can discern
those points that appeared novel, because the term always described
a complex intellectual understanding of the nature of buddhahood.
Before I launch into my study of the three esoterica I should ex-
plain why I choose to translate sanmi as "three esoteric a," rather than
the more common "three mysteries." The use of "three mysteries"
derives ultimately from the French trois mysteres used early on by
Tajima RyUjun and, in places, by Etienne Lamotte.
7
Lamotte,
nevertheless, did not always find that translation suitable because he
also uses trois secrets. Perhaps following Lamotte, Yamasaki Taik6
and Kenneth R. White translate the concept more or less consistently
as "three secrets."g Although Hakeda Yoshito uses "three mysteries,"
by contrast, Kiyota Minoru prefers the more Sanskritic reconstruc-
terms) (3 vols., Tokyo: Tokyo Shoseki 1975; 1 voL rpt. Tokyo: Tokyo Shoseki 1999)
490b-<:, s.v. "sanmitsu." Lii Jianfu 3!fill has examined the "three esoterica" in early
Tantric literature (viz. T 848). See his Zhongguo Mijiaoshi <P (History of Eso-
teric Buddhism in China) (Beijing: Zhongguo shehui kexue chubanshe 1995) passim, esp.
97-99,242-244.
6 McBride, "Is there really 'Esoteric' Buddhism?" (cf. n. 1 above); and McBride,
"DharaI,ll and Spells in Medieval Sinitic Buddhism," Journal of the International Associa-
tion of Buddhist Studies 28/1 (2005) 85-114.
7 See Tajima RyUjun, Etude sur le Mahavairocana-siitra (cf. n. 5 above) 93-94, pas-
sim, and Etienne Lamotte, trans., Le traite de la grande vertu de sagesse de Niigiirjuna
(MahaprajfiClpClramitiisClstra), 5 vols. (Louvain: Institut orientaliste, Universite de Louvain
1944--1981) 1:19-20.
8 Lamotte, trans., Le tmite (cf. n. 7 above) 1:560; Yamasaki Taiko, Shingon: Japanese
Esoteric Buddhism (Boston: Shambhala 1988) 30--32, 106--122. Kenneth R. White, in his
The Role of Bodhicitta in Buddhist Enlightenment Including a Translation into English of
Bodhicitta-sastra, Benkemmitsu-nikyoron, and Sammaya-kaijo (Lewiston, N.Y.: Edwin
Mellen Press 2005), typically employs the translation "three secrets," although he also
calls them "three mysteries" in a few places (pp. 199,441). White also translates
na (kaji :lJQf,f) as "three secrets practice," which I [md problematic.
308 RICHARD D. MCBRIDE, II
tion tri-guhya or his translation "three secret acts."9 More recently,
with the rise of Daoist and Chan/Zen studies, scholars have also be-
come accustomed to translating the word xuan 15: as "mystery," as in
chongxuan 15: (twofold mystery) and sanxuan .::::.15: (three myster-
ies).10 Both sanmi and sanxuan should not be the "three mysteries"
because mi (secret, intimate, confidential) and xuan (dark, arcane,
profound, subtle), though similar in some respects, ultimately invoke
different meanings and I think that xuan is more
"mysterious." My more technical translation of "three esoterica"
allows for a wide range of meaning to be added to the term. Further-
more, it also permits a linguistic link to the related polemical terms
mi * (esoteric) and xian (exoteric), which figure in any
discussion of scholastic and Tantric Buddhism in East Asia.
I will not treat the views of the Japanese monk Kukai in this
study. Kukai's views on the three esoterica are varied, complex, and
deserving of their own study, but they do not jibe with my present
purposeY My reasons for this are twofold. First, I do not find the
position compelling that Kukai preserves unaltered the views of his
Chinese master Huiguo ;!5f!: (746-805), with whom he studied a
mere six months. Second, Kukai did not participate in the continental
9 Yoshito S. Hakeda, Kakai: Major Works (New York: Columbia University Press
1972) 87-93, 97, 152, 225-235, 251; Kiyota Minoru, Shingon Buddhism: Theory and
Practice (Los Angeles and Tokyo: Buddhist Books International 1978) 175-176; and
Kiyota, Tantric Concept of Bodhicitta: A Buddhist Experimental Philosophy (An Exposition
based upon the Mahavairocana-siltra, Bodhicitta-sastra, and Sokushin-jBbutsu) (Madison:
South Asian Area Center, University of Wisconsin-Madison 1982) 153.
10 See, for instance, Robert H. Sharf, Coming to Terms with Chinese Buddhism: A
Reading of the Treasure Store Treatise (Honolulu: University of Hawai 'i Press 2002) 47-
71.
11 See, for instance, Kongo hannya haramitsukyo kaidai T
2201, 57. Ie, 2e; Nihonkyo kaidai B ;$:1i&MJI!, T 2211, 58.4c; Kongojo kaidai itJMJUUiMJI!,
T 2221, 61.4b-c; Rishukyo kaidai MUIlBIi&iJf'JI!, T 2236, 61.612a; Himitsu mandara
jajiishinron 1, T 2425, 77.303a, roll 10, T 2425, 77.361b; Sokushin
jobutsugi (six editions) T 2428, 77.381c, 383a, 385a, 386a, 387b-c, 389c, 393c, 391b,
395b, 400a-b (Hakeda, Kakai: Major Works [cf. n. 9 above] 225-234); ShOji jissogi
:lUlU!, T 2429, 77.401c (Hakeda, Kakai: Major Works, 239); and Unjigi Pf'*t%, T 2430,
77.405b (Hakeda, Kakai: Major Works, 257); Heianjo taijo tenno kanjomon
T 2461, 78.1a, 2a, 2e; and Benkenmitsu nikyo ron T 2427,
77.375a (Hakeda, Kiikai: Major Works, 152).
discourse on intellectual Buddhism aftel: his return to Japan. As Abe
Ryiiichi has shown, he created a new Buddhist discourse. No one on
the continent was familiar with his work and they probably would
have been surprised at some of his conclusions because they reflect
Kiikai's (successful) attempt to change the views of the proponents
of Nara's *' exegetical schools (Nanto Bukky6
Furthermore, because later Shingon proponents placed many doc-
trinal arguments and positions in Kukai' s mouth I am sufficiently
skeptical of including him here.
12
The origins of the concept of "three esoteric a"
The first scripture to mention the three esoterica individually is the
Miji jin 'gang jing (The Satra on the Vajrasattva Esoteric
Traces),13 which was translated by (Pahu :t:-iif, ca.
12 See Abe, The Weaving of Mantra (cf. n. 2 above) 152-154; Omura Seigai :k;J;1iffim,
Mikkyi5 hattatsushi (History of the development of Esoteric Buddhism) (To-
kyo: Kokusho Kank5kai 1918; rpt. Tokyo: DaitB Shuppansha 1972) 373-375 and passim;
Yamaori Tetsuo lir:lfit!i'ti, "Sannin no nitt5s5 to komikky5" ':::::,AO) Ar.ll'1111 i:
(Three monks who went to Tang China and old Esoteric Buddhism) in Chingo kokka to
jujutsu: Nihon Bukkyi5 no hajimari $I\!ilOO* i: u5t#J: it '? (State protection
and spell craft: The origins of Japanese Buddhism), ed. Uehara ShBichi ...t@Jg- and
Yamaori Tetsuo (Tokyo: Shiieisha 1989) 143-146. Misaki Ry5shil .:::::wi\jll:!.m, Taimitsu no
kenkyt7 tJifO)1iifJE (Research on Tendai Esoteric Buddhism) (Tokyo: S5bunsha 1988)
146-150, gives an overview of these problems in a short discussion of the problems
regarding the late terms junmitsu I.\lliif (pure esotericism) and zi5mitsu ltif (mixed esoteri-
cism).
13 Although cited by KumarajIva, could the Miji jin 'gang jing really date from the time
of the later Bodhiruci (d. 727)7 The evidence suggests that such a sfitra did exist because it
was known to KumarajIva, Lushan Huiyuan 1i!L!J:fl,J! (334-417), and Sengzhao 1111!l
(374---414). The earliest references to the scripture are from the early fifth century.
KumarajIva quotes from or alludes to the satra four times; see Dazhidu lun :k1!i' JjJ{oj I, T
1509, 25.59a, roll 26, T 1509, 25.248b; roll 57, T 1509, 25.466b; and roll 88, T 1509,
25.684a-b. Huiyuan and KumarajIva discussed it in their correspondence; see Jiwnoluoshi
fa dayi 1, T 1856, 45.125c; roll 2, T 1856, 45. 130b. Sengzhao refers to
it twice in his Commentary on the Vimalakfrtinirdefa Satra; see Zhu Weimoji jing tlj,l)J),'
5, T 1775, 38.37Ia7-1O; and roll 9, T 1775, 38.404b. The scripture was also known
to and used by scholars of the sixth century. Sengyou 1111#5 (445-518) was familiar with a
Miji jing, in five rolls, also known as Miji jin 'gang jing, in either five or seven rolls. He
gives the date on which the translation was completed as the eighth day of the tenth month
of the ninth year of the Taikang .:kll reign period (October 18,288), and he included it in
265-316)14 and later collected in the Dabaoji jing (Mahii-
ratnakiita, T 310) by Bodhiruci (putiliuzhi i?fmAtx, fl. 693-727):15
At that time [the Bodhisattva] Quiescent Thoughts (Jiyi again ques-
tioned the Vajrasattva Esoteric Traces (Miji jingang lishi
"The Tathiigata possesses a few things that are secret and important. They
are those things that all iriivakas and all pratyekabuddhas are unable to ac-
cess. How much more so for the ilk of the common masses and the be-
nighted [icchantika]! Well done!"
Esoteric Traces desired to think of a blissful abode [sukhavihiira]. He
promulgated the Tathiigata's secret and important [things] everywhere.
All the living beings assembled together and desired to hear them. The
Vajrasattva Esoteric Traces addressed Quiescent Thoughts saying, "Listen
up! Listen up! Ponder this well. I will now discourse on the Tathiigata's se-
cret and important [things]. There are three things (sanshi .=.JJ). What do
we call these three? The first is called body-esoterica (shenmi 5:tW), the sec-
ond is called speech-esoterica (yumi and the third is called mind-eso-
terica (yimi
This scripture goes on to provide many details on the Buddha's
body-esoterica. In particular, it repeatedly refers to the Buddha's
ability to manifest his physical majesty everywhere. In essence the
scripture introduces these three types of esoterica to rationalize many
of the more unbelievable aspects of the Buddha Sakyamuni's minis-
try and supernormal powers as described in the fully developed leg-
his catalog Chu sanzang jiji 2, T 2145, 55.7bI8. Bodhiruci (Putiliuzhi 'lttlHiIi:
iil;;, ca. 508-535) alluded to portions not discussed by previous scholars in his translation
of the Mile pusa suwen jing iun 3, T 1525, 26.245a23-25. The siitra
was also used by Jizang (549--623) in his Weimo jing yishu 6, T 1781,
38.983b; and Guanding JilJi (531-632) in his Daban niepanjing shu 19, T
1767, 38.147c. Hence, the evidence strongly suggests that a Mijijing did exist prior to the
later Bodhiruci's inclusion of it in the Dabaojijing.
14 For the biography of see Gaoseng zhuan 1, T 2059, 50.326c2-
327a12; see also Eric Ziircher, The Buddhist Conquest of China, 2nd ed. (Leiden: E. J.
Brill 1972) 1:65-70.
15 For the biography of Bodhiruci see Song gaoseng zhuan 3, T 2061,
50.720b4-c12.
16 Mijijingang lishi hui (The convocation of the Vajrasattva Esoteric
Traces), in Dabaoji jing 10, T 310, 11.53b9-15; see also Stanley Weinstein,
Buddhism under the T'ang (Cambridge: Cambridge University Press 1987) 44, 49.
end of the Buddha's life. The scripture does not develop the meaning
of these three 'types of esoterica in any systematic manner, however.
The Vajrasattva Esoteric Traces' intent, it seems, is to cause the
of the crowd to accept these logical inconsistencies as as-
pects of the wondrous power of the Buddha. This scripture is a suit-
able beginning in that does not use the term "three eso-
terica" (sanmi). Instead he calls them "three things."
As with many seminal terms and topics of interest among Maha-
yiina Buddhists' in the Sinitic cultural sphere, the locus classicus of
the term is, in fact, found in the first roll of KumarajIva's (Jiumo-
luoshi 343-413) Dazhidu lun (Treatise on the
Great Perfection of Wisdom, T 1509), attributed to Nagarjuna
(Longshu ca. 50-150 C.E.):17
Just as the buddha body is immeasurable, bright and clear sounds and echoes
are also immeasurable. Morality, meditation, wisdom, and so forth, all the
meritorious virtues of all the buddhas are immeasurable. Likewise, in the sa-
tra on Esoteric Traces, the three esoterica (sanmi) are mentioned. In this
[scripture] they are described in great detail. Furthermore, when the Buddha
was first born he fell to the ground and walked seven paces. From his mouth,
by himself, he uttered speech. When his speech ended he was then silent just
like all infants and he neither walked nor spoke. He ate mother's milk for
three years. All of his mothers raised him and he grew up gradually.
Nevertheless, buddha bodies are without number and exceed [the number of]
all world systems. Because there are living beings they [buddhas] appear like
ordinary people. This is because when ordinary people are born their bodies
divide all roots and their [capacity for mental] consciousnesses is imma-
ture.
ls
17 See Lamotte, trans., Le traite (cf. n. 7 above). On many different names by which
this text was known in medieval China and on the attribution of the text to Nagiirjuna see
Paul Demieville' s review of the second volume of Lamotte's translation (originally pub-
lished in 1950), in Choix d'etudes bouddiques (1929-1970) (Leiden: E. J. Brill 1973) 470,
n. 1,475-476. For the biography of Kumarajlva see Gaoseng zhuan 2, T 2059,
50.330a-333a; see also Kenneth Ch'en, Buddhism in China: A Historical Survey (prince-
ton: Princeton University Press 1964) 81-83. For the problem of Nagarjuna's existence
and dating in Indian literature see Joseph Walser, "Niigarjuna and the Ratniivali: New
Ways to Date an Old Philosopher," Journal of the 1nternational Association of Buddhist
Studies 25/1-2 (2002) 209-262.
18 Dazhidulun 1, T 1509, 25.59a2-8; cf. Lamotte, Le traite (cf. n. 7 above) 1:19-20.
At this stage of the project, Kumarajiva conceptualizes the three eso-
terica as one of the virtues of a buddha that enable him to do things
with his body, speech, and mind that are beyond the capabilities of
ordinary living beings and alludes, of course, to the satra treated
above. Later in the work, Kumarajiva further develops the idea of
the three estoerica when he analyzes a scriptural passage stating that
the spiritual power of the Buddha Sakyamuni is his ability to appear
in the SaM world system in the desire realm to preacp. the Prajiia-
paramita (Perfection of wisdom) to the bodhisattva-mahasattvas.
Nevertheless, the Buddha's real preaching is accomplished without
speech. Kumarajiva explains that the Buddha's method of preaching
is produced from his awakened observation - but awakened observa-
tion is a coarse thing. He elucidates further:
The Buddha does not move, but constantly abides (ru A, lit. "enters") in
dhyana-meditation. Because of causal connections due to meritorious virtues
of previous lifetimes, the edges of his body emit sounds and resonate with
objects like echoes. Just like a heavenly skilled musician [gandharva] pro-
duces sounds spontaneously and, furthermore, just as a gives a
person whatever he wants instantly - if he wants clothing and quilts, food
and drink, or music, personal lusts and necessities, he acquires them all
spontaneously. The Buddha is also like this. From the edges of his body, all
the pores of the skin, there are sounds in accordance to his thoughts that
preach the Dharma spontaneously. The Buddha does not reflect upon these
things and also he does not distinguish between them. It is just like that
which he preached in the Sutra on the Vajrasattva Esoteric Traces: "The
Buddha possesses three esoterica - body-esoterica, speech-esoteric a, and
mind-esoterica." All gods and humans neither comprehend nor know them.
[If] there is a single assembly of living beings, some see the Buddha's body
as yellow gold, white silver, or all the colors of miscellaneous treasures. [If]
there are people who see the sixteen-foot body of the Buddha, some see one
li, ten Ii, a hundred, a thousand, ten thousand, a hundred thousand, up to
something boundless, measureless, everywhere in space. Things like this,
and so forth, are called body-esoterica.
As for speech-esoterica, [if] there are people who hear the sounds of the
Buddha's voice at one Ii, there are those who hear them at ten Ii, a hundred,
a thousand, ten thousand, a hundred thousand, numberless, measureless,
everywhere in space. In an assembly, some hear [the Buddha] preach about
giving liberally [dana], others hear him preach about observing the precepts,
and others hear him preach about enduring abuse, making seminal progress,
dhyana-meditation, and wisdom. [It is] like this up to the twelve sections of
the canon.
19
In a dharma assembly of 80,000, what each person hears is in
accordance to their mind. This is called speech-esoterica.
2o
Kumarajiva then goes on to illustrate how the sounds of the Bud-
dha's voice may be heard anywhere and everywhere in the universe
by relating a story about the Buddha's disciple Mahamaudgalyayana
(Mulian Jl). This monk famous for his psychic power (shenzuli t$
fEj]) desired to know if the sounds of the Buddha's voice are near
or far. So, at this instant he manifested his mental power and pro-
duced an immeasurable number of buddha world systems. He report-
edly heard the sounds of the Buddha's voice no different than if he
were nearby. In the world systems he produced the buddhas and the
people in the great assemblies were very large in stature. Because of
this Mahamaudgalyayana stood on his bowl out of respect. The
disciples of those buddhas insultingly asked those buddhas about the
small people
21
of other world systems like Mahamaudgalyayana.
They wanted to know where the insects on the heads of these people
came from and why they were allowed to walk around wearing the
19 The twelve sections of the canon (shierbu jing also shierfen jing +=*
are (1) sLltra (xiuduoluo scriptures, discourses), (2) geya (qiye fi1:lt, long
poems), (3) vytikara!1a (shoujijing :':rr:'.Ii.lJl:, scriptures about those receiving prophesies of
future buddhahood), (4) gtithti (jiatuo 11mWt;, verses, songs of praise), (5) udtina (youtuona
song of joy, utterance), (6) nidtina (yinyuanjing l29#lM[, scripture on the causal
connections or origins of something), (7) avadtina (apotuona parables, meta-
phors, stories, illustrations), (8) itivfttaka (rushiyujing scriptures in which the
Buddha tells of the deeds of his disciples and others in previous lives), (9) jtitaka (ben-
shengjing 1js::i:Ii.lJl:, stories of the previous lives of the Buddha), (10) vaipulya (guangjing fiIi
broad scriptures, general scriptures), (11) adbhuta-dhanna (weizengyoujing '* 'i\' fjli.lJl:,
scriptures telling about the inconceivable spiritual penetrations of the Buddha), (12)
upadesa (lunyijing treatises and exegeses). See Dazhidu lun 33, T 1509,
25.306cl6-20.
20 Dazhidu [un 10, T 1509, 25.127c, cf. Lamotte, Le traite (cf. n. 7 above) 1:559-560.
21 The large sentient beings in the other world systems referred to beings of ours and
other world systems - such as Mahamaudgalyayana in this story - in a derogative manner
as "small people" (xiaoren IJ\ A) because their stature was considerably smaller than the
people in the other systems. The implication is that beings who reside in world systems
that are considerably impure possess bodies that are small in size and the lengths of their
lives are comparatively short.
clothing of sramalJas. Those buddhas responded saying, "Do not
take these people lightly. From here, in the eastern region, crossing
over measureless buddha lands, there is a Buddha called Sakyamuni.
Here is that Buddha's disciple who specializes in psychic power."
That buddha told Mahamaudgalyayana, "You are investigating
where the sound of the Buddha crosses over measureless, hundreds
of thousands of kalpas - but one is able to ascertain their bounds and
limits. Once again, the Buddha leaves the world to resolye the doubts
of living beings; hence, he preaches the Dharma. Here he does not
respond to the difficulty of explaining how this is possible just like
he does not respond to the question how does one remove ignorance?
The Buddha is also like this. He does not respond to the question
what is Buddha by answering."22
For KumarajIva, the concept of the "three esoterica" is essentially
a convenient way to talk about the inconceivable properties of
buddhahood. Although living beings continually produce the "three
types of karma" (sanye Skt. tri-karma): referring to karma pro-
duced via the body (shenye kaya-karma) , speech (yuye
vak-karma), and the mind (yiye what a buddha
produces with his body, speech, and mind is incomprehensible by
ordinary, rational thought. Perhaps because a buddha acquiesces to
the non-production of dharmas and does not infuse a misplaced no-
tion of self into his actions - hence, in this sense he does not "act"
like ordinary beings - what a tathiigata does fundamentally tran-
scends what ordinary people can do and, thus, can be used like an
expedient means (jangbian 1J{, upaya) to preach the Buddha-
dharma. The three esoterica are also related to the emerging doctrine
of the three bodies of the buddha (sanshen and how the true
body (zhenshen or Dharma body (jashen t:!i'l', dharmakaya)
produces other types of bodies in response to the needs of living be-
ingS.
23
22 Dazhidu fun 10, T 1509, 25.127c-128a; cf. Lamotte, Le traite (cf. n. 7 above) 1:560-
562.
23 See, for instance, Sharf, Coming to Terms with Chinese Buddhism (cf. n. 10 above)
100-111; and David W. Chappell, "Chinese Buddhist Interpretations of the Pure Lands,"
in Buddhist and Taoist Studies I, Asian Studies at Hawai'i 18, ed. Michael Saso and David
THE MYSTERIES OF BODY, SPEECH, Al'ID MIND 315
Kumarajlva's translation of the Rem-vang jing 1=.:J.& (Book for
h[;mane kings, T 245) also introduces other concepts that will be
conflated with the three esoterica by later scholiasts. In this connec-
tion, the scripture says: "Of the Tathagata's three types of karma,
there is no limit to his virtue.,,24 Furthermore, "The so-called three
types of karma are the same as morality (jie itlt), the same as sight
(jian .%), and the same as learning (xue By linking this
specialized conceptualization of karma to a buddha or bodhisattva's
"virtue" (de 1i), the concept of the three esoterica is primed to frame
the true meaning of nirvalJa to a Mahayana bodhisattva.
The Sui and early Tang period
In the late sixth century, Tiantai Zhiyf6 used the concept of the
"three esoterica" to describe the fruits attained by a bodhisattva
practitioner who magnifies his practice continually and cultivates the
virtues associated with buddhahood. In his Miaofa lianhua jing
wenju (Passages from Lotus Sutra of the Sublime
Dharma, T 1718), Zhiyi concretized the one Buddhist path leading to
nirvalJa in terms of the three actions, which calls to mind the three
virtues, the three esoterica, and other lists of three. The three eso-
terica are the fruit of cultivating the three types of karma appropri-
ately:
In summary, these three actions
27
are the path to nirviil}a. In summary, with
respect to the condition of the three virtues (sande = iliff,) in practice, the
W. Chappell (Honolulu: University of Hawai'i Press 1977) 23-53.
24 Renwang bore boluomijing 1=xJl*i'ffi'!.ilHl1lliif 1, T 245, 8.828a3.
25 Renwang bore boluomijing 2, T 245, 8.831b6.
26 For the biography of Zhiyi see Xu gaoseng zhuan 17, T 2059, 50.564a-568a; see
also Leon Hurvitz, Chih-i i\'l'ifi (538-597): An Introduction to the Life and Ideas of a Chi-
nese Buddhist Monk, Melanges chinios et bouddhiques (Bruxelles: 1 'Institut Beige des
Hautes Etudes Chinoises 1962), and Ch'en, Buddhism in China (cf. n. 17 above) 303-313.
27 The three actions (sanxing .==:: 1'r) are essentially interchangeable with the concept of
the three types of karma (sanye). They refer to body-actions (shenxing speech ac-
tions (kouxing piT), and mind actions (yixing :\t.iT). See Dazhidu lun 36, T 1509,
25.325b21-26.
condition is designated as "peace and bliss" and the, path is designated as
"practice. "
The Treatise on the. Great [Perfection of Wisdom] says: "The bodhisattva
from his initial arou'sal of the bodhicitta lfaxin constantly
the path of practice [leading to] nirvli:l)a.,,28 When the seed [is planted], he
uses this dharma of the three practices, [and] instructs about the three types
of karma (sanye) by practice. Because the three types of karma are pure,
they are pure in the six roots.2
9
If the six roots are pure, by giving rise to the
marks and resembling. the understanding [of an enlightened being], they en-
ter into reality. When they are brought to fruition they are called buddha
eyes, ears, and so forth. The seed is called the practice of calmness and the
fruit is called the virtue of severing (duande Iltfitti). The seed is called the
practice of insight [visualization] and the fruit is called the virtue of wisdom
(zhide t'1i&\). The seed is called the practice of compassion and the fruit is
called the virtue of kindness (ende
Furthermore, if the seeds are called the three types of karma, the fruits are
called the three esoterica (sanmi). When the seed [is planted] compassion
leads to benefiting others by means of the three types of karma. When it
comes to fruition, it is called the inconceivable transformation of the three
wheels (sanlun Like this, when in visualization, he does not discrimi-
nate between things. All dharmas possess the quality of peace and bliss. All
living beings will then [attain] great nirvlil)a, quiescence from which it is
impossible to return. Practicing in the wrong path lfeidao [the mun-
dane world], he completely accomplishes the path to buddhahood. This then
is the incomparably bright practice of peace and bliss. This practice accords
with the meaning of nirvlil)a.
That [the Lotus Satra] says: "Again, [if] there is a single practice, this is a
. tathligata practice. A tathligata is a person and peace and bliss is a dharma.
Tathligatas are people of peace and bliss. Peace and bliss is the tathligata's
dharma." Speaking of it in summary, its meaning is not different and its
particulars also are not different. 30
28 Typically Da lun is a reference to the Dazhidu [un. This is not a direct quote;
however, it bears similarities with the following passages: Dazhidu lun 41, T 1509,
25.362c27-363a26 and Dazhidu [un 54, T 1509, 25.448b28-c3.
29 T)J.e six roots (liugen ;\":m.) are those of the eyes, ears, nose, tongue, body, and mind
that respond to sense objects.
30 Miaoja lianhuajing wenju 8, T 1718, 34.1l8b27-c18.
Zhiyi implies that when one arouses the bodhicitta and becomes a
bodhisattva that places him on the path toward nirvii':la. The cultiva-
tion of appropriate bodhisattva practices of body, speech, and mind
is like planting a seed that will eventually and inevitably bear fruit
causing that person to acquire the three esoterica described in detail
in the Dazhidu fun. Zhiyi introduces the concepts of the "three vir-
tues" and "three wheels" as roughly interchangeable with the three
esoterica. The three wheels are defined as referring not only to the
karma produced by body, speech, and mind,3! but also to magical
power (shenzu t$ ,IE), preaching the Dharma (shuofa and
remembrance and recollection (yinian which bear a close
resemblance to the spiritual penetrations (shentong t$iffi) acquired
by advanced bodhisattvas and manifested by enlightened beings. The
spiritual penetrations come in lists of five or six, and include the
ability to work miracles, supernormal hearing, the ability to read
minds, recollection of one's past lives, the ability to discern the
previous lives of others, and comprehension that one's spiritual state
is no longer plagued by any form of defilement.
33
The spiritual
31 See Dabaojijing jcWfj\*, 52, T 310, 11.303c9-11; ro1186, T 310, 11.493b17-24;
Dafangguang fa huayan jing shu jc1JJlI(f:lll*M#,i!j[ 16, T 1735, 35.621a4--7; and roll 19,
T 1735, 35.642b8-9.
32 See Dafangguangfo huayanjing jc1JJlI(17Il*Mfi' 6, T 278, 9.435c13; roll 31, T 278,
9.596b27-28; and Dafangguang fa Iwayan jing shu 11, T 1735, 35.574b15-16. Later,
during the Song period they were described as an enlightened being's spiritual
penetrations (shentong), or powers of bodily transformation, which is associated with the
body; his discriminating mind (jixin llc,L.'), which is able to understand the thoughts of
others, which is associated with mind (yi); and his oral powers of teaching and warning
(jiaojie or correct teaching (zhengjiao associated with speech (kou). See
Jin'gang jing zuanyao kandingji 1, T 1702, 33.178c18-21.
33 The five spiritual penetrations (Ch. wu shentong E1$lm, wutong Elm, Skt. paiica-
abhijl1iifz) are the 1) divine eye (divyacakeus, tianyan tong 2) divine ear (divya-
srotra, tianer tong ;JC:&:lm), 3) knowledge of the thoughts of others (paracittajiiiina, taxin
tong ft!l,L.,lm), 4) recollection of former incarnations (pl7rvanirviisiinusmrti, suzhu tong tEl
ftlm), 5) deeds leading to magical power and release (rddhivimokeakriyii) or direct experi-
ence of magical power (rddhisiikeiitkriyii, shenjing tong :j<$ilm). See Apidamo da piposha
lun ([Abhidharma-JMahiivibhiieii) 411, T 1545, 27.728b12-24;
727b22-24. The six spiritual penetrations (Ch. liu shentong 7\1'$lm; Skt. earjabhijiitih) are
1) psychic power (rddhividhijiiiina, shenZLt tong 1'$ )Elm), magical power; 2) heavenly ear
(divyasrotrajiiiina, tianer tong ;JC:&:lm) , supernormal hearing; 3) cognition of others'
penetrations are marks or proofs of an advanced. bodhisattva and aid
him in liberating beings.
In the Guanyin x.uanyi (Arcane meaning of Avalokites-
vara, T 1726), which is reportedly a discourse spoken by Zhiyi that
was recorded by his disciple Guanding (561-632),34 Zhiyi is
even more explicit in his practical application of the three esoterica
and his conflation of them with the spiritual penetrations:
Next, as for spiritual penetrations, if he desires to convert others he displays
the three esoterica. Spiritual penetrations are displaying the material body
[rilpakaya] and, by expedient means, displaying thoughts similar to living
beings.
35
Preaching the Dharma is displaying vocal [karma] in accordance to
its various sounds. This is the second order of converting others. Making
offerings to all the buddhas forms one's conduct. Not only flowers, incense,
and the four things [food, clothing, bedding, and medicine] are offerings: in
accordance with cultivated conduct is the offering of the Dharma. With re-
spect to offerings, the greatest, the siltras say: "If you follow my words, then
make offerings to the Buddha."36 He practices by requesting to teach. This
forms his personal conduct. As for perfecting living beings, this forms the
conversion of others. Among the four dignified rites of the bodhisattva,37 he
still does riot forget living beings. How much more so does he enter all ap-
thoughts (paracittajfilina, taxin tong 1t!!.,L.'Jili), the ability to read minds; 4) recollection of
past lives (parvanirvlislinusmrtijlilina, sllming tong fEl1l'l:rJili), 5) heavenly eye
jlilina, tianyan tong * HIllJili) , the ability to discern the previous lives of others; and 6)
cognition of the extinction of outflows loujin tong ooillJili), a state in
which one is no longer plagued by any form of defilement. See Apidamo da piposha [un
102, T 1545, 27.530aI8-blO; and Dazhidu lun 28, T 1509, 25.264a-266b; see also
Lamotte, Le traite (cf. n. 7 above) 4:1809-1838. By means of the spiritual penetrations a
bodhisattva purifies his buddhaketra; see Mohe zhiguan 2a, T 1911, 46.14a-b.
34 For the biography of Guanding see Xu gaoseng zhuan 19, T 2060, 50.584a25-
585bll. See also Chen Jinhua, Making and Remaking History: A Study of Tiantai Sectarian
Historiography, Studia Philologica Buddhica Monograph Series no. 14 (Tokyo: The
International Institute for Buddhist Studies 1999); Linda Penkower, "In the Beginning ...
Guanding lim (561-632) and the Creation of Early Tiantai," Journal of the International
Association of Buddhist Studies 23/2 (2000) 245-296.
35 Literally, "[beings possessing] passions/affections" (qing
36 Cf. Da niepanjing 5, T 374, 12.396a2.
37 The four dignified rites of the bodhisattva (siweiyi II!! ffil(; {j'ji:) refers to all the actions
and practices of a bodhisattva, but particularly walking (xing 'iT), resting (zhu 11), sitting
(zuo and lying down (wo See Dafangguang fo huayanjing 5, T 278, 9.424a28.
pro aches to Dharma and pure buddha lands. All [bodhisattvas] liberally
benefit all living beings. Hence, this one expression forms the conversion of
others.
38
The three esoterica are what the advanced bodhisattva Avaloki-
tesvara actually manifests when he appears in the mundane world
and performs actions to save and benefit living beings. Bodhisattva
actions are the three esoterica because they lead living beings to ac-
cept the Buddhadharma. Zhiyi's position, mediated by Guanding,
bears some surprising similarities to that which will be expounded
by the early Tantric masters a century later. A bodhisattva, such as
Avalokitesvara, uses the three esoterica when he does what the Bud-
dha would do to convert others. If he manifests in physical form or
gives of himself to benefit beings, or if he makes offerings to the
Buddha, that is body-esoterica. If he uses his voice to preach the
Dharma, that is speech-esoterica.
In his own writings, Guanding equates the three esoterica to the
supreme dharma or superior teaching, the final teaching of the Bud-
dha as described in (Tanwuchen {t;f!lHli, 385-433)39
"Northern Translation" of the Mahayana Mahiiparinirviir,ta Satra. In
his Daban niepan jing shu (Commentary on the Ma-
hiiparinirviir,ta Sidra, T 1767), Guanding explains that although there
are three esoterica, they are no different than the four virtues of con-
stancy, joy, personality, and purity. These are the fundamental
characteristics possessed and taught to living beings by the Buddha:
Next, as for the Abhidharma (duifa this, verily, is the three esoterica.
Life-force (ming in) is the mind-esoterica, the pacification of physical power
[lust] is body-esoterica. Non-obstruction is They [the three
esoterica] are also called the four virtues.
40
Mind is the virtue of constancy.
38 Guanyin xuanyi 2, T 1726, 34.888b20-28.
39 For the biography of see Gaoseng zhuan 2, T 2059, 50.335c-337b;
see also Ch'en, Buddhism in China (cf. n. 17 above) 88,114.
40 :The four virtues (side are constancy or permanence (chang 'ill); joy (Ie
personality, the soul or the self (wo :f,Ii;); and purity (jing *). "Northern
Translation" of the Mahayana MahaparinirvQ(la Satra discusses these four terms as a
group 135 times beginning with Dahan niepanjing 1, T 374, 12.365a21; see
also Dahan niepanjing shu 1, T 1767, 38.44b5-6.
320 . RICHARD D. MCBRIDE, II
Body is taking pleasure in the self. Speech is verily-the virtue of purity. If
one can obtain this [state of] mind, all dharmas everywhere are nothing but
these five. The Buddha is [also] endowed with these five; hence, he bestows
them on people. He bestows the five constancies
41
also without exhaustion.
What I mean by "without exhaustion" is verily the five fruits.42
Although Guanding's numbered list games become muddled because
he conflates the three esoterica with the four virtues and also the five
constancies and five fruits, it is clear that he them as
representative not only of appropriate bodhisattva practice but also
of the manifestation of the fruits leading to buddhahood. Guanding
uses the concept of the three esoterica to correspond to three aspects
of the attainment of enlightenment in another passage in this
commentary:
The old one [Faxian's translation of the Nirva1J-a Satra?] says this chapter
illuminates the three esoteric a and does not illuminate the three virtues. It
also says this chapter illuminates the three virtues and does not illuminate
the three esoterica. Furthermore, it says that it only illuminates the three eso-
terica, which verily explain the three virtues. It explains that speech-eso-
terica is prajiia, it explains that body-esoterica is the Dharma body, and it
explains that mind-esoteric a is liberation. The [understanding of these] three
depend on [each other] but are not the same, and yet are three expressions.
Presently my elders use these three expressions to extinguish [reliance on]
41 The five constancies (wuchang n 11\) are (1) humaneness (ren 1=), which is equated
to not killing; (2) righteousness (yi which is equated to not stealing; (3) propriety (Ii
t!), which is equated to not misusing sex or participating in licentiousness; (4) wisdom/
knowledge (zhi which is equated to not drinking alcoholic beverages; and (5) faith
(xin which is equated to not using false speech. See Guang hongming ji 3, T
2103, 52.107b22-25.
42 Daban niepan jing shu 3, T 1767, 38.57c6-10. The five fruits (wuguo n*, SkI.
paiicaphaIani) are (1) fruit ripening divergently (yishuguo which refers to the
idea that pleasure and wholesomeness are in different categories and that one's present
situation accords in pain and pleasure with past wholesome and unwholesome deeds; (2)
fruit of the same order (deng/iuguo which refers to wholesomeness reborn from
previous wholesomeness; (3) present position and function fruit (tllyongguo ffl *),
which refers to the rewards of meritorious actions from previous lives; (4) superior fruit
(zengshangguo :tfli 1:*), which refers to one's position arising from previous seminal
progress and spiritual capacity; and (5) fruit in freedom from miscellaneous bonds
(ZaxigliO which refers to the fruit of nirv[l1:za. See Apidamo jushe fun
(Abhidharmakosa) 17, T 1558, 29.91a25-b8.
literary descriptions. Nevertheless, they are named the three esoterica also
[because] they are mutually similar. Following many classifications I regard
them as the three esoterica.
43
In spite of the fact that neither Faxian's nor transla-
tions use either the term "three esoterica" or "three virtues," the
point of Guanding's rhetoric is obvious. Although the three esoterica
are often equated to the three virtues, they are, in fact, a slightly
different way of describing the characteristics of an enlightened be-
ing. Each of the three concepts equated individually to the three eso-
terica correspond to different ways of describing the ineffable goal
of bodhisattva practice. The attainment of prajiiii is the new goal of
the Mahayana bodhisattva described verbally in the Prajfiaparamita
literature and other Mahayana siltras. The Dharma body - the
dharmakiiya or dharmadhiitu - the universe - is the very bodily form
of reality as it is with all of its endless transformations. And libera-
tion is, for lack of a better way to express it, the state of mind en-
joyed by advanced bodhisattvas. These three esoterica are what
bodhisattvas seek and use in their practice of converting others.
Guanding describes this clearly in the final passage we will observe
from his commentary on the NirviiIJa Sfitra:
If there are bodhisattvas who dwell peacefully below, the second [issue I
treat here] broadly illuminates converting others in four [practices]. First,
transforming their bodies to be like the Buddha; second, returning to their
original location; third, severing other's delusion; and fourth, manifestation
of the three esoterica.
This manifestation, furthermore, is of four kinds: first, speech-esoteric a; sec-
ond, body-esoterica; third, doubly illuminating speech-esoterica; the former,
verily, is a different adaptation to all sounds and the present one illuminates
a different adaptation to all dharmas; and fourth, rnind-esoterica.
44
In other words, bodhisattvas are expected to manifest the three eso-
terica while they abide in the world and it is what they use to convert
beings.
43 Dahan niepanjing shu 9, T 1767, 38.86c28-87a4.
44 Dahan niepanjing shu 22, T 1767, 38.167a9-13.
The Chinese Madhyamaka (Sanlun exegete Jizang
(549-623)45 uses the concept of the three esoterica as a heuristic de-
vice to describe the symbolic and practical importance of the name
of A valokitesvara in the famous "Gateway to everywhere" chapter
(pumen pin of the Lotus Satra. Rather than regarding the
three virtues and the three esoterica as different aspects of the same
thing, in his Fahua xuan fun (Arcane treatise on the Lotus
Satra, T 1720), Jizang combines them and introduces th,e concept of
the "virtues of the three esoteric a" (sanmide to describe the
latent power in the name of the bodhisattva:
Fourth is one pair of name virtues. Avalokitesvara is said to have three kinds
of names. The "Gateway to everywhere" chapter calls them the virtues of the
three esoteric a (sanmide). The three names are: first, the sound "observer of
the world" produces wholesome verbal karma in living beings; second, the
thought "observer of the world" produces wholesome mental karma in living
beings; and third, the body "observer of the world" produces wholesome
physical karma in living beings.
The virtues of the three esoterica of the "Gateway to everywhere" chapter
are: first, knowing the thoughts of others everywhere; second, preaching the
Dharma everywhere; and third, [manifesting] spiritual penetrations every-
where. If the three names, it is said, produce the three wholesome [actions]
in living beings, then they produce wholesome meaning in their environs.
The virtues of the three esoterica then respond to living things and exhaust
wholesomeness.
46
The Lotus Satra does not actually use the term "virtues of the three
." esoterica." Nevertheless, as we have seen before, the concept of the
three esoterica had become a catch phrase to describe the powers,
capabilities, and responsibilities of bodhisattvas. In his Fahua yishu
(Commentary on the Lotus &"ltra, T 1721), Jizang uses the
three esoterica in roughly the same manner used by Kumarajlva - to
explain the special powers possessed by advanced bodhisattvas and
buddhas. He invokes the three esoterica in the tenth of ten general
45 For the biography of Jizang see Xu gaoseng zhuan 11, T 2059, 50.513c-515a; see
also Ch'en, Buddhism in China (cf. n. 17 above) 132-134.
46 Fahua xuan lun 10, T 1720, 34.447b20-26.
meanings of the doctrine that the Buddha dwells constantly in purity
arid meditation:
Tenth, the three esoterica benefit living beings. It is said that mind-esoterica
is entering meditation, body-esoterica is emiting light, and speech-esoteric a
is preaching the Dharma. Furthermore, if one is quiescent, it is used for leav-
ing [sm!1sara]. If used being quiescent, [it] becomes entering [samadhi]. For
this reason, the function of quiescence is non-hindrance; hence, it is a name
for leaving and entering.
47
He then goes on to describe how the act of entering samiidhi is the
manifestation of mind-esoterica. In other words, the three esoterica
are manifestations of the Buddha in samiidhi responding in accor-
dance to circumstances. Entering samiidhi is thus understood as the
"good medicine" causing a practitioner to cultivate unexcelled bodhi.
In his Niepan jing youyi (Musings on the Nirvii]Ja Su-
tra, T 1768), Jizang not only reconfirms the connection between the
three virtues, three esoterica, and the three types of karma but ex-
plains how the three esoterica demonstrate that the bodhisattva has
transcended the cycle of rebirth and death (saf!Lsiira):
Furthermore, I will clarify that the three virtues are explained as the three
esoterica of the Tathagata. Hence, Kasyapa asked desiring that the Buddha
would explain the "subtle esoterica" (weimi in detail and preach it to
living beings. Just like this, the three esoterica and the class of the four
marks
48
are an example. Also, these match the three types of karma of ordi-
nary people because the three types of karma of ordinary people are not eso-
teric (mi). Furthermore, by matching them to the three hindrances of sal'(lsa-
ra
49
one illuminates the three virtues of nirval}a. By matching them to the
hindrance of recompense [the inability to hear the Dharma due to bad rebirth
because of unwholesome karma] one clarifies the Dharma body. By match-
ing them to the hindrance of karma [the inability to accept the Dharma be-
47 Fahuayishu2, T 1721, 34.469al-4.
48 The class of the four marks (sixiang pin here probably refer to birth (sheng
!:t.), abiding existence (zhu it), change (yi and extinction (mie
49 The three hindrances of sal!lsiira (sanzhang Skt. iivarartatraya) are the hin-
drance of karma 6'ezhang the hindrance of defilements (jannaozhang and
the hindrance of recompense (baozhang also called the hindrance of ripening
(yishuzhang See Apidamo jushe lun 17, T 1558, 29.92b23-24.
cause of language, behavior, mind], one clarifies ligeration. By matching
them to the hindrance of defilements, one clarifies prajfiii. Because sa1'[lsiira
is merely the three hindrances, nirvii1Ja only has the three virtues.
5o
Once again we return to the idea that the three esoterica .manifested
by buddhas and the three types of karma produced by ordinary living
beings are correlated. But to Jizang the three esoterica have another
meaning apart from referring to the powers of body, speech, and
mind produced by advanced bodhisattvas and buddhas. As with
Guanding, they also compare with the three virtues or Mahayana-
style enlightenment - Dharma body, prajfiii, and liberation.
In his lin' guangming jing shu (Commentary on the
Sutra of golden light, T 1787), Jizang is even more explicit in his
promotion of the three esoterica as a heuristic device to classify and
explain the significance of the preface to the sutra and its relation to
the rest of the sutra:
"Vulture Peak in Rajagrha is like the constant Sakyamuni. .. "
There is a difference, in particular, between what is introduced in the preface
(xu J'F) and the remaining scriptural prologue (yujingxu Merely take
the mystical endowment (mingjia of the three esoterica as the pro-
logue. The three esoterica are the esoterica of body, speech, and mind (shen
kou yi mi r:l When he wants to preach this sutra, his body resides in
Vulture Peak and explains the superiority of this sutra. Verily, this is the
body-esoterica. His mind recollects the name of dharmas and protective
maintenanceS! of the four buddhas up to procedures for repentance, and so
forth. Verily, this is mind-esoterica. Furthermore, he desires to preach these
deep procedures [dharmas] in the future. This, verily, is speech-esoterica.
Take this as a form of faith in the mystical endowment of the three esoterica.
Hence, the form of faith is suspicious, resonates with the buddhas of the four
directions, 52 and preaches on the fruit of limitations of lifespan. Furthermore,
50 Niepanjing youyi 1, T 1768, 38.237a15-21.
51 Protective maintenance (huchi iiiM', Skt. anuplilanli) is a by-product of practicing
prajfilipliramitli. It is a type of recollection that enables bodhisattvas not to begrudge their
physical bodies and their lives. See Dazhidu lun 74, T 1509, 25.578a28-blO.
52 The buddhas of the four directions according to the Suva77;laprabhlisa Satra are
Alajobhya (Achu in the East, Jeweled Marks (Baoxiang JU) in the South, Amita-
yus (Wuliangshou in the West, and Subtle and Miraculous Sounds (Weimiao-
the form of faith in the three esoteric a resonates with dreams and preaches
procedures of repentance, and so forth. This is because the Four Heavenly
Kings and so 'forth made a vow to magnify the sLltra up through preaching
the chapter on forsaking the body. It is said that the three esoterica are pro-
moted in the prologue.
53
Because the current preface to the lin' guangming jing SA (Su-
van:taprabhasa Satra, Satra of golden light, T 663) by the Chan
monk Cijue %tJl: dates to the Five Dynasties period (ca. 907-960)
and because the passage Jizang quotes at the beginning is not found
in any extant version of .the scripture, I am not sure whether Jizang
refers here to an earlier preface that circulated with the satra or to
the long gatha that comprises the prologue to the satra (jingxupin
r&).54 Although the concepts of body and mind (shen yi) appear
toward the end of the gatha, nowhere is there a direct reference to
the idea of the mystical endowment of the three esoterica. Neverthe-
less, the three esoterica serves conveniently and successfully as a
classification device.
Like Jizang, the Chinese Yogacara (Faxiang monk Ci'en
Kuiji (632-682),55 a disciple of Xuanzang (ca. 600-
664), uses the concept of the three esoterica to explain how the Bud-
dha appears to do things although, in actuality, he is constantly in
meditation. Also, like Jizang, in his Miaofa lianhua jing xuanzan
t:Ji:J:tJ&"E"Jr (Arcane praise of the Lotus Satra of the sublime law, T
1723), Kuiji emphasizes the three esoterica as comprising the tenth
and final proof of this situation:
Question: The Buddha has his mind in complete meditation and his conduct
[is to] dwell permanently in meditation. He manifests majesty by not giving
rise to meditation on extinction. What does he need now enter?
Answer: There are ten meanings. First, entering and leaving, he moves
following connections because in stillness he benefits living beings. Second,
sheng jlijjryil.\') in the North. See Jin'guangmingjing I, T 663, 16.335bI2-13.
53 Jin'guangmingjingshu I, T 1787, 39. 160c17-25.
54 Jin'guangmingjing I, T 663, 16335b5--c15; see 335c12 for body and mind.
55 For the biography of Kuiji see Song gaoseng zhuan *il'lJ{'1I!{$ 4, T 2061, 50.725b--
726c; see also Ch'en, Buddhism in China (cf. n. 17 above) 320-321.
if he does not enter meditation, he emits light witholJt cause and manifests
auspicious signs. If he does not enter meditation he fears that there will be
no auspicious signs of buddhahood. Third, when he wants to preach the
Dharma he shows that he deliberates opportunity. Fourth, because he mani-
fests the bizarre profundities of the Dharma he enters meditation and insight
causes reverence. Fifth, manifesting wisdom necessarily attests to principle
(li :1); entering meditation he is actually able to arouse it. An instructor later
teaches him causing him to cultivate meditation: hence, [producing] wisdom.
Sixth, showing meditation is a fullness of wisdom. the Dharma
shows a fullness of wisdom. This is because entering meditation manifests a
fullness of meditation. Seventh, due to the regulations of the final age [of the
Dharma], persons who preach the Dharma must first calm their minds them-
selves. Eighth, he shows wholesome pensiveness and the marks of intelli-
gence. Also, he causes other people to examine phenomena. Ninth, he enters
meditation, manifests auspicious signs, and gives rise to three questions and
answers. If not this then he will ask questions without Maitreya and so
forth. 56 Tenth, he manifests the three esoterica. This is because entering
meditation is mind-esoterica. Emiting light, and so forth, is body-esoterica.
Preaching the Dharma is speech-esoterica.
The ode says: accordance, auspiciousness, examination, profundity, instruc-
tor, fullness, pensiveness, question, esoterica.
57
By associating the threeesoterica to entering meditation, emitting
light, and preaching the Dharma, the evidence suggests that although
Jizang and Kuiji belonged to different exegetical traditions, they
shared a common understanding of the meaning of the three eso-
terica as indicative of the enlightened state of advanced bodhisattvas
and buddhas.
The monk Wonch'lik (613-696) of the Korean state of Silla
fJT*I was a fellow disciple of Xuanzang with Kuiji at Ximingsi IW
56 The context alluded to in this passage seems to refer to the belief, commonly held
among medieval Buddhist scholars and proponents of Yogacara especially, that by enter-
ing into samadhi one can ascend to heaven, see Maitreya, ask him questions, and
receive his clarifications regarding their doubts concerning the complex and often
contradictory doctrines contained in Mahayana Buddhist scriptures. See Etienne Lamotte,
Historie de Bouddhisme Indien (Louvain: PublicationsUniversitaries 1958) 787.
57 Miaofa lianhuajing xuanzan 2, T 1723, 34.679b3-15.
.. in the Tang capital Chang'an his Inwang-gyong so 1=.:
(Commentary on the book for humane kings, T 1708),
Wanch'uk suggests that the three esoterica is the initial theme cov-
ered by the first gathii in Kumarajiva's translation of the Renwang
jing,59 although more direct connections may be found at the end:
"Of the Tathilgata's three types of karma, there is no limit to his vir-
tue."60 Nevertheless, Wanch'uk locates the three esoterica even in
epithets applied to the Buddha:
The first two phrases eulogize the Buddha's three esoterica. The first one of
these phrases eulogizes the esoteric a of the Buddha's body, which contains
all virtues. Because [he] is honored and esteemed by living beings he is
called "World-Honored One." Because his marks are good, perfect, and full,
and [because] he leads and guides living beings, [he] is called "The Guide"
(tosa, Ch. daoshi Because his body does not decay and is similar to
adamant (hlmgang, Ch. jin'gang vajra) it is called the "essence of ada-
mant" (kllmgangch'e, Ch.jin'gangti
Later phrases eulogize two other kinds of esoterica. The first four characters
eulogize his mind-esoteric a (llimil, Ch. yimi), because it is internally realized
true thusness and, marked by the ability to forsake, it is called quiescence of
mental practice.
The later three characters eulogize the speech-esoterica (omit, Ch. yumi). It
is heard the same both far and near, without being obstructed and it is called
"turning the wheel of the Dhmma." If you want more detailed distinctions go
to the Satra on the Three Esoterica (Sanmijing
The Sanmi jing alluded to by Wanch'uk no longer exists. Excerpts
from this scripture - the full title of which was probably Rulai sanmi
jing :fio* - (Satra on the Tathagata's three esoterica) - are pre-
served, however, in two passages from the Boredeng lun shi
(PrajfiapradfpamaZamadhyamakavrtti, Analysis of the treatise
58 For the biography of Wanch'uk see Song gaoseng zhuan 4, T 2061, 50.727b5-14;
see also Shotaro Iida, "The Three Srnpas of Ch'ang An," in Papers of the First Interna-
tional Conference on Korean Studies (Seoul: Academy of Korean Studies 1979) 484-497.
59 Renwang bore boluomijing 1, T 245, 8.827b8-828a8.
60 Renwang bore boluomijing 1, T 245, 8.828a3.
61 Inwang-gyiJng so 1=:EMll:i9iE 2A, T 1708, 33.393a8-15; HPC 1.68bI7--c2.
328 D. MCBRIDE, II
on the torch ojprajfiti, T 1566).62 The original tre;ltise was putatively
written by Nagarjuna and then explored by the Bodhisattva Bha-
vaviveka (Bhavya, Fenbieming d.u.). It was translated by
Prabhakaramitra (Borepomiduoluo d.u.) into Chinese
in 632.63 In particular, the second passage quoted from the Satra on
the three esoterica refers to several wondrous qualities of the Bud-
dha's body. Because no other exegete refers to this siUra it was
probably not well known.
In his Haesimmil-gyong so (Commentary on the Sal'fl-
dhinirmocana Sutra),64 W6nch'uk alludes to the three esoteric a to
explain the existence of various transformation bodies (hwasin, Ch.
huashen 1t::t, Skt. nirmtilJaktiya), reward bodies (suyongsin, Ch.
shouyongshen ::ffl::t, Skt. saf!lbhogaktiya), buddha bodies and bud-
dha lands, spiritual penetrations, Brahma sounds (pomum, Ch. janyin
and so forth:
Mental dharmas are trifling and forsake wisdom. They do not resonate with
explanations of the One. The one sound, and so forth, verily, is the body-
and speech-esoterica of the three esoterica of the Tathagata. With respect to
body-esoterica, it is like it says in the first roll of the Treatise on the {Great]
Peifection of Wisdom:
When the Buddha first turned the wheel of the Dharma it resonated with and
was maintained by the bodhisattvas. When it came from others, they desired
limited buddha bodies. Ascending and transcending empty space and
measureless buddha lands he arrived in the Lotus World System (hwasang
segye, Ch. huashang shijie ..ttt!J'l'-). Because he saw the thusness of the
Buddha's body, he preached a gatha and eulogized a larnent.
65
62 Boredeng lun shi 11, T 1566, 30.108c23-29; and roIll3, T 1566, 30. 121a20-b4. The
Torch of Prajiia later became one of the principal commentaries adopted by the Nara
schools in Japan. See Abe, The Weaving of Mantra (cf. n. 2 above) 214.
63 For the biography of Prabhiikaramitra see Xu gaoseng zhuan 3, T 2060, 50.439c26-
440c3; see also Mochizuki, Bukkyo daijiten (cf. n. 5 above) 5:4277c-4278a. For a biogra-
phy of Bhavaviveka see Mochizuki, Bukkyo daijiten 3:276lc-2762a.
64 Although Wanch'uk was considered an outsider in China, his commentary on the
Sal!ldhinirmocana Satra (Satra on freeing the underlying meaning) was influential later in
Tibet. See Matthew T. Kapstein, The Tibetan Assimilation of Buddhism: Conversion, Con-
testation, and Memory (New York: Oxford University Press 2000) 78-83.
65 Haesimmil-gyong so 1, HPC 1:131cl8-23.
At this point Wonch 'uk quotes verbatim the discussion of the three
esoterica as found in the Dazhidu [un's precis of the Satra on Va-
jrasattva Esote'ric Traces.
66
The intellectual heritage and exegetical
importance of the Dazhidu lun continued through the seventh cen-
tury. No matter what other directions exegetes directed their
interpretations of this concept, KumarajIva's work remained the ba-
sis.
At this point, let us summarize the views held by the exegetes of
the Sui and early Tang period. The three esoteric a refer to the three
types of karma, the three virtues, three wheels, and so forth, which
are produced and manifested by buddhas and bodhisattvas. In order
to explain the term "three esoteric a of the Tathagata" (rulai sanmi ~ D
*':=:!&;), medieval Buddhist scholars connected the three esoterica to
several numbered doctrinal conceptions on the nature of buddha-
hood. The three esoterica are related to the spiritual achievements
and acquisitions of advanced bodhisattvas - A valokitesvara in
particular. It is representative of the enlightened state and the powers
associated with enlightenment. Although its role was not emphasized
previously, the three esoterica appears to have been a term whose
importance was emerging in mainstream Sinitic Buddhism to de-
scribe the nature of buddha hood.
The High Tang Period
Because the mature Japanese Esoteric tradition emphasizes the ritual
replication of the body, speech, and mind of the Buddha as a key
component of the Vajrayana, I originally expected that the concept
of the three esoterica would receive its most explicit explanations in
the writings of the early Tantric masters on efficacious ritual prac-
tices. Subhakarasirpha (Shanwuwei ~ M , { l t , 637-735)67 and Yixing's
-iT (683 or 673-727)68 Da Piluzhena chengfo jing shu (Com-
66 Haesimmil-gyong so 1, HPC 1: 132a-b; cf. Dazhidu lun 10, T 1509, 25. 127c-128a.
67 For the biography of Subhakarasirp.ha see Song gaoseng zhuan 2, T 2061, 50.714b-
716a; see also Chou Yi-liang, "Tantrism in China," Harvard loumal of Asiatic Studies 8
(1945) 251-272.
68 For a study on Yixing that provides a compelling argument that Yixing was actually
mentary on the Sutra on Vairocana's attaining bud,dhahood, T 1796),
for instance, uses the term more than twenty times, suggesting that it
was a major theme in practices the early Tantric master sought to
promote during the first half of the eighth century. The sutra on
which their study is based, their translation of the Da Piluzhena
chengfo jing (also called Dari jing A B *, T 848), completed in 725,
however, only uses the term "approach of the three esoterica" (sanmi
men twice in the seventh roll.
69
This is significant because
the seventh roll is comprised of ritual procedures not found in the
later Tibetan translation of the sutra.
70
Nevertheless, body, speech,
and mind are alluded to repeatedly throughout the SLttra.1
1
The open-
ing of the sutra alludes to important thematic concepts associated
with the three esoterica:
Because of the empowerment (jiachi ;/JON, adhi.ythana) of the Tathagata Vai-
roc ana, there was an exhibition of supernatural power (jenxun 11m, vikrf-
(iita), the storehouse of the inexhaustible ornamentation of the body (shen
wujin zhuangyanzang Like this, there was [another] exhi-
bition of supernatural power, the storehouse of the inexhaustible orna-
mentations of the universality of speech and mind (yu yi pingdeng
It is not produced from the body, or speech, or mind of the Buddha
Vairocana. All places (chu sthana) arise [from it]; [and as for its]
destruction, its fringes are unattainable.
Yet, all of Vairocana's bodily acts (shenye), all of his vocal acts (yuye), and
all of his mental acts (yiye), [and] all places and all times are proclaimed as
true words (zhenyan if( i3, mantra), spoken phrases of Dharma (daojufa i!!-iiJ
born in 673 and not 683 see Jinhua Chen, "The Birth of a Polymath: The Genealogical
Background of the Tang Monk-Scientist Yixing (673-727)," T'ang Studies 18/19 (2000-
2001) 1-39.
69 The concept of the "approach of the three esoterica" (sanmi men =ffi'F5) is alluded
to twice; see Da Piluzhena chengfo shenbianjiachi jing 7, T
848, 18.51c10, 52b2.
70 For an English translation of this chapter see Chikyo Yamamoto, trans., Mahi'ivairo-
cana-Siitra (New Delhi: International Academy of Indian Culture and Aditya Prakash an
1990) 175-199; see also Rolf W. Giebel, trans., The Vairocaniibhisa1'(lbodhi Satra (Berke-
ley: Numata Center for Buddhist Translation and Research 2005) 227-277.
71 There are 10 references to body, speech, and mind clustered, more or less, in two
rolls of the satra. See Da Piluzhena chengfo jing 1, T 848, 18.1a24, 6a25-b8; and roll 5, T
848,18.31bl-20,36c8.
TIIE MYSTERIES OF BODY, SPEECH, AND MIND 331
. i*) in the realm [of living beings] who passions (youqing jie 1Hf

Once again body, speech, and mind are linked with supernatural
powers. In this instance, the miraculous power is attributed to Vairo-
cana. Following the Avatarrzsaka Sutra, the sutra repeatedly uses the
concept of universality (pingdeng That these attributes of Vai-
roc ana are, in fact, a common heritage of all living beings and of the
nature of the universe itself.73 Furthermore, this passage alludes to
the idea that the Tathagata has inexhaustible body, speech, and mind,
which also seems to draw upon themes found in the Avatarrzsaka
Sutra.
74
Although it is not described clearly in the sutra, there are
hints of connecting the three types of karma to particular ritualized
practices, such as mantras.
Several passages in the commentary describe the three esoterica
in detail. Although there is a new frame emphasizing mantra,
SubhakarasiIpha and Yixing's understanding of the three esoterica is
actually quite similar to the views held by the great exegetes of the
Sui and early Tang period in the late sixth and seventh centuries:
Entering the approach of mantra (zhenyan men Jl may be summarized
as being comprised of three phenomena: The fIrst is the approach of body-
72 Da Piluzhena chengfo jing 1, T 848, 18.1a25-b2; cf. Tajima, Etude sur Ie
rocana-satra (cf. n. 5 above) 44-45, 53-54.
73 I translate the Chinese term pingdeng as "universality" and "universal." This
term is often employed as a translation for such Sanskrit words as stimtinya (universal,
equal, common, joint), stimya (equipoise, equality), and stidhiira1;za (universality, common
to all), and so forth. The concept of pingdeng is used 517 times in the 60-roll version of
the Huayanjing, T 278; 636 times in the 80-roll version of the Huayanjing; and 20 times
in the Da Piluzhena chengfo jing, T 848. Although the concept of the "universality of
body" (shen pingdeng) is found twice in Huayan jing, T 278, and 9 times in Huayan jing,
T 279, the concepts of universality of speech and mind are not found in either translation
of the Huayan jingo
74 Body, speech, and mind (shen yu yi) are alluded to 56 times in the Huayanjing, T
279. See, in particular, the chapter on "Pure Practices" (jingxing pin Huayanjing
14, T 279, 1O.69b2O--c22, which emphasizes the fundamental purity of body, speech, and
mind. In the Tibetan version of the sutra, Alex Wayman says this idea is found in the
"Arising of the Tathiigata" chapter; see The Enlightenment of Vairocana (cf. n. 5 above)
49. See also Stephen Hodge, trans., The Mahti-Vairocana-abhisalJ1bodhi Tantra, with
Buddhaguhya's Commentary (London: RoutledgeCurzon 2003) 49.
332 . RICHARD D. MCBRIDE, II
esoterica, the second is the approach of speech-esoteric a, and the third is the
approach of mind-esoterica. These phenomena will be explained in detail be-
low. The practitioner should regard these three expedients (jangbian) as
three kinds of karma that are pure of themselves and [he] will verily be
empowered by the three esoterica of the Tathagata and he will be able [to at-
tain buddhahood] in this life, satisfy the [ten bodhisattva] stages and the [ten]
piiramitiis, not have to pass through successive numbers of kalpas, and culti-
vate in preparation all antidotal practices. Hence, the Large perfection of
wisdom satra (Dapin ::kff'Q) says: If there is a bodhisattva, when he initially
arouses the mind [bodhicitta] he verily ascends the bodhisattva levels and at-
tains [the stage of] non-backsliding. When someone initially arouses the
mind he verily attains unsurpassed bodhi and then turns the wheel of the
Dharma.
Nagarjuna regarded these as distant practices. A person riding a goat will ar-
rive after a long, long time. A horse, then, provides a difference in speed. If
someone rides a person with spiritual penetrations, he will arrive where he is
going in the instant he gives rise to the thought. He is not able to say when
he gave rise to the thought or how he arrived. The marks of the spiritual
penetrations, and so on, do not resonate with or produce doubt, which,
consequently, is the profound purpose of this Slttra.
75
The three esoteric a are still related to the three types of karma, but
they are inherently pure; they are expedients; and they are associated
with the transcendent nature of buddhahood that transforms beings.
Rather than being something that the practitioner does on his own,
the three esoterica are a way of referring directly to the spiritual
penetrations of the Buddha. By relying on these powers - by being
empowered by them - the practitioner is spiritually transformed into
an enlightened being. The powers associated with the three esoterica
are universal and inherently pure. Universality is one of the primary
themes of the sutra and resonates with the popular doctrines of the
inherence of Buddha nature (foxing iiJt't1) in all living beings, the
Tathiigatagarbha (rulaizang ~ Q * ~ ) . This concept is described more
fully in the following passage, in which we can see how the previous
75 Da Piluzhena chengfo jing shu 1, T 1796, 39.579b27-c7; cf. Tajima, Etude sur Ie
Mahavairocana-satra (cf. ll. 5 above) 91-92.
understanding of the three esoterica is extended to indicate their
manifestation in the practitioner through ritual and empowerment:
What was the Dhanna preached by the Buddha at this time? Verily it is the
Dhanna Approach of the triple universality of body (shen), speech (yu), and
mind (yi). Whatever the nature of the Tathagata's three kinds of kanna, they
all arrive at the edge of the sublime extreme of primordial reality. Body is
equivalent to speech. Speech is equivalent to mind (xin). It is similar to the
great ocean's extending the same saline flavor everywhere it goes. Hence, it
is said to be universal. The tenn in Brahmic [Pralqit?] is pada (Ch. botan,
Kor. paltam ifi-ft) ...
Thus, empowennent by the esoterica (mimi ;jiJb1f,<) of body, speech, and mind,
which are universal, is the approach of entry. It is said to be the esoteric seal
[mudra] of body universality (shen pingdeng the true word [mantra]
of speech universality (yu pingdeng and the sublime visualization
of mind universality (xin pingdeng Because they are expedients one
is caused to see the reward body (shouyong shen sal'flbhogakiiya) of
empowennent.
Like this, the reward body of empowerinent, verily, is Vairocana, the omni-
present body. The omnipresent body, verily, is the body of knowledge of the
universality of the practitioner. For this reason, those who abide in this
vehicle go without going and arrive without arriving, which is the name of
the phrase "universality." All beings that have entered therein have, in actu-
ality, neither penetrated it nor been penetrated by it. Hence, it is called uni-
versality. The Dhanna Approach of Universality is the main intent of this
.
The reward body of empowerment represents what Buddhists gener-
ally consider to be the type of buddha that they worship; the type of
buddha represented in icons and images. It is the buddha who resides
in a buddha land displays the physical marks of his
position, and is constantly in meditative absorption (samiidhi).
Through the process of equating concepts, because of their uni-
versality, the commentators link this reward body not only to Vairo-
cana, the symbolic representation of the universe as it is in buddha
form, but also to the universality of knowledge residing in the
76 Da Piluzhena chengfo jing shu 1, T 1796, 39.583a12-15, 20--27; cf. Tajima, Etude
sur Ie Mahavairocana-satra (cf. n. 5 above) 93-94.
practitioner. In essence, the practitioner is transformed into an icon
for the period of ritual practice, which alludes to a new dimension to
practice and understanding. Can the universality of knowledge be
anything other than an allusion to the Tathagatagarbha?
The language of the commentary suggests that the three esoterica
are an expedient and also spiritual power used by Vairocana. This
power transforms and purifies the three types of karma produced by
living beings. In other words, the power of this expedient turns mun-
dane human karma into the acts of an enlightened being. This power,
ultimately, is none other than the inherent Buddha nature that resides
within the practitioner. Nevertheless, the power of the expedient is
dependent upon recognizing the innate purity of original nature. An-
other passage in the commentary relates:
As for bringing expedients to their ultimate conclusion, this means that the
myriad practices are perfect, reach the extreme, and are impossible to aug-
ment. The power to resonate with/respond to material objects, examine
exhaustively, and be competent [lit. "able in phenomenal affairs"], verily, is
ghee,77 the profound fruit, the source of the three esoterica. Furthermore,
those who purify the bodhicitta may be compared to pure gold. Their origi-
nal nature is bright and clean, forsaking all excesses and evil. Great compas-
77 "Ghee" (tihu, Jpn. daigo 1\Ik1\i\ll) is clarified butter. Traditionally an offering poured
over sacred icons, it means the "cream" (Skt. mmJ';la) and symbolically suggests the good-
ness of the Buddha and the excellence of man's talents. Essentially, "ghee" functioned as a
referent to buddhahood, enlightenment, the cream or best of what the Buddhadharma had
to offer. For instance, in a passage from his Arcane Meaning of the Lotus Satra of the
Sublime Dharma, which uses "ghee" 29 times, Zhiyi said, "Hence, it is said that the Bud-
dha/buddhahood is like ghee .... If you are able to cultivate the eightfold [path] and then
see your Buddha nature it is called obtaining ghee .... By means of the principle and wis-
dom of the middle path you become ghee." Zhiyi also explains that the perfect teaching
(yuanjiao as he describes it, is also "ghee." See Miaofa lianhuajing xuanyi i!Yi'!Ji
5, T 1716, 33.739c-740b. Another extensive discussion of ghee is found in the
final roll of the same work. Here Zhiyi says that those who embody the perfect and sudden
[teaching] (yuandun and become awakened, verily, are ghee; see roll 10, T 1716,
33.807a21-22. Also, "Acquiring the acquiescence to the non-production of dharmas is
performing like ghee"; see roll 10, T 1716, 33.807b6. "Again, as for what we refer to as
'ghee,' it is after the taste of the throngs [of sentient beings]; nirvQT}a is designated as
ghee;" and, of course, the lotus is ghee; see roll 10, T 1716, 33.808a19-20, a25. Surpris-
ingly enough, to Zhiyi, even drhats, pratyekabuddhas, and buddhas are like ghee; see roll
10, T 1716, 33.81Ob26.
sion is like learning an ingenious skill. Taking all manner of medicinal drugs
and practicing cures up to a "mirror penetrates softness, pliability, and self-
existence" [are all examples of expedients]. Expedients are like the comple-
tion [siddhi] of ingenious arts. What one creates is all completed according
to one's wishes.
78
.
This rhetoric on the original cleanliness and purity of one's original
nature and that all practices lead toward this goal sounds surprisingly
similar to what will emerge in Chan discourse later. In fact, corollar-
ies to the three esoterica, the three mysteries (sanxuan -= %) and the
three essentials (sanyao =JO developed within the Chan tradi-
tion during the Five Dynasties and Ten Kingdoms n (907-
978) and Northern Song ::it* (960-1127) period.
79
A description of
a practical procedure for developing the three esoterica is only al-
luded to at best, although the commentators are convinced that they
describe it in detail:
Now, this approach of mantra is comparable to memorizing and chanting
(chisongzhe which is also like this. It is described in detail in the
text below. In accordance to cultivating and practicing the three esoterica,
you attain and complete all peculiarities and inconceivable phenomena. Even
though you scrutinize among the connections of each and every thing, in the
end you forsake four-part [differentiation].80 The Dharma, and so on, is like
this. [It is] no different than the pure mind.
Sl
A practitioner develops the three esoteric a in the same manner that
he learns and practices mantras: memorizing and chanting. The by-
product of this practice is the cessation of discriminating mental con-
78 Da Piluzhena chengfo jing shu 1, T 1796, 39.587a27-b4; cf. Tajima, Etude sur Ie
Mahiivairocana-sL7tra (cf. n. 5 above) 97-98.
79 See George Keyworth, "The Body, Speech, and Mind of the Patriarchs: The Three
Mysteries and Three Essentials of Chinese Chan Buddhism," a paper delivered at the
Western Conference of the Association for Asian Studies, Phoenix, Ariz., October 11,
2003. He is currently revising the paper for publication.
80 Four-part differentiation (sijufenbie i1!l1:iJ51-JjiJ) is the standard Mahayana intellectual
conceptualization of phenomena as being existent, non-existent, neither, or both; or
phenomenal, noumenal, neither, or both; double, single, neither, or both. See Fayuan
zhu/in 5, T 2122, 53.306a.
81 Da Piluzhena chengfo jing shu 3, T 1796, 39.607a3-9.
336 RICHARD D. MCBRIDE,II
ceptualization and the awakening of the pure mind. Practices involv-
ing ma1J-{iala are described briefly in the following passage:
I now also desire to .continue the works of the Tathagata by following that
which has been practiced by the Buddha. For this reason, draw this mm)ljala.
Even though I have not yet attained everything like unto a Tathagata, never-
theless, because I am empowered by the three esoterica of Vairocana, I will
also be able to manifest the Dharma body and universally (pu collect a
great assembly of all For this reason, you should also mani-
fest evidence [of your buddhahood] and cause all demons to be unable to op-
press or dwell within yoU.
82
SUbhiikarasiIpha and Yixing describe the three esoterica as playing a
dual role in this commentary. On the one hand they are described as
something outside of the practitioner, the three esoterica of Vairo-
cana, which purifies and empowers him. On the other hand, by
means of the ma1J-{iala the practitioner is supposed to ritually recog-
nize his dominance over an interior realm populated by demons. 83
This duality ,is transcended, however, because all enlightened beings
are no different than Vairocana at the moment of awakening. It is the
inherent Buddha nature within that, when it is activated, empowers
the practitioner. Within this symbolic context, demons represent
defilements. By activating the three esoteric a the practitioner mani-
fests evidence of his buddhahood, symbolically represented as the
suppression of demons. Hence, defilements are subdued immediately
and as inherently powerless.
In his Zunsheng fading xiu yugafagui yi
(Ritual procedures for cultivating yoga by means of the Paramount
Buddha Crown [spell], T 973), SubhiikarasiIpha describes numerous
ritual procedures using dhiira1J-l, mantra, ma1J-{iala, and homa. The
ninth chapter, "Manifesting evidence of yogasiddhi [yogic accomp-
lishment by means of] the Paramount Mantra" (zunsheng zhenyan
82 Da Piluzhena chengfojing shu 4, T 1796, 39.619c6--10.
83 On the practice of dominating an interior realm populated by demons see Ronald M.
Davidson, Indian Esoteric Buddhism: A Social History of the Tantric Movement (New
York: Columbia University Press 2002) 113-144.
zheng yuga xidi contains an instructive pas-
'sage:
The three types of kanna, verily, are the three esoterica. The three esoterica,
verily, are the three bodies [of the Buddha]. The three bodies, verily, are the
wisdom of the Tathagata Mahavairocana. If you' acquire, in this manner, the
body of Vairocana and if you attest that the Dhanna realm everywhere mani-
fests fonn bodies [n7pakilya], it is the same as a Dhanna realm [dharma-
dhatu] and the same as a svabhava [essential nature]. Outside of the One
Mind, moreover, there is not one material object that can be established. All
buddhas [possess] the mark of emptiness. Emptiness [is] also the markless
mind, [and it] is the same as emptiness; hence, those who cultivate [the prac-
tices of] the yogin are also the same as the essence.
In an instant, the length of a single thought, you transcend the three absurd
graspings
84
and transcend trikalpilsarrzkhyeya [worth of] practices. When you
initially arouse the mind [bodhicitta] you will then achieve complete
buddhahood (zhengjue iEJi, saf!lbodhi). Verily, this is the body of siddhi
(xidi accomplishment, completion). This, among the markless siddhi,
is the dhanna of the most superior siddhi.
85
The language of this passage is surprisingly straightforward and
clear. Employing the concept of universality, the practitioner's own
karma of body, speech, and mind produces the three bodies of the
buddha, which is merely another name for Vairocana. Although un-
stated, the concept of acquiescence to the non-production of dharmas
(wushengfa ren Skt. is alluded
to by such positions as no material forms can be established outside
of the mind and all buddhas possess the mark of emptiness.
Subhakarasir!ilia's explanation that the achievement of buddhahood
is as immediate as the arousal of the bodhicitta - in which the three
esoteric a play a central role - is actually quite similar to some as-
84 The three absurd graspings (sanzhi ::::il.) are grasping the concept of self, grasping
the concept of the dhannas, and the delusion of ignorance. See Da PilLlzhena chengfo
shenbianjiachijing 1, T 848, 18.3a.
85 ZLlnsheng foding XiLl YLlgafagLli yi 2, T 973, 19.380a25-b3.
86 Obtaining the "acquiescence to the non-production of dhannas" (wLlshengfa ren
i*B) 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 sLloshLlO
jing 1, T 475, 14.539a, 540c; roll 2, 14.546a.
pects of the doctrinal position held by Huayan proponents. In-
stead of suggesting that all ten or fifty-two bodhisattva stages are
interconnected, Subhakarasirpha promotes a position influenced by
the rhetoric of subitism. In the realm of reality, complete enlighten-
ment is instantaneous. This is the most superior of all spiritual attain-
ments.
The third great Tantric master, Amoghavajra (Bukong /G3::, 705-
774),87 adds new attributes to the three esoterica that ,accord with
aspects of Tantric doctrine and practice he promotes. Just as Subha-
karasirp.ha emphasized universality with respect to the esoterica of
body, speech, and mind, Amoghavajra accentuates its constancy and
firmness. In his Dale jin'gang bukong zhenshi sanmeiye jing bore
. boluomiduo liqu shi /G 3:: - $:El@
(Analysis of the principle and intention toward Prajfiiipiiramitii
and the Sutra on the samaya [coming together] of great bliss, vajra
non-emptiness and true reality, T 1003), for instance, he likens it to
vajra or adamant:
Vajra (jin'gang adamant) attests to attaining the state of buddha hood in
which aIr dhannas are self-existent. One attains and attests to the vajra
[adamantine nature] of the three esoterica of body, speech, and mind. Within
the storehouse consciousness [alayavijiiana] one cultivates the [religious]
path, defilements [kle.sa], and habits [vasana]. "Finnness" accords with vajra
and is difficult to destroy. One should employ the samadhi on the vajra-
knowledge of great emptiness and attest to and attain the Dhanna body, the
universal radiance of glorious brightness (guangming bianzhao
the Tathagata Vairocana.
88
My understanding of this passage is that enlightened beings should
manifest vajra-like samiidhi. The primary purpose of ritual involving
the body, speech, and mind is to develop the samiidhi that will render
the practitioner - actually the practitioner's mind - firm, immove-
87 For the biography of Amoghavajra see Song gaoseng zhuan 1, T 2061, 50.712a-
714a; see also Chou, "Tantrism in China" (cf. n. 67 above) 284-307.
88 Dale jin 'gang bukong zhenshi sanmeiye jing bore boluomiduo Jiqu shi
2, T 1003, 19.607b25-28.
able, and bright in nature, making him thereby functionally no differ-
ent than Vaircicana.
Amoghavajra's translation of the Wuliangshou ntlai guanxing
gongyang yigui (Ritual procedures for
making offerings related to the visualization practice of the Tathii-
gata Amitayus, T 930) suggests another way in which the concept of
the three esoterica came to be understood in the nascent Tantric
tradition: It was used as a key phrase encapsulating the approach to
Buddhist practice that uses dhara1J,l and mantra:
At that time the Bodhisattva Vajrapani resided in the midst of the great
assembly of the Buddha Vairocana. Arising from his seat he joined palms
and respectfully addressed the Buddha saying, "World Honored One, for the
sake of living beings plagued by unwholesome karma in world systems
characterized by sundry defilements (zaran saJ?1klesa) in the future
during the age of the decline of the Dharma (mofa ?Kit), I will preach the
dharal},l of the Buddha Amitayus [by which one] cultivates the approach of
the three esoterica (sanmi men), attests to [the power of] the buddhanusmrti-
samadhi (nianfo sanmei 2iiJll =1liK), acquires rebirth in the Pure Land, and
enters the proper position/level of the bodhisattva.
89
One does not acquire rebirth in that k:jetra by means of a few meritorious
[deeds] and expedients devoid of wisdom. For this reason, if one relies on
this teaching and cultivates this practice with correct contemplation
(zhengnian iE2), he will assuredly be reborn in the highest grade of the
highest class of the world-system SukhavatI and attain the initial stage [of
the bodhisattva path].
If one is either a householder or a renunciant who desires rebirth in the Pure
Land, he should first enter the mal'}rJala and obtain consecration
Thereafter he should receive the Ritual guide on chants and intonations
(Niansong yigui from his master. On either a superior site or
89 The proper level or position of the bodhisattva (zhengwei iE 1Jr, Skt. niyiima,
samyaktva). See Weimojie suoshuo jing 1, T 475, 14.542b; roll, 2, 545c; roll
3,553c.
90 The Niansong yigui referred to here, if it does not refer to the current text, probably
refers to the Wuliangshou ndai niansong yigui (Ritual guide on
chants and intonations of the Tathiigata Amitiiyus), twelve sheets in one roll, translated by
Amoghavajra, which, though now lost, was listed repeatedly in catalogs prepared by
Yuanzhao (fl. 794): Da Tang zhenyuan xukaiyuan shijiao lu *H!'i
residence paint, wipe, and cleanse [the area] and set up a square altar. Spread
a heaven-shaped [round] covering on top and hang banners around the
circumference. The top of the altar is divided into an eight[ -sectioned]
ma,:uj.ala. Grind white sandalwood incense paste and smear it on the
positions of the holy [deities]. Enshrine an image of AmWiyus on the
western face of the altar. He [the aspirant] sits on the east of the altar. He sits
facing west opposite the image. I either spread a reed mat or sit with my legs
on a small table.
Three times a day scatter various types of flowers, bum various types of
incense, and place two queqie M1ilO-dishes. Either use bowls or cups and
utensils of gold, silver, bronze, stone, earthenware, tile, and so forth. Fill one
that has not yet been used with perfumed water and place it on top of the
altar. Enshrine four worthy [wish-fulfilling] vases on the four horns
[comers] of the altar. Bum incense, light lamps, daub incense, and [make
sure that] food and drink, in accordance to their power, are set out. Empower
(jiachi [the objects] one by one and attentively make
offerings.
The practitioner [should] wash or bathe daily and put on new clean clothes.
[If he cannot] he should use the empowerment of the true word (zhenyan
jiachi 1JJQ ff, mantra-empowerment) to render them clean. In pensive
visualization all beings possessing desires are originally clean and pure: for
all are covered over with transient dust. [If] they have not awoken to the
truth, they are deluded and have lost bodhi. They are drowning in life and
death [saJ?1Sara] and endure immeasurable suffering.
For this reason, I preach the empowerment of the three esoterica (sanmi
jiachi =ffi'JJQff). Now both self and others may attain cleanliness and purity.
Taking a lotus flower in one's two hands and joining the palms one should
then intone the 'True word that purifies the three kinds of karma' (jing sanye
zhenyan 1) three times."91
This guide provides one of the most succinct descriptions of ritual
procedures for cultivating the three esoterica and ritual empower-
ment by means of the three esoterica. What is striking is the similar-
ity between the stated goals and benefits of this ritual and the
1, T 2156, 55.749bI2; roll 3, T 2156, 55.767b20; ZhenYllan xinding shijiao mulu
1, T 2157, 55.772b29; follI5, T 2157, 55.879c18; roIl 21, T 2157, 55.924c12-
13; and ro1127, T 2157, 55.1011aI4-I5.
91 Wuliangshou rulai guanxing gongyang yiglli 1, T 930, 19.67b--c.
purposes of mainstream Sinitic Pure Land practice.
92
Although the
rhetoric of empowerment (jiachi) pervades, the fundamental goal is
rebirth in Sukhavatl and attainment of the aspirant's proper bodhi-
sattva level. It is merely a repackaging of mainstream Mahayana
ritual with flashy new terminology and ritual accoutrements. Effi-
cacy (lingyan is the critical issue here. The name of the mantra
the aspirant is supposed to chant is also significant because the
chanting of the spell is said to ritually purify the three types of karma
produced by the aspirant, thus transforming them from ordinary
karma into the three esoterica. It is both the "dhiiranl of the Buddha
Amitayus" and the "True word that purifies the three kinds of
karma." In this manner the aspirant ritually becomes a bodhisattva
by the power of the dhiirm}l/mantra and is assured rebirth in the Pure
Land. In this connection we can see that the concept of the three eso-
terica as used in the early Tantric literature of the High Tang period
encapsulates the universal approach of the Mantrayanaor Vajrayana
within the larger context of mainstream Mahayana Buddhism. In
other words, the three esoterica becomes a catch-phrase to refer to
the putatively more efficacious approach utilizing mantra, dhiirm}l,
ma1}rjala, and so forth.
Amoghavajra's most straightforward statement on the meaning of
the three esoterica is found in his translation of lin' gangding yuga-
zhong fa anouduoluo sanmiao sanputi xin lun
(Treatise on producing the thought of anuttarii-
samyaksaY[lbodhi in the yoga of the vajra-crown [vajrasekhara] , T
1665).93 This undated discourse, erroneously attributed to Nagarjuna
92 For more on this ritual manual and its procedures see Charles D. Orzech, "A Tang
Esoteric Manual for Rebirth in the Pure Land: Rites for Contemplation of and Offerings to
Amitayus Tathagata" (forthcoming in a festschrift for Roger Coreless). I would like to
thank Charlie for sharing his study of this interesting text with me.
93 Abe Ryiiichi translates the title as Discourse on the Enlightened Mind; see The
Weaving of Mantra (cf. n. 2 above) 255. Ken White calls it the Bodhicitta Sastra and pro-
vides an annotated translation of the entire text in his The Role of Bodhicitta in Buddhist
Enlightenment (cf. n. 8 above) 209-235. I agree with White that the text was spuriously
attributed to Nagarjuna (ca. 150-200 C.E.), but wonder why he holds to an Indian heritage
for the document? See White, The Role of Bod hie itt a in Buddhist Enlightenment, 13 n. 5, 33.
(Longshu ca. 150-200),94 was probably composed in China
because it alludes to Subhakarasirp.ha and Yixing's Da Piluzhena
chengfo jing ShU.
95
The title suggests it is an exegesis on the vajra-
sekhara literature, although the body quotes a passage from only one
such text towards the end (Jin'gangding yugajing T
865).96 It may date to after the rebellions of An Lushan i<t1:ili and
Shi Siming (ca. 755-763), In my opinion, the short treatise
was crafted for a non-specialist audience - probably Tang scholar-
officials - because it briefly describes the meaning of ritUal practices
from the Da Piluzhena chengfo jing in a straightforward manner and
squarely within the conceptual and practical framework of the bodhi-
sattva path found in mainstream Mahayana scriptures, by quoting
and alluding to such works as the A vatalflsaka Satra, Nirvii/'}a Sutra,
Mahiiprajfziipiiramitii Sutra, and Lotus Sutra. Its explanation of the
three esoterica comes toward the end:
Generally, practitioners who cultivate yogic visualization (yuga guan ftu{ho
im) should necessarily cultivate all the practices of the three esoterica (sanmi
xing and give evidence of and awaken to the meaning of "complet-
ing the body of five marks."97 With respect to the three esoterica, the first is
94 There is a great debate as to whether Nagarjuna 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. For the problem of Nagarjuna's existence and dating in
Indian literature see Joseph Walser, "Nagarjuna and the Ratnavalf" (cf. n. 17 above).
95 Jin'gangding yugazhong fa anouduoluo sanmiao sanputi xin lun, T 1665, 32.574a23;
cf. Da Piluzhena chengfo jing shu 14, T 1796, 39.723b.
96 Jin'gangding yugazhong fa anouduoiuo sanmiao sanputi xin lun, T 1665, 32.
574c11-13. The Jin'gangding yugajing is another name for the Jin'gangding yiqie rulai
zhenshi she dasheng xianzheng dajiaowang jing
!.\ (Sarvatathagatatattvasaligraha), in three rolls. The portion paraphrased in this essay is
from rollI, T 865, 18.207c7-1O.
97 "Completing the body of five marks" (wuxiang chengshen is also called
"completing the body of five dharmas" (wufa chengshen and "completing the
body of five turnings" (wuzhuan chengshen The five marks are (1) the
penetrating and accomplishing mind (tongda xin iffiii,l,'), (2) [cultivating] the bodhicitta
([xiu]putixin (3) [achieving] the vajra mind ([cheng] jin'gang xin
(4) [giving evidence of] the vajra body ([zheng] jin'gang shen and (5) giving
evidence of unsurpassed bodhi (zheng wushang puti Ultimately, manifesting
the five marks was tantamount to "achieving the body of the Tathagata Vairocana in the
THE MYSTERIES OF BODY, SPEECH, AND M1ND 343
body-esoterica (shenmi) , like binding seals (qiyin mudrii) and sum-
moning the saintly assembly here. The second is speech-esoterica (yumi),
like secretly chanting true words (zhenyan, mantra). The passages are clear
and distinct and devoid of obstruction and error. The third is mind-esoterica
(xinmi), like abiding in yoga, mutually resonating with the white and pure
full moon and visualizing (guan) bodhicitta (putixin *!lfAJ,).98
This passage is the most succinct statement on the nature and pur-
pose of the three esoterica in the early Tantric tradition. The three
esoteric a are clearly labeled as the ritualized practices of making
mudra, chanting mantra, and abiding in yogic meditation and are
said to produce the ultimate goal of enlightenment - "completing the
body of the five marks." Although concise, when seen in the context
of other passages in siitras and commentaries of the time, it is far
from being a representative statement about the three esoterica.
Amoghavajra's views presented here later became the orthodox posi-
tion in the mature Esoteric traditions in Japan. I think it is both
important and ironic that a discourse probably composed as a simpli-
fied statement to promote aristocratic and official support for his
form of Mahayana Buddhism would ultimately exert so much influ-
ence on the understanding of three esoterica.
Mid and late Tang Period
By the mid-Tang period the meaning of three esoterica as used in the
growing literature used in the circles of early Tantric practitioners
appears to have become relatively fixed. This is because there was
no need for translators and scholars to explain what the three eso-
terica were or referred to in either the newly translated siitras or in
prefatory or exegetical materials. Not only was the meaning fixed
but the concept of the three esoterica was presented as one of the
primary teachings of the Tantric movement. This is the way the Silla
Vajra Realm (jin'gangjie See lin'gangding yugazhong fa anouduoluo sanmiao
sanputi xin lun, T 1665, 32.574b17-20.
98 lin'gangding yugazhong fa anouduoluo sanmiao sanputi xin lun, T 1665,
32.574bIl-16.
344 RICHARD D. MCBRIDE, IT
monk Hyech'o (704-d. after 780),99 one of the recognized
disciples of Amoghavajra, uses the concept of the three esoterica in
his preface to the Dasheng yuga jin' gang xinghai Manshushili qian-
bei qianba dajiaowang jing
(Satra of the king of the great teaching, the thousand armed
thousand alms-bowl carrying Mafijusrf who possesses an ocean of
yoga vajra nature in the Mahiiyana, T 1177A). Hyech'o considers
himself one who has been consecrated to transmit the three esoteric a
by covenant. 100 Amoghavajra's translation of this siitra, which
Hyech'o calls for short the Satra of the thousand alms-bowl carrying
Mafijusrf, refers to the three esoterica in a variety of constructions:
an approach of the three esoterica (sanmi men), as it is found in
Subhakarasirpha and Yixing's translation of and commentary to the
Da Piluzhena chengfo jing; and in lists of euphemisms for Buddha
nature: "minds of the thus-come buddhas, the three esoterica, the
thirty branches,101 the vajra-wisdom mirror, the way of the sages,
and the ocean of the bhatatathata [the all-containing immaterial
nature of the dharmakaya]," and so forth. 102 Another example is the
Tang Emperor Daizong's (r. 762-779) preface to Amogha-
vajra's retranslation of the Renwang jingo Here the concept of the
three esoterica is again used as a term encompassing the Tantric
approach to Mahayana practice: "The trepitaka [master of the Tripi-
99 See Daizong chaozeng sikong dabian zheng guangzhi sanzang heshang biao zhiji
3, T ZIZ0, 5Z.844bl-3; for a brief biography of
Hyech'o see Yang Han-sung, et al., trans., The Hye Ch'o Diary: Memoir of a Pilgrimage to
: the Five Regions of India (Berkeley: Asian Humanities Press 1984).
100 "Taesung yuga kumgang songhae Mansusilli ch'onbi ch'onbal taegyowang kyong
so" HPC 3.381c9, 38Za3.
lOt The thirty branches (sanshizhi =+;0 is a term found only in this satra, where the
term appears 18 times, not counting Hyech'o's use of it in the preface. It always appears in
lists following the "approach of the three esoterica" (sanmi men) and usually precedes the
concept of samiidhi (sanmodi See Dasheng yuga jin'gang xinghai Manshushili
qianbei qianba dajiaowang jing 1, T 1177A, ZO.725c1Z-13, 727cZ-3; roll 5, T 1177A,
ZO.750a9, 750aZ5, 750aZ9; ro116, T 1177A, ZO.75Z, c9-1O.
102 "Taesling yuga kumgang songhae Mansusilli ch'onbi ch'onbaI taegyowang kyong
so," HPC 3:38Za3; T 1177A, ZO.724cZO--Zl; see also Dasheng yuga jin'gang xing/wi
Manshushili qianbei qianba dajiaowang jing I, T 1177A, ZO.725cZ-3, 727cZ-3, roll Z, T
1177A, ZO.731aZ9-bl, and so forth.
THE MYSTERIES OF BODY, SPEECH, Ai'ill MIND 345
taka] studies and researches the two truths [absolute truth and
coiwentional truth] and teaches and transmits the three esoterica."103
The monk Liangben (717-777)104 was also a disciple of
Amoghavajra and served on his team that executed the retranslation
of the Renwang jing in 765-766.105 This exegete served as a bridge
between mainstream Sinitic intellectual Buddhism and Amoghava-
jra's Tantric tradition. As in commentaries on Kumarajlva's trans-
lation above, Liangben's Renwang huguo bore boluomiduo jing shu
(Commentary on the Peifection ofwis-
dom sutra for humane kings who wish to protect their states, T 1709)
locates the concept of the three esoterica in references to the three
types of karma:
Question: The Buddha does not have frivolous thoughts. They are com-
pletely and entirely fixed [in meditation] (ding 1). Why then does this scrip-
ture repeat the word "enter" [as in enter samiidhi]?
Answer: The Buddha does not have frivolous thoughts. Because he is an
instructor he manifests wisdom in accordance to meditation. For this reason
he showed his first entrance [into samiidhi]. For this reason he dually culti-
vates meditation and wisdom and he neither falters nor moves [from
samiidhi]. When he enters meditation he emits light and, in regard to preach-
ing the Dharma, the three types of karma (sanye), he instructs, are verily the
three esoterica (sanmi).
The Tathligata constantly [abides in the] Dharma and reveres prajiiii. He
desires to discourse and show respect for the rites. Profound samiidhi only
the Buddha alone attains; hence, it is claimed to be "profound." As for
samiidhi, this is said to be "universal maintenance" [samiidhi] (dengchi
fif). Its substance is separate from sinking or shaking; hence, it is said to be
"universal." It solely fixes on one position; hence, it is said to be "main-
!O3 "Renwang huguo bore boluomi jing xu" T 246, 8.834b8.
[04 For the biography of Liangben see Song gaoseng zhuan 5, T 2061, 50.735a26-c22.
[05 For historical background to the translation of this satra and detailed study and
translation of this satra see Charles D. Orzech, Politics and Transcendent Wisdom
(University Park, Penn.: Pennsylvania State University Press 1998). In this book Orzech
refers to Liangben as Liang-pi; the logograph ben # can also be read as bi andfen.
tained." Because its universality is maintained it is also-said that the meaning
of samtidhi is the same as "ground/stage/bhami" (di if!.).106
As before, the Buddha's power to emit light, preach the Dharma,
and abide in prajfia for the benefit of beings is described as the three
esoterica. Liangben holds that the Buddha is absorbed in a constant
state of samadhi in his pure land and that this samadhi is universally
maintained or, in other words, pervades everywhere. The three eso-
terica are the three types of karma of the Buddha because the Bud-
dha has never really left samadhi in his pure land. He is grounded
there and by means of the three esoterica seems to do the things he
does. His teaching skills, compassion, and even his first entrance into
samadhi are all a show: He has never left samadhi. He is grounded in
the unaltered stage or bhumi of buddhahood. This point is reiterated
later, when he comments on the following passage from one of the
scripture's gathas:
The abodes of the three worthies
107
and ten saints
108
are fruition rewards.
109
A buddha only is the one person that dwells in a pure land.
All living beings have temporary abode rewards
[Yet] when they ascend to the adamantine source they dwell in pure lands.
The virtue of the Tathagata's three kinds of karma (sanyede is
immeasurable.
Consequently all living beings equally receive his compassion.lIo
106 Renwang huguo bore boluomiduo jing shu 2, T 1709, 33.449a4--16.
107 The three worthies (sanxian ':::::Jl:) are bodhisattvas, prior to the bhiimis, who have
achieved the level of the ten abidings (shizhu +i), the ten practices (shixing +11'), and
the ten transferences (shi huixiang +l rtJ). See Dasheng yizhang I7A, T 1851,
44.788b27-28.
108 The ten saints (shisheng are bodhisattvas above the sage of the ten transfer-
ences (shi huixiang). See Renwang bore boluomijing 1, T 245, 8.827bI2; 828al.
109 A fruition reward (guobao Skt. phala, vipaka) is retribution for good or evil
deeds, implying that different conditions in this life or any other life are the fruits of seeds
sown in one's previous life or lives. See Miaofa lianhuajing fr!);lt;;Ji1tf..& (Saddharmapu(z-
tjarfka Siitra) 7, T 262, 9.56a8-9.
110 Renwang huguo bore poluomiduo jing 1, T 246, 8.838aI-3; for an alternate transla-
tion see Orzech, Politics and Transcendent Wisdom (cf. n. 105 above) 234. The same
passage in the earlier recension of the siitra (T 245) reads: "The abodes of the three wor-
thies and ten saints are fruition rewards / A buddha only is the one person that dwells in a
Liangben interprets the meaning of the virtue of the Tathagata's
three kinds of karma as follows:
Because the eulogy speaks of 'virtues,' the foregoing virtues of the three
types of karma are verily the approach of the three esoterica (sanmi men). In
this function of affairs, verily, [they are] the three extraordinary occur-
rences.l1l
The three extraordinary occurrences (san shidao Skt. trllJi
priitiharyiilJi) refer to three miraculous ways that bodhisattvas re-
spond to living beings suffering in hell (diyu :tiMli\) and work to save
them. For the most part these are manifestations of the inconceivable
power of the spiritual penetrations. The first is the employment of
spiritual transformation or magical performance (shenbian shidao 1$
rddhipriitiharya) demonstrating the bodhisattva's ability to
manipulate the phenomenal world because of his acquiescence to the
non-production of dharmas. The second is the use of mind-reading
(jixin shidao iideaniipriitihiirya) in order to observe the
spiritual state of the audience so as to gage his remarks appropriately
to encourage them to convert. Third is the miracle of admonition
(jiaojie shidao anusiisanapriitiharya) through which the
bodhisattva's compassionate mind reaches out to strike a chord with
a listener causing the latter t'o destroy his desires for mistaken views
and practices that lead to unwholesome consequences.l12 Liangben's
position is fundamentally and strikingly mainstream Mahayana,
especially for a disciple of Amoghavajra commenting on a putatively
"Tantric" or "esotericized" recapitulation of the Renwang jingo His
equating the three esoterica to the three extraordinary occurrences
demonstrates that the inclusive spirit of the earlier intellectual
approaches to defining this concept were preserved within the cadre
of monks associated with the early Tantric masters.
pure land / All living beings have temporary abode rewards / [Yet] when they ascend to
the adamantine source they dwell in pure lands / The virtue of the Tathagata' s three kinds
of kanna (sanyede) is boundless (wuji f!\T;@) / I now, by the light of the moon, worship the
three treasures [Buddha, Dhanna, sQ/igha]." See Renwang bore boluomiduo jing 1, T 245,
8.828al-3.
III Renwang huguo jing shu 3, T 1709, 33.427a5-8.
112 See Apidamo jushelw! IWJ 27, T 1558, 29. 143c7-144a23.
The Tiantai monk Zhanran #frrrit (711-782), I13 heir to the exegeti-
cal tradition of Zhiyi and Guanding, followed the example of schol-
ars of the times in locating the three esoterica in the earlier sLltras. In
his Weimo jing We shu (Concise commentary on the
Vimalakfrti Satra, T 1778), Zhanran comments on the first giitlui-
verse (jiesong in the first chapter, "Buddha Land" (joguo pin
J'b), of Kumarajlva's translation of the Vimalakfrtinirde.sa Satra
(T 475). The second half of the giithii uses the terms "ten powers"
(shili + jJ) and "spiritual power" (shenli 1$ jJ) repeatedly, alluding
to the Buddha's employment of his spiritual penetrations in preach-
ing the Dharma. Zhanran's analysis begins in this part of the verse:
Question: The Great Saint, over the next fifteen lines,114 specifically de-
scribes and praises the three esoterica of the Buddha in three passages. In the
first two lines he praises the body-esoterica. In the next six lines he praises
the speech-esoterica. And in the next seven lines he praises the mind-eso-
terica.
The three esoterica verily are the inconceivable transformations (busiyihua
/F}JMHt) of the three wheels. The three are not lost and the three cannot be
protected. The throngs who take refuge in the Great Saint and the Dharma
King are those who praise the superior response of the Dharma bodyy5
A passage from his Fahua wenju ji (Record of literary
passages from the Lotus Satra, T 1719) explains what Zhanran
means by equating the three esoterica to the three wheels:
In order to differentiate their changing functions, it is said that the three
types of karma, and so forth, are called by the different names of the types of
karma, three esoterica, three wheels, and the three virtues. Mind-esoterica
113 For the biography of Zhaman see Song gaoseng zhuan 6, T 2061, 50.739b9-740a16.
For more on Zhanran see Chen Jinhua, "One Name, Three Monks: Two Northern Chan
Masters Emerge from the Shadow of Their Contemporary, the Tiantai Master Zhanran i!
P& (711-782)," Journal of the International Association of Buddhist Studies 22/1 (1999) 1-
91. See also Chen, Making and Remaking History (cf. n. 34 above).
114 Zhanran reports that it is the part of the verse beginning withjinfeng shizun 4-*t!:!:
#; see Weimo jing We shu 2, T 1778, 38.587b8; cf. Weimoji suoshuo jing 1, T 475,
14.537c25-538a14.
115 Weimo jing We shu 2, T 1778, 38.587b18-22.
(yimi) is namely prajJ1a. Speech-esoterica (koumi) is namely liberation.
Body-esoterica (shenmi) is namely the Dharma body.l16
His position is essentially the same as Zhiyi's.
The diction of a passage in Zhanran's short exegesis titled Shi
buermen (The ten non-dual approaches, T 1927) strongly
suggests that, with respect to his understanding of the three esoteric a,
he was somewhat influenced by the literature translated or composed
by Subhakarasirpha and Amoghavajra. The primary thrust of the
exegesis is to demonstrate the ultimate non-duality of reality, the
"consummate interfusion and entry into the sublime" (rongtong
rumiao Although these terms, consummate interfusion, in
particular, seem to indicate a Huayan scriptural or exegetical setting,
they are actually found repeatedly in the writings of Zhiyi and other
early exegetes, including Amoghavajra. In this connection, the three
esoterica or three types of karma are ultimately interchangeable with
the Dharma body and Vairocana:
8. The non-dual approach of the three kinds of karma. With respect to the
approach of transforming others, phenomena (shi $) are divided into the
three esoterica. Accordingly, the principle of material objects (wuli ac-
quire the deSignation "not the same." The mind-wheel (yilun
"thought") mirrors potentiality and the two [other] wheels estab-
lish/engender transformation. The presently-appearing body preaches the
Dharma and does not have a hair's difference [from the true Dharma body].
With reference to this body, distinctions lie between the true (zhen Jli,) and
the false (ying 1l\I.\). With reference to the Dharma, distinctions lie between
expediency (quan ti) and reality (shi .).
If the two bodies are different, why then do we call it the "Dharma body"? If
the two types of preaching are perverse, why then do we say "all complete
the Path to Buddhahood"? If only the Dharma body responds and does not
condescend to the [mundane] world and if there is only the Path to Buddha-
hood, who exhibits/bestows the Three Vehicles? The body is still a non-body
and the preaching [is] necessarily a non-preaching. Body and speech are
universal (pingdeng) and are equal to that mind-wheel. Mind forms are all
thus, they transform without a [pre-determined] plan. Are there not a hun-
dred spheres in one mind? All these spheres are nothing but the three types
116 Falnta wenjuji 9, T 1719, 34.317a6--8.
of karma. A sphere still is a recollection (nian 2", "thought"); so how are the
three types of karma peculiar? The function of a result never deteriorates; its
cause must be designated a result [of some other cause]. If you believe in
causes and results, you should know that the three esoterica are the basis.
The hundred spheres and the three types of karma are fully prepared in
emptiness (kong Y:), the provisional (jia &), and the mean (zhong 'P).
Hence, by employing the designation that is appropriate everything goes to
fruition. Each and every response form, each and every word sound are
nothing but the hundred spheres and the three types of karma all complete.
Transformation repeatedly begets transformation: Thus it is said! Hence, one
recollection is every mind and already possesses the nature of principle
Clixing :FJlHi) and the ocean of marks of the three esoterica. One mote of dust
recompenses form and the same resides in the original principle Vairocana.
Forthwith they are named as the three indistinguishables. This, by means of
self and others, completes the approach of non-duality.1l7
In Zhanran's own words the three indistinguishables (san wuchabie
- are those of the mind (xin JL.'), the Buddha (fo 1Jt), and
living beings (zhongsheng ffR:1:.).118 This is a grand statement on the
fundamental non-duality of phenomena with respect to the nature of
ultimate reality. The three esoterica, which in the beginning are
equated to the three types of karma, are quickly substituted for the
nature of all manifested phenomena. Yet these manifested phenom-
ena - people, things, objects, ideas that have meaning in provisional
contexts - are merely mental phantasms produced by the mind and
are ultimately indistinguishable from body, speech, and the realm of
reality. This is essentially the same as the refrain of the Huayan
tradition's cosmic ecology (but which was also shared in most of the
mature exegetical traditions as well as Chan): all in one and one in
all. The three esoterica are the cause of the transformations of all
things and yet they are indistinguishable from Vairocana, the nature
of reality, because they are ultimately empty of self-nature.
117 Shi buermen 1, T 1927, 46.704a26-b11.
118 Fahuaxuanyi shiqian 6, T 1717, 33.858c2.
What are the three esoterica?
We are now ready to return to the questions that have guided our
investigation of the concept of the three esoterica. First, how did
medieval Chinese Buddhist intellectuals and early Tantric practitio-
ners understand the significance of this term? The evidence is clear
that the three esoterica recalled much more than the ritualized ap-
proach of mudrii, mantra, and yogic meditation. In medieval Sinitic
Buddhist scripture and exegesis the three esoterica have many
snugly-packed layers of meaning. Although single layers of meaning
can be peeled off, no one layer can account for the wide range of
semantic usage. Many of the major scholar-monks of the medieval
period sought to locate traces of the three esoterica in scripture and
used them as a heuristic device to talk about unexplainable aspects of
the Buddhadharma and, especially, the nature of the actions of
enlightened beings such as bodhisattvas and buddhas.
The idea of the three esoterica first appeared in an early Maha-
yana scripture in the early fourth century, the Miji jingo Its basic con-
tours, however, were actually set in Kumarajlva's Dazhidu [un, be-
cause most exegetes alluded to this work's presentation of the three
esoteric a as referring to the inexplicable spiritual powers of the Bud-
dha as a result of his constantly abiding in samiidhi and manifest in
his preaching the Dharma. Kumarajlva also linked the three esoterica
to the concept of the virtue of the three types of karma. Zhiyi built
upon this foundation by using the three esoterica as a heuristic de-
vice to talk about the three types of karma, three virtues, three
wheels, which he associated with the spiritual penetrations, and gen-
eral bodhisattva actions employed in preaching the Dharma. Guan-
ding expands upon his master's position by describing the three eso-
terica as bodhisattva powers and virtues and connecting them to
other lists, such as the four virtues, five constancies, five fruits, and
four bodhisattva practices, which all describe either the enlightened
state or the method by which a bodhisattva converts others. In Guan-
ding is also found an interpretation of the three esoteric a shared by
Jizang: that the three esoterica refer to prajfiii, the Dharma body, and
liberation: all terms to describe absolute reality.
352 RICHARD D. MCBRIDE, IT
Jizang applied the three esoterica directly to the Bodhisattva.
A valokitesvara to speak of his spiritual penetrations and wholesome
actions to convert living beings. He suggests that they refer to the
bodhisattva's entering meditation, emitting light, and preaching the
Dharma: a view he shares with Kuiji. Jizang compares them to the
three virtues, but adds another list, the four marks, to explain their
meaning. He also refers to a mystical endowment of the three eso-
terica by which buddhas and bodhisattvas preach siitras and convert
beings by making their message fit the audience. W<Snch'uk also
explains the three esoterica as associated with the spiritual powers
used in preaching the Dharma, referring directly information in
KumarajIva's Dazhidu lun. He also connects them to the adamantine
(vajra) nature of the Buddha's body and as what enables buddhas to
tum the wheel of the Dharma while remaining in samadhi.
This leads us to the question: Did its meaning in Sinitic Buddhist
literature change prior to its use in Tantric literature and ritual? The
simple answer is yes. In medieval Sinitic Buddhism, the three eso-
terica came to represent not only the method but the means of bodhi-
sattva practice: the way by which bodhisattvas both resonate with
and manifest the Dharma body. In other words, the term was used to
account for the essence and function of enlightenment. The three
esoteric a were not completely redefined in the newly emerging Tan-
tric literature. The Tantric masters use slightly different terms but
the semantics are essentially the same. Furthermore, there is no liter-
ary evidence of disputes over the meaning of the term. New meaning
is simply added to the term. The whole of the preexisting exegetical
tradition's understanding was preserved and new meanings were
assimilated with the translation of new siitras and the production of
new commentaries. Some were conceptual like universality (ping-
deng) , as found in Subhiikarasilpha and Yixing's translation of and
commentary on the Da Piluzhena jing, and others were practical.
Perhaps because the practical definition applied by Amoghavajra,
equating the three esoteric a to the mudra, mantra, and yogic medita-
tion, is the most simple to understand and put into practice it later
became a preferred understanding of the term. Among exegetes,
however, this was merely one of a many of ways of describing the
term and certainly not the most common.
THE MYSTERIES OF BODY, SPEECH, A.."ID MIND 353
The early Tantric master SubhakarasiIpha and Yixing attempted
to . remove the mystery surrounding the three esoterica by clearly
labeling them as expedients by which one is empowered to attain
buddhahood in this life. Like previous exegetes, the two scholar
monks also directly associated them with arousing the bodhicitta and
turning the wheel of the Dharma. Although they still refer to the
three types of karma, Subhakarasirpha also combines them with
concepts that resonate more fully with his message: sfinyatii
(emptiness), sar[tbodhi, and siddhi. By employing the concept of
universality, he made more concrete ideas developed in earlier
exegesis: their being what resonates with and responds to things,
their ability purify the bodhicitta, and their explanation as the
Dharma body and Vairocana himself. The three esoterica are put into
action in practices involving malJt;lala and mantra, which cultivate a
pure mind and provide for interactions with all things. Following
Kumarajlva, Vajrabodhi saw the three esoterica as associated with
constancy and firmness, the vajra-like nature of samiidhi. They still
refer to the three types of karma, of course, but they also refer to
Vairocana, the Dharma body. Vairocana is not an exterior entity; he
is the buddha within that is manifest when the practitioner cultivates
the three esoterica.
In the mid-Tang period, Hyech'o described the three esoterica in
language recalling Buddha nature: the mind of the thus-come ones,
the way of the sages, the vajra-wisdom mirror, and the Dharma
body. Liangben shows that the ideas of the earlier exegetical tradi-
tion still had currency by combining the newly employed concept of
universality with older ideas of entering samiidhi, abiding in prajfiii,
emitting light, and preaching the Dharma. They are still the three
types of karma and their association with the spiritual penetrations is
found in his explanation of them as the three extraordinary occur-
rences. Zhanran merges ideas found in the writings of Zhiyi (the
three wheels) and Guanding (prajfiii, liberation, the Dharma body)
with concepts introduced by the early Tantric masters (universality,
Vairocana, the Dharma body) - all within the context of the com-
monly shared exegetical frame of consummate interfusion and entry
into the sublime (the realm of reality), the nature of principle.
All of the Buddhist exegetes linked the three esoterica with the
three types of karma produced by Tathagatas, the three virtues, the
three wheels, and so forth. The workings of these threeesoterica are
ultimately inexplicable and beyond human understanding. Fbr this
reason, I think that they understood them as representing the per-
fectly interfused reality of emptiness in active form and as resonating
with mainstream Mahayana doctrinal concepts such as the Dharma
body (dharmakaya), the dharmadhiitu (the realm of reality), and
Vairocana. The three esoterica are expedients (upaya) and are also
directly linked to the spiritual penetrations manifested by buddhas
and bodhisattvas. These powers are associated not only with the
marks of spiritual attainment but also with the way these Buddhist
figures preach the Buddhadharma and convert living beings. The
strong connection to the spiritual penetrations may explain why it
was an increasingly important topic in exegesis and account for its
importance in nascent Tantric literature.
The meaning of the three esoterica eludes simple definition. Be-
ing empty of individual meaning, it was freely interconnected and
interfused with any and all terms defining the nature of buddhahood.
The term was conceptualized as transcending duality and, ultimately,
as the perfect manifestation of the universality that enables living
beings to be fundamentally and functionally not different from
Vairocana, or the nature of the universe as it is. As a pithy catch
phrase it came to represent the inherent potential for buddhahood in
all things, as well as the buddhahood manifested in the actions of all
Jiving beings, and recalls what Buddhist scholiasts typically and
. more familiarly render as "Buddha nature" or the "Tathagatagar-
bha." In this connection it is indelibly tied to the great secret or eso-
teric teaching of the Mahayana, the one mind, the one vehicle, the
Mahayana vision of reality, and, because it is a product of constantly
dwelling in samadhi, the bodhisattva's "acquiescence to the non-
production of dharmas."
What are the implications for the study of Tantric or Esoteric
Buddhism in China? This research questions some assumptions that
have been made about the nature and contributions of Tantric Bud-
dhism in East Asia. First, Sinitic Buddhist scholiasts had a sophisti-
cated understanding of the ineffability of the Dharma body - Vairo-
THE MYSTERlES OF BODY, SPEECH, AND MIND 355
cana, the dharmadhiitu - long before the arrival of the Tantric mas-
ters, so the claim cannot be advanced that this was a Tantric doc-
trinal innovation.
1l9
Also, the concept of the body, speech, and mind
of the Buddha, the three esoterica, is not a unique or defining
characteristic of Tantric Buddhism in China without some important
qualifications. In other words, we cannot make a strong case for
Tantric Buddhism as a distinctly different tradition using exegetical
materials or in intellectual history because what we see in the litera-
ture is also highly influenced by or indicative of the Huayan rhetoric
of universality and consummate interfusion. Nevertheless, efficacy
in ritual and Indianness are key themes. It is presented as more
efficacious than mainstream Sinitic Buddhist cultic practices to
which it bore many superficial similarities, such as the use of
dhiira1Jf and other types of ritual procedures to accomplish particular
purposes. It was fresh from India and seemingly promoted as such.
Do we underestimate the importance of its exoticness to its Chinese
audience? Although agreement on a tighter definition of Tantric or
Esoteric Buddhism may be impossible for specialists, I think the
evidence for the role of the three esoterica in mainstream inter-
pretations of Buddhist practice strongly encourages us to limit it to
include the replication of the body, speech, and mind of the Buddha,
within the narrower context of rituals involving mantra, mudrii, and
meditation perfonned in new specialized ma1Jtj.ala under the auspices
of a master to produce enlightenment immediately. A broad defi-
nition, in my opinion, makes it hard to distinguish from mainstream
Sinitic Mahayana.
119 See, for instance, Matsunaga Yiikei t ~ f t l l jJ., "Esoteric Buddhism: A Definition,"
in Mikkyo: KobO Daishi Ktikai and Shingon Buddhism (Bulletin of the Research Institute of
Esoteric Buddhist Culture, special issue, K6yasan: Research Institute of Esoteric Buddhist
Culture, K6yasan University, October 1990) 23-40, esp. p. 25: "In The Differences be-
tween Exoteric and Esoteric Buddhism Kiikai states that Esoteric Buddhism differs from
the earlier Nara schools and the Tendai school in regard to the following four points: 1) it
asserts the possibility of becoming a Buddha in this very body; 2) it recognizes that the
dharma-kaya may preach; 3) its asserts the effability of the state of buddhahood; and 4)
the benefits deriving from its teachings are superior to those of Exoteric Buddhism."
ON DOCTRINAL SIMll.,ARITIES BETWEEN
STHIRAMATI AND XUANZANG
RIDENORl SAKUMA
O. Introduction: Reasons for comparing Sthiramati and Xuanzang
The foundations of the system of doctrinal theories in the Chinese
Faxiang y:t;J: school lie in the Cheng weishi lun translated
into Chinese by Xuanzang 2:* (with the wishes of his disciple Kuiji
being said to have been also strongly reflected in this transla-
tion). In the Rosso school of Japan too, efforts were made to
preserve this tradition as faithfully as possible. Furthermore, in the
traditions of the Faxiang school it is believed that the Cheng weishi
[un was compiled on the basis of several commentaries on Vasu-
bandhu's Trif(lsikii, with Dharmapala's interpretations being deemed
to represent the legitimate interpretation. The Cheng weishi lun is
not, in other words, a translation of a single scholar's commentary,
and the legitimate view was determined by picking and choosing
among several diverging views. The same method had been used
when Xuanzang translated the Buddhabhiimisiistra ten years earlier,
with the interpretations of Bandhuprabha being deemed to represent
the legitimate interpretation among those of several other scholars.
In the Faxiang school, the views of various Indian scholars were
assessed in accordance with their treatment in the Cheng weishi lun.
With the views of Dharmapala being deemed to represent the legiti-
mate view, the ideas of other scholars were recorded and judged to
be not legitimate, and one gains the impression that Sthiramati in
particular was an important target of criticism. This can also be in-
ferred from Fukaura Shobun' s detailed study of the Cheng weishi
[un, in which he remarks more than once that Sthiramati was the
scholar who stood on a par with Dharmapala.
1
I Fukaura gives, for instance, the following explanation (1954, vol. 1: 341): "Were
one to seek a great figure comparable with Dharmapli:la among the ten great Y ogacara
Volume 29' Number 2 2006 (2008) pp. 357-382
358 HIDENORI SAKUMA
Here we need to consider the question of how to deal with
Dharmapala. While we can accept that the Cheng weishi tun was
compiled from a position that regarded Dharmapala' s views as
legitimate, there survives no commentary on the TriY[lsika by
Dharmapala himself in either the original Sanskrit or a Tibetan
translation. In addition, to the best of my knowledge the only other
work attributed to Dharmapala survives only in Chinese translation.
2
If translations can be said to reflect the aims of the translator, this
means that there exist no sources by which we can ascertain
Dharmapala's true intent. If, as the traditional view would have us
believe, Dharmapala died at the early age of twenty-nine and Sila-
bhadra, thought to have been one year older, succeeded him as head
of Nalanda University and met Xuanzang when he was over one
hundred years old, it must be assumed that Dharmapala's all-impor-
tant system of doctrinal theories to be seen in the Cheng weishi tun
scholars, one would indeed have to point to Sthiramati. But his style of scholarship, as has
already been mentioned, differs completely from that of DharmapaJa and adopts the stance
of the merging of essential nature and external characteristics, recognizing the identity of
phenomena and thusness." It has been recognized in Japan too that Sthiramati has
traditionally been understood as standing in opposition to Dharmapala. But the assertion
that Sthiramati' s philosophical tendencies are the same as those of Paramartha, who trans-
lated the Mahiiyiinasal!lgraha and Mahi'iyiinasraddhotpiidasiistra, is no more than specula-
tion on the part of Fukaura, and there is a strong possibility that Sthiramati's works were
modified when being translated into Chinese as a result of the Chinese predilection for
Tathagatagarbha thought. It is questionable whether it is valid to go beyond the reflection
of this predilection in the Shelun Jlilffil school and link it to Sthiramati in India. In my
experience, it is difficult to fmd any proof in extant commentaries by Sthiramati of Fu-
kaura's assertion that Sthiramati'sstyle of scholarship, characterized by the merging of
essential nature and external characteristics, was taken over by Paramartha and developed
into a doctrine asserting that all beings have one and the same nature. It should be noted
that Sthiramati is not mentioned in the main text of the Cheng weishi lun and appears only
in the afterword in a reference to the ten great bodhisattvas "Dharmapala, Sthiramati, and
so on."
2 Tsukamoto et al. (1990: 174-175) mention a commentary on the CatuftSataka by
Dharmapala which survives only in Chinese translation. Apart from this, the Cheng weishi
lun baosheng lun P.!GII/tm&!ii1i'!lf1:.ilffil (T. 31, no. 1591) and Guan suoyuan lun shi
(T. 31, no. 1625), both translated by Yijing are also attributed to Dharmapala. To the
best of my knowledge, these too have not survived in the Sanskrit original or Tibetan
translation. In addition, as is also noted by Tsukamoto et al. (ibid.: 362), among the ten
scholars said to have written commentaries on the Tril!lsikii, only Sthiramati's is extant,
and it survives, moreover, in the original Sanskrit.
SIMILARITIES BETWEEN STHIRAMA TI AND XUANZANG 359
had been transmitted by Silabhadra. But in the doctrinal theories that
I have studied. to date this has not been the case. For example,
whereas the correspondences between the eight consciousnesses (vi-
jfiana) and four knowledges (jfiana) , regarded as a matter of com-
mon knowledge in Faxiang doctrine, are firmly entrenched in the
Cheng weishi [un, they are still in a transitional stage in Silabhadra's
Buddhabhfimivyakhyana, which survives in Tibetan translation. How
is one to comprehend the fact that something which ought to have
been finalized at the time of Dharmapala is still in a transitional
stage in Silabhadra's writings? We may have to conclude that Dhar-
mapala did not give any thought to the correspondences between the
eight consciousnesses and four knowledges.
Sthiramati (Anhui meanwhile, is said to have been based at
Valabhl and to have been a contemporary of Dharmapala.
3
But the
scholar mentioned by Xuanzang alongside GU1).amati (Dehui in
the Datang xiyu ji in his accounts of Nalanda (9.3.5) and Valabhi
(11.8.4)4 is not Anhui but Jianhui In the Datang Daciensi san-
zang fashi zhuan .:=.iU:ftjjJ{t his name is given as An-
hui. Among works included in the TaishO edition, the author of the
Dacheng fajie wuchabie lun (T. 31, nos. 1626 &
1627; neither translated by Xuanzang) is given as Jianhui, while the
author of the Dacheng apidamo zaji [un (T. 31,
no. 1606; translated by Xuanzang) and Dacheng guang wuyun lun *
(T. 31, no. 1613; translated by Divakara) is given as An-
3 This is based on the Cheng weishi [un shuji (T. 43: 231cI9ff.): :::.'lti':
;>g;jfC",f,i*!l'o llPg*Jt*o iiFPOCm

Details about the "land of Palapi (Valabhi, Vallabhi)" f::UILE<e. can be found in the Da-
tang xiyu ji (T. 51: 936bI6ff.; cf. Mizutani 1999: 318). Sthiramati has been
associated with ValabhI on account of an inscription issued by Guhasena II in A.D. 588-
589 (see Shizutani 1979: no. 177; Tsukamoto 1996: Wa!a 7), according to which the king
made a donation to the Bappapadlya temple founded by Sthiramati at ValabhI, and this
Sthiramati has been identified as the commentator Sthiramati. According to the Cheng
weishi [un shuji quoted above, Sthiramati was a contemporary of Dharmapala and came
from the "land of Luoluo" in south India. "Luoluo ,f,i,f,i" corresponds to GujaratI
La(a=La!a, which was the name of an ancient kingdom affiliated to ValabhI. It is also evi-
dent from sources cited in Law 1976 and Dey 1927 that Wa!a refers to ValabhI.
4 Mizutani 1999: 168 & 321.
360 . HIDENORI SAKUMA
hui. While a detailed examination of this state of affairs will be omit-
ted here, the original Sanskrit equivalent of both Jianhui and Anhui
may be considered to have been Sthiramati.
5
Further, a possible
point of contact between Sthiramati and Xuanzang would have been
Shengjun Jl%.'ij[ (*Jayasena), a contemporary of Dharmapala who is
mentioned in the Datang Daciensi sanzang fashi zhuan together with
GUI).amati in connection with both Nalanda and Valabhl. It is re-
corded that Shengjun, under whom Xuanzang studied or with whom
he spent time, had studied under Sthiramati and SIlabhadra, and
since it is not stated that Xuanzang actually met Sthiramati, it would
be reasonable to assume that there was no direct contact between
Xuanzang and Sthiramati.
6
A point worth noting is that nowhere in
these works is it stated that the ideas of Xuanzang were in conflict
with those of Sthiramati. JUdging from the inscriptions at ValabhI,
there was not just one scholar named Sthiramati.
7
But in order to
simplify the following discussion, I shall proceed on the assumption
that the Sthiramati who wrote a commentary on the Mahayanasatra-
larrzkiira, the Sthiramati referred to in inscriptions at Valabhl, and the
Sthiramati under whom Jayasena studied were all the same person,
and that he was, moreover, an important figure at Valabhl, which
5 Tsukamoto Keisho (1996: Wapi 1) already equates Dehui and Jianhui with Gu-
I).amati and Sthiramati respectively. In Li Rongxi's translation of the Datang xiyu ji in-
cluded in the BDK English Tripilaka, Jianhui is rendered as "Sthiramati" (Li 1996: 284,3
& 343,4). Likewise, in Li's translation of the Datang Daciensi sanzang fashi zhuan Anhui
is also rendered as "Sthiramati" (Li 1995: 126,21). Previously, Hirakawa Akira (1979: 14)
had already suggested that Jianhui might be the same person as Anhui: On the assumption
that this view has become established in academic circles, I have therefore decided to re-
gard both Jianhui and Anhui as Chinese equivalents of Sthiramati.
6 Datang Daciensi sanzangfashi zhuan (T. 50: 244a7ff.):
*U*fUfl-llo
The person by the name of Shengjun re-
ferred to here is thought to be Jayasena, who according to Xuanzang studied under
Sthiramati and SIlabhadra. In connection with the original Sanskrit equivalent of Sheng-
jun, we fmd in the Datang xiyu ji (T. 51: 920a15) the words and
this Jayasena may be assumed to refer to the person under whom Xuanzang studied. The
Datang xiyu ji (T. 51: 899a13) also mentions a king by the name of Prasenajit whose
name is also rendered in Chinese as Shengjun but he was a
legendary figure and differs from the person under whom Xuanzang studied.
7 Tsukamoto 1996: 527, @ & @.
SIMILARITIES BETWEEN STHIRAMATI AND XUANZANG 361
ranked with Nalanda as a centre of Buddhist learning. On this basis,
I shall set abol,lt ascertaining the fact that the theories presented in
Sthiramati's commentary on the Mahiiyiinasfitriilaf!lkiira and the
theories deemed to be legitimate in the Cheng weishi lun bear a close
resemblance to each other.
Now, even if Dharmapala was a contemporary of Sthiramati,
there would still seem to be difficulties in immediately equating the
views deemed to be legitimate in the Cheng weishi lun with those of
Dharmapala when one considers that none of his writings have sur-
vived in the original and one also takes into account the passage of
time within the confines of Nalanda from Dharmapala to Silabhadra
and then to Xuanzang. But if one posits a picture pitting Sthiramati,
based at Valabhl, against Dharmapala, based at Nalanda,8 it seems
strange that Sthiramati should have already completed the systemiza-
tion of the correspondences between the eight consciousnesses and
four knowledges.
At any rate, a scheme of correspondences between the eight con-
sciousnesses and four knowledges cannot be found in Silabhadra's
writings but does exist in Xuanzang's translations. It thus seems that
this theory was either formulated by Xuanzang, who had an extra-
ordinary enthusiasm for Abhidharmic systemization, during the
course of translation or else he knew of the interpretation given in
Sthiramati's commentary on the Mahiiyiinasfitriilaf!lkiira and
8 It seems to me that this kind of confrontational schema does not accord with the ac-
tual situation in India at the time. The Mahayana movement is thought to have spread with
considerable speed from southern India to northern India, and when one considers the re-
mains, etc., centred on Gandhara and extending over a wide area, it is inconceivable that
there would have been any clear-cut segregation between so-called Hlnayana and Maha-
yana such as we tend to posit today. Because Xuanzang saw divisions between schools
when he viewed his longed-for India from the vantage point of China, he probably as-
sumed that Nalanda and ValabhI stood opposed to each other. One should rather also take
account of the fact that Xuanzang travelled to many places in India, including ValabhI.
Hakamaya Noriaki once wrote that it was Xuanzang's disciples who first began saying
that Dharrnapala and Sthiramati were divided on all matters, and that one can find in-
stances suggesting that there were in fact surprisingly close connections between the two
(Kuwayama and Hakamaya 1981: 238). I go one step further and consider there to be a
close relationship between not DharrnapljJa, but Xuanzang, and Sthiramati.
362 HIDENORl SAKUMA
adopted this as the legitimate interpretation. Xuanzang would
naturally have known of Prabhakaramitra's translation of the Mahii-
yiinasiitriilaYJ1kiira.
9
Here I shall assume that since comparatively few
typically Chinese interpretations seem to have found their way into
the Chinese translation of the MahiiyiinasiitriilaYJ1kiira, unlike in the
case of the MahiiyiinasaYJ1graha, Xuanzang did not produce a new
translation. That being so, it is probably safe to suppose that
although Xuanzang translated neither the verses of the Mahiiyiinasu-
triilaYJ1kiira together with Vasubandhu' s commentary, which together
constituted an important work, nor the commentaries by Asvabhava
and Sthiramati, he was fully cognizant of their content.
Similar evidence can in fact also be found in connection with sev-
eral other doctrinal theories. In order to make clear the thread of my
arguments, I shall therefore in the following proceed on the assump-
tion that the views deemed to be legitimate in the Cheng weishi lun
were not advocated by Dharmapala, but were actually propounded
by Xuanzang at the instance of Kuiji. As for Sthiramati, I shall put to
one side his connections with ValabhI and Dharmapala and focus
solely on the content of the commentaries attributed to him, which I
shall consider to represent Sthiramati's theories.
1. Focal points for a comparative examination of the doctrinal
theories of Sthiramati and Xuanzang
When undertaking a comparative study of the doctrinal theories of
Sthiramati and Xuanzang, it is necessary to indicate the criteria on
which such a study is based. In the case of Sthiramati, I consider the
Sanskrit originals and Tibetan translations of works attributed to him
(the latter of which may be regarded as word-for-word translations)
9 The Mahayanasfitralal!1kara is frequently cited as a doctrinal authority in treatises
composed by Xuanzang's disciples: Kuiji, Cheng weishi lun shuji (T. 43: 599b21ff.); Hui-
zhao Cheng weishi lun liaoyi deng (T. 43: 809cl8ff.); Zhizhou
Jll!J, Cheng weishi lun yanmi (T. 43: 976alOff.). It is clear from the Datang
Daciensi sanzang fashi zhuan that Xuanzang received instruction from Jayasena in the
MahayanasatralalJ!kara (T. 50: 244a21ff.:
:!lI:fiiil/F{ti!E.@IH
to represent his views. In the case of Xuanzang, on the other hand, I
consider the Mahiiyiinasa1!lgraha, Buddhabhiimisiistra, and other
. works translated by him to represent his views. To regard Tibetan
translations as the equivalent of originals while viewing Chinese
translations as expressions of the thought of their translator Xuan-
zang may seem to indicate a lack of consistency in my criteria. But
grounds for equating Xuanzang's translations with his own thought
can be found in the findings of several researchers,JO and this ensures
the validity of my criteria. Nonetheless, it is true that there is a
difference between the criteria, and scrupulous care will be taken in
the treatment of all material.
Next, I wish to mention the doctrinal theories I shall use as indi-
ces in my comparative examination. I shall focus on the following
three topics, regarding which I have already achieved some results in
past investigations.
1. Correspondences between the four knowledges and eight
consciousnesses
11
2. Correspondences between the four knowledges and three
bodies
12
3. The formation of the five-gotra system
13
10 The following research may be cited as corroboration of this. Basing himself on a
comparison of Xuanzang's Chinese translation and the Tibetan translation of Asvabhava's
commentary on the Mahtiyanasal?lgraha, Hakamaya (1969) pointed out early on that
whereas the correspondences between the eight consciousnesses and four knowledges are
clearly indicated in Xuanzang's translation, they are not found in the Tibetan translation.
This article has been reprinted in Hakamaya 2001 (490-503) with the addition of many
subsequent research findirigs, and considerable depth has been added to his observations.
For a history of research on this subject, cf. the supplementary section of Hakamaya 2001.
11 See Sakuma 1983, 1984, 2002.
12 See Sakuma 1982, 1987.
13 See Sakuma 2007a, 2007b.
364 IDDENORI SAKUMA
1.1. Similarities between Sthiramati and Xuanzang as seen from Cor-
respondences between the four knowledges and eight conscious_
nesses
The correspondences between the four knowledges and eight
consciousnesses are not mentioned in either the verses of the Maha-
yanasiitralalflkara or the prose commentary attributed to Vasuban-
dhu, both preserved in Sanskrit,14 nor are they mentioned in the Ti-
betan translation of Asvabhava's commentary on the Mahayanasam-
graha.
15
Jn Silabhadra's Buddhabhfimivyakhyana (preserved in Ti-
betan) we find evidence of a transitional stage in the formulation of
these correspondences.
16
The finalized scheme of correspondences is
14 MSA(Bh) IX.67 (F: 38,18-23; L: 46,15-19): buddhajiiiinavibhiige dasa I
iidarsajJ1iinam acala/?I trayajJ1iinal!l tadiiSritam I eva
ca II 67 II caturvidha/?I buddhiiniiJ?1 jiiiinam iidarsajl1iina/?1 samatiijJ1iil1aJ?1
naI!l ca I iidariajiiiinam acalaJ?1 trD!i jJ1iiniini tadiiSritiini caliini I
15 See n. 10. The relevant passage in the Tibetan translation begins as follows (Haka-
maya 2001: 496): rnam par shespa'i phung po gyur pas ni me long Ita bu dang I mnyam
pa nyid dang I so sor rtog pa dang I bya bas grub pa'i ye shes la dbang 'byor pa thob ste I.
Correspondences with the eight consciousnesses are not mentioned in any subsequent pas-
sages either. It is obvious from the material cited by Hakamaya that the correspondences
between the four know ledges and eight consciousnesses are given in the corresponding
Chinese translation by Xuanzang (see n. 17).
16 Nishio 1940, voU: 120,17-121,15: rnam pa gcig tu na dngos po shes pa dang I de
dmigs pa zhes bya ba zlas dbye ba yin te I gnas nganlen mtha' dag gi gnas kun gzhi mam
par shes pa gnyen po'i stobs kyis gnas ngan len ma Ius pa dang bral bas yongs su gyur pa
me long Ita bu'i ye shes zhes bya ba gzhan gyi dbang dag pa zhes tha snyad gdags pa sems
kyi rnam par TtOg pa thams cad med pa'i ngo bo la 'di ni dngos po tsam mo zhes spyi'i
rntl1n par sgro btags nas dngos po 'i sgra brjod do II me long Ita bu 'i ye shes dmigs par bya
ba dang I dmigs pa mnyam pa 'i mam pa yang gnas ngan len gyi gnas yongs su gyur na I de
ltar mam par bzhag go II
dngos po de shes pa ni dngos po shes pa ste I de la dmigs pa zhes bya ba'i tha tshig go I de
yang mnyam pa nyid kyi ye shes yin no II
de'i rjes la thob pa dag la 'jig rten pa rang gi rtog pa yongs su gcod pa 'i mam pa gang yin
pa de'i spyod yul yang gzhan gyi dbang gi ngo bo nyid yongs su {{yur pa yin no II ries las
thob pa'i ve shes de ni so SOT rtog pa'i ye shes kho na yin te I de rang gis rtogs pa la so SOT
rtog pa'i tshe I de gnyis yul yin pa'i phyir TO II de la 'am de gnyis la dbang ba ste I 'di de la
zad mi shes pa'i mtshan nyid yod pas zhes bya bar tshig mam par sbyar TO II 'dis mtshon
par byed shes par byed pas na zhes byed pa'i byed pa por byas pa'i phyir ro II
dngos pa shes pa de dmigs pa de la 'am dngos po shes pa dang I de dmigs pa de gnyis la
dbang zad mi shes pa'i mtshan nyid ces bya ba'i tha tshig ste I 'dis ni lam gnas yongs su
SIMll..ARITIES BETWEEN STHIRAMATI AND XUANZANG 365
found in Xuanzang's translations of the BuddhabhamiSiistra
17
and
Cheng weishi fun, 18 and the same theory appears in Sthiramati's
commentary on the Mahiiyiinasiltriifarrtkl'ira.19 These points were
touched on briefly in the previous section.
gyur yang bstan pa yin no II mon mongs pa cart gyi yid yongs S<! gyur na I me long ita bu'i
ye shes chos kyi dbyings ia dl11igs nas I11nyal11 pa mid kyi ye shes skye ste I de bzhin yongs
su dag na de ia yang dbang zad mi shes pa 'thob po II
The correspondences between iilayavijfiiina and iidarsajfiiina and between and
samatiijfiiina are clearly defined here, but it is not clear which consciousness is trans-
formed into pratyavekCfiijfiiina, and there is no explanation here or elsewhere regarding
krtyiinuCfthiinajl1iina. The corresponding passage in Xuanzang's translation of the Buddha-
bhamisiistra (T. 26: 324b4ff.) reads as follows: f!!ll5f1:I19%\'llP:kt'i' J3
'1i-5Jo
J7<Jffi
fHfjgm 7 J!\Ii5JjJiJ;!;y:o i':t%\' 1!f 0 0
S3 mil 7':J i:tm

fEl
5J31J J3
0
Here the original would seem to have been been modified, and it is stated that the other
two knowledges are connected to the six consciousnesses, although the translator did not
go so far as to state which knowledge is connected to which consciousness.
17 An explicit indication of the relationship between the four knowledges and eight
consciousnesses is found in the following passage (T. 26: 302b29ff.):
%\'tJ!!,c"o
i8!Jilf!=jiA J3

t J!!,c"o ImmP.ltrMi-fflf'Fw:o
No variants have been reported for this passage. It is thus evident that it presents the legiti-
mate view of the Faxiang school in an unadulterated form. The corresponding section is,
moreover, completely missing in the Tibetan translation of STIabhadra's Buddhabhamivyii-
khyiina. One is thus compelled to accept that this passage was added by Xuanzang.
18 Having resolved the question of the relationship between the Abhidharmic catego-
ries of consciousness, belonging to the category of the mind, and knowledge, belonging to
the category of mental attributes, by stating, "consciousness is associated with the mind"
the Cheng weishi iun continues as follows (T. 31: 56b2ff.): t:kn
frol'1zlffif1joo In other words, it treats
the correspondences between the two as if they were self-evident.
19 SeizQ Bunten Kenkyukai 1979: 32,4ff. (D. 113b3ff.; P. 128a3ff.): yang na gzugs
dang I tshor ba dang I 'du shes dang I 'du byed dag dang I mam par shes pa brgyad ia yod
pa'i stong pa nyid dag na chos kyi dbyings mam par dag par 'gyur ro II mam par shes pa
366 HIDENORI SAKUMA
In addition, it is also clear that the correspondences found in Pra-
bhakaramitra's Chinese translation of the MahayanasutralalJ1kara
20
and Xuanzang's Chinese translation of Asvabhava's commentary on
the MahayanaSal?lgraha
2i
differ from those given by Sthiramati and
in Xuanzang's translations of the Buddhabhumisastra and Cheng
weishi lun. In the former group the correspondences are alayavijiiana
- adarsajiiana, manas - samatajiiana, manovijiiana -
jiiana, and five active consciousnesses - while in
the latter group the correspondences are alayavijiiana - adarsajiiana,
manas - samatajiiana, manovijiiana - and five
active consciousnesses - Since I have already
demonstrated elsewhere that originally the former set of correspon-
dences would have been the more natural interpretation, I shall not
go into any further detail here.
22
Important in this regard is the fact that Prabhakaramitra' s transla-
tion of the MahayanasutralalJ1kara and Xuanzang's translation of
Asvabhava's commentary on the MahayanasalJ1graha were trans-
brgyad las kun gzhi dag na me long Ita bu'i ye shes su gyur ro II nyon mongs pa'i yid dag
na mnyam pa nyid kyi ye shes su 'gyur ro II yid kyi mam par shes pa dag na so sor kun du
rtog pa'i ye shes su 'gyur ro II mig nas Ius kyi bar du mam par shes pa Inga dag na bya ba
grub pa'i ye shes su 'gyur te I ye shes bzhi dang chos kyi dbyings mam par dag pa Inga
thob pa la gnas gzhan du gyur pa Inga zhes ba'o II Similar explanations can also be found
elsewhere in the same work.
20 T. 31: 606c23ff. - }\.
-=-'1fWl.!@io IZ9
'1ff1=:$!@io {iiJJJ,$:o -=-!@iItiMco

!@i 0 This represents the reading of the old Song edition, the oldest manuscript
used by the editors of the Taish6 edition when editing this text.
21 T. 31: 438a13ff. -
ili!mB!o it
j!\UJ-5JiJhlitlJffi','tJtlo
aifo n-NitJ!l:fli:i)ifii1j!fl'o
0 5JGt/o';{Jilli[o 1t-:k1f'P
i3 :(:E{1=Jtlo 0 'If it +::!J--i:lJJ iltJilo
This too represents
the reading of the old Song edition.
22 See Sakuma 2002, based on the Sanskrit text of the Mahayanan7tralay[!kara and
other works.
lated earlier than Xuanzang's translations of the BuddhabhUmisiistra
and Cheng weishilun.23 The legitimacy ofthe correspondences is ex-
plained in the Buddhabhamisiistra, but the only reason given is the
weak reason that the former set of correspondences is irrational be-
cause it does not follow the regular order, while the latter set repre-
sents the legitimate view because it follows the regular order.24 Why
would Xuanzang have been compelled to give such a reason? If
Sthiramati had prior to this set forth this latter set of correspon-
dences, it would mean that Xuanzang rejected the views of SIla-
bhadra and Prabhakaramitra, the latter of whom is thought to have
studied at Nalanda,25 and emended it on the basis of Sthiramati's
view. If Sthiramati's view should prove to have been formulated
around the same time as Xuanzang advanced this view, it would be-
come necessary to rethink Sthiramati' s dates. Such is the positional
relationship between Sthiramati and Xuanzang as deduced from our
first index.
1.2. Similarities between Sthiramati and Xuanzang as seen from
correspondences between the four know ledges and three bodies
In order to simplify things, I first wish to confirm the following
facts. The purity of the Dharma-realm (dharmadhiituvisuddhi) was
added to the four know ledges as a distinguishing feature of the state
23 According to Kuwayama Shoshin (Kuwayama and Hakamaya 1981: 49ff.), Xuan-
zang would have met Prabhakaramitra shortly before his departure for Central Asia and
India and would have obtained from him information about these regions and about Sna-
bhadra at Nalanda. I too believe that this is highly likely to have been the case. As is noted
by Hakamaya (ibid.: 195), it may be safely assumed that Prabhiikaramitra translated the
Mahiiyanas!7traia1Jlkara during Xuanzang's twenty-year absence from China. A list of
works translated by Xuanzang with their dates can be found in the same work (252ff.).
24 T. 26: 302c7ff. -
jIl', One cannot help feeling that the citing of the fact that the order does not
follow the regular order as the reason for rejecting this correspondence is an extremely
weak reason, and it is to be surmised that Xuanzang too was unable to fmd any other
legitimate reason for doing so.
25 Xu gaoseng zhuan T. 50: 439c26ff. -
mlIl, git-i5:iiJ<J.m, Jlti5::J'tl?, ;;js:*UfrJ::Ef!E, +ID:ti:ll:,
+i1it{'%,
k*, 1Z91rt/FJtfif+
368 HIDENORI SAKUMA
of Buddhahood, and together these are referred 10 as the five
dharmas (or five elements). The purity of the Dharma-realm is
characterized as principle and the four know ledges as wisdom.
Originally the five dharmas and three bodies represented different
schemata, and the process of their development also differed. The
four know ledges of the Buddha appear in their finalized form al-
ready in the Sanskrit verses of the MahiiyanasatralaJ?Lkara (IX.67-
76). Since there are no earlier passages indicative of the develop-
ment of this concept, the circumstances of its establishment are not
known. As for Buddha-bodies, the basic theory until then had posited
two bodies, namely, the physical body (ntpakaya) and the Dharma-
body (dharmakaya). With the emergence of the Yogacara school, a
three-body theory consisting of the dharmakaya or svabhavakaya,
the saJ?Lbhogakaya (enjoyment-body), and the nirma7Jakaya (trans-
formation-body) came to be advanced from the standpoint of
Mahayana Buddhism. Because of the use of the two different terms
dharmakaya and svabhavakaya, these came to be treated as two
separate bodies, resulting in effect in a four-body theory. In later
times, the four-body theory developed into a five-body theory and
other theories of multiple Buddha-bodies.
26
The four knowledges and three bodies are mentioned in chapter 9
of the MahiiyanasatralaJ?Lkara, with the purity of the Dharma-realm
being discussed in verses 56-59, Buddha-bodies in verses 60-66, and
the four knowledges in verses 67-76. But there is no mention of any
correspondences between them in either the verses or Vasubandhu's
and Asvabhava's commentaries, and they appear only in Prabhakara-
mitra's Chinese translation of the MahiiyanasatralaJ?Lkara and in the
Tibetan translation of Sthiramati's commentary (SA Vbh). Apart
from this, the correspondences between the four knowledges and
three bodies are also described in the Tibetan translation of Sua-
bhadra's Buddhabhamivyakhyana.
27
26 I have previously discussed the development of Buddha-body theory in the direction
of theories of multiple bodies on the basis of the "Dharmakiiya Chapter" in the Abhisama-
yiilal!lkiira; see Sakuma 1992a, 1992b, 1994.
27 Because the relevant passages in these works are all lengthy, and also because com-
plex procedures are necessary to demonstrate the correspondences, the passages wiII not
The correspondences between the five dharmas and three bodies
are clearly described in Sthiramati's SAVbh. In SAVbh IX.60 they
are explained with reference to iiSrayaparilvrtti: illayavijfiilna turns
into ildarsajiiilna and is associated with the dharmakilya, which also
corresponds to the svabhilvakilya; turns into samatil-
jiiilna and manovijiiilna into and these are associ-
ated with the sa1"!lbhogakilya; and the five active consciousnesses
tum into which is associated with the nirmill}a-
kilya.
In Prabhakaramitra's Chinese translation of the Mahilyilnasutril-
laT!lkilra these correspondences are indicated in X.53ff., correspond-
ing to IX.59ff. in the Sanskrit text. Prabhakaramitra presents the
relationship between the eight consciousnesses and four know ledges
in a form different from that of Sthiramati and Xuanzang, and it may
be summarized in the following, manner: illayavijiiilna turns into
ildarsajiiilna and into samatiljiiilna, and these are
associated with the dharmakilya; the five active consciousnesses tum
into which is associated with the sa1"!lbhogakilya;
and manovijiiilna turns into which is associated
with the nirmill}akilya.
In the case of STIabhadra's Buddhabhumivyilkhyilna, on the other
hand, in which the correspondences between the four knowledges
and eight consciousnesses have not been finalized, one must posit the
following relationships. First, it is stated that illayavijiiilna turns into
ildarsajiiilna and into samatiljiiilna, but no relationships
are posited between the other consciousnesses and knowledges.
Under these circumstances, the correspondences between the five
dharmas and three bodies are as follows: the purity of the Dharma-
realm and ildarsajiiilna are associated with the svabhilvakilya (=
dharmakilya), samatiljfiilna and are associated with
the sa1"!lbhogakilya, and is associated with the
nirmill}akilya.
be quoted here. Reference should be made to my earlier studies on this subject (Sakuma
1982, 1987, 1989).
370 . IllDENORI SAKUMA
Thus, the correspondences between the five dharmas and three
bodies differ from one text to another, and in content they are even
more complicated than has been indicated in the above.
That being so, how are these correspondences treated in Xuan-
zang's translations of the Buddhabhamisiistra and Cheng weishi lun?
First, in the case of the Buddhabhiimisiistra it is difficult to compre-
hend even the gist of the relationship between the five dharmas and
three bodies. If one assumes that Xuanzang considered the .connec-
tions between the two with reference to Silabhadra's Buddhabhiimi-
vyiikhyiina, it is to be surmised that he decided that it would be diffi-
cult to deal with the relationship between the five dharmas and three
bodies, and also the eight consciousnesses, with the consistency of
Abhidharmic categories. It is obvious, in other words, that Xuanzang
was rather perplexed about the relationship between the five dhar-
mas, three bodies, and eight consciousnesses when he translated the
Buddhabhiimisiistra.
How much clearer, then, had the relationship between these three
become when Xuanzang translated the Cheng weishi lun ten years
later? ill the Cheng weishi lun, the relationship between
jfiiina and the three bodies is by no means clear, but it can be
generally inferred that the purity of the Dharma-realm is associated
with the svabhiivakiiya, iidarsajfiiina with the self-enjoyment body,
samatiijfiiina with the other-enjoyment body, and
na with the nirmiiT}a-kiiya. is subtly related to both
the other-enjoyment body and the nirmiiT}a-kiiya,28 but I cannot go
into details here.
29
The concepts of self-enjoyment body and other-enjoyment body
had in fact already appeared in Xuanzang' s translation of the Bud-
dhabhiimisiistra,30 but it was only in the Cheng weishi lun that they
28 T. 31: 56c29ff. -

f!:I 0
29 See Sakuma 1987.
30 T. 26: 294b3ff. -
$Ii! f!:I 'tm
were to some extent clearly utilized in explaining the relationship
between the five dharmas and three bodies, and they result in effect
in a four-body theory. The four-body theory is discussed at great
length in the AbhisamaytilaY[lkara, and possible connections with this
work raise some interesting questions. But the AbhisamayalaY[lkara
was not translated into Chinese, and I shall not delve any further into
this subject since it would lead us away from the question at handY
As for the two concepts of self-enjoyment body and other-enjoy-
ment body, it is in fact possible to detect intimations of the former in
Sthiramati's SAVbh.Unfortunate1y the procedures necessary for
demonstrating this are somewhat involved, and limited space does
not allow me to reproduce them here. Reference can be made to my
previously published study on this subject.
32
If my above analysis is correct, it is possible to infer the follow-
ing process. The five dharmas and three bodies initially developed as
two separate theories, but by the time of Sllabhadra and Prabha-
karamitra correspondences between the two had been established.
Xuanzang had doubts about his teacher Sllabhadra' s views in terms
of Abhidharmic categories, and his solution could be easily ex-
plained were one to assume that he borrowed the notion of the self-
enjoyment body and the schema of correspondences between the
four knowledges and eight consciousnesses from Sthiramati. This is,
of course, no more than a possibility, but in light of the investi-
gations I have conducted until now, it would seem natural to me to
view the situation in this fashion. Such is the positional relationship
between Sthiramati and Xuanzang as deduced from our second
index.
{jjJJ1JJJ:: o lit!
Also
294b14ff.: 'tJfl !!iil!l:i;]'i!.-- {J!l.'tJfl
There are no corresponding passages in the Tibetan transla-
tion of SIlabhadra's Buddhabhiimivyiikhyiina.
31 See Sakuma 1992a, 1992b, 1994.
32 Sakuma 1987: esp. 394.
372 HIDENORI SAKUMA
1.3. Similarities between Sthiramati and Xuanzang as seen from the
formation ofthefive-gotra system
One theory propounded by the Faxiang school in China and Japan
that became the cause of much debate with other schools was the
five-gotra system presented in the BuddhabhLimisiistra translated by
Xuanzang. When one traces its content back to India, one can cer-
tainly detect a process whereby the part of the five-gotra system
relating to beings without any possibility of attaining Buddhahood
and the part relating to the three vehicles gradually merged. It is to
be surmised that Mahayana Buddhism advanced the idea of the three
vehicles of the sriivaka, pratyekabuddha, and bodhisattva out of a
need to assert its legitimacy vis-a.-vis Mainstream Buddhism, but ini-
tially beings without any possibility of attaining Buddhahood were
not juxtaposed to the three vehicles. The question of gotra (lineage)
was simply discussed quite separately from the idea of three vehicles
in terms of beings with the possibility of attaining Buddhahood
(gotra) and beings without any possibility of attaining Buddhahood
(agotra). It would appear that these two separate groupings were
first brought together as five categories in Sthiramati's SAVbh.
The overall current of thought leading to the five-gotra system
can be understood in the following manner. Initially, the vehicle
among the three vehicles to which a practitioner belongs is not deter-
mined, and if one supposes that his association with one of the vehi-
cles is gradually determined in the course of his practice, then the
initial stage corresponds to the indeterminate lineage and the stage
when his lineage has been determined corresponds to one of the
three vehicles. Therefore, the indeterminate lineage and the three
vehicles are not parallel categories. The question of gotra and agotra
had already been raised from the time of the YogiiciirabhL""imi, and it
can also be readily inferred that there was some connection between
the state of having the possibility of attaining Buddhahood (gotra)
and the three vehicles. But it was in Sthiramati's SA Vbh that agotra
is first presented alongside the indeterminate lineage and the three
vehicles. The LankiivatiirasLitra is often considered to provide a
theoretical basis for the five-gotra system, but as is indicated in the
Yuqielun ji it was known from an early stage that the
Lmikiivatiirasatra was unsuitable as the theoretical basis of the five-
gotra system.
33
In light of the above, I wish to show the process
leading to the five-gotra system with reference to the "Gotra
Chapter" in the MahiiyiinasatriilaT(lkiira.
In the Sanskrit text of the MahiiyiinasiUriilm!lkiira, the first ten
verses of the "Gotra Chapter" explain the existence of gotra and the
eleventh verse explains the absence of gotra, or agotra. Within this
overall framework, the verses necessary for the establishment of the
five-gotra system were verses 6 and 11. It would presumably be safe
to assume that originally there was no intention in either the verse
section or Vasubandhu' s commentary to forge a direct link between
these two verses.
A verse on the distinction between the kinds [oflineages]:
The lineage may be determinate or indeterminate, shaken or unshaken
By conditions. This distinction between lineages is, in brief, fourfold. (v. 6)
In brief, lineages are fourfold. They are determinate and indeterminate, and
these are in [that] order unshaken and shaken by conditions. (Mahayanasil-
tralaY[lkara III.6)34
As can be seen in this verse, there are lineages that are determinate
and others that are indeterminate. Since the term "three vehicles" is
used in Vasubandhu' s commentary on verse 2, "determinate" means
belonging to one of the three vehicles. "Indeterminate," on the other
hand, means that the practitioner, under the guidance of a teacher, is
still in a state of vacillation regarding his lineage. This later became
the independent category of "indeterminate lineage," but it is
unlikely to have been regarded as an independent category at this
stage. This verse simply gives expression to the process of practice,
that is, to the fact that there are both those who, under the guidance
of a teacher and so on, are no longer in a state of vacillation and
those who are still vacillating in the midst of their practice.
33 On the subject of the above process, see Sakuma 2007a.
34 MSA(Bh) m.6 (F: 21,14-18; L: 11,20-24): prabhedavibhage slokaf:z. niyattiniyata1Jl
gotram ahiiryaql haryam eva ca / pratyayair gotrabhedo 'yaql samiisena caturvidha(l / / 6 / /
samiisena caturvidhaql gotram. niyatiiniyata1rz tad eva yathiikramarrz pratyayair ahiiryarrz
haryarrz ceti.
374 HIDENORI SAKUMA
In contrast, it is evident that in Asvabhava's commentary
(MSAT)35 and Sthiramati's SA Vbh
36
these four categories have
clearly come to be treated as the three vehicles of the sriivaka, praty-
ekabuddha, and bodhisattva and, independent of these, an indetermi-
nate lineage.
Next, Asvabhava's MSAT and Sthiramati's SAVbh begin their
explanations of Mahiiyiinasutriila1!1kiira III. I I in the following man-
ner. Asvabhava's commentary on verse 11 begins by commenting
directly on Vasubandhu' s commentary with the statement "Where it
says, 'In this sense "he who does not have the quality for parinir-
35 MSAT ID.6 (D. 51b6-52a3; P. 58b6-59a2): rigs nges pa ni nyan thos dang I rang
sangs rgyas dang I sangs rgyas kyi rigs su nges par gnas pa gang yin pa ste I nyan thos
nyid thob (D: 'thob P) par nges pa'i rigs gang yin pa de ni nam yang rang sangs rgyas
dang sangs rgyas nyid 'thob pa'i rgyur mi 'gyur ro II de bzhin du rang sangs rgyas dang I
sangs rgyas kyi rigs dag kyang sbyar bar bya'o II
ma nges pa ni (em.: pa'i DP) rkyen gyi dbang gyis nyan thos dang rang sangs rgyas dang I
sangs rgyas kyi (D: kyis P) rigs rnams kyi rgyur 'gyur te I dper na ri'i phyogs gang dag la
(P: las D) gdon mi za bar gser 'ba' zhig 'byung gi I dngul 'ba' zhig kyang ma yin la I zangs
'ba' zhig kyang ma yin pa de Ita bu yang yod ia I phyogs gang zhig 'jim gong drii ba la
sogs pa'i bcos legs (P: lags D) bya ba'i dbang gyis (D: gyi P) gdon mi za bar res 'ga' gser
'byung la I res 'ga' dngulla sogs pa 'byung bar yang yod pa de ita bu' 0 II
de nyid kyi phyir rigs nges pa ni I rkyen rnams kyis mi 'phrogs la ma nges pa ni 'phrogs pa
yin no II
36 SAVbh ID.6 (D. 45a4-45bl; P. 49a3-49bl): de ia rigs nges pa ni gang nyan thos su
rigs nges par gnas pa dang I rang sangs rgyas su rigs nges par gnas pa dang I sangs rgyas
,sit rigs nges par gnas pa ste I nyan thos su rigs nges par gnas pa yang rigs des nyan thos
kyi byang chub nyid 'thob kyi ji itar byas kyang nams kyang rang sangs rgyas kyi byang
chub dang I sangs rgyas su 'thob pa'i rgyur mi 'gyur ro II rang sangs rgyas kyi rigs nges
pa yang rigs des rang sangs rgyas kyi byang chub nyid thob kyi ji itar byas kyang nams
kyang nyan thos dang sangs rgyas kyi byang chub 'thob pa'i rgyur mi 'gyur ro II sangs
rgyas kyi rigs can yang rigs des sangs rgyas kyi byang chub nyid 'thob (D: thob P) kyi ji
ltar byas kyang nams kyang nyan thos dang rang sangs rgyas kyi byang ch!,b tu mi 'gyur
ba'D II
rigs ma nges pa ni rkyen gyi dbang gis nyan thos dang rang sangs rgyas dang sangs rgyas
(D: dang sangs rgyas, missing in P) kyi rigs gang yang rung ba cig gi (D: gyis P) rgyur
'gyur te I nyan thos kyi dge ba'i bshes gnyen dag gis bsgral na ni nyan thos kyi rigs can du
yang 'gyur I rang sangs rgyas kyi dge ba'i bshes gnyen gyis bsgral na ni I rang sangs rgyas
kyi rigs can du yang 'gyur I byang chub sems dpa'i dge ba'i bshes gnyen gyis bsgrai na ni
sangs rgyas kyi rigs can du yang 'gyur ro (D: gyur ba'o P) II
va7Ja" is meant by "he who dwells in no lineage''', "37 but Sthiramati
adds: "Where it says 'a verse on the distinction of the lineage-less,'
having earlier explained the lineage of the sravaka, the lineage of the
pratyekabuddha, the lineage of the bodhisattva, and the indetermi-
nate lineage, it now explains the lineage-less."38 Whereas Asvabhava
clearly refers to the lineages of the three vehicles and the indetermi-
nate lineage in his commentary on verse 6, but does not link them
directly to the verse on agotra, there is clear evidence in Sthiramati's
commentary of an intent to create a scheme of five gotras. Here one
can discern the manner in which the scheme of five gotras gradually
evolved.
In Xuanzang's translation of the Buddhabhumisastra this scheme
developed into five categories consisting of the lineages of the three
vehicles, the indeterminate lineage, and the lineage-less, the last of
which was simplified to mean those without any possibility of attain-
ing Buddhahood.
39
In the Tibetan translation of STIabhadra's Bud-
dhabhumivyakhyana there is no mention whatsoever of these ideas.
Originally, MahayanasutralaY(lkara III.II referred to two kinds of
agotra, namely, those who are unable to attain Buddhahood at the
present point in time but will be able to after a certain period of time,
and those who will never attain Buddhahood.
40
It is to be surmised,
therefore, that in order to simplify the five gotras, Xuanzang
37 MSAT m.ll (D. 52blf.; P. 59a8f.): don 'di la ni rigs med pa la gnas pa yongs su
my a ngan las mi 'da' ba'i chos can yin par bshad do zhes bya ba na.
38 SA Vbh m.ll (D. 48a3ff.; P. 52b3ff.): rigs med pa la mam par dbye ba'i tshigs su
bead pa zhes bya ba la I gong du nyan thos kyi rigs dang I rang sangs rgyas kyi rigs dang I
byang chub sems dpa'i rigs dang I rigs ma nges pa bshad nas I da ni rigs med pa 'chad de I
39 T. 26: 298a12ff. - -r,fJ'ft*H:o
tto 1m /fJE'ft*-t!Eo tfj i:!t;IJi'tll'ft*'l1:o 71J3IJl.l
:li:JltrIm'ft*'l1:o tfji:!t;IJ
l.'tlllZ9il!co

40 MSA(Bh) m.ll (F: 22,21-23,3; L: 12,19-13,2): agotrasthavibhiige slokal;. ekantiko
duscarite 'sti kascit kascit samudghiititasukladharma I amokeabhiigfyasubho 'sti kascin
nihlnasuklo 'sty api hetuhfnal; 1111 II
376 HIDENORI SAKUMA
restricted the meaning of agotra to those without any possibility of
attaining Buddhahood.
To sum up, the five categories of the five-gotra system do not ap-
pear in the verses of the Mahiiyiinasutriilarphira or Vasubandhu's
commentary; in Asvabhava's commentary the content of III.6 was
clarified in the form of the lineages of the three vehicles and an
indeterminate lineage, and Sthiramati further linked this verse to
agotra mentioned in III.ll; Xuanzang, it is to be surmised, simpli-
fied the content of agotra by restricting it to those without any
possibility of attaining Buddhahood and thus brought to completion
the five-gotra system, regarded as one of the distinguishing features
of Faxiang doctrine. Here too one senses greater similarities between
the doctrin:i.l theories of Sthirarri.ati and Xuanzang than between
those of other scholars.
2. A comprehensive assessment: by way of conclusion
On the basis of the data on the three doctrinal theories summarized
above, I wish to focus here in particular on the doctrinal similarities
to be observed between Sthiramati and Xuanzang. The doctrin:i.l
theories selected here for the purpose of comparison represent of
course just one part of the theories of the Y ogacara school, and
therefore it is not my intention to apply the conclusions reached be-
low to the entire body of Y ogacara theories. The correspondences
between the four lmowledges, the eight consciousnesses and the five-
gotra system taken up in the above are doctrinal theories that in the
Faxiang school of China and Japan are treated as if they are self-ex-
planatory, but they were not necessarily clearly defined in India, and
therefore they are unlikely to have been centr:i.l of the
Y ogacara school. In point of fact, the correspondences between the
four lmowledges and eight consciousnesses as clarified in the Cheng
weishi lun and the correspondences between the five dharmas and
three bodies, clarified to a certain degree in the Cheng weishi lun, are
not mentioned at all in the Sanskrit text of Sthiramati's commentary
on the Tri1'[!sikii, on which the Cheng weishi lun would natur:i.lly have
been based. Since it is to be surmised that Xuanzang would have
been motivated by. different aims in the case of the five-gotra sys-
tern, it is not surprising that this is not mentioned in Sthiramati' s
commentary on the Tril'!tSika. But in verses 29 and 30, which discuss
asrayaparav!tti, the body of emancipation (vimuktikaya) , and the
dharmakaya, Sthiramati neither mentions the four know ledges nor
touches on the three bodies.
41
In addition, there is no mention of the
four knowledges in Xuanzang's translation of the Abhidharrna-
42 attributed to Sthiramati, nor do they of course
appear in the Sanskrit said to be the
work of Jinaputra, although the question of its authorship has not yet
been resolved.
43
This means that there is a need to consider why
Xuanzang should have attributed it to Sthiramati. Likewise, there are
no references to any such ideas in Sthiramati' s commentary on the
Madhyantavibhaga.
44
At any rate, when one considers that in his
commentary on the Mahayanasutralaf!lkara Sthiramati would seem
to display an enthusiasm for using his encyclopaedic knowledge to
systematize the doctrinal theories of the Y ogacara school in line with
Abhidharmic categories, it seems strange that he makes no mention
whatsoever of the four knowledges in his commentary on the
Trif!lsikii. Assuming that, as is currently estimated, he lived during
the period between 510 and 570, would he have mentioned the four
knowledges and discussed their relationship with the eight con-
sciousnesses only in his commentary on the Mahiiyiinasutriilaf!lkiira
among the voluminous commentaries he composed during his life-
time simply because the four knowledges happened to be mentioned
in the verse section of the If that were the
case, then Sthiramati's failure to mention the four knowledges and
three bodies in his commentary on the Trif!lsika could be explained
by the fact that they do not figure in the verses of the Trif!lsikii.
41 Cf. TV 29-30. As far as I can see, there is no discussion of these correspondences
anywhere in Sthiramati's commentary.
42 This is based on a search of the SAT and CBET A electronic versions of the text.
43 Cf. my index to the (Sakuma 1996). On the question
of its authorship, see Schmithausen 1969: 100, note y.
44 Cf. Yamaguchi Susumu's index (Yamaguchi 1966).
378 . RIDENOR! SAKUMA
There is one further moot point, namely, the fact that up until the
time of Xuanzang's translations of the Buddhabhamisiistra and
Cheng weishi lun one cari trace in the Tibetan translation of the
Buddhabhumivyiikhyiina (thought to be the work of Silabhadra), Pra-
bhakaramitra's Chinese translation of the Mahiiyiinasutriilarrzkiira,
and Xuanzang's Chinese translation of Asvabhava's commentary on
the Mahiiyiinasarrzgraha a process whereby the doctrinal theories of
the four knowledges and eight consciousnesses gradually .merged
and their correspondences were developed. The same process can be
seen in the correspondences between the five dharmas and three
bodies, and if one recognizes a similar process with regard to the
five-gotra system too, the theories presented in Sthiramati's com-
mentary on the Mahiiyiinasutriilarrzkiira tum out, as is evident from
our above investigations, to have overly close similarities to the
theories ultimately formulated by Xuanzang. It might be suggested
that one should consider Sthiramati's commentary on the Mahiiyiina-
sutriilarrzkiira separately from all his other works and regard it as the
work of someone else with the same name, but it is not such a simple
matter. When one also takes into account developments in the idea of
asrayapariivrtti and questions relating to the treatment of the trisva-
bhiiva theory, it becomes exceedingly complicated. Therefore, it is
desirable to reach a conclusion here with the qualification that it
applies only to the topics dealt with in the above. With such a quali-
fication, it may be assumed that the relationship between Sthiramati
and Xuanzang in the realm of philosophical thought was closer than
we have until now imagined. With this as my conclusion for the time
being, I wish to bring this paper to a close.
Abbreviations and references
Dey, Nundo La!. 1927. The Geographical Dictionary of Ancient and Mediaeval
India. London: Luzac.
Fukaura ShObun 1954. Yuishikigaku kenkyu [Studies in
Yogiiciira doctrine], 2 vols. Kyoto: Nagata BunshOd5 7k
Hakamaya Noriaki 1969. "Genj5-yaku ShOdaijoron shaku ni tsuite -
Chibettoyaku to no hikaku ni yom ichi k5satsu -"
SIMILARITIES BETWEEN STIllRAMATI AND XUANZANG 379
- T-<V - [Xuanzang's translation
of the MahiiyiinasaT[lgrahopanibandhana: Compared with its Tibetan
translation]. Indogaku Bukkyogaku Kenkyii -1/' F:{bWt1iJfYE 18-1, pp.
140-141. Repr. in Hakamaya 2001, pp. 490-503.
_ 2001. Yuishiki shiso ronko [Studies in Yogaciira thought].
Tokyo: Daizo Shuppan *_Il!t&: ..
Hirakawa Akira 1jZ)llrw. 1979. Indo Bukkyoshi -1 /' [History of Indian
Buddhism], vol. 2. Tokyo: Shunjiisha
Kuwayama SMshin and Hakamaya Noriaki 1981. Genjo 12::
[Xuanzang]. Tokyo: Daizo Shuppan *_Il!t&:.
Law, Bimala Churn. 1954. Historical Geography of Ancient India. Paris: Societe
Asiatique de Paris.
Li Rongxi, tr. 1995. A Biography of the Tripitaka Master of the Great Ci'en
Monastery of the Great Tang Dynasty. BDK English Tripi!aka 77. Berkeley:
Numata Center for Buddhist Translation and Research.
_ tr. 1996. The Great Tang Dynasty Record of the Western Regions. BDK
English Tripi!aka 79. Berkeley: Numata Center for Buddhist Translation and
Research.
Mizutani Shinjo tr. 1999. Daito saiiki ki [Datang xiyuji],
vol. 3. Tokyo: Heibonsha 1jZfLt.
MSA(Bh): F: Funahashi Naoya :liH1!iii;l;J'Illi;, ed.,
Nepiiru shahon taishO ni yoru Daijo shOgonkyoron no kenkyii
J:: 7.) [MahayiinasiitriiIaT[lkiira (Chapters I, II,
ill, IX, X) revised on the basis of Nepalese manuscripts], Tokyo: Kokusho
Kankokai 1985; L: Sylvain Levi, ed. and tr., Mahiiyiina-siitrii-
laT[lkiira: Exposi de Ia doctrine du Grand Vihicule seion Ie systeme
Yogiiciira, Paris, 1907.
MSAT: *Mahiiyiinasiitriilarjzkiiratfkii by Asvabhava. Peking ed., no. 5530;
Derge ed., no. 4029.
Nishio Kyoo WfJRQ$. 1940. Butchikyoron no kenkyii [A study
of the Buddhabhilmi-vyiikhyiina], 2 vols. Tokyo: Kokusho Kankokai lEilHU

Sakuma Hidenori 1982. "GoM to sanshin no musubitsuki" C:
= [On the connection between the five elements and the three
bodies, with special reference to the Buddhabhiimivyiikhyiina]. Indogaku
Bukkyogaku Kenkyii FPlt'{bWc'1iJfYE 31-1, pp. 124-125.
380 HIDENORI SAKUMA
1983. "Shichi to hasshiki no ketsugo kankei - sono seiritsu -katei -"
- .:.to)P)(;:lz:l&im. - [On the development of the connec-
tion between the four jfiiina and eight vijfiiina]. Indogaku Bukkyogaku Ken-
kya 32-1, pp. 178-179.
_ 1984. '''Chi' to 'shiki' - ryosha no ketsugo kankei to sono seiritsu katei -
- - [Jfiiina and vi-
jfiiina: On the development of their interconnection]. Buzan Gakuho :I': $
28/29, pp. 125-141.
_ 1987. "'Sanshin'to 'gohO' - ryosha no ketsugo kankei to sono seirltsu ka-
tei" < .=:!it> <3i7't > - .:.to)P)(;:lz:l&im. [The five ele-
ments and the three bodies: On the development of their interconnection]. hI
Takasaki Jikido hakushi kanreki kinen ronsha: Indogaku Bukkyogaku ronsha
. -1:...-- [Felicitation volume for
Professor Jikido Takasaki on the occasion of his sixtieth birthday], pp. 387-
411. Tokyo: Shunjiisha JW:tkt.
_ 1989. "Genjo ni okeru 'busshin' no atsukaikata" Q < {f,.:!it> 0)
tbiv'j] [Some aspects of Xuanzang's treatment of Buddha-bodies]. Bukkyo
Bunka {f,.W:X{[:; 25, pp. 94-108.
_ 1992a. "Genkan shOgon ron daihasshO 0 meguru hIdo shochiishakuka no
bunrui - sanshinsetsu to shishinsetsu -" m A1if 00 <. Q
-1 :...-- - .=:!itIDI. 1I9:!itIDI. - [The classification of the
"Dharmakaya Chapter" of the Abhisamayiilal'{lkiira by Indian commentators:
The threefold and fourfold buddhakiiya theories]. Shitennoji Kokusai Bukkyo
Daigaku Bungakubu Kiyo 24, pp. (1)-
(30).
_ 1992b. "Genkan shOgon ron 0 meguru sanshinsetsu guriipu ni yoru daiis-
,;shO daijiishichige kaihen no keii" 00 <. Q .:=.:!itIDI. -7. lv--:/,
I::. J:: Q [The detailed process of the modifica-
tion of the 17th verse in the fIrst chapter of the Abhisamayiilarrzkiira by the
"threefold buddhakiiya theory" group]. hI Mano Ryakai hakushi shOju kinen
ronsha: Hannyaharamitta shiso ronsha
[Felicitation volume for Professor Ryiikai Mano on the
occasion of his seventieth birthday], pp. 183-194. Tokyo: SankibO Busshorin

_. 1994. "The Classillcation of the Dharmakaya Chapter of the Abhisama-
yiilal'J'lkiira by hIdian Commentators: The Threefold and the Fourfold Bud-
dhakiiya Theories." Journal of Indian Philosophy 22, pp. 259-297.
_ 1996. Sansk!"t Word-Index to AbhidharmasamuccayabhiifYam edited by
N. Tatia with the Corrigenda. Tokyo: The Sankibo Press.
SIMILARlTIES BETWEEN STHIRAMATI AND XUANZANG 381
2002. "ChUgoku, Nihon Hosso kyogaku ni okeru shiki to chi no ketsugo
kankei - ruin sareta dairokushiki -> joshosachi, gogenshiki -> myokanza-
tchinoseit6sei"q:rOO
[The connection be-
tween the jiiiina and vijiiilna in the Fa-hsiang school in China and Japan: On
the rightness of the sealed-up connections - sixth vijiiilna -> krtyilnu:frhiina-
jiiiina and first five vijiiilna --> pratyavek:fiijiiiina]. In Kimura Kiyotaka haku-
shi kanreki kin en ronshii: Higashi Ajia Bukkyo to sona shahen *ttr'A'#t.w:
. :$: 7 :; 7 {L -t 0) )l!iJ ill [Felicitation volume for
Professor Kiyotaka Kimura on the occasion of his sixtieth birthday], pp. 65-
86. Tokyo: Shunjiisha *t!ct.
_ 2007a. "In Search of the Origins of the Five-Gatra System." Journal of
Indian and Buddhist Studies 54-3, pp. 1112-1120.
_ 2007b. "Yugashijiron ni mirareru jobutsu no kanosei no nai shujo" Wllliiitm
[On sentient beings with no
possibility of attaining Buddhahood mentioned in the Yogilciirabhami].
Tetsugaku Shiso Ronsha . (Institute of Philosophy, Univer-
sity of Tsukuba) 32, pp. 130-156.
_ 2007c. "Gosho kakubetsu no genryii 0 tazunete"
[In search of the origins of the five-gotra system]. In Kato Seiichi hakushi
koki kinen ronbunshii: Shingon mikkyo to Nihon bunka
S [Shingon Esoteric Buddhism and Japanese
culture: In honour of Dr. Kato Seiichi on his seventieth birthday], vol. 2, pp.
265-305. Tokyo: Nonburusha /
SA Vbh: * SiitrillaY(lkilra-vrtti-bha:fya by Sthiramati. Peking ed., no. 5531; Derge
ed., no. 4034.
Schmithausen, Lambert. 1969. Der Nirviil}-a-Abschnitt in der Viniscayasa,!lgra-
der Yogiicilrabhiim*. Osterreichische Akademie der Wissenschaften,
Philos.-hist. Klasse, Sitzungsberichte, 264. Bd., 2. Abh. Wien: Oster-
reichische Akademie der Wissenschaften.
Seizo Bunten Kenkyiikai 1979. Seizo bunken ni yom Bukkyo
shiso kenkyii, daiichigo: Anne-zo "Daij6 shi5gonkyoron shakuso" - Bodai-
bon (l) .m 1 %
iE,i)U - (I) [Studies in Buddhist thought based on Tibetan texts, 1:
Sthiramati's &itrillaY(lkiira-vrtti-bha:fya, "Bodhi Chapter" (I)]. Tokyo:
SankibO Busshorin rJJfr)JHIIJ:fi1*.
Shizutani Masao 1979. Indo himei mokuraku -1 [A
catalogue of Indian inscriptions]. Kyoto: Heirakuji Shoten
382 RIDENORI SAKUMA
Tsukamoto KeishO 1996. Indo Bukkyo himei no kenkyii..{ P'{.b.Wc1i
[A study of Indian Buddhist inscriptions], vol. 1. Kyoto: Heirakuji
Shoten
Tsukamoto KeishO Matsunaga Ytikei and Isoda Hir.ofumi
1331!!l:>c, eds. 1990. Bongo butten no kenkyii III: Ronsho hen
J:e III [A descriptive catalogue of Sanskrit Buddhist literature, ill:
Treatises]. Kyoto: Heirakuji Shoten
TV: Sylvain Levi, Vijfiaptimiitratiisiddhi, deux traites de Vasubandhu, Vil'{lsatikii
accompagnee d'une explication en prose et Tri1'{likii avec Ie commentaire de
Sthiramati, Paris, 1925.
Yamaguchi Susumu ill 1966. Kanzo taishO Benchabenron, ttl Chabenfim-
betsuron shakuso bonpon sakuin Jlf.t
i9fE5iIS**51 [The Madhyantavibhiiga in Chinese and Tibetan, with an index
to the Sanskrit text of the Madhyantavibhiigapkii]' Tokyo: Suzuki Gakujutsu
Zaidan &fl*,**fJl1ffl.

Volume 29 Number 2 2006 (2008) pp. 383400
CONCEPTS, INTENSION, AND IDENTITY IN TIBETAN
PHILOSOPHY OF LANGUAGE
JONATHAN STOLTZ
Indian Buddhist philosophers task of accounting for the meaningful-
ness of language is perhaps most famously addressed in Digngas
doctrine of apoha, which claims that general terms do not possess
meaning by referring to positively existing universals, but instead
impart meaning through a negative process of exclusion. As Mark
Siderits points out, this Buddhist theory of language forced [Dign-
ga and his followers] to say that there is more to linguistic meaning
than just reference.
1
Siderits own contention is that, beyond refer-
ence, Buddhist philosophers were forced to recognize sense the
mode of presentation of the reference of an expression as a distinct
element of meaning.
2
There is something importantly correct about
Siderits remarks, but to the extent that Buddhist philosophical theo-
ries differ from school to school and time period to time period,
more needs to be said about Buddhist theories of word meaning.
In this article I will examine one highly localized set of develop-
ments to the Buddhist doctrine of word meaning that was made by
twelfth and thirteenth century Tibetan Buddhist epistemologists
primarily schooled at gSa phu Monastery in central Tibet. I will
show how these thinkers developed the notion of a concept (don
spyi) in order to explain how it is that words are capable of applying
to real objects, and how concepts can be used to capture elements of
word meaning extending beyond reference to real objects. In particu-
lar, I will focus on the developments made by Phywa pa Chos kyi
se ge in the middle of the twelfth century, as well as on reactions to
those developments by Sa skya Paita in the first half of the thir-

1
Siderits (1986), p. 343.
2
Ibid, p. 331. Italics mine.
JONATHAN STOLTZ 384
teenth century. My investigation will begin by explaining how Ti-
betan Buddhist epistemologists make use of concepts in order to ac-
count for how it is that words are capable of applying to particular
objects. I will then focus on one particular area where this appeal to
concepts plays a central role in Tibetan accounts of word meaning
namely, with regard to the relation of sameness or identity (gcig pa).
Finally, I will tie this discussion to the views of the Tibetan polymath
Sa skya paita, showing how he reacts against earlier Tibetan
theories of word reference.
1. bKa gdams epistemologists on intentional objects and inten-
sional objects
It is often asserted that Buddhist philosophers uphold the view that
certain objects or experiences are ineffable, unable to be communi-
cated in language. While many claims are made by Indian and Ti-
betan Buddhists about the limitations of language, one particular
claim that was important within the Buddhist epistemological tradi-
tion was the declaration by the traditions forefather, Dignga, that
verbalization can only occur in conceptual thought thought that
takes general entities, and not concrete particulars, as its object.
Given this foundational assertion, how then can Buddhist
epistemologists account for the referential nature of language? How
is it that words can refer to objects in the external world?
In order to see how bKa gdams Tibetan epistemologists were
able to make sense of the referential capacity of language, it is
important to locate these thinkers theories of language within their
general accounts of conceptual thought, for it is out of a more gen-
eral account of cognitive objects that Tibetan thinkers construct their
theories of language. Even though the Indian and Tibetan Buddhist
epistemological traditions accept Digngas root assertion that
conceptual thought takes general entities (smnyalakaa) which
are unreal as its objects, thinkers as early as Dharmakrti sought
also to explain how conceptual thought can be grounded in reality. A
major step is taken by the eighth century philosopher Dharmottara,
TIBETAN PHILOSOPHY OF LANGUAGE

385
who delineates two different aspects of the objects of cognition.
3

These items are refined by Tibetan thinkers at gSa phu Monastery
in central Tibet in such a way as to expose both the phenomenal na-
ture of thought and the intentional nature of thought. According to
rog Blo ldan es rab and Phywa pa Chos kyi se ge, two major 11
th

and 12
th
century thinkers from gSa phu Monastery, conceptual
thought bears on two distinct kinds of objects.
4
The object that di-
rectly appears to a persons mind in conceptual thought is a mentally
constructed i.e., non-objective phenomenal object (gzu yul or
sna yul). Phywa pa provides a name for these phenomenal objects
of conceptual thought. He calls them concepts (don spyi). The Ti-
betan term don spyi or at least its Sanskrit equivalent, arthasm-
nya is present in the works of Dignga, but this particular role as
the phenomenal object of conceptual thought appears to begin with
Phywa pa. While Phywa pa and his followers accept the existence of
concepts, conceptual thought isnt typically about concepts. Thought
typically bears on, or is directed toward, external reality.
5
The ob-
jects that conceptual thought is directed toward are called intentional
objects (en yul) where these intentional objects are not mental
constructions, but are, rather, (typically) objects or properties in the
external world.
This distinction between phenomenal objects and intentional ob-
jects is quite sensible. When I think about something like the Potala
Palace, what directly appears to my mind is a concept, or maybe
even an image, of it. Nevertheless, my thoughts are not about the
concept of the Potala, my thoughts are about the Potala Palace itself.
Similarly, in the Buddhist context, when a person in conceptual
thought infers the existence of a fire on a hill from having perceived
smoke, even though what directly appears to the persons mind is the
concept of a fiery hill, the inference is not about a concept. Rather,
the inference is about a real fire on a real hill. As the thirteenth cen-

3
Dharmottara, 179a4: es pa rnams kyi yul ni rnam pa gis te / gzu ba da en pao /.
4
See Phywa pa (2006b), 1b11a, and rog (2006), 8b9a.
5
Conceptual thought neednt necessarily be directed toward external reality. Thought
can be about mental episodes (es pa) or about concepts (don spyi).
JONATHAN STOLTZ 386
tury Tibetan scholar Sa skya Paita (henceforth, Sa pa) sarcasti-
cally notes,
6

When someone says, Pour water into pot!, I [automatically] understand the
object of application [i.e., the intentional object] to be a real particular, and
dont ask, [Should I] pour the water into a real pot or [should I] pour the
water into the concept of a pot?
This distinction between the phenomenal objects and intentional ob-
jects of conceptual thought is projected by early Tibetan epistemolo-
gists onto their theory of language. Phywa pa and his followers agree
that when a person utters a word like snake, what the word directly
signifies (dos kyi brjod bya) is the concept of a snake. Yet, these Ti-
betan thinkers also hold that, insofar as language is intentional and
attempts to describe reality, the word snake has as its intentional
signification (en pai brjod bya) a real, objectively existing snake.
7

It is a phenomenal object a concept that is directly signified by
words, but the words are not about the phenomenal object. Instead,
words are directed toward, or denote, the real object that is the
words intentional signification.
8

We can also, by drawing a comparison to Western philosophical
theories of language, identify an important link between directly and
intentionally signified entities. Georges Dreyfus understands the
distinction to be comparable to Gottlob Freges famous distinction
between sense and reference.
9
The concepts that are directly

6
Sa skya paita, p. 122: bum par chu blugs ig ces brjod pa naa ra mtshan gyi
bum par chu ldugs sam don spyii bum par ldugs es mi dri bar jug yul don ra mtshan
id go ba yin te.
Admittedly, Sa skya paita puts forward this quote within the context of how language
can be efficacious, and not exactly so as to distinguish phenomenal objects from inten-
tional objects. As such, he speaks of the object of application (jug yul), which, in concep-
tual thought is equivalent to the intentional object (en yul).
7
In Phywa pas Tshad ma yid kyi mun sel and Tshad ma rnam es kyi grel pa there is
no full discussion of this distinction, though the distinction is clearly presupposed when he
uses the terms. My explanation comes primarily from Sa skya paitas much fuller
discussion of the terms.
8
Of course, all this is consistent with there being a failure of reference. Sometimes I
can use a term, intending to signify a real object, and yet fail to do so.
9
Dreyfus (1997), pp. 26174. Cf. Frege (1980).

387
signified by words are taken to be analogous to Freges senses,
whereas the real particulars that are intentionally signified are said to
be analogous to Freges referents. Now, for Frege, it isnt simply that
word meaning consists of two freestanding elements. Rather, he
contends that it is a words sense that determines its referent. Each
sense determines a unique referent. As such, identifying the referent
of an expression is typically accomplished by grasping the expres-
sions sense. Much the same idea appears to hold true in the Bud-
dhist context as well. Words are capable of intentionally signifying
real particulars by virtue of the relation that those particulars bear to
the concepts that are directly signified by the words.
10

By thus distinguishing the intentional signification of words from
the words direct signification, these Tibetan thinkers believe that
they have successfully accounted for how language can apply to real-
ity i.e., how particulars can be effable while still holding true to
Digngas root assertion that verbalization can only occur in mental
states that take universals (smnyalakaa) as their objects. In
addition to particular objects in the external world, there are also
mentally constructed concepts. It is these concepts that help link lan-
guage to objectively existing particulars. Having established this, we
can now examine a more specific role that concepts play in Tibetan
Buddhist theories of language, in the development of a philosophical
account of identity.
2. Identity and intensionality
To the extent that both Tibetan Buddhist thinkers and Gottlob Frege
held the belief that a fully adequate theory of language requires an
appeal to some sort of intensional element either a concept (Tib:
don spyi) or sense (Ger: Sinn) it is not surprising that we can find
similarities in these thinkers views on identity. In fact, the stated
goal of Freges essay On Sense and Reference is precisely to ex-
plain just what we mean when we assert identity statements. His pa-
per begins with the question of whether identity statements assert a

10
See Siderits (1987) pp. 3423 and Tillemans (1986) p. 215.
JONATHAN STOLTZ 388
relation between objects, or between names or signs of objects.
11

Similarly, in the Tibetan Buddhist context much ink has been spilled
in trying to explain the complex nature of the identity relation, and
just like in Freges case, Tibetans attempt to make sense of identity
by appealing to something like a terms sense or intension. One nota-
ble contrast, however, involves the relata of the identity statements
themselves. Freges central concern is with proper names, whereas
Buddhist thinkers are focused principally on identity statements con-
structed with general terms. The central question that I will be
addressing in what follows is that of what conditions twelfth and
thirteenth century Tibetan epistemologists, and Phywa pa Chos kyi
se ge in particular, held to be necessary and sufficient for the truth
of identity statements involving two general terms.
Tibetan philosophers describe various types of identity relations
identical essence (o bo gcig), identical substance (rdzas gcig),
identical extension (don gcig), and so forth. Many of these forms of
identity have been carefully discussed in articles by Tom Tillemans
(1983 and 1986), and will not be addressed here. My concern in this
section of the essay is with the relation of identity simpliciter, and
focuses on the simple question: when is an identity statement con-
structed with two general terms true? Operating under the presump-
tion that identity simpliciter (gcig) in the Tibetan Buddhist context is
an identity between qualities or dharma (chos gcig), the question can
be stated: when do two general terms express the same quality?
A natural reply to this question by Indian Grammarians would be
to claim that such identity statements are true if and only if the two
general terms pick out the same universal. This option is not open to
Buddhist philosophers, however, as they generally reject the exis-
tence of universals. Given this rejection, one might be tempted to
conclude that what a general term expresses is the class of particular
objects formed through a process of exclusion excluding all those
objects to which the general term does not apply. This is, after all,
the heart of Digngas apoha project. Were this interpretation cor-
rect, it would follow that statements of identity are true whenever the

11
Frege (1980), p. 56.

389
two general terms are coextensive. Indian and Tibetan Buddhist
epistemologists overwhelmingly maintain, however, that two differ-
ent qualities can be coextensive, thus showing that a terms exten-
sion alone is insufficient for determining a qualitys identity condi-
tions. That two different qualities can be coextensive is seen from
the fact that there are non-trivial inferences involving coextensive
qualities. When, for example, a person infers that sound is imperma-
nent because it is a product, the two qualities, impermanence and be-
ing produced, are extensionally equivalent. Yet, the fact that this
inference is non-trivial implies that these two terms possess distinct
meanings. Needless to say, it is apparent that the meaning of a gen-
eral term must indeed consist of more than the extension of that
term.
It is at this point where the developments made by the twelfth
century Tibetan epistemologist Phywa pa Chos kyi se ge help to
provide a novel explanation of what this element of meaning is over
and above the terms extension. As shown above, in addition to the
particular objects falling within the exclusion classes of general
terms, twelfth century Tibetan thinkers claim that for each term there
is also a corresponding, mentally constructed concept (don spyi), and
it is these concepts that provide us with something like intensional
elements of meaning. Identity and difference of these intensional ele-
ments are captured with the relations of identical connotation (ldog
pa gcig) and different connotation (ldog pa tha dad).
12
Two quali-
ties satisfy the identical connotation relation just in case the con-
cepts that arise in ones mind when conceiving of the two qualities
are identical.
13
Expressed linguistically, this is to say that the concept

12
The translation of ldog pa as connotation is, quite obviously, far from literal.
Nevertheless, it is meant as a good approximation of the role played by the term in this
particular philosophical context. Translating ldog pa as distinguisher or isolate, as
some have done, does produce a somewhat more literal translation, but does so by using
words in English that have no standard philosophical meanings.
13
This equivalence is never directly stated by Phywa pa himself. There are several
places where he expresses half of this equivalence, by stating that ldog pa tha dad entails
don spyi so sor sna pa. For example, in Tshad ma yid kyi mun sel 6b3: ldog pa tha dad
pas don spyi so sor sna par gyur te. (This line appears within a hypothetical opponents
argument, but Phywa pa never questions the veracity of the claim.) In dBu ma ar gsum

JONATHAN STOLTZ 390
that dawns in a persons mind when she cognitively entertains (and
grasps the meaning of) the one term is the same as the concept that
dawns when she grasps the other term.
14
When this condition is not
met, the two qualities stand in the different connotation relation.
The further question that needs to be addressed is how appeals to
concepts and the identical connotation relation can help solve the
problem of when two general terms pick out the same quality. A
natural thought at this point is that satisfaction of the identical
connotation relation might be sufficient for identity simpliciter. This
would mean that if two terms give rise to the very same concept
that is, if they are intensionally/connotatively identical then the two
terms express the same quality. My own belief is that this is, in fact,
the most philosophically tenable position to hold. While this view
fits well with the commonly held belief that connotation determines
denotation, it is not entirely obvious whether Tibetan philosophers
actually accept this view, or whether they have an even more strin-
gent requirement for identity simpliciter.
3. Ultra-intensionality
Tom Tillemans (1986) claims that Tibetan philosophers and here,
he is referring principally to dGe lugs thinkers uphold a thesis of
ultra-intensionality for terms each term has a different intension
from every other distinct term. Tillemans derives this thesis by
analyzing the conditions under which Tibetans assert the identity

gyi sto thun 2a1: rdzas gcig id la ltos sa log sa tha dad las log pa la ltos pas don spyi so
sor ar te sgro dogs tha dad bsal nas es par bya ba bum pai byas pa da mi rtag pa ltar
o bo gcig la ldog pas byas pai tha dad paam / i da ljon pa ltar chos gcig id la rnam
gras pai sgras tha dad lta bur brjod pa yin. In this longer quote, it is not only clear that
ldog pa tha dad entails don spyi so sor sna ba which means, when stated in the
contrapositive, that sameness of concepts (don spyi gcig) entails sameness of connotation
(ldog pa gcig), it is also reasonable to think that sameness of connotation entails sameness
of concepts.
14
In some sense, these two phrasings of the same connotation relation are not iden-
tical. The difference here arises in connection with the fact that there are terms those like
son of a barren woman and horn of a rabbit that give rise to concepts but do not ex-
press qualities (dharma) at all.

391
relation. Tillemans claim is essentially that, for any two qualities
and ,
(a) is identical to ( da gcig yin)
if and only if
(b) is identical to .
Simply put, two terms express the same quality if and only if the two
terms are themselves identical. Thus, on this account it is true that
(e
1
) Cordate is identical to cordate
but false that
(e
2
) Cordate is identical to renate.
Despite the fact (assuming, rather obnoxiously, that it is a fact) that
the general terms cordate (creature with a heart) and renate
(creature with a kidney) have the same extensions, the fact that these
two terms are not identical gives us grounds for concluding that
Cordate is identical to renate is false.
Expounding on this a bit more, it is important to note that there
are two distinct readings possible for sentence form (b). On the first
reading, we can take the expressions in quotes and to be
instances of what Frege calls indirect reference, where the indirect
referent of a term is taken to be the ordinary sense of that term.
Replacing Freges term sense with Carnaps term intension, on
this first reading of (b) we get,
(b
1
) the intension of is identical to the intension of
On the second reading of (b), what this condition requires is that the
general terms and be two tokens of the very same sign. That
is, (b) is claimed not to be stating an identity of the intensions of
and , rather, it is stating the identity of the signs and them-
selves. On this reading we get,
(b
2
) The sign is identical to the sign .
There is no question that these two readings (b
1
) and (b
2
) are
not logically equivalent. Condition (b
2
) entails (b
1
) but not vice
versa, since two distinct terms can possess the same intension, as for
example with the words horse and steed. We may suppose along
JONATHAN STOLTZ 392
with Frege that the words horse and steed have not only the
same extension but also the same intension. There may be aesthetic
reasons for preferring the word steed over horse in certain con-
texts, but in point of meaning the two terms are identical. Neverthe-
less, the two signs horse and steed are obviously distinct.
Since the two conditions, (b
1
) and (b
2
), are logically distinct, the
question of which of the two conditions is actually the right one for
Tibetans to adopt for identity simpliciter, (a), is a pressing concern.
Should we hold, as I suggested earlier, that connotative sameness is
sufficient for sameness simpliciter, or should we say that two terms
pick out the same quality only if the two terms are themselves identi-
cal?
Tillemans contention is essentially that Tibetan philosophers do
not distinguish these two different interpretations of condition (b).
He believes that Tibetan thinkers uphold the view that (b
1
) is, in fact,
equivalent to (b
2
), for he takes them to adopt the view that two dis-
tinct terms cannot possess the same intension. This is precisely the
thesis of ultra-intensionality: each term possesses a different inten-
sion from every other term. As Tillemans puts it,
For a Fregean bachelor and unmarried man would express the same
sense (intension). But for a Buddhist, they would have different sense be-
cause the names are different. Hence, I will speak of the Buddhist entities as
being ultra-intensional.
15

Indeed, there is a large quantity of textual support corroborating
Tillemans belief that Tibetan philosophers of the dGe lugs sect view
language as ultra-intensional. Tillemans gives an extensive quote
from the Collected Topics textbook of the nineteenth century writer
Phur lcog pa,
16
in which the author not only sets condition (b
2
) as a

15
Tillemans (1986), p. 213.
16
In the Tibetan exile community this author Yos dzin phur lcog pa blo bza tshul
khrims byams pa rgya mtsho is commonly called Phur bu lcog. This rendering of his
name is derided as a linguistic corruption by scholars in Tibet itself where he is usually
called Phur lcog pa or, sometimes, Phur bu lcog pa. In any case, this epithet is given
to the author due to his having supposedly lived, from ages 5 to 10, in the Phur bu lcog
mountain range just north of Se ra Monastery.

393
necessary condition for identity viz., because if [two items] are
identical, then they must be identical in both name and referent
17

but also contrasts this with cases in which two qualities (dharma,
chos) are taken to be distinct precisely because a person can validly
ascertain the one without ascertaining the other. But this ability to
ascertain one quality without ascertaining a second quality that is
coextensive with it is generally rooted in those two qualities having
distinct concepts/connotations associated with them. In thus drawing
the contrast between two elements being identical in both name
and referent and having different connotations at least indirectly
suggests that Phur lcog pa took conditions (b
1
) and (b
2
) to be equiva-
lent.
This view was certainly not universally accepted by twelfth and
thirteenth century Tibetan philosophers, however. There are several
cases of twelfth century Tibetan philosophers unmistakably dis-
tinguishing condition (b
1
) from (b
2
). In two different texts
18
Phywa
pa Chos kyi Se ge briefly describes a number of different forms of
identity and difference. In each case, he claims that it is possible to
have two distinct terms that pick out the very same quality/dharma
(chos gcig id la rnam gras pai sgras tha dad). Moreover, it is
apparent from the context in which these claims are made that he
takes these two terms to have exactly the same intensions.
19
That is,
the very same concept dawns in ones mind when thinking about
each of the two terms. Phywa pas example of two distinct words
having the same intension are i and ljon pa both of which
carry the same meaning as the English word tree. One of Phywa
pas successors, the author of The Summary of the Quintessence of
Epistemology (Tshad mai de kho na id bsdus pa), makes a similar
claim.
20
He states that there are cases in which two distinct terms

17
Yongs dzin (2000), p. 31.
18
For the quote from the dBu ma ar gsum gyi sto thun see footnote 13. Tshad ma yid
kyi mun sel 11b8: i da ljon pa ltar ldog pa cig id la rnam gras kyi sgras brjod pa.
19
These cases are contrasted with cases in which the terms have distinct connotations
(ldog pa tha dad) and distinct concepts arising.
20
Klong chen rab byams, p. 51: ldog pa da dos po gis ka tha dad med pa zan da
bos lta bu. This text is credited to the fourteenth century scholar Long chen pa. However,

JONATHAN STOLTZ 394
have the same connotation (ldog pa) and refer to the same material
thing(s) (dos po). As an example, he offers the terms zan and
bos, both of which carry the meaning of the English word food.
These illustrations of a theory of word meaning in which distinct
linguistic signs possess the same connotation or intension go a long
way toward helping us to understand some key issues in early Ti-
betan philosophy of language. First, we can see that, pace Tillemans,
not all Tibetan philosophers accept the thesis of ultra-intensionality.
It would have been a remarkable historical fact about Tibetan
philosophical thought had they never recognized the possibility of
two distinct terms possessing the same meaning. Nevertheless,
Phywa pas recognition of this possibility appears to have largely
fallen on deaf ears with his Tibetan successors. While a close
examination of the philosophical writings of other twelfth and thir-
teenth century Tibetan thinkers may yet yield many more examples
of a rejection of ultra-intensional word meaning, Tillemans is quite
right that in the more recent philosophical texts of the Tibetan
epistemological tradition many of the most famous authors do adopt
this thesis of ultra-intensionality.
21
Another moral we can draw from
this discussion is the fact that, in Phywa pas opinion, the actual
requirement that must be met in order for two terms to pick out the
same quality (simpliciter) is that the two terms be intensionally
identical. The two terms need not themselves be identical. In this

as van der Kuijp (2003) has documented, this attribution is highly dubious. Many
contextual clues suggest that this work was composed shortly after Phywa pas famous
Tshad ma yid kyi mun sel.
21
One contemporary writer, however, is very explicit in his rejection of the ultra-
intensionality thesis. In a book (derived from his MA thesis) written by a young Tibetan
scholar named Bum skyabs a very provocative account of concepts (don spyi) is put for-
ward. The author straightforwardly proclaims that multiple terms can give rise to the same
concept. His example is of the five terms gtan tshigs, rtags, rgyu mtshan, phyogs
kyi chos, and de sgrub kyi bsgrub pai chos, all of which carry the meaning of the Eng-
lish terms evidence or reason. Explicitly, he claims, (Bum skyabs, p. 43): don spyi
gcig la mi du ma jug pa yod de / rtags rigs kyi skabs su gtan tshigs da / rtags / rgyu
mtshan / phyogs kyi chos / de sgrub kyi bsgrub pai chos rnams la don spyi gcig las med
kya mi la yod pa lta buo //.

395
way, Phywa pa adopts the view that (b
1
) is the correct requirement
for identity simpliciter, not (b
2
).
Finally, this discussion provides us with clear examples of early
Tibetan philosophers making use of concepts in order to capture ele-
ments of word meaning extending beyond extensionality. Not only
can two different terms express the same quality by virtue of those
two terms giving rise to the same concept two different qualities
can be distinguished from each other by way of these associated con-
cepts. Thus, even when two terms are coextensive such that any
object instantiating the one term also instantiates the other term, and
vice versa the qualities they express can be distinguished in a per-
sons mind through the different concepts to which they give rise.
4. Sa skya Paita on concepts and reference
Stepping back from the specific case of identity and returning to the
central theme, what all of this serves to show is that by employing
the notion of a concept (don spyi) bKa gdams epistemologists were
able to explicate certain elements of cognition and language use in a
clear and interesting manner. In what follows I will raise one diffi-
culty with this bKa gdams Tibetan account as it plays out in the
writing of the thirteenth century polymath Sa pa. The claim by
bKa gdams Tibetans that word meaning consists of both direct
signification and intentional signification makes it possible to ex-
plain how Buddhist philosophers can connect language to external
reality without falling into the Indian Grammarians trap of granting
that words directly express particulars or universals. Sa pa cannot
accept the details of the bKa gdams account, however, for he rejects
the existence of concepts (don spyi).
22
Insofar as this is the case, Sa
pa cannot accept that concepts are the entities directly signified by
language. In fact, Sa pa explicitly denies that there is anything at all
that words directly signify. He states, Ultimately, what is called
signified (brjod bya) is not established at all.
23
This would not be a

22
For more on this, see my article, Stoltz (2006).
23
Sa skya paita, p. 121: don dam par na brjod bya es bya ba grub pa cia med de /.
JONATHAN STOLTZ 396
big deal, except that, as was pointed out above, words are capable of
successfully applying to real things only indirectly. On the bKa
gdams account, language directly signifies concepts, and these con-
cepts in turn make it possible for language to apply indirectly to real
particulars. Once one eliminates concepts from this theory of lan-
guage, as Sa pa desires to do, it is difficult to see how words could
still be capable of intentionally signifying real things.
What we have here is an apparent conflict between Sa pas pre-
ferred accounts in the philosophy of mind and philosophy of lan-
guage. When presenting his theory of mind and mental content at the
beginning of his Tshad ma rigs gter (Epistemology: Treasure of
Reasoning), Sa pa goes through pains to argue that concepts (don
spyi) do not exist. Yet, when he presents his own positive theory of
language later in the same text, Sa pa wishes to make use of the
bKa gdams distinction between direct signification and intentional
signification, even though the former requires an appeal to a cate-
gory of entities whose existence he rejects. The challenge that Sa pa
faces is that of explaining how language is capable of intentionally
signifying particular objects given that concepts do not exist that
is, given that there is nothing at all that language directly signifies.
Sa pas preferred method for dealing with this challenge is to
look back to the Indian Buddhist tradition. Whereas Phywa pa Chos
kyi se ges mode of operation is to account for the meaningfulness
of language by inventing a new role for concepts (don spyi), Sa pa
rejects this bKa gdams pa creativity and instead seeks to provide an
explanation of word meaning that is modeled after that given by the
Indian Buddhist masters. That said, the actual explanation given by
Sa pa blends elements of the traditional Indian view together with
elements of the bKa gdams pa account. On the one hand, he repeat-
edly acknowledges that there are, in fact, no objects directly signi-
fied by language (since the entities directly signified by language
would be concepts, whose existence he rejects). Yet, at the same
time, when he attempts to explain how it is that language can be di-

397
rected toward real things and how words can be used efficaciously,
he repeatedly appeals to concepts. Sa pa states,
24

Insofar as conceptual thought is essentially mistaken,
An image, which is the object [indicated by] words (sgra yi don), dawns.
Through that one longs for an external object,
And that is posited as what the words signify.
The view presented here is quite similar to his predecessors. Two
different sorts of entities are at work. There is both a mentally con-
structed image (or reflection) dawning in ones mind and a real
external object. As becomes clearer in subsequent passages, how-
ever, the image dawning in ones mind is none other than a concept
(don spyi).
25
For Sa pa, this concept is required in order for lan-
guage to be applicable to reality, for he contends that it is only by
virtue of conceptual thought mistakenly conflating concepts with
real objects that language can be directed toward real particulars.
This exposes even more clearly the conundrum that Sa pa faces. He
wants to hold that concepts do not exist, and yet at the same time he
wants to hold that language is capable of applying to reality only be-
cause of a mistaken conflation of concepts with external objects. But
how can concepts be conflated with real particulars if concepts dont
even exist? It seems that Sa pa cant have it both ways. I will con-
clude by offering one possible explanation for how Sa pas theory
of language can be rendered sensible, and how this explanation
shows quite clearly that Sa pa rejects the notion that external reality
is ineffable.
The explanation that I think Sa pa would offer in reply to the
above problem is one involving a nuanced understanding of the
ontological status of concepts. Ive mentioned above that Sa pa
spends the first several pages of his Tshad ma rigs gter arguing
against the existence of concepts. This is true, but it also glosses over
important details about the very notion of existence. For Sa pa, to

24
Ibid, p. 1212: rtog pa o bos khrul pa la // sgra yi don gyi gzugs bran ar // de la
phyi rol don du en // de ni sgra yi brjod byar btags //.
25
Ibid, p. 126: jig rten pa rnams don spyi la ra mtshan du khrul nas ra mtshan id
bla ba da dor ba la jug pas na tha sad rgyun ga la chad /
JONATHAN STOLTZ 398
say that something is existent (yod pa) is equivalent to saying that it
is an object of epistemic evaluation (gal bya). Thus, what he is
specifically denying is that concepts could be objects of epistemic
judgment. This does not preclude the possibility that concepts exist
in a more minimal way. Concepts might exist as mere fleeting im-
ages in ones mind; images incapable of being evaluated in judg-
ment. By granting such a minimal status to concepts it would then be
possible for these entities to be mistaken for real objects in the exter-
nal world, thus making Sa pas theory of language comprehensible.
Ultimately, what this amounts to is an account of the efficacy of lan-
guage that on the one hand calls upon concepts in order to explain
how the mind works so as to provide semantic content to language,
but on the other hand does not postulate the genuine existence of
concepts.
Why is it, though, that Sa pa argues against the existence of con-
cepts? It is in answering this question that our investigation comes
full circle. Sa pa offers two arguments against the existence of con-
cepts, with his second argument much more detailed and sophisti-
cated than the first. This second argument focuses in on the absurd
consequences that he believes would follow were one to accept the
existence of concepts. At the heart of his argument is the contention
that inasmuch as concepts are mind dependent entities, they are,
therefore, private objects, incapable of being communicated to oth-
ers. Sa pa states, If they are dependent solely on an individuals
mind, although one could, through introspection, express [concepts]
to another person, they would not understand.
26
Language could not
be used to communicate the existence of these private concepts to
others. In short, Sa pas argument is essentially that concepts do not
exist because if they did theyd be ineffable. What this means is that,
on Sa pas account, far from it being the case that language is
incapable of referring to real particulars, just the opposite is the case.
Words can be used to reference real things, but cannot be used to
reference mentally constructed items such as concepts.

26
Ibid, p. 42: ra ra gi blo kho na da brel na ra rig pa bin du gan la bstan kya
go bar mi gyur ro //

399

Bibliography
Bum skyabs. 2000. Bod kyi gtan tshigs rig pai spyi don go bder brjod pa.
Chengdu: Sichuan Peoples Press.
Dharmottara. Tshad ma rnam par es pai grel bad. D: 4229.
Dreyfus, Georges. 1997. Recognizing Reality: Dharmakrti's Philosophy and its
Tibetan Interpretations. Albany: SUNY Press.
Frege, Gottlob. 1980. On Sense and Reference, in P. Geach and M. Black
(eds.), Translations from the Philosophical Writings of Gottlob Frege.
Oxford: Blackwell.
Klo chen rab byams pa (attributed). 2000. Tshad mai de kho na id bsdus pa.
Chengdu: Sichuan Peoples Press.
van der Kuijp, L. 2003. A Treatise on Buddhist Epistemology and Logic
Attributed to Klong chen rab byams pa (13081364) and its Place in Indo-
Tibetan Intellectual History. Journal of Indian Philosophy 31: 381437.
rog Blo ldan es rab. 2006. Tshad ma rnam es kyi grel ba. In bKa' gdams
gsu 'bum phyogs sgrig, vol. 1. Chengdu: Sichuan Peoples Press.
Phywa pa Chos kyi se ge. 2006a. dBu ma ar gsum gyi sto thun. In bKa'
gdams gsu 'bum phyogs sgrig, vol. 7. Chengdu: Sichuan Peoples Press.
Phywa pa Chos kyi se ge. 2006b. Tshad ma rnam es kyi grel pa. In bKa'
gdams gsu 'bum phyogs sgrig, vol. 8. Chengdu: Sichuan Peoples Press.
Phywa pa Chos kyi se ge. 2006c. Tshad ma yid kyi mun sel. In bKa' gdams
gsu 'bum phyogs sgrig, vol. 8. Chengdu: Sichuan Peoples Press.
Sa skya paita Kun dga rgyal mtshan. 1989. Tshad ma rigs pai gter gyi rtsa
ba da grel pa. Qinghai: Tibet Peoples Press.
Siderits, Mark. 1987. The Sense-Reference Distinction in Indian Philosophy of
Language. Journal of Chinese Philosophy 14, 331355.
Stoltz, Jonathan. 2006. Sakya Pandita and the Status of Concepts, Philosophy
East and West 56, 567-82
Tillemans, Tom. 1983. The Neither One nor Many Argument for nyat and
its Tibetan Interpretations. In Contributions on Tibetan and Buddhist Relig-
ion and Philosophy, eds. E. Steinkellner et H. Tauscher, 305320. Vienna:
Arbeitskreis fr Tibetische und Buddhistische Studien Universitt Wien.
Tillemans, Tom. 1986. Identity and Referential Opacity in Tibetan Buddhist
apoha Theory. In Buddhist Logic and Epistemology, Studies in the Buddhist
JONATHAN STOLTZ 400
Analysis of Inference and Language, eds. B.K. Matilal and R.D. Evans, 207
227. Dordrecht: D. Reidel.
Yos dzin phur lcog pa blo bza tshul khrims byams pa rgya mtsho. 2000. Rigs
lam phrul lde. Gansu Peoples Press.

REPORT ON THE XyTH CONGRESS OF
THE INTERNATIONAL ASSOCIATION OF BUDDHIST STUDIES.
TOM J.F. TILLEMANS
General Secretary IABS
The Xyth Congress of the International Association of Buddhist
Studies (lABS) was held at Emory University, in Atlanta, Georgia,
from the 23
rd
until the 28
th
of June, 2008. Participation in IABS con-
gresses has been steadily increasing over the years, and this was no
exception to that trend: four hundred twenty-five people attended
over the six day period. The congress was very capably prepared by
a planning committee chaired by Professor Sara McClintock, and
benefitted from the support of staff and students of the Department
of Religion of the Faculty of Arts who helped with the considerable
tasks involved. The organization was absolutely exemplary.
The academic program began with a convocation ceremony at
which the President of the IABS, Professor Oskar von Hintiber,
delivered his address intitled "Hoary past and hazy memory. On the
history of early Buddhist texts." Over the next days there were
morning and afternoon sessions of paper presentations, approxi-
mately two hundred fifty presentations in total. These were
complemented by several evening events - notably a program of
Tibetan sacred music and dance performed by the monks of Drepung
monastery and the construction of a sand mm:u;lala - as well as recep-
tions, excursions, a general meeting and a memorable farewell
dinner.
Sections and panels
As in other recent congresses of the lABS, paper presentations were
assigned to either 'sections' on the various broad themes treated in
Buddhist Studies or to 'panels' centered upon specific topics. Panels
were organized by one or more conveners, who were responsible for
Volume 29 Number 22006 (2008) pp. 401405
402 TOM J.P. TILLEMANS
determining content and structure. Sections were organized by the
conference planning committee, who grouped together papers that
were submitted individually. Here is a list of the panels and their
convenors:
1. Tibetan Scholasticism in the 11th and 12th Centuries. Pascale Hugon,Ka-
zuo Kano, Kevin Vose.
2. Chinese Buddhist Meditation Practices and Chan. John McRae, Eric
Goodell.
3. Universal Claims, Postcolonial Frames: An Interdisciplinary Session on
Bodhgaya. Tara Doyle, David Geary.
4. The Academic Discipline oj Buddhist Studies in North America. Charles
Prebish.
5. Indian Buddhist Metaethics. Martin Adam.
6. Rebirth oj Buddhist Scholasticisms in Twentieth-Century China. Eyal
Aviv, Jason Clower.
7. Analyzing and Advancing Buddhist Philosophy: The Two Truths. Jay Gar-
field, Jan Westerhoff.
8. Miracles and Superhuman Powers in Buddhism. David Fiordalis.
9. How Theraviida is Theraviida? Peter Skilling.
10. Mahiimudrii: Approaching the Great Seal. Roger Jackson, Lara Brait-
stein.
11. New Research in Avadiina Literature. Andy Rotman.
12. Buddhism in the Writings oj Ippolito Desideri, S.J. (1684-1733). Trent
Pomplun.
13.Theory and Practice oj Healing, Medicine and Longevity in Buddhism.
Robert Mayer, Geoffrey Samuel.
14. Mahiiyiina Sutras. Jonathan Silk.
15. Beyond Simple Constructions oj So-called Humanistic Buddhism (renjian
jojiao): Reinterpreting Modernity in 20''' Century Buddhism in China and
Taiwan. Esther-Maria Guggenmos.
16. Representations oj Brahmins and Brahmanism in Early Buddhist Litera-
ture. Brian Black.
17. Buddhist Monumental Sculpture and Architecture. Cameron David War-
ner.
REPORT ON THE XV
TH
CONGRESS OF THE lABS 403
18. Madhyamaka and Yogacara Models of Truth or Reality in Indo-Tibetan
Buddhism. Klaus-Dieter Mathes.
19. "Terms of Art" in Indian Esoteric Buddhism. David Gray, Christian
Wedemeyer.
20. Gandhiiran Manuscripts and Gandharan Buddhism. Richard Salomon.
21. Wonhyo (617-686) and his Comprehensive Buddhist Philosophical
Vision. Charles Muller.
22. Buddhist Funerary Cultures: Art, Text, Ritual, Performance. Justin
McDaniel, Pattatorn Chirapravati. .
23. Borobudur in International Perspective. Hudaya Kandahjaya.
24. Humor and Buddhism. Shayne Clarke.
25. Buddhist Theories of Self-Awareness (svasaTpvedana): Interpretations
and Critiques. Birgit Kellner.
26. Reexamining Tibetan AutolBiographical Writing. Sarah Jacoby, Andrew
Quintman.
27. Recovering Anew the Lotus SCUra's Originality as a Religio-Philosophi-
cal System. Joseph Logan.
28. Pioneer Translators, Missionaries, and their Transmitted Texts. Mariko
Namba Walter.
29. Yogacara Buddhism: Approaches. Tao Jiang.
30. Facets of Esoteric Buddhism and the Tantras in East Asia. Charles
Orzech.
31. Santideva and the Bodhisattvacaryiivatiira. Ani Kunga Chodron.
32. Buddhist Commentarial Traditions: Transdiscursivity and Textual
Production in Buddhist Intellectual History. Mark Dennis.
33. Uses and Abuses of the "Modem" in Tibet and Tibetan Buddhism.Leigh
Sangster, Robbie Barnett, Laura Harrington.
The sections were as follows:
1. Early Buddhism
2. Himalayan Buddhism
3. Textual and Philological Studies
4. Vajrayana Buddhism
5. Narrative Studies
6. Buddhism and the West
404 TOM J.F. TILLEMANS
7. Buddhist Art
8. Vinaya Studies
9. Contemporary Developments in Buddhism
10. Hermeneutics, Scholasticism, and Commentarial Techniques
11. Dunhuang Studies
12. Indian and Tibetan Buddhist Philosophy
13. Ethnographic Studies
14. South Asian Buddhism
15. Buddhist Theories of Mind and Meditation
16. East Asian Buddhism
17. Logic and Epistemology
18. Technology and Resources
19. Southeast Asian, Inner Asian, Korean and Mongolian Buddhism
20. East Asian Buddhist Philosophy
21. Mahayana Buddhism
The combination of sections and panels once again seems to have
successfully provided an organizational structure that represented the
diversity of the field and allowed for specialized discussions within
it.
lABS business
Here is a summary of the various resolutions passed at the lABS
Board Meeting and at the General Meeting of the members.
1. The IABS decided to accept the invitation of Dharma Drum University /
Chung Hwa Institute of Buddhist Studies in Taiwan to host the next
lABS congress (June 20-25, 2011, Iinshan, Taiwan).
2. The lABS Board decided to raise dues moderately, i.e., by 10 USD for
students, 15 USD for regular members and 20 USD for institutions.
This increase will take effect at the beginning of 2010.
3. The financial statement for 2007, as prepared by KPMG Accountants
(Lausanne), was approved and will appear in the next issue of the
IIABS.
REPORT ONTRE XyTH CONGRESS OF THE lABS 405
4. Ernst Steinkellner, Tilmann Vetter and Richard Gombrich were named
honorary members of the lABS in recognition of their important
contributions to the field of Buddhist Studies.
5. The General Meeting ratified the following amendment to the lABS
constitution:
"The dissolution of the lABS shall be treated in accordance with Article
XII. If the lABS should be dissolved, its financial assets will revert to a
recognized not-for-profit international association to be used for the
continuation of scientific activities that are, as far as possible,
compatible with the Objectives outlined in Article II of the lABS
Constitution. "
The amendment was presented to the lABS General Meeting on Friday,
June 27,2008 and received the members' unanimous approval with no
abstentions. It now figures as Article XIII in the lABS Constitution (see
http://www.iabsinfo.net/iabs. pdf).
It remains for me, in the name of the lABS, to thank again the
organizers of the XVth Congress for their excellent preparation. The
Congress represented the "state of the art" of Buddhist Studies and
bodes well for the future of this flourishing field.
IABS TREASURER FINANCIAL REPORT
mROME DUCOR, Treasurer IABS
Below are reproduced the financial statements for the IABS prepared
by KPMG Accountants, Lausanne, Switzerland, April 28, 2008.
These statements were approved by the Board at its meeting in
Atlanta, Georgia, June 23, 2008.
1:1 1:1 1:1
International Association of Buddhist Studies
Lausanne
i. Balance sheet as of December 31. 2007
2.' Profit and loss statement from January 1st, 2007 to December 31! 2007
April 28, 2008 FGJlh
Volume 29 Number 22006 (2008) pp. 407--409
408
.JEROME DUCOR
International Association of Buddhist Studies
Balance sheet as of December 31, 20.0.1
Current assets
Petty cash
BCV, account no L 0.983.51.02
BCV, account no C 0983.51.04
BCV, account no Z 5042.0.9.62
Withholding tax to recover
Total cilrrent assets
TOTAL ASSETS
LIABILITIES AND CAPITAL
Short-teRn liabilities
Accrued Dabilities
Capltal-FDrtune
Fortune as January 1 st
Result for
Fortune as December 31
TOTAL LIABIUTIES AND CAPITAL
31.12.2007
CHF
97.0.5
2'133.25
50.'529.10.
1'589.85
18.03
54'367.28
54'367,28
4'033.35
4'033.35
66'190.22
-15'856.29
50'333.93
54'367,28
31.12.2006
CHF
10.0..75
1'0.69.90.
67'731.55
3'30.0.85
16.87
72'219.92
72'219.92
6'0.29.70
6'029.70
80'839.07
-14'648.85
66'190..22
72'219.92
lABS TREASURER FINANCIAL REPORT
International Association of BUddhist Studies
Profit and loss statement
for the period from January 1st. 2007 to December 31,2007
Dues, sub., back issues and donations
Interests income
Exchange gain
Total income
Gross salaries
Social charges
Office supplies
Communication
Postage
Subscription
Accounting fees
Computer fees
Printing, books
Bank fees
Exchange loss
Total expenses
Result for the period
2007
CHF
35'060.44
3.39
59.90
35'123,73
-14'007.90
-1'004.10
-1'926.20
-240.00
-3'627.72
0.00
-1'464,70
..{i42.20
-23'110.60
-267.86
-4'688.74
50'980.02
15'856.29
409
2006
CHF
25'581.06
46.67
0.70
25'628,43
-14'131.50
-1'009.20
-31.85
-240.00
-3'180.61
-86.70
-2'130.50
0.00
-12'939.05
-568.12
-5'959.75
-40'277.28
-14'648.85
NOTES ON CONTRIBUTORS
Junjie CHU is an Indologist and Tibetologist, and teaches Indo-
Tibetan Buddhist philosophy, Sanskrit and classic Tibetan in the
Department of Indology and Central Asian Studies at the University
of Leipzig, Germany. His current research centers on the Yogacara
system, especially its epistemological theories and its relationship to
the Sautrantika system.
David HIGGINS is a doctoral candidate at the University of Lausanne,
Switzerland, in the Section de Langues et Civilisations Orientales
and a Sessional Lecturer in Religious Studies at the University of
Victoria, Canada. His current research concerns the phenomenologi-
cal distinction between dualistic mind (sems) and primordial aware-
ness (ye shes) as was articulated and defended in the Tibetan rNying
rna tradition from the 7th to 14th centuries. His dissertation will offer
a critical study of rNying rna views on mind and attempt to clarify
their complex relationship to earlier Madhyamaka and Y ogacara
views.
Richard D. McBRIDE, IT, is an assistant professor of history at Brig-
ham Young University-Hawaii. He earned a Ph.D. in East Asian
Languages and Cultures (Buddhist Studies) from UCLA in 2001. He
is the author of Domesticating the Dharma: Buddhist Cults and the
Hwaom Synthesis in Silla Korea (University of Hawaii Press, 2008)
and is currently completing a manuscript on the cult of Amitabha
and Pure Land Buddhist scholiasts in Silla Korea titled Aspiring to
Enlightenment. He was a Fulbright senior researcher in Korean his-
tory during the academic year 2007-2008.
Jonathan STOLTZ is an assistant professor of philosophy at the
University of St. Thomas. He specializes in contemporary philoso-
phy of logic as well as Tibetan Buddhist epistemology. His current
research examines epistemological developments made by thinkers
412
in the bKa' gdams and Sa skya schools of Tibetan Buddhism, as well
as philosophical debates between these two schools of thought. His
most recent articles include "Sakya Pandita and the Status of Con-
cepts" (2006, Philosophy East and West) and "Gettier and Factivity
in Indo-Tibetan Epistemology" (2007, The Philosophical Quarterly).
Hidenori SAKUMA is a professor at the Institute of Philosophy of the
Graduate School of Humanities and Social Sciences, University ,of
Tsukuba, Japan. He is the author of Die Asrayaparivrtti-Theorie in
der Yogaciirabhiimi, Teil I & II (1990), and Sanskrt Word-Index to
the Abhidharmasamuccayabha/iyam edited by N. Tatia with the
Corrigenda (1996). The main focus of his research is the history of
changes in Y ogadira theories as they spread from India to China and
Japan.
New fy01tl<L! if
j F
.. ' .......... : ...... r:It .. ....... 1 . .1. '.J .. a. i A . . . A .... / ..
;is
t' .',-'e';"- . ./'_,- -<,'. - i"', - - -;.- -. -"" -<.,
SOOALLY ENGAGED BUDDHISM
Sallie B. King
DIMENSIONS OF AsIAN SpooruALriY
ANCIENT CHINESE DIVINATION
Stephen 1.. Field
EMINENT NUNS: Women Chan Masters
of SeventeenthcCenturyChina
Beata Grant
MAKING TRANSCENDENTS
Ascetics and Social M"moryin Early
Medieval China
Robert Ford Campany
THE CHINESE CLASSIC
OF FAMILY REVERENCE
A Philosophical Translation of the Xiaojing
Henry Rosemont, Jr. & Roger TAmes
THE SHAOLIN MONASTERY
History, Religion, New ill paper
and the Chinese Martial Arts
Meir Shahar
CAMBODIAN BUDDHISM
History and Practice
Ian Harris
New in paper
BODHISATTVAS OF THE FOREST
AND THE FORMATION OF THE
MARA YANA: A Study and Translation
of the Rtitrapaiapariprcchii-sfttra
Daniel BOilcher
StuDIES CN THE BUDDHisT TAADmoNs
Published in association-with the Institute for the Study
of Buddhist Traditions, University of :Michigan
Incest and Schism in Indian Buddhist
Legend and Historiography
Jonathan A. Silk
DEATH AND THE A:FrERLIFE
INJAPANESE BUDDHISM
Edited by Jacqueline 1. Stone
& Mariko Namba Walter
THE FOUR GREAT TEMPLES
Buddhist Art, Archaeology, <'l!ld Icons
of Seventh-Century Japan
Donald F. McCallum
IMPERIAL POLmCS AND SYMBOLICS
IN ANCIENT JAPAN
The Tenmu Dynasty, 650-800
Hennan 00111S
WHEN TENGU TALK
Hirata Atsutane's Ethnography
of the Other World
Wilburn Hansen
THE RECORD OF LINJI
Translated by Ruth Fuller Sasaki
Edited by Thomas Yah!; Kirchner
NANZAN LrnRM.Y OF AsIAN RELIGION AND CULTURE
SOTO ZEN IN MEDIEVAL JAPAN
William M. Bodiford New in paper
5ruorES IN EAST AsiAN BUDDHISM, No.8
Published in association with the Kuroda Institute
WWW.UHPRESS.HAWAII.EDU ORDER TOLL FREE 1-888-UHPRESS
The International Association of Buddhist Studies
Prof. Oskar von Hiniiber (Freiburg)
President
Prof. Cristina Scherrer-Schaub (Paris)
Vice-Presiden t
Prof. Tom Tillemans (Lausanne)
General Secretary
Dr. Jerome Ducor (Lausanne)
Treasurer
Regional Representatives:
Prof. Janet Gyatso (Cambridge, MA.)
Prof. Kazunobu Matsuda (Kyoto)
Prof. David Seyfort Ruegg (London)
Mr. Peter Skilling (Bangkok)
Directors at large:
Dr. Nalini Balbir (Paris), Prof. Georges Dreyfus (Williamstown, MA.)
Prof. Robert M. Gimello (Notre Dame, IN), Prof. Paul Harrison (Los Angeles)
Dr. Helmut Krasser (Vienna), Dr. Li-ying Kuo (Paris)
Prof. Ulrich Pagel (London), Prof. Richard Salomon (Seattle)
Prof. Lambert Schmithausen (Hamburg), Mr. Peter Skilling (Bangkok)
Prof. Jikido Takasaki (Tokyo)
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 institutions
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-lOOI 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 Intemationale 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 Euro cheque 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 the previous year. Payments other than direct bank
transfers should be sent to Dr. J er6me Ducor, Dept of Oriental Languages and
Cultures, Anthropole, University of Lausanne, CH-1015 Lausanne, Switzerland.

Jiabs 29-2

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Jiabs 29-2

Uploaded by

Copyright:

Available Formats

JIABS

Journal of the International

You might also like