Sunteți pe pagina 1din 184



Gregory Schopen .
Indiana University
Bloomington, Indiana, USA
Peter N. Gregory
University of Illinois
Urbana-Champaign, Illinois, USA
Alexander W. Macdonald
Universite de Paris X
Nanterre, France
Steven Collins
Indiana University
Bloomington, Indiana, USA
Roger Jackson
Fairfield University
Fairfield, Connecticut, USA
Ernst Steinkellner
University of Vienna
Wien, Austria
Jikido Takasaki
University of Tokyo
Robert Thurman
Amherst College
Amherst, Massachusetts, USA
Number 1
This Journal is the organ of the International Association of BUddhist
Inc. I.t is .governed .b( the of th.e accepts
contnbutlons pertammg to BuddhIst StudIes m all the vanous discipline '. t;IY';i
as philosophy, history, religion, sociology,
psychology, studIes, etc. The JIABS IS publIshed tWIce yearlYinarN:
summer and wmter. ' ... e .
Manuscripts for publication (we must have two copies) and
concernin.g articles should be submitted to fIABS editorial office
address below .. refer to the gmd:lmes for contributors
JIABS prmted on the InsIde back cover of every Issue. Books for review shoiild::
be sent to address below. The Editors cannot guarantee to
revIews of unsolICIted books nor to return those books to the
The Association and the Editors assume no responsibility for the
expressed by the authors in the Association's Journal and other
Andre Bareau (France)
M.N. Deshpande (India)
R. Card (USA)
B.C. Cokhale (USA)
Gregory Schopen
clo Dept. of Religious Studies
230 Sycamore Hall
Indiana University
Bloomington, IN 47405
Jacques May
Hajime Nakamura
John C. Huntington (USA)
P.S. Jaini (USA) E. Zurcher
Both the Editors and Association would like to thank Indiana Univeii"
sity and Fairfield University for their financial support in the p r o d u c ~
tion of the J ournal.,",
The Editors wish to thank Mr. Kevin Atkins for his invaluable help:
in the preparation of this issue. '
Copyright The International Association of Buddhist Studies 1988
ISSN: 0193-600X
Indexed in Religion Index One: Periodicals, American Theological Li::
brary Association, Chicago, available online through BRS (Biblio:
graphic Retrieval Services), Latham, New York, and DIALOG Infor-<
mation Services, Palo Alto, California_
Composition by Publications Division, Grote Deutsch & Co., Madison, WI 53704 ..
Printing by Thomson-Shore, Inc., Dexter, MI 48130.
Four Levels of Pratztya-
Samutpada According to the
Fa-hua hsuan i, by Carl Bielefeldt 7
the Possibility of aN onexistent Object
of Consciousness: Sarvastivadin and
Theories, by Collett Cox 31
Upaya in the,
by Edward Hamlin 89
Sanskrit in the Kalacakra Tantra,
byJohnNewman 123
New Fragments of Buddhist Sanskrit Manuscripts
from Central Asia, by Richard Saloman and
Collett Cox 141
Reflections on R.S.Y. Chi's Buddhist Formal Logic,
by TomJ.F. Tillemans 155
Divine Revelation in Pali Buddhism, by Peter Mansfield
(Charles Hallisey) 173
Studies in the Buddhist Art of South Asia, ed. A. K. N arain
(Robert L. Brown) 175
Traditions of Meditation in Chinese Buddhism,
ed. Peter N. Gregory (Henrik H. Sorensen) 179
Fo.ur Levels of
to the F a-hua hsuan l
.1-",>')' "
t,the historical status of the Madhyamika school in China and
its famous doctrine of the middle way, is not easy
iograsp.l On the one hand, of course, all the major traditions
t&fEast Asian Buddhism claim Nagarjuna as a patriarch and
to embrace his madhyamaka teaching; on the other hand,
for a few scholastics of medieval China and early Japan,
fliohe of the East Asiatic traditions have identified themselves
:4ir'ectly with the Madhyamika school or made the works of its
;efotmder-let alone of its later representatives, of whom they
largely ignorant-the primary textual basis of their sys-
The Madhyamaka-karikas, which modern scholarship has
;'$ade so famous in the West, may have been basic reading for.
r;inost well-educated Buddhists, but it rarely attracted prolonged
and it is probably fair to say that most who read it
,aid so less in search of ultimate answers than in preparation for
were considered more sublime expressions of the
J, Already in the fifth century, even as the Karikas and other
treatises of the major Indian schools were becoming avail-
table in China, scholars there were turning their attention to the
of the relationship among these schools; and by the
and seventh centuries, when the country was learning the
flew literature of the Y ogacara, they were creating their own
'original syntheses of the Indic materials. While the content of
these new systems inevitably owed much to the imported sastm
literature, their structure was often built on indigenous inter-
pretative categories-like substance and function (t'i yung) , prin-
ciple and phenomena (li shih), sudden and gradual (tun
and the like; and while they could not fail to take into
the. doctrine and its
catIOn m Madhyamlka, they were more msplrea by certains-P:0
. 1 1 Ch . 11 h utras:
o partlcu ar anty_ m . espeCla< y t ose-like the
and Avata'f[lsaka
pressed a posItIve mterpretatlOn of the absolute, as aSiinya
the dharrna-kaya, and tathagata-garbha, and so on, and that offei&"i
hope a single great vehicle, or ekayana, in which all
BuddhIsm could be resolved. The so-called Three Treatise (S<!
lun) school of Chi-tsang, supposed to represent East
Madhyamika, was itself such a synthetic system. .Th;
Of these new Chinese systems, none was more characteristi1
of the age nor more influential than that of the great sixth-cell];
tury T'ien-t'ai scholar Chih-i (538-597). Inspired as it
the Lotus Sidra, none was more committed to the higher Bud";
dhism of the one vehicle. Yet probably none was mote
tic to (what its author took to be) the insights of Nagarjunajs;
middle way. In what follows, I want to explore some feature{.
of this system - in particular its famous schema of doctrinal
classification (p'an chiao) - to give a sense of how it sought
incorporate the teachings of the middle way into its
the one vehicle. Rather than try here to discuss the schemain;l
the abstract, I shall focus on a single concrete example - a cor:6
sample, as it were - of how Chih-i's system actually
in the analysis of a specific Buddhist doctrine; I shall then goJ
on to make one or two more general observations aboutihe;
principles at work in the example.
The doctrine I want to use for this sample is the famous.
Buddhist teaching of prat'itya-sarnutpada, or conditioned origina-
tion, especially as this is expressed in the classical formula of.
the twelvefold chain of causation. Few doctrines are more
erable or more centrally placed in Buddhist tradition than the.
dvadafanga-pratftya-samutpada. It was, after all, supposed tobe,
the insight into the truth of this chain that most occupied
Buddha himself as he sat on the bodhi-marpja; and the sutras
etirnes said that to see this t'ruth was itself to see the dharma
affldto see the Buddha.
.Nagarjuna himself, though of course
for emptmess are based more on the general
) IS . 1 .. h h fi d f
of re atlVlty t an on t e speo lC cause an e ect re-
',pfl .c ld h .
of.the twelvelo c am, to have
the anCIent formula of the cham qUlte senously and de-
1 d' . . 3 Y f 11 h' b bi f
... ..)..;t.ed severa ISCUSSIOns to It. et, or a t IS, pro a y ew
would seem less immediately susceptible to interpreta-
as an expression of the sort of supreme Mahayana en-
by Chih-i. Buddhist contemplative tradition had regu-
:t\larly consigned the investigation of the twelvefold pratitya-samut-
'i::ipada -:- along mindfulness of reflecti?ns, on
ui>jinpunty, and lIke - to the lowly, medltatIOns
,{intended as antIdotes to unwholesome states.
IndIan commen-
,;Ctj;lwrs on Nagarjuna (including Pingala, whom Chih-i read) had
rf:tended to dismiss his discussion of the chain as merely conven-
(sarrz,vrti) teaching, intended for the edification of the
No less than the Lotus Sidra itself (at least in
;:j!;.Kumarajlva's version) identified the doctrine as a teaching in-
?itended for the relatively unsophisticated understanding of the
Hence it is crucial to Chih-i's vision of the intel-
and ethical coherence of the one great vehicle that he
;'::beable to show why this doctrine was so central to the tradition
SiaIid hO-N, despite appearances to the contrary, it could function
at the highest levels of the religion,
Chih-i's extensive corpus contains quite a few discussions
;:'of pratitya-samutpada, many of which reflect traditional ways of
the twelvefold chain. In his influential organization
Y2p contemplative technique, for example, he treats meditation
'Qil conditioned origination as one of the five techniques for
; generating wholesome states (shan ken);7 like the Lotus Sidra, he
\associates the twelvefold chain with the pratyeka-buddha-yana.
. Yet he also has a more exalted reading of the chain that extends
.: its significance across the entire range of the buddha-dharma,
'from the basic teachings of the Hlnayana through the supreme,
. perfect enlightenment of the Buddha himself. For my purposes
:;.here, the most important example of such a reading occurs in
'the Miaoja lien-hua ching hsuan i, his extended commentary on
.;the "dark import," or deeper meaning, of Kumarajlva's version
.of the Saddharma-pur.u;larika. The work is largely organized
aro.und a analysis of of
Chmese title of the sidra. In Its second fasCicle, m the Conte'
of his discussion of the first word of the title, Chih-i
six. objects of Buddhist (ching-miao) , as the
whICh he takes up the docirme of prat'itya-samutpada. 9.<5
Chih-i divides his interpretation of prat'itya-samutpada int>;
four or levels: of understanding, to: which he
the followmg rather unWIeldy names: (1) conceIvable origination;;'
and cessation (ssu-i sheng mieh), (2) conceivable non-origination::.
and non-cessation (ssu-i wu-sheng wu-mieh) , (3) inconceivabl{;
origination and cessation (pu-ssu-i sheng-mieh), and ( 4) inconceiv_'"
able non-origination and non-cessation (pu-ssu-i wu-sheng wu-',
mieh). As the names suggest, the a:-e in two groups}
of two: first, prat'itya-samutpada IS divIded mto the conceivable'
and inconceivable; then, each of these is sub-divided into origi-:i;
nation and cessation and non-origination and non-cessation ..... ' .
The hermeneutical categories of the conceivable (cintyii) , or
what can be grasped by the reason, and its opposite
are common, of course, not only throughout Chih-i's writings,
but in Buddhism in general. This epistemological dichotomyis.';
identified by Chih-i here with what is more properly a religious>
or moral distinction between the mundane (chieh-nei; laukika)'
and transmundane (chieh-wai; lokottara). These terms derive
from the traditional Buddhist distinction between the state of
those dominated by the defilements (yu-lou; siisrava) and the
pristine state of the iirya, who has attained the anasrava stages.
Thus, Chih-i's analysis of prat'itya-samutpiida begins with a distinc-
tion between two spheres of application or understanding of
the doctrine-that of the defiled world of ordinary experience, ...
and that of the immaculate world of the advanced adept.lo
Each of these spheres is again divided into two, according
to two ways of treating them-in terms of origination (utpiida) ....
and cessation (nirodha) , and in terms of non-origination and
non-cessation. These two kinds of treatment, says Chih-i, are'
intended for those of dull (tun) and acute (li) faculties respec-',
. tively. Though he does not elaborate the point here,
near the end of his discussion to the terms "phenomena" (shih)
. and "principle" (li) indicates that he also identifies the two with.
these metaphysical notions, commonly used in T'ien-t'ai and
other Chinese exegesis for the Buddhist categories of sarrtvrti-;
(su ti), or conventional truth, and paramartha-satya (chen ti),
u.ltimate truth. II Thus, both the mundane and trans mundane
W\eres can be discussed for the dull in the more easily under-
terms of the phenomena that comprise them, and for the
in the more subtle terms of the principle that underlies
phenomena. These identifications, then, allow Chih-i to
pratftya-samutpada on four levels of discourse: (1) mundane
::vphenomena, (2) mundane principle: (3) transmundane
and (4) transmundane pnnClple;. and we can expect
!itrelationshlp among the four such that (1) IS to (2) as (3) IS to
#,{1-). As we shall see, this relationship is central to the T'ien-t'ai
;(p'an chiao system.
, ':{
;:'1 Well over half of Chih-i's discussion of pratitya-samutpada is
with his first level of understanding, that of conceiv-
Uble origination and cessation. Since this represents what he
;'i(:onsiders the lowest understanding, the space devoted to it
i;'might seem somewhat surprising, and one might have expected
;'hiffi to move quickly on to the higher and more sublime realms
:i{)f interpretation. In fact, however, the attention paid here to
lithe details of the basic teaching appears quite characteristic of
approach. Elsewhere in his writings as well, it is precisely
:;;the lower teachings that seem to receive the most detailed and
:'thorough treatment, while the higher understanding is often
,;passed over quite quickly. In one sense, of course, this imbalance
'may be inevitable, since the lowest level is usually, as here, con-
)cerned with the more detailed scholastic teachings of the
but it is also suggestive of the importance Chih-i
. placed on a firm grounding in the basic doctrines of Buddhism
, and a measure of the conservative, classical approach he took
,'to the religious life. This approach gives to his teaching a strong
sense of what the Chinese like to call "gradualness" (chien) as
opposed to the flashier "sudden" (tun) style that is often held
. up as more characteristic of East Asian Buddhism.
, The section on the first level of pratftya-samutpada is com-
of two parts: a general explanation and a discussion of
',some additional considerations. The former provides a basic
12 JIABS VOL. 11 NO.1
definition of the doctrine and identifies it as the charact .;tH!ii'r;'
d d
.c l' h d" , h erlstlc"
un mg 01. casua Ity t a.t IstmgUls es Buddhism
the theones of the non-BuddhIsts. . tn'
This [teaching of pratftya-samutpiida] differs from
infidels (wai-tao) , They falsely maintain that phenomena ori 0,
-, . " , gInatet'
from Isvara, or from nature (shzh-hszng; prakrtz,) or from at 'J)f,;
(wei-ch'en; a1',lu), or from male and female, or without cause,
, fl h' d d'hh' .eSe
v( s B
se th,eones 0 n[d
Wf]lt t e prmciple of
tao- z, . ut t IS correct octnne 0 pratftya-samutpiida dif"i
from such false notions, It holds simply that ignorance
in the past produces in the perverted mind (tien-tao
paryasta-citta) the predispositions (sa'f[lSkiira), which bring
the present fruit of in the s.ix destinies in
ways accordmg to [whether one s karma IS] good or evil,I2'j
On the basis of the Smrtyupasthana-sutra, Chih-i draws
analogy between the first three members of the twelvefold.;
chain-avidya, sar(tskiira, and vry'nana-and a painter, his paint;
and his picture: the ignorant mind is like a painter, using th;;
various shades of good and evil karma to produce the rebirth,
consciousness (pratisandhi-vijnana) in the six destinies, 13 He theh'
summarizes the first level by saying that the chain
through the three times like a wheel, the members arising and'
ceasing again and again in moment after moment-hence,
designation "twelvefold pratZtya-samutpada of
cessation," >:};
In the rather lengthy section devoted to additional consid-
erations, Chih-i takes up several traditional topics oIl,;'
the twelvefold chain that we find in the abhidharma literature,':
including various approaches to the distinctions between
dent production (prat'itya-samutpada) and dependent origination,
(prat'itya-samutpanna) , the division of the twelve members into
the three times (san shih) and their application to the maturation:
of the individual, the simultaneous occurrence of the twelve id,
a single moment, the cause and effect of the first and last
bers respectively, the members occurring in each of the three
loka (san chieh), and so on.14

The second section deals with the pratftya-samutpada of con-
non-origination and non-cessation, intended for those
,.:tel I . I h h d" .
facu ties. n s arp contrast to t e prece mg, It IS very
direct, stating simply that all members of the twelvefold
:t'h'ain are like empty space (hsu-k'ung), like an apparition (huan-
and ungraspable (pu-k'e-te). Chih-i does not
to gl:: here any for the emptiness of con-
entitles but merely CItes the Suvarr;,a-prabhiisa-siitra to
';fi;e effect that avidya does not exist of itself but only in depen-
/!dence on deluded ideas (wang hsiang; vikalpa), or false thinking
ssu-wei; Thus, he leaves it to :he
;,reader to supply the major premiSe-l.e., that dependently eXlst-
entities are empty-and the conclusion-that, therefore,
is empty. He then covers the remainder of the chain by
out-as the popular simile has it-that, just as the
;inagician produces elephants, horses, necklaces, and people,
;Which the deluded take to be real, so avidya magically produces
.,sfue karma of the six destinies. Finally, by means of another
simile, Chih-i explains the religious significance of
"{rYon-origination and non-extinction: "When one realizes that
hhe vine [he has taken for a snake] is not a snake, fear of it will
that originate, and not originating, it will not cease. This is called
;/the twelvefold pratitya-samutpada of conceivable non-origination
The third section, that dealing with the pratitya-samutpada
;{)f inconceivable origination and cessation, is perhaps the most
}nteresting and difficult. This level of interpretation is said to
.refute the "lesser" (hsiao) understanding and reveal the "greater"
(ta), teaching the transmundane dharma for the sake of those of
both dull and acute faculties. The discussion here concerns the
cittamatra teaching that the mind is the cause of all dharmas. This
teaching is introduced by a quotation from the Avatar[Lsaka-siitra,
which employs the same painting simile we have seen in the
first section: "The mind is like a painter producing the various
14 JIABS VOL. 11 NO.1
five skandhas. Throughout all worlds, there is nothing
d d f h
. d "17 I 1" h' pro;o,\
. rom t e mm. n exp w It means
mmd to produce phenomena, ChlP-I fIrst cItes two confli'
views of the alaya-vi:jrulna: "Some 'say' that the alaya
all is the. true conscious.ness
the alaya producmg all dharmas IS the unsmkmg conscious .'{r,:,!
(wu-mo shih) that is neutral (wu-chi; avyakrta) and
Neither Chih-i nor his famous Chan-jan
here the proponents of these two VIews, but the text does ret/;:!;
us to another discussion of them in the author's Mo-ho
from which it wo:xld appear that he assigns
to the so-called TI-lun and She-Iun schools-I.e., the SIxth ceitl
tury Chinese exegetical traditions emphasizing, in the
case, Vasubandhu's commentary to the Dafabhumika-sutra (Shih1;
ti ching lun) and, in the latter, the Mahiiyana-sar(lgraha
sheng lun). In the same discussion, Chih-i rejects the
both schools, arguing in effect that the former
tifies citta with the ultimate dharmata, while the latter
account for any relationship between the tWO.
he says in our text, comes from attachment to the reality
svabhiiva (hsing), which leads to a satkaryavada
casuality akin to the Sar:p.khya theory of the evolution of
world from prakrti (ming-ch'u)-an understanding we
ready seen Chih-i reject in the first section.i;'!
Having thus dismissed these two views, Chih-i goes oni
state what he holds to be the correct understanding of the Bud-"
dhist teaching that the mind produces the dharmas. . .,.
Not by themselves, not by another, not by both, and not without'
cause [do the dharmas arise.] According to these four
[the production of the dharmas by the mind] is inconceivable. Yeti;
given the conditions of the four siddhiinta, fPrat'itya-samutpiida]'
can still be explained.
. "
Here Chih-i employs the opening verse of the Karikiisto'
establish that the occurrence of dharmas is inconceivable-i,e.,\
that they have only provisional reality and in their own
are ungraspable.
Their occurrence, he says, is like the arising
of images in a dream: though we say that the dream produces,"
images, the nature of the dream itself cannot be grasped;
though we say that avidya produces the other members of
chain, the status of avidya itself cannot be grasped
of the four propositions (ssu chu; catU$koti). 22 Nevertheless,
with the four siddhanta (ssu hsi-t'an), or heuristic
of the Buddha's teaching, we can still discuss the arising
and the rest of the twelvefold chain from the mind
93 .
-.. .
thIS remmder that the teachmg of trans mundane phe- .
pratitya-samutpada is established only as a device for
of the practitioner, Chih-i proceeds to a consideration
actual content of this teaching as it applies to those ad-
bodhisattvas of the anasrava-dhatu, who, although freed
Wr()lll the mundane realm of the klesas, still transmigrate in the
(i-sheng shen). Here he relies on the Ratnagot-
doctrine of the four spiritual obstacles to ultimate
(yuan; pratyaya), causes (yin; hetu), origina-
utpada), and cessation (mieh; nirodha)-to draw out
significances of the twelvefold chain.
. '<


"Conditions" refers to avidyii; ["cause"] to sar[lSkiira; "origination"
to niima-rupa and the rest of the five [present effects]; (the three
members, tr$r;ii, upiidiin, and bhava, are to be understood as
above;) "cessation" to jiiti and jarii-marar;am. These twelve are
numerically the same as those of the mundane [prat'itya-samut-
piida] , but their meaning is very different.
Chih-i then uses the traditional division of the twelvefold
';i:hain into klesa, karma and vastu (or duhkha) to show the relation-
between its members and the Ratnagotravibhaga's doctrine
\8fthe four higher inverted views (tien-tao; viparyasa) - impurity
(pu-ching; asuddhi) , selflessness (wu-wo; anatman), suffering (k'u
quMha), and impermanence (wu-ch'ang; anitya)-that still charac-
.terize the understanding of even the advanced bodhisattva: the
of condition (i.e., avidya, etc.) prevents the realization of
clmrity; the karma of cause (i.e., sar[lskara, etc.), the realization of
sdfhood; the vastu of origination (i.e., vij'nana, etc.), the realiza-
Jion of bliss; [the vastu of] cessation (i.e., jara-mararJam), the
;tealization of permanence.
16 JIABS VOL. 11 NO.1
Finally, our text briefly consip.ers the fourth and
level of
and non-cessatIOn. Here quotes
effect that twelvefol:I IS Itself the
nature (jo-hszng).27 The IdentIfIcatIOn IS worked outby mean'<,.:
between the three divisions of the twelvelomil
cham-mto klesa, .karma and vastu-and the three
causes, under whIch the Nirva1J,a-sutra treats the bUddha-ni:
ture-the cause of apprehension (liao yin), the cause of
tions (yuan yin), and the cause proper (cheng yin)-i.e., the
lectual, ethical, and metaphysical causes identified with
mokJa, and dharma-kaya respectively.28 the klda
(i.e;, vijiiana, tr)1J,a and upadana) are assoCIated with
karma members (sar(l,skara and bhava) with mokJa; and the
ing, the vastu members, with dharma-kaya. On the basis
correspondences, the three divisions of the chain are furthei;
identified with the four gu1J,as (ssu te) of nirva1J,a taught
same sutra. The argument runs somewhat as follows: the klesas"
are themselves bodhi; bodhi is by definition free from
hence, the kleSas are themselves the ultimate purity (ching; suddhi)i
or nirva1J,a. In like fashion, karma is identified with the ultimate';
self (wo; atman) of nirva1J,a,and vastu with the bliss (lo; sukhd)1'
and permanence (ch'ang; nitya).29 ....
In the sections immediately following his analysis of
four levels of pratitya-samutpada, Chih-i goes on, in a pattern
typical of his exegetical methods, to make several general points,
about the character and significance of this analysis. In a section
on "distinguishing the coarse and subtle" (p'an ts'u miao), he
makes explicit its hierarchial structure, pointing out that, while.
there are no levels in the object (ching) of pratitya-samutpada
itself, there are more or less profound understandings of the
object, which, like the famous Nirva1J,a-sutra simile of the refine-
ment of milk, progress from the "coarse" to the "subtle." From
the perspective of the fourth level, corresponding to the ultimate
'(shih) of the middle way taught in the one vehicle of the
all three of the lower levels are "coarse" because
expedient (ch'iian).30 .
succeeding section, on "opening the coarse to reveal
(k'ai ts'u hsien miao), reminds us of another, rather
implication of the one vehicle: that the classification
teachings as coarse is itself based on a coarse under-
. Chih-i here quotes the line from the Lotus Sutra in
the Buddha says, "My dharma is subtle (miao) and difficult
" Since all three of the lower teachings are part
buddha-dharma, the argument goes, it follows that even
as they are the expression of the Buddha's miracu-
are subtle and inconceivable.
his last section, on "discerning the mind" (kuan hsin) ,
. points out the religious implications of his analysis: to
one moment of ignorance (wu-ming) is itself enlighten-
(ming). Each moment of thought contains all twelve mem-
. of the chain, and, since these members are ultimately the
virtues of nirvar;a, to discern them is itself to discern inher-
'permau'eIHX, bliss, selfhood, and purity. In such discern-
the mind constantly abides in the womb of the aryas (sheng
from which it is destined to emerge into full enlighten-
." Such, in outline, is Chih-i's multi-faceted account of the
chain. His basic notion that the doctrine of pratftya-
./sarnut;paGta could be thus distributed over several levels of in-
is by no means without precedent: the Nirvar;a-sutra
for example-one of Chih-i's favorite scriptures and the
.pne he cites as authority for his final section-has its own four-
' ( ~ i ~ r e d division of the doctrine, associated with the four types of
;!;Buddhist adept: sravaka, pratyeka-buddha, bodhisattva, and
buddha;33 similarly, the Ta-chih-tu tun, which East Asian tradition
)ttributes to Nagarjuna, and which regularly provides the source
for so much of Chih-i's material, identifies three types oftwel-
vefold chain: for the prthagjana, for those on the two vehicles
and bodhisattvas not yet established in emptiness (anutpattika-
18 JIABS VOL. 11 NO.1
dharma-k$anti), and for the advanced bodhisattva. 34 Such
precedents no doubt provided inspiration for the
passage, but .the structure of this
ment remams Chlh-I s own and de. arly reflects the part'.
. . I f h I . b h h leu at
pnnClp :s o. IS arger prOject to nng t e w ole of
mto a smgle coherent system. Here I want to add just am:Bf
about those principles as they relate to the
In recent years, there has been some debate about
natur.e of Chih-i's s'ystem and its relation
doctnnal schemas through whICh It was taught by later
but there is no dou?t that most
Important feature of that system IS the doctrme-of the so-eal1d\.2
"four teachings" (huaja ssu chiao ).:..-that divides the
dharma into ''pitaka'' (tsang) , "common" (t'ung) , "distinct"
and "complete" (yuan). Put very briefly and
first corresponds to the Hinayana teaching on dharmas, intende4;1
for friivaka and pratyeka-buddha vehicles, the second to the basic!;'.
Mahayana teaching of funyatii that Jeads onto the
yana, the third to the advanced Mahayana teaching of
that is "distinctive" of the bodhisattva-yana, and the last
"complete," perfect understanding of the Buddha that
transcends and unifies all the other teachings. It should be ir
mediately obvious that such a fourfold division stands behlnd%
our pratftya-samutpiida passage; and, in fact, though he
make the connection in our text, later on in the same
Chih-i explicitly assigns his four levels of interpretation to
four teachings.
While we may (or may not) want to applaud Chih-i'sin .. ::i;:
genuity in bringing pratftya-samutpiida into accord with his own
system, S0 far removed in time, space, and spirit from the
Buddhist formula of the twelve nidiina, for the Tien-t'ai master;
himself, this success must have seemed only a natural elaboration'!
of the passage on dependent origination by Nagarjuna that is'
supposed to have provided the metaphysical basis for the system,"
This is the verse, Madhyamaka-kiirikiis 24: 18, in which, according;
to Kumarajlva's translation, it is said,
Phenomena produced by causes and conditions,
We declare to be empty;
, are called pr<?visional names, .
3l,Or, agam, they are what IS meant by the mIddle way.37

';' particular understanding of this famous passage (the only
as far as I know, in which the Kiirikiis actually uses the term
is to th<; for
teachlllgs of the FIrst PatrIarch of T len-t aI, HUl-wen (fl.
;,550), and to have been handed down to his disciple, Hui-ssu
'1i,515-577), by whom it was transmitted to Chih-i. The under-
in question is formulated by the T'ien-t'ai teachers as
doctrine of the three truths (san ti). This
understands Nagarjuna's verse to be describing con-
1aitioned phenomena in terms of three levels of truth: the empty
!(i(k'ung; sunyatii), the (chia; prafiiapti), and the middle
!{(thung; rnadhyama). Very brIefly .the three truths can be
somewhat as follows: condItIOned dharmas, when un-
on the fitst level, are empty of svabhiiva; on the second
this emptiness is seen not as the total negation of the
but as the affirmation of their conditionality, or "provi-
real status; on the third level, the two categories of the
and provisional are understood to be nondifferent-or,
,{pli,t in other terms, the ultimate truth about the dharmas is un-
to lie in a middle ground, free from the extremes of
'%le empty and provisional.
What is most immediately striking and most peculiar about
reading of Nagarjuna's verse is that it seems to isolate his
;;tamous teaching of sunyatii as but one pole, set off from both
conditioned phenomena of the first line and the provision-
':aIly established phenomena of the third, an extreme that is itself
be overcome, or resolved, by his middle way. Such an under-
2standing of emptiness, quite common in East Asian exegesis, is
expressed in the other formula by which Chih-i most fre-
,quently discusses the movement among the three truths. This
adopted from the P'u-sa ying-lo ching, an important sutra gen-
;erally thought to have been written in China in the fifth century,
which explains the relationships among the three truths in terms
;()f (1) entering the empty from the provisional (ts'ung chia ju
k'ung), (2) entering the provisional from the empty (ts'ung k'ung
Uk chia), and (3) the ultimate middle way (chung-tao ti-i i).38
Though at first glance it appears here that the empty and
20 JIABS VOL. 11 NO.1
the provisional are merely two opposing poles, Chih-i
read the of first line in the
correspor:dmg to the condltloned',ph.enomena m the
of NagarJuna's verse; hence he gIves the formula a
dialectical character, such that it moves through four ,,9'!Yl;:
ments"-froIn the thesis of dharmas, to the antithesis of
h h h
h h' f h . ,'-. yalall;
t roug a Ig er synt eSlS 0 t e two m praJnaptl, to the fi,,:1fc,'
. (f . d h .. 1)' ,lUab
synt eSlS /_ 0 ax: ,; e . m. madhyair:t;z',
C.learly,,, sur:ya.ta IS" the m thIS dIalectic,
hlghe: pnnCIple, as Chlh-l calls It m our example, that
both mto and beyond the transmundane phenomena ofth":
Maha.rana. As it is not merely opposite of the dharmfs$
but, lIke the provlSlonal, must functIOn on two levels:
opposition to the dharmas as their mere negation; and
as the higher negation of the opposition that accounts for tRci:
ultimate unity of the two poles. Whether or not
himself would still recognize himself in this dialectical
tation of his verse, it is probably possible to read it as an interest;,t
ing extension of his reminder that emptiness is also emptYof
, 'i;
There is one final feature of Chih-i's
teaching to which I should like to call attention in closing. If'
the metaphysical levels we have seen here are supposed
supplied by Nagarjuna's verse, the logic of these levels and th
religious significance attached to them seem to come from what
strikes me as an extremely interesting coalescence in Chih-i's'
thought of two ancient Buddhist formulae. The first is the
known rhetorical device of the catwkoti or "four propositions.:'
This device, through which the speaker is thought to exhaust
all meaningful positions on a topic, was of course much
preciated by Nagarjuna, and in fact we have seen his use ofiC
quoted in our example to establish the merely provisional nature,
of citta-matra. As in this case, Nagarjuna tends to use the formula
to negative ends, as a means to the refutation of others' views. 40,
Chih-i, however, also has a more positive, more metaphysical
reading of the four propositions that assigns to each member a
of truth exactly corresponding to our four levels of pratitya-
.ttpiida. Th?ugh .he. does not the J?as-
relatlOnshlp IS made explICIt elsewhere ill hls wntmg,
each of the teachin?s is assigned .a basic metaphysical
that pztaka bemg the common
emptmess (k ung, here ObvIOusly eqmvalent to non-
"ileac . d . b h d h .c . h 4
the Istmctasserts ?t ; an t : per.lect, nelt er. I
recent years, the catu$kotz has occupIed a number of com-
who have been particularly disturbed by the third
.. . b h h fl h h 1 f
:cra 'fourth mem ers, w lC seem to con lCt WIt t e aws 0
and of the excluded middle respectively.42 It
be. it his own awareness of such logical conflicts
Chlh-l to assIgn these two members to the realm of the
Yet, whatever we may say of other uses of the
Chih-i's actual application of it to his fbur teachings
resolves these difficulties. The apparent contradiction of
"ithethird proposition disappears in the distinction, central to
teaching, between paramiirtha and sar(tvrti realms
i'bLdiscourse-a distinction we have seen reflected in Chih-i's
that, while the mind itself cannot be ultimately estab-
it can still be treated for heuristic purposes as the cause
The "distinct" teaching here is precisely that
;iform of Buddhism that seeks to "straddle" the two realms of
;aiscourse, for the sake of the advanced bodhisattva, who, though
established in emptiness, still needs to cultivate the
of the transmundane path. The fourth propo-
>sition, according to Chih-i, does indeed transcend the two-
>v:a1ued logic presupposed by the law of the excluded middle; it
,'Floes so on the basis of the "complete" teaching of the buddha
itehicle, which specifically posits a higher "middle" ground to
the predicates "being" and "emptiness" do not apply .
. ' As interesting as Chih-i's hierarchical reading of the catu$koti
;may be in itself, perhaps more striking is the way in which he
'is able, once the formula is read in this way, to lay its logical
pattern over the seemingly quite unrelated spiritual hierarchy
by the traditional Buddhist model of the miirga. In this
gverlay, as should be apparent from our example, the assertion
pfbeing, characteristic of the first level of teaching, is associated
:'with the laukika path; its denial, with the insight into emptiness
<that leads one to the lokottara plane; the higher affirmation of
22 JIABS VOL. 11 NO.1
both being and its denial then guides the bodhisattva alo'
upper bh'L7.mis of the bhavana-marga; and the still higher
of both leaves him at the ultimate way of the
of the BuddhaY . . Path;!
In this kind of overlay, then, the spiritual developme .: .... : ... :' '.'
each individual adept on the stages of the marga is but an
., f h d 1 f B ddh' . . . nee,,;
m mIcrocosm 0 t e eve opment 0 Uism Itself-a
as it were, of ontogeny Or to
matter from other sIde, the kmd .of analysIs of the we have seen III of
samutpada enVISIOns (If I may be allowed thIS confusIon of ancie'?;:1
tongues) a sort of "meta-marga," in which both the
lationships of the various conflicting Buddhist
the concrete historical development of the disparate
doctrinal literature recapitulate the inherent metaphysical ana;;:
spiritual structure of the one great vehicle on which eachiridi}!':
vidual Buddhist must make his way to the final goal
hood promised by the Lotus Sidra');;0
l. A version of this paper was originally presented to the
"Middleism: Nagarjuna and His Successors," Fifteenth Annual
on South Asia, University of Wisconsin, Madison, 1986. I should like to
my thanks to Profs. Roger Jackson and Jose Cabez6n for comments offCr'ed';
at that reading.vii
2. E.g., Maha-sarrtnipata-sutra (Ta-chi ching, T.397: 13a15-16). :"'_>".-+:,;;;
3. E.g., in chapter 26 of his Madhyamaka-karikas (Chung
T1564:36b-c); and in his Pratitya-samutpada-hrdaya-karikas (Yin-yuan
see TI651-1654). For an excellent study of Nagarjuna's treatment of the"
chain in the latter, see Kajiyama Yuichi, "Chukan ha no juni shi engi
Bukkyo shiso shi 3 (1980), 90-146. '.;.
4. Especially in the common schema of the five samatha
known as the wu ting-hsin kuan, among which meditation on
is recommended as an antidote to moha. A discussion of these practices can
be found in Ominami Ryusho, "Go teishin kan to go mon zen," in Sekiguclii:
Shindai, Bukkyo no jissen genri (1977), 71-90. . J
5. Chung lun, TI564:36bI8; similarly, Bhavaviveka's
vrtti (Pan-jo teng lun shih, T.1566:131bI3). .'.
6. T.262:3c23-24. Here, as elsewhere in the literature, the dvadasiinga-y
pratitya-samutpada is set in contrast to the four aryan truths, taught for the';.
sr(ivakas, and the six paramitiis, intended for the bodhisattvas. For a generah,'
'nt of some of the various interpretations given the twelvefold chain
, literature, see Mitsukawa Toyoki, "Daijo butten ni mirareru
inEngi no kenkyu, Bukkyogakukenkyu (toku-shlt) 39-40 (1985), 19-49.
See, e.g., his popular Hsiao chih-kuan, T.1915:469clOff. The five
to'the wu tmg-hsm kuan.
.g., in his Tz'u-ti ch'an men, T.1916:480cl5.
" 9. T.1716:698b29ff. This section has recently been translated by Paul
nson, in his "The Two Truths Controversy in China and Chih-i'sThreefold
<!,th Concept" (Ph.D. dissertation, University of Wisconsin, 1985), 530-561;
'''slations appearing herein below are my own. The other five objects dis-
".d by Chih-i in this section are the ten "suchnesses" (shih ju) of the Lotus
, the four [aryan] truths.(ssu til, the two truths (erh til, the three truths
and the one truth (z tz).
In technical T'ien-t'ai parlance, chieh-nei refers to all states within
worlds (san chieh), prior to the elimination of the so-called chien-ssu
(i.,e., the darsana-heya and bhiivana-heya kleiavara1Ja, as distinguished
f;!r'orU the two types of jiieyavara1Ja that Chih-i calls ch'en-sha and wu-ming).
;li%::JIL E.g., at 700a17.
12. 698c6-11. All the views of the infidels here, with the exception of
and mother" (ju mu), appear in Pingala's list of false doctrines corrected
teaching of Chung lun, T. 1564: Ibl8ff.
13. Paraphrase of Nzen-chu chmg, T.721:135aI7.
14. 698c28-699b28. For a discussion of some of Chih-i's material in this
see Nitta Masaaki, "Chugoku Tendai ni okeru inga no shiso," in
Shis'o Kenkyukai, ed., Bukkyo shiso 3: Inga (1978), 253-272.
15. After Chin kuang-ming ching, T.663:340bI5, with some omissions;
':f6rthe literature on this notion that the twelvefold chain depends on false
see Mitsukawa, op cit., 35-44.
16. 699c7-8.
17. After Hua-yen ching, T.278:465c:26-27.
yC'., 18. 699cl4-16. The term wu-mo here comes from one traditional in-
of alaya as alaya, "not sinking."
19. Mo-ho chih-kuan, T.1911 :54a23-b6. The argument 'of this passage
seem to be that, if, following (what Chih-i takes to be) the Ti-lun
we identify the source of phenomena with dharmata, which is neither
nor object, then we cannot explain in what sense it is citta that produces
:'phehomena; on the other hand, if, following the (reputed) She-lun position,
i1Ve identify this source with an alaya distinct from dharmata, we cannot explain
:<the relationship between the dharmas and the dharma nature.
: 20. 699c20-22.
21. Chung lun, T.1564:2b6.
;: 22. Chih-i applies the to the dream in the Mo-ho-chih-kuan
:;(T.l911:54b8ff), to show that the dream cannot be understood as arising from
,the mind of the dreamer, the condition of sleep, both or neither. The dream,
"inthis analogy, is to the mind as the alaya is to dharmata ..
23. And see the parallel passage at Mo-ho chih-kuan, T.19ll:54c7. The
siddhanta here derive from the Ta-chih-tu lun (T.1509:59b18ff), in which
24 JIABS VOL. 11 NO.1
it is said that the Buddha uses four types of teaching: the worldly ( h }."i .
to encourage practice; the individual (wei-jen), to promote virtue; I
, h") 'I d h d' 1 " anti
tUI-C Z, to counteract eVl ; an t e car ma meanmg (tl-i i), to bri '"".
enlightenment. The Fa-hua hsuan i (686bff) the four at som n
interpreting them according to ten different aspects and relating th e ell
, . ern to
four aryan truths, the four levels of practitioner, and so on. For discu <
,. h' K k "'f "Sh' h" SSlOn:',
t lpS, awa atsu lV, amoru, 1 s ltan gl to kyoso
Seklguchl, Bukkyo no JZssen genrz, 303-318. '
24. For the Ratnagotravibhaga teaching (of which Chih-i's
see Chiu-ching i-shengpao-hsing lun, 1 :830a28f[ cl _
1 S dISCUSSion IS marred by the fact that throughout he conSistently substi"Jtf(!
. ('" k") r h' - , ' h'
szang mar lor t e sastra S yzn- slang.. ...... ;
25. 700a4-7 (the parenthetical clause here is Chih-i's). The
being established i.n passage upon the common
the twelvefold cham mto three dlvlsIOns (san tao or san lun) distributedgv'ii
the three times, (This tripartite division, found in the
kosa, etc., is also employed by Nagarjuna in his PratZtya-samutpiida-hrd
Chih-i's schema here can be shown as follows: .
pratyaya: kleSa: 1) avidya 8)

9) upadana
hetu: karma: 2) sar[!Skara 10) bhava
utpada: vastu: 3) vijnana
4) nama-rupa
5) Ja(iayatana
6) sparia
7) vedana
26. 700a7-16; paraphrasing the Sastra at T.1611:830bI3ff.
27. Ta-pan nieh-p'an ching, T.374:524b7.
28. For this teaching, see T.374:530aff.
29. 700a16-27, The gU7Jas here are, of course, the four original
paryasas. The relationships among the members of these lists can be
as follows:
liao yin: bodhi: kleSa:
yuan yin: karma:
chengyin: dharnza-kaya: vastu:
1) avidya
9) upadana
2) sar[!Skara
10) bhava
3) vijnana
4) nama-ntpa
5) ;a(iayatana
6) sparia
7) vedanii
11) jati
12) jara-marar;am
31. 700b22ff; the Lotus Sutra line occurs at T.262:6cl9.
'32. 700c5ff.
33.T.374:524b2. The Sutra goes on here to identify the twelvefold chain
the Buddha-nature, with paramiirtha-sunyatii, with the middle way, Bud-
ood and nirviir}a.
34:. 'T.l509:a27ff.
35. The arguments have been brought together in Sekiguchi, Tendai
kuno kenkyu (1978).
36. 709b; and see Chan-jan's sub-commentary, Hsuan i shih-chien,
717:S48b26, ell, etc. In the T'ien-t'ai system, the second teaching is "com-
"u"to all three vehicles, in the sense that all realize sunyatii-the two vehicles,
oligh what is called "analytic emptiness" (hsi-k'ung); the bodhisattva, through
"utial emptiness" (t'i-k'ung). .
37. T.1564:33bll.
"'38. T.1485:1014bI9-21.
. 39. E.g., at Kiirikiis 12:10-11, 8:8. Chih-i's notion here that Nagarjuna's
die way corresponds to the higher, self-negating function of emptiness
iM*gflects the sort of statements one finds in the Nirviir}a Sutra: e.g., "The
is called emptiness in its cardinal meaning (ti-i i k'ung;
!hi: emptine:s is wisdo.m (chih-hui). The e.mptiness
of here conSIsts m .not seemg eIther emptmess ?r (pu-
.. To see everythmg as empty and not to see It as non-empty IS not
called the middle way." (374:524bI2ff) A general discussion of the
i:tbhiuese notion of the middle way as a third, higher truth appeared some
in .the. of this . ir: Whalen "Non-duality of ;he
Truths m SmItIc MadhyamIka: Ongm of the ThIrd Truth," 2:2 (19/9),
For more specific comparison of the T'ien-t'ai and Indian interpreta-
of Karikiis 24: 18, see Nakamura Hajime, "Chud6 to kukan," in Yuki ky8ju
Bukky8 shis8 shi ronshu (1964),139-180.
;3.;40. The closest study of Nagarjuna's uses of the in the Kiirikiis
:ihas been done by Tachikawa Musashi; see, e.g., his recent Ku no k8z8: Churon
1,"41. E.g., in Ssu-chiao i, T.1929:73a. This sort of hiararchic reading of
i:tiie in the Kiirikiis is not without its Indian parallels. Candraklrti, for
uses it to explicate verse 18:8, the only passage in which Nagarjuna
employs the four propositions in an affirmative sense to claim that
;theBuddha teaches that everything is real (shih; tathya) , not real (jei shih; na
:;tathya), both and neitHer. The first, says CandrakIrti, is intended to impress
'hhe worldly with the Buddha's complete knowledge of the world; the second,
'!O cure the believer of his belief in realism; the third, to distinguish the ordinary
.arid enlightened views; the last, to free the advanced practitioner from the
'grial traces of the iivarar}as. (Prasannapadii 370-371)
, 42. While the logic of the catu5koti has been discussed by Robinson,
Jayatilleke, etc., perhaps the clearest statement of the basic logical problems
n,vas given by Frits Staal, in Exploring Mysticism (1975); for a discussion of the
'actual implications of the schema in Mahayana literature (and additional bib-
liography on the topic), see David Ruegg, "The Uses of the Four Propositions
'.of the and the Problem of the Description of Reality in Mahayana
26 JIABS VOL. 11 NO.1
Buddhism," Journal of Indian Philosophy 5: 1-2, (9-1211977), 1-71.
43. In Chih-j's standard terminology, the second member of the
corresponds to the wisdom eye (hui yen) .of the two vehicles that
jiwtii (i-ch'ieh chih); :he third member repTesents the dharma eye (fa yen;
bodhtSattva that achIeves miirga-}iiatii (tao chung chzh); and the last is the
cient buddha eye (fo yen) that has realized sarviikiira-jiiatii (i-ch'ieh chung
A. Names and Terms
cheng yin \11
ch'en-sha It '/Y
chen shih t!.
chen ti
chieh-nei i}Jl
\., '......::.
ching :.,
ching 1"t
ching -miao j:-l: *y
' --r:fii.'
ch'uan 1t
chung i '<f 1,
erh ti :;..
fa yen
fei shih
fo-hsing lit'I'!.
fo yen
fu mu
hsiang "#l
hsiao ta ,j'K.
hsing r\'! ,
hsu-k'ung 11. \1
hua-fa ssu chiao
huan-hua ill
hui yen
i-ch'ieh chih -tl} fU'
i-ch'ieh chung chih -u7 ;ft1\i
i-shenK 1-
. ...-;''k
1 tl ;)1" , ",
k'ai ts'u hsien miao iry[
k'u t ...... ' .. .
kuan hsinll'G
k'ung \'):
liao yin J III
Ii shih "ff .f-
miao ,h.y
ming DR ,

p'an chiao{'jl(
p'an ts'u miao Ii 4-1/
pieh F'J
pu-ching /),:t
pu-k'e-te 1-"01t
pu-k'ung /)' 't
. pu-shan ssu-wei .,;
pu-ssu-i sheng-mieh
pu-ssu-i wu-sheng wu-mieh ",
wei-jen JJ:;!,--
wu-ming uA
wu-mo shih
wu ting-hsin kuan

yin \{l
yin-hsiang \17-;\:8
yu 11
yuan ".zJi ..
yuan III
yuan yin \1]
yu-lou ilj ---rfi
B. Authors and Titles
Bukkyo Shiso Kenkyukai 1M ,Bukkyo shiso: Inga

Bukkyo shiso shi Ill( "%):. 'j:..
Chin kuang-ming ching i"t, .
Chiu-ching pao-hsing lun'W.. l.,--*'l'j,t My
Chung lun
Engi no kenkyu ,Bukkyogaku kenkyu (tokushu)
, . U
Fa-hua hsuan i .
Hsiao chih-kuan
Hsuan i shih-chien ff
Hua-yen ching -t rnd,if
Kajiyama Yuichi ,"Chukan ha no juni shi engi .' .
kaishaku" <t iWL.;;\K.dJ t:::..
Kawakatsu Mamoru H\ f%- , "Shi shit an gi to kyoso ron"
Miaoja lien-hua ching hsuan i
Mitsukawa Toyoki '1u)l\ t t ,"Daijo butten ni mirareru juni
-K -t /11>. \:'-:If 7 ft.-"1- f::.
Mo-ho chih-kuan P\f%4 Y: 'tt
Nakamura Hajime '>fHfu, "Chudo to kukan"
NittaMasaaki Jw1l1tt ,"ChugokuTendainiokeruingano
r.. ';j. 1:'-h'Ii} 11/ lQJ
Ominami Ryusho "Go teishin kan to go mon zen"
Pan-jo teng lun shih
P'u-sa ying-lo ching
Sekiguchi Shindai \I 11... ,Bukkyo no jissen genri X fl 'I.

Sekiguchi Shindai, Tendai kyogaku no kenkyu 1 ... %K"} cfJ IW,.
She ta-sheng lun th;t.:1J;:.Mr
Shih ti ching lun t :l1!!...illi U
Ssu-chiao i \1J11/r'f.-
Ta-chi ching t ......
Tachikawa Musashi ,Ku no koZOi Churon no ronri
'If a, tl
Ta-chih-tu lun 1..
Ta-pan nieh-p'an ching -r ... ffZfl llli:l1l
....-------- Now Available ----__
Heinrich Dumoulin
Translated by James W Heisig and Paul Knitter
This is the newly revised and greatly expanded edition of the
claimed history of Zen Buddhism, updated to take into
the wealth of historical studies published during the past
The first of a two-volume set, this book offers a detailed account of .....
the history, development, ideas, and image of the Zen school of
Mahayana Buddhism. Here, Heinrich Dumoulin explores the ear-
liest beginnings of Zen, from its roots in the Buddhist and Yoga
traditions of ancient India to the influence of Taoism on its forma-.
tion and development in China. He also invokes the .
personalities of Zen Buddhism, among them Sakyamum
Huineng, Lin-chi, Ma-tsu, Shih-t'ou, and others. In addition:
Dumoulin discusses the many and varied expressions of Zen in
the art and culture of China.
Also included are six appendices: a list of abbreviations; a chrono-
logical table; Chinese characters; genealogical tables; an index of
names; and an index of subjects.
This superb study captures the essence of Zen Buddhism, clarify-
ing its origins and revealing a richer picture of its history as a
Volume Two, on Japan and Tibet,
will be published in 1989.
349 pages g $14.95, paper
ISBN 0-02-908260-9
Also available in cloth, $35.00
ISBN 0-02-908270-6
At bookstores or call 1-800-323-7445
to order by VISA or MasterCard
A Division of Macmillan, Inc.
8661'hird Avenue New York, N.Y. 10022
AU: R. Oppedisano
Prices subject to change.
the Possibility of a Nonexistent Object
iGonsciousness: Sarvastivadin and
"Collett COJC
In the first five centuries e.E., both Buddhist and non-
'(Whist philosophical schools increasingly turned to the
f .klysis of perception and specifically of the locus and existential
objects of perception. These schools! elaborated their
on the dynamics of the perceptual process as a whole
f:t@6ugh an examination of seemingly minor issues. Among
the question of whether or not a nonexistent object can
perception, and the explanations offered for the per-
of objects of questionable existential status such as illu-
and dream images, had significant ramifications for their
of ordinary external or internal perception and
l:ibgnitive functioning. On the one hand, admitting that nonexis-
objects can stimulate the arising of perception not only
the existential status of the objects of ordinary per-
but also jeopardizes the possibility of certain knowledge.
other hand, demanding that all perception depend only
existent objects makes it extremely difficult to account for
perception of these objects that have questionable existential
ii;sta:tus .
. :: Within Buddhism, this issue of a nonexistent object of per-
was extensively treated in northern Indian Abhidharma
These discussions not only reveal the position of Buddhist
t;t\bhidharma schools, but also provide the indispensable back-
and context for understanding the epistemological po-
of the later Buddhist logicians.
The Sarvastivada
schools have particular importance be-
32 JIABS VOL. 11 NO.1
cause their positions best represent the two logically co
views on this issue. The Sarvastivadins hold that all
an existent obj.ect, .whil) th: admit
In certaIn cases, the object IS nonexlstent.
ThIS differen" ,Acl
. .' . ICe .of!
OpInIOn reflects a broader dIsagreement concerning .' li,e;
of perceptual process and its relation to
cognItIve functIOns, such as memory and conceptual thou h. ..... ....: ...i.i
O h
h' h '. g t
ntIs, as on many ot er Issues, t e opposmg VIews orih'i,)
and generated a c?mplex and rich)
of argument. As wIll. be shown, recurrent and'
detaIled arguments can be reduced to two basIC
veloping a defensible model of the perceptual process, and
counting for the perception of objects of questionable
status. The important texts that present theSarvastivadin pOS!-'"
tion include the early Sarvastivadin Abhidharma cano
commentaries,? and the later Sarvastivadin expositoty}
works, notably Vasubandhu's Abhidharmakosabhii5ya, Sanghabh_.r
dra's Nyayanusara, and the Abhidharmadlpa.
The com'}!
mentaries as t?e p.rimaI'Y;
sources for vlews.
Hanvarman s
is also a valuable source for views often identical to those
presented elsewhere. II;;:,
Like many of the controversies between the Sarvastivadins':
and their debates about percepti6tii
often seem to revolve around minor, obscure, and inherited
doctrinal issues. Closer inspection, however, shows that these>:
debates, including those over perception, are actually
according to two fundamental disagreements .. The first
the way in which constituent factors of experience (dharma) are,
thought to exist.
The Sarvastivadins argue that factors exist ai
real entities (dravya) in the three time periods of past, present,
and future. As such, they are defined as intrinsic nature;
(svabhava), characterized by a particular inherent characteristic
(svalak)a1J,a). Given appropriate causes and conditions, these
existent factors manifest a particular activity (karitra), which theIl'
defines them as present. However, since factors also exist as
past or future, they are capable of serving as conditions in those,
states as well. Sanghabhadra defines this past and future fune:
. tioning of a factor as capability (samarthya), thereby distinguish-
ing it from that factor's activity (karitra), which occurs onlyin
the equa.te a factor:s with
esent actIvIty. One cannot meamngfully dlstmgmsh a fac-
nature from its activity, and thereby speak of its
in the past or future. Further, they argue, factors do
." . 1 d . f' . . h . r
:*);li'exist as ISO ate umts o. mtnnslC nature t at mam.lest a
activity through the influence of other isolated condi-
the the process of causal interrelation
fact of experience; the fragmentation of this process
factors possessed of individual existence and unique
is only a mental fabrication.
second fundamental area of disagreement between the
fS'rvastivadins and the concerns the dynamics of
The Sarvastivadins allow both successive and
Mfuultaneous models of causation: certain causes (hetu) or con-
(pratyaya) arise prior to their effects, while others, which
a supportive conditioning efficacy, arise simultaneously
them. The however, allow only successive
a cause must always precede its effect. These basic
about the nature of existents and causality con-
set the framework within which the Sarvastivadins
conducted their debates.

;II.iThe Sarvastivadin Model of Perception
order to construct their model of perception, the north-
Abhidharma schools begin from the description of
lperception found in the scripture. There, a given type of percep-
(vry'nana) is said to arise in dependence upon
organ and an object.
Both the sense organ and the .
are necessary conditions; if either is lacking, perceptual
:S9nsciousness will not arise.
There are six such sense organs
six corresponding objects, referred to as the twelve sense
spheres (ayatana) , which together with their six corresponding
of perceptual consciousness constitute the eighteen ele-
:tneuts (dhatu), of which all experience is composed. These eigh-
elements include the five external objects, the five exter-
directed sense organs, and the five corresponding types
pf externally directed perceptual consciousness. Internal mental
34 JIABS VOL. 11 NO.1
awareness is also analysed according to the model of e t
. h' f Xe
sensory perceptIOn: t e prevlOus moment 0 perceptu I,'
sciousness, which serves as the mental organ, and mental
. h . . f 'd' acto
con ItIOn t e ansmg 0 a correspon mg moment of III . :7i.
. 16 enta,!ii'
erceptua consCIousness.' . .t .. ,ini:
..... ; ..
In their attempts to clarify aspects of the perceptual
1 f b
. h . Abh'dh' oceSS;!
e t l?,UOUS m t e s.cnpture, 1 texts
exammatIOn of perceptIOn on three questlOns: 1) what nas.tL.:i.}! ....
. h b' h .!te'h'
ower 0 sensmg teo 'lect: t e sense orllan, pe.rceptual .. ..... >;.,6." . '.'
J U con'.,.,,;;
sciousness, or some other mental faculty; 2) what is the cna"r ';iM
ter of mental perceptual and how does it diff6#J
from the five externally dIrected types of perceptual conscioti.;Sf;,
ness; and 3) in what sense do the sense organ and object
.for the arising of perceptual and
IS the speCIfic character of the object perceived? In their
gent answers to these questions, the northern
Abhidharma schools developed different models of the
For the Sarvastivada school, the perceptual process
with the sense organ (indriya) , or basis (asraya) that senses.Sfi,
grasps an object-field (vi5aya) appropriate to it. A given
organ grasps an object-field, only when supported by
consciousness;17 nevertheless, this function of grasping
ject-field is attributed only to thesense organ, and not to
tual consciousness, or to some other thought
(caitta)18 associated with perceptual consciousness.
The fun8-;';
tion of perceptual consciousness consists simply in being aware,
of (vij"anati), or generically apprehending (upalabdhi) the natute;
of the object-field grasped by the sense organ.
In this way,
. function of perceptual consciousness is distinguished from thafi
of its associated thought concomitants (caitta). Perceptual
sciousness generically apprehends the nature of a particular"
object-field: for example, visual perceptual consciousness grasps.;
an object as visible material form. The associated thought con-;
comitants, however, grasp the particular characteristics of the'
object-field: for example, whether that object is pleasant or un
pleasant, male or female, and so on.
In other words, perceptual'
consciousness apprehends only the particular characteristic of
an object-field in its generic category as a sense sphere (ayatana-
svalak5a'/fa): for example, as form, sound, and so on. It does not
"ehend the distinguishing particular characteristic of a given
"""",iP,f,t' field as an individual real entit,y (dravyasvalaksana) within
f, l be
C - , . .
category. These individual particular characteristics
only by the thought 22
of the five externally dIrected sense organs IS restncted
to one object-field: the eye can grasp only
material form, the ear only sound, and so on. The object-
the sense exists real entity
as a prOVIsIOnally eXIstmg compOSIte (ho-ho, samagrz,
Further, the appropriate. sense organ a 1?ar-
only when both are m the present tIme penod.
;!rliepresent sense organ and present object-field then serve as ,
for the arising of a corresponding simultaneous in-
of perceptual consciousness.
When apprehended in the its associated
rtI1pught concomItants, the object-fIeld (vz$aya) IS referred to as
i;fue.object-support (iilambana).25
When the Sarvastivadins assert that the externally directed
'fhse organ, the external object-field, and the resulting exter-
directed perceptual consciousness must be present in the
"Jarne moment, they assume a simultaneous model of condition-
tirig. Indeed, to support their contention that conditions may
;arise simultaneously with their effect, the Sarvastivadins cite the
statement that perceptual consciousness arises in de-
'pendence upon two conditions.
The Sarvastivadins further
"invoke the fact of direct perception as proof of the simultaneity
:?fthe sense organ, object-field, and perceptual consciousness.
:Indirect perception (pratyak$a) , a momentary external object-
field is grasped by a momentary externally directed sense organ
and apprehended by an equally momentary instance of one of
the five externally directed types of perceptual consciousness.
;'This is possible only if the object-field, sense organ and percep-
Nal consciousness are simultaneous.
, Mental perception differs from external perception in sev-
eral significant respects. The mental organ (manas), which con-
ditions the arising of a present moment of mental perceptual
Consciousness, is defined as the immediately preceding moment
of perceptual consciousness, regardless of its type.
That is to
say, any of the six varieties of perceptual consciousness may be
designated as the mental organ for a subsequent moment of
36 JIABS VOL. 11 NO.1
mental perceptual consciousness. Unlike the other five
. . . exter"'"
sense thIS precisely
It IS past, cannot be saId to perform Its dlstmctIVe (ka . '>;:7i
. h b' . f h ntra)."
o sensmg or graspm!S teo 0 t e mome@'\
of con.scIOusness .. Instead, It SImply
door, or Immediately contiguous condItIOn (samanant"i.\
pratyaya) f?r the of the present moment of
tual conscIOusn.ess, which then apprehe.nds the
Therefore, .unlIke the five externally sense organs
correspondmg types of perceptual conSCIOusness, the priormen_;'
tal organ and its resultant present mental perceptual
ness are not simultaneous, and do not necessarily share the
Nevertheless, the two requisite conditions
. . . r".
the ansmg of a present moment of mental perceptual
ness, that is, a basis (iisraya) and an object-support (iilambana)s
are still provided through the past mental organ and the
Mental perceptual consciousness also differs from the five'
externally directed types of perceptual consciousness in its model
of operation. Mental perceptual consciousness not only
prehends the particular characteristic of an object-field in its I.'
generic category, for example, visible material form like the

color blue, but also apprehends the designation, "this is blue.;' . ;
Thus, unlike the five externally directed types of
consciousness, mental perceptual consciousness operates by:
means of designation (adhivacana), or namesY .
In addition, mental perceptual consciousness is distin-
guished from the five externally directed types of perceptuaIi
consciousness on the basis of the differennypes of conceptual:
thought (vikalpa) with which each is associated. According. t6
the Sarvastivadins, there are three types of conceptual thought:
1) simple conceptual thought, or conceptual thought in its intrin-'
sic nature (svabhiivavikalpa) , which is identified with initial
quiry (vitarka);33 2) conceptual thought through discrimination
(abhinirupat;avikalpa); and 3) conceptual thought through
lection (anusmarat;avikalpa). Even though the five externally di-
rected types of perceptual consciousness are said, by tradition,
to be without conceptual thought (avikalpika), the Sarvastivadins
interpret this as indicating that only the last two types of concep-.
tual thought, that through discrimination and that through re-
, are absent.
Each moment of perceptual conscious-
is associated with both insight (prafiia) and mindfulness
rti). When they are associated with mental perceptual con-
'6usness they are strong and are identified, respectively, with
'hceptual thought through discrimination and conceptual
when and mind-
are assoCIated wIth the fIVe externally dIrected types of
rceptual consciousness, their activity is weak; therefore, the
c,:./g'rtesponding types of conceptual thought are said to be ab-
Nevertheless, since inquiry (vitarka) still characterizes
five externally directed types of perceptual consciousness,
can still be said to have the first variety of conceptual
in its intrinsic nature. By contrast, moments of mental
consciousness. associated with insight and
are charactenzed by all threevanetles of conceptual

offers a further explanation of the charac-
"without conceptual thought (avikalpika)" as it is
to the five externally directed types of perceptual con-
A given type of perceptual consciousness can be said
'1:lit4have conceptual thought under two conditions: 1) that a given
i'il;fype of perceptual consciousness can apprehend, within one
an object-field of more than a single category, or 2)
a series of many moments of the same type of perceptual
can occur with regard to the same object-support.
:,/,0The five externally directed types of perceptual consciousness
l;:;fail to meet these two conditions: they apprehend only a present
:t;'9bject-field of a single category, and a subsequent moment of
,.,the same type of perceptual consciousness cannot apprehend
':;that same object-support. However, since mental perceptual
.,jconsciousness is unrestricted with regard to both the category
,.'and time period of its object-field, it may apprehend an object-
;2 field of more than a single category in one moment, and several
;',;fnoments of mental perceptual consciousness can apprehend
i:.the same object-support. Therefore, Sanghabhadra concludes
it can be said, in agreement with tradition, that only mental
J;perceptual consciousness has conceptual thought.
, . Further, the scope of the object-field of mental perceptual
i(:()nsciousness is much broader than that of the five externally
'i,directed types of perceptual consciousness. Within the tradi-
tional classification of eighteen elements (dhatu), the
of the mental organ and mental
dharma element, or all constituentlfactors (dharma) not
in any of the other five object-field categories. The ten
directed sense organs and their corresponding types of
tual consciousness are restricted to present obj
single category. The mental organ and merital
sciousness have no such restriction. Mental
sciousness can apprehend all factors (dharma)
of the eighteen categories of elements. Therefore, the
nal object-fields may be both their own
tive perceptual consciousness and mental perceptual
ness. The other thirteen elements, that is, the six sense
the six types of perceptual consciousness, and the
ment, which includes the three unconditioned
(asa'f[l,Skrtadharma), are apprehended only by mental
Mental perceptual consciousness also
prehend factors of any of the three time periods, past,
or future.
Therefore, mental perceptual consciousness,
unrestricted in both the category and time period of its
field, is said to be capable of apprehending all factors.
. In addition to these eighteen categories of u. .......
tors, which exist as -real entities (dravyasat) in the
periods,the scope of the object-field of mental ....
sciousness includes composite entities (ho-ho, sa.magrf,
whose existence is merely provisional (prajiiaptisat).40
composites are apprehended only. by conceptual
through discrimination (abhinirupa1J,avikalpa) , they are
ject-field of mental perceptual consciousness alone.
III. The Model of Perception
The Dar!?tantikas also accept, as a provisional rlp"rrtnt1l
the Sarvastivadin model of perception as involving a sense
object, and perceptual consciousness, hu.t they differ from
Sarvastivadins on the following points: 1) the process .
which perception occurs; 2) the temporal relation among.
provisionally designated sense organ, object, and .
consciousness; and 3) the nature of the object perceived.
the first point of disagreement, concerning the process
notes that according to the
ntika view, It IS not the sense organ, but rather the collo-
n(samagri) of sense organ, ,and so that can be said
iihse or percelveY The and the
elaborate upon this model ofpercep-
tfrerception, like all experience, can be described only
as consisting of individual factors possessing
Hue activities; in .case of perception, as in. all
kal relations, there eXIsts no dIstmct agent or cause possessmg
bwnactivity of producing a distinct effect.
Instead, there is
Slfi{plya stream of experience, or more precisely, a stream of
effect (hetuphalamatra). These provisionally designated
kf@iVidual causes and effects can be said to have activity only in
that they constitute a conventionally existing colloca-
rlt?r'(of factors.
In the experience of perception, words such
organ, object, or perceptual consciousness can be used
1Srtlyfiguratively to refer to moments abstracted from the causal
as a whole; there is no single factor that perceives or
that are perceived ..
the ninth chapter ofthe Vasuban-
a model of the perceptual process which, though
explicitly to the similarly refuses to
il!Icate distinct activity to any of the components through which
fth'eprocess is described:
. In that case, when it is said in the scripture that "perceptual
consciousness (vijnana) is aware (vijanati)," what does perceptual
consciousness do? It does not do anything. Just as it is said that
the effect conforms to the cause since it attains its existence
(iitmalabha) through similarity (sadrsya) [to its cause] even without
doing anything, in this way also it is said that perceptual con-
sciousness is aware since it attains its existence through similarity
. [to its object] even without doing anything. What is [this that is
referred to as] its "similarity"? It is the fact that it has the aspect
of that [object]. For this reason, even though that [perceptual
consciousness] has arisen due to the sense organ, it is said to be
aware of the object-field and not of the sense organ. Or, just as
the series of perceptual consciousness is the cause with regard
to a given [moment of] perceptual consciousness, so there is no
fault in saying that perceptual consciousness is aware, since one
can apply the word "agent" to the cause,
Thus, for Vasubandhu,perceptual consciousness should
. d f h' . . . not
mterprete as a actor avmg umque actIVIty: that
awfareness of a d
0fbject-field. Tfhe
re ers to a causa senes o moments 0 conscIOusness that '.'
. h h . 1 f h . f d arIses>'"
WIt t e partlcu a:r .aspect 0 w. at IS re erre to as its
One can also provlsIOnally descnbe perceptual
. h h' d' . h .. ess as
aware m t e sense t at It con ItIOns t e ansmg of subseq.j .. 'J.!.
f 1
Th f uenv
0 perceptua conSCiousness. ere ore, as in
model, Vasubandhu suggests that one
sharply the of object from that
perceptual that IS saId to apprehend it;
one must view perceptIOn as a causal
response to this model
ceptIOn IS sImple: though all do indeea't.
anse from a collocatIOn of causes and condItIOns, each
within the collocation has a distinct particular characteristic ctnd):j
activity.46 Similarly, even though perception results from a cql-.B
location, the existence of its individual causes and
real entities each having a distinct intrinsic nature and activity
may be proved through scriptural references and argument. .:.
On the second point of disagreement concerning theierri-:,;
poral relation among the provisionally designated
of perception, the reject the
claim that, in the case of the five externally directed typesQf/(
perceptual consciousness, a simultaneous temporal relation ob:;t
tains among the sense organ, object-field, and perceptual corf-'
sciousness. Their rejection is a consequence of their refusaltct(
accept any type of simultaneous causal relation: the ..
claim that there is no possibility of a relation of producer
produced (janyajanakabhava) between factors that are simultane-;:'
ous (sahotpanna). If such simultaneous causal relations were pos- .
sible, then the generative factor (janakadharma) would be without
any generative capability, since the factor that it supposedly
produces arises simultaneously with it. Therefore, the factor
that is designated as the generative cause must exist at a time
different from (bhinnakiila), that is, specifically prior toits effect.
Consequently, the two provisional conditions for the arising of
perceptual consciousness, the sense organ and the object, must
exist prior to, not simultaneously with their effectY .
This refusal to accept the simultaneity of the sense organ,
and the perceptual consciousness results in a model
as a process. 1his model is attrib-
to the master Sri:iata.
The sense
and the object-field in the first moment condition the
... ... .. of perceptual consciousness in the second moment.
with the assemblage of the sense organ, object-
and the
;RlZlits-feelmgs (vedana) , concepts (sarrtJna) , and vohtlOn
in the third and subsequent
i,\I{i,;C\.Both Sanghabhadra and Vasubandhu cntICIze Snlata's suc-
model of perception. In their view, it results in a multi-
of cognitive functioning, in which the various
activities such as perceptual consciousness, feelings, and
that occur in the same moment ?ave different object-sup-
They claim that, to Srllata's ar: .obje<;:t-
and sense organ present m one moment A condItlOn the
of the corresponding perceptual consciousness of that
(p,articular in the subsequent r.n0ment "B." For exam-
form. and the eye m one
the ansmg of visual perceptual conSClOusness of Its
object-field. in the next moment. This assemblage
or collocatlOn (siimagr'i) of these three over two mo-
acts as a cause to produce feelings with regard to that
:6riginal object-field in the third moment "C." However, in this
moment another object-field and sense organ, for
"example, sound and the ear, occur and condition the arising of
perceptual consciousness in the third moment "C."
a'his auditory perceptual consciousness in this third moment
I'g" would have sound as its object-support, while the concurrent
;thought concomitant, feeling, would be supported by the prior
yjsual object-field. In this way, moment after moment, percep-
t)lal consciousness and its associated thought concomitants
have different object-supports. This model then con-
tradicts the Sarvastivadin provision that perceptual conscious-
ness, or thought, and its associated thought concomitants must
share the same object-support.
This first criticism of the Darstantika position is valid only
if one accepts the Sarvastivadin model of cognitive functioning
through both thought (citta) and thought concomitants (caitta).
Each moment of experience contains one factor of thought
42 JIABS VOL. 11 NO. I
(citta) , or perceptual .in addition
least ten mcludmg
cepts, volItIOn, and so on. Smce thought and thought
tants exist as distinct real entities with different particular
teristics and activities, they can exist simultaneously and fun
. d d 1 . h .. h h . C IOU,-
m epen ent y WIt one restnctIOn: t ose t at Occur
must function having the same
DarHantlkas, however, assert that thought concomItants
exist as entities distinct from thought or perceptual
ness. They claim that the various mental functions
by these supposed thought concomitants are actually function";,;
of thought itself.
Therefore, each of the cognitive
indicated by the so-called thought or thought concomitant:Sj1"!'
occur only successively. 52 A particular object-field and
organ in one moment "A" would give rise to
scious?ess in subse.quent moment,-:v
:vould then
duce, m succeSSIOn, vanous mental functIOns With regard torhat, ,
object-field. Thus, from the perspective, the
tivadin criticism that perceptual consciousness and its associated";.:
thought concomitants have different object-supports is
founded. ';"
, ,'<'f.'
Throughout the Nyiiyiinusiira, Sanghabhadra raises a second";'
criticism of this successive perceptual model, a criticism
reflects the controversy concerning the possibility of a nonexis ...
tent object of perceptual consciousness. If perception is
sive, as the claim, then even in the case of the
externally directed types of perceptual consCiousness, the object-,\\
field would be past when its corresponding
sciousness arises.
The must then explain whya:;f
given moment of perceptual consciousness takes as its SUppatt'!
only the immediately preceding object-field, and not all
object-fields. If the claim that a present moment,'
of perceptual consciousness perceives only its own cause, thatJ
is, the immediately preceding moment, then they must explai
why an object-field of the distant past is not also consideredtoi.
be its cause. They might respond that the object-field of
immediately preceding moment is the cause because it alone'
has a connection (sarrtbandha) with that present moment of per:,
ceptual consciousness. However, since the immediately preced-,
ing object-field, like that of the distant past, is, in their
nonexistent, how can they justify a special "connection"
consecutive, moments? might defend
by replymg that the of the
precedmg moment acts as a condItIOn when that
moment of perceptual consciousness is on the point
'Sll . h' S ' h bh d '
In t IS case, ang a a ra argues, the
their initial claim that a present moment of percep-
perceives only past objects; for their reply
that a future visual perceptual consciousness, that is, one
i,e . ,.
is about to anse, perceIves a present object-field. Thus, the
theory of a successive perceptual model requires
explanation for the unique character of the immediately
feceding obj.ect-field, a that distinguish,es it fr?m all
:tpiher past objects and speCIfies It as the only possIble object of
In hIS concludmg cntICIsm, Sanghabhadra argues that the
'%proponents of this successive perceptual model have made their
[ppsition completely untenable by rejecting the existence of past
future factors.
When the sense organ and object-field exist,
Illieir corresponding perceptual consciousness has not yet arisen,
hence does not exist; when perceptual consciousness arises,
:)!the sense organ and object have already passed away, and hence
longer exist. Since no causal interaction can be established
::between a factor that exists and one that does not exist, the
,'previous sense organ and object-field can have no causal effect
perceptual consciousness. Thus, in Sanghabhadra's opin-
:ion, this successive model of perception leads to the conclusion
I either that perception occurs without its two requisite conditions,
contradicting the scripture, or that perception is con-
:ditioned by nonexistents, which, from the Sarvastivadin perspec-
'live, is absurd.
In either case, the position results
''in a denial of direct perception,56 and an implicit admission that
;,all perception depends upon a nonexistent object.
elV. The Possibility of a Nonexistent Object of Consciousness
The Sariputrabhidharmasastra is one of the first northern
(Indian Abhidharma texts that explicitly raises the issue of the
possibility of a nonexistent object of knowledge or perceptual
44 JIABS VOL. 11 NO.1
consciousness. 57 Regarding the possibility of knowledge
without an existent object-field, the text offers two
1) such knowledge is not possible;' or 2) particular kno ...i;.
. wed";:
of past and future factors can be saId to have a none
. .. f XIstenr"
object-field. The eXIstentIal status 0 past and future
o! knowledge is also discussed in the first .fascicle
Here, th.e author argues for the eXistence of
and future agamst an opponent,
allows the eXIstence only of present and of unconditio :,;,
(asarrtskrta) factors.
In defending his view, the authorcttF
numerous scriptural passages that refer either to the
activity of past factors, or to the
both past and factors. The author, m usmg these passages'f
to support the of past and future factors, implicitlY.")
assumes that only eXistent factors can exert causal efficacy,and;';
that knowledge or perception only with an
support. Maudgalyayana replIes that thought WIthout
tent object-support is indeed possible: precisely, that
which depends upon past and future factors.
If this is the
the author responds, the definition of thought or
consciousness given in the scripture must be rejected. Perceptu(ilk;
consciousness is defined as intentional awareness; that is, as that}
which is aware (vijiiniiti) of visible material form, sound, and.sO';
on up to mental factors (dharma). If the object-support wefkt;,
nonexistent, there would be no object of awareness and aware.!
ness itself would be impossible. Further, the scriptural passage'j;
stating that perceptual consciousness arises on the basis of
conditions, the sense organ and the object-support, would be.
contradicted. If a nonexistent object-support were allowed, these.
two conditions would not be present. Here the author again
assumes that only. existent factors can function as conditions. . .....
The Mahiivibhii$ii further develops the arguments of the
ViJiiiinakiiya; it supports the position that perception and knowl-
edge depend only upon an existent object-support,52 and that,
only actually existing entities can function as conditions.
opinion of the Mahiivibhii$ii is evident in an argument with the
and other schools concerning whether instances of
knowledge (jiiiina) or its objects are more numerous. 54 For the
Mahiivibhii$ii, all knowledge depends upon an existent object.
Further, knowledge itself can become an object for subsequent
ts of knowledge. Therefore, the objects of knowledge
numerous. However, the Daqtantikas apparently con-
instances of knowledge more numerous, since they assert
knowledge can depend upon nonexistent object-fields, in-
illusions, sky-castles, circles made from whirling fire-
and mirages.
, and other cases of nonexistent object-fields given by
tikas indicate that by the time of the Mahiivibhii$a,
of possible objects of knowledge or perceptual con-
whose existence was disputed exceeded that of simply
and future factors. For example, the cites the
as rejecting the existence of objects of mistaken
65 such as the snake that is cognized in place of the
rope, or the human being in place of the pillar, or the
that is seen to exist within one's own body The
also reject reflections and echoes,66 dream images,67
(maya) and magical creations (nirma1J,a),68 negative ex-
, such as impermanence,69 and denials.
In the opinion
such examples do not prove that knowledge
perceptual consciousness may depend upon a nonexistent
Instead, the concludes the con-
: because such things act as supporting conditions in the
. n of perception, there must in each case be some exis-
object-field. .
Among the Abhidharma texts, the
. Nyayanusara, and
fAbhidharmadzpa all contain extensive discussions of the possibility
:'ofa nonexistent object of perceptual consciousness. In these
texts, as in the Vijiianakaya, the impetus for raising this issue is
controversy concerning the existence of past and future
. factors. Each text, regardless of its particular stance on this
controversy, appeals to both scriptural references (agama) and
arguments (yukti) as reasons to support its position. The similar-
ity between the reasons and examples employed by the Abhidhar-
Nyayanusara, and Abhidharmadzpa, which have
documented historical connections, with those cited in the
TattvasiddhiSastra suggests a shared store of arguments and scrip-
tural references on the topic, a common source, or intentional
Among the reasons offered by these texts in support of the
existence of past and future factors, particular importan
, kId eelS'
accorded to the fact tnat now e geor perceptual conscious "Ii
d 1
. b' 7 . ness .
. epends on y an eXIstent 0 1 For
III the Tattvaszddhzsastra, out of the nmeteen reasons
existence of past and future factors presented by the
seven require the existence of an object of perceptual
The four are relevant 1) since thougJti
IS produced only wIth regard to factors that eXIst; 2) since menrn
perceptual consciousness takes the immediately past
of perceptual consciousness as its basis (iisraya) and may
upon future factors as its object-support (iilambana), if past ana'
future factors did not exist, mental perceptual consciousne;s"/
would have no basis or support; 3) since ordinary mental
tual consciousness cannot ap.prehend the fiv.e external
fields when they are present, If past factors dId not exist,
lection of those object-fields would be impossible; and 4) sint6
thought and thought concomitants cannot know
factors associated with them, or their co-present causes, these,:
various factors can only be known when they are past byi:'
subsequent moment of thought.
The Abhidharmakosabhii$ya
offers four reasons in suppbrfk
of the existence of past and future factors, two of which concedL
perceptual consciousness and its object-support: ,I) accordil1g!!
to scripture, "there is the arising of perceptual consciousness in.
dependence upon two,"75 that is, the sense organ and the objectS:
support; and 2) according to argument, since perceptual COh-',
sciousness operates only when there is an existent object-field,
if past and future factors did not exist, perceptual consciousness
of past and future factors would have a nonexistent object-sup:'
port, and hence, would not arise."
The Nyiiyiinusiira and Abhidharmadipa, even though pat:
terned closely on the Abhidharmakosabhii$ya, display a striking
similarity to the Tattvasiddhisiistra in their treatment of the exis-
tence of past and future factors. The Tattvasiddhisiistra begins!
its discussion with the following observations.
There are people who claim that factors of the two time
[of past and future] exist, and [others who claim that they] do
not exist. [Question:] For what reasons are they said to exist;fof
what reasons are they said not to exist? [Response:] Those [who
state that they] exist, [claim that] ifo factor exists, thought is
pro(iuced with regard to it. Since one is able to produce thought
with regard to factors of the twotime periods [of past and future],
one should acknowledge that they exist. [Question:] You should
first state the characteristic [i.e., definition] of existence. [Re-
sponse:] That range (gocara) upon which knowledge operates is
referred to as the characteristic of existence. .
For Sanghabhadra
also, establishing the existence of past
future factors first requires defining or stating the charac-
of existence (sallakJa1Ja, sattvalakJa1Ja). Once this defining
of existence is understood, he claims, the existence
f{qfpast and future factors will be universally accepted. Some
,ireachers, he notes, define the characteristic of existence as that
has already been produced and has not yet passed away.
Sanghabhadra, this is simply to identify existence with the
and thereby to assume, a priori, that past and future
:;Jtctors do not exist. Instead, Sanghabhadra offers the following
that will include factors of all three time periods:
be an object-field that produces cognition (buddhi) is the
:Ytiue characteristic of existence." Similarly, the Abhidharmadipa, 79
:i'defines the characteristic of existence (sattvalakJa1Ja) as "that of
iC,{yhich the indicative mark (cihna) is considered by cognition,"
explains it as follows:
An objective thing, whose own form is established by intrinsic
nature, is said to exist as a real entity when one observes its
defining characteristic determined by an observation of factors,
which is free from mistaken aspects.
Thus, the Nyayanusara and the Abhidharmadipa, like the oppo-
nent in the TattvasiddhiSastra, define existence as that which
,Serves as the object of cognition.
, These texts, however, admit several categories of existence,
'and hence, several categories of possible objects of cognition.
! first broadly distinguishes between existence as
a real entity (dravyasat), equated with absolute existence (para-
marthasat), and existence as a provisional entity (prajfiaptisat) ,
:equated with conventional existence (sar(l,vrtisat). Sanghabhadra
subsumes within these two categories of existence a third cate-
gory of relative existence (apekJa) recognized by some teachers,
including the author of the Abhidharmadfpa.
.. The category of includes factors
vIsIble matenal form or feelmg, whIch produce cognition
out depending upon anything else. These real entities exi tvi1;,
several modes (bhava): specifically, present existence as
nature (svabhava) activity (karitra), iillW
past or future eXlstence as mtnnslC nature alone. Since th';'':
. . . h h f . .IS
mtrmslC nature, w et er past, present, or uture, can serve',,!,.
the object-support for knowledge, past and future factors
can be said to exist.
The second category entities, such as apgt'
or an army, produces cogmtlOn only In dependence upon
entity. dep:ndence is twofold: 1) dependence
real entltles, as m the case of a pot, whICh
fundamental material elements (mahabhuta) of which it is
and 2) dependence first upon other provisional entities,
secondarily upon a real entity, as in the case of an army, whicH
depends first upon its human members, and finally upon
ultimate factors of which humans are composed.,.
This Sarvastivadin definition of existence in terms of
that give rise to cognition has significant implications for the'.
dispute concerning the possibility of a nonexistent object-sup:
port of perceptual consciousness. Since an entity's status as ah'
object-support condition for the arising of perceptual conscious?
ness is the very criterion by which the existence of that entitY.
is established, no such object-support can, by definition, be
nonexistent. However, it is important to note that the object
perceived may exist in different ways. As the Mahavibha$a makes,
clear, all conditions must actually exist as real entities, and the'
object-support, as one such condition, must also so exist.
Nevertheless, provisionally designated entIties may also become
the objective content of mental perceptual consciousness. Does
this then imply that the object-support condition may exist only
provisionally? The answer lies in the definition of provisional
existence: all provisional entities depend primarily or secondar-
ily upon a real entity. Thus provisional entities, exclusive of
their actually existing bases, cannot serve as the object-support
condition for the arising of perceptual consciousness. Instead,.
, the real entity upon which provisional entities depend serves as
the object-support.
t?F':;Objects Whose Existence is Disputed
the deny that conditions must exist as
they .reject Sarvastivadin of existence
;;,.:idconsider theIr use of It to prove the eXIstence of past and
factors groundless. This objection is pte-
in the Nyayanusara:
"This [definition] also does not yet
the true characteristic of actual existence because [we J
nonexistent are able serve as
that produce cogmtIOn. The Mahavzbha$a, Tattvasld-
Nyayanusara, and Abhidharmadipa all provide examples
of cognition claimed to be nonexistent:
1) products
error, such as two moons, and products of mistaken
such as a circle from a whirling firebrand, a
a hUI?an or the concept of. self; 2)
'ii'objects perceived m certam medItative states; 3) dream Images;
images, echoes, illusions, and magical creations; 5)
having a nonexistent object including: a) certain
t:ilegations, such as nonexistence, or the prior nonexistence of
b) affirmative expressions referring to unattested and
impossible objects, such as the horn of a hare, and
logically contradictory objects such as the thirteenth sense
:phere (ayatana) , or the son of a barren woman; 6) past and
Nture objects either cognized through inferential memory and
';;ilnticipation, respectively, or perceived directly.
';, The Sarvastivadins respond to these examples by indicating,
:in each case, the existent object-field that supports perception,
'and hence, cognition.
L;'.- ,
]'Sensory Error and Mistaken Cognition
Sensory error,87 such as the visual distortions produced by
ophthalmic disorders, or the image of two moons, results from
faulty sense organs and does not imply a nonexistent object-field.
For example, a visual sense organ afflicted by ophthalmic disor-
ders does grasp existent visual material form, albeit unclearly.
This then results in mistaken cognition with regard to that exis-
tent object-field. In the case of the image of two moons,
Sarighabhadra explains that the visual sense organ and that
initial moment of visual perceptual consciousness depend upon
or see the single existent moon. However, the clarity of per
tion is influenced by the sense @rgan, which is a conditio
equal with the object-field in the arising of perceptual
Th f h d
. d 'r h . OUs.
ness. ere ore, t e etenorate state 01 t e vIsual sense or::'
produces an unclear visual perceptual consciousness,
sul.ts in the cognition of two N
object-field, the smgle moon, actually eXIsts. ThIS IS eviddir
because no such cognition of the moon, confused or
arises where the moon is not found.
?f mistaken cog?ition
also not. arise
an eXIstent object-field. The CIrcular form m whIch a whirlin'
firebrand appears, or the human form in which a pillar appeart
do not, in themselves, exist as real entities (dravya). However
the cognition that apprehends them does have an existent object;
field: the individual points of light comprising the apparent
circle, or the form of the pillar. Similarly, regarding the
that the self exists in one's own body (satkayadr$ti) , 89 the existent
object-field is the five appropriating aggregates (upadanal!
kandha), which are then mistakenly cognized as self (atman),
as what belongs to self (atmiya). '1,
The Sarvastivadin explanation of these instances of sensor):
error and mistaken cognition assumes that cognition
either correct, that is consistent with the object-field, or mist<i,-
ken, that is deviating from the true character of the object-field
due to certain intervening conditions. However, whether
or mistaken, cognition only arises if supported by an existent
object-field. The status of cognition as correct or mistakenis
determined by whether or not that cognition apprehends the
object-field through a correct or a mistaken aspect (ahara). For
example, the conditioning influence of a visual sense organ
afflicted with an ophthalmic disorder causes the visual
field to be grasped unclearly, and produces cognition (buddhi)
characterized by a mistaken aspect (viParitakara). Similarly, cog-
nition of a whirling firebrand has the mistaken aspect of
larity, and cognition 6f the five appropriating aggregates ha.s
the mistaken aspect of self and what belongs to self.
in none of these cases does the object-field itself, in its true
nature, possess these mistaken aspects, nor is it nonexistent-gl;
Instead, error resides in the aspect of cognition through which
the object-field is apprehended.
this Sarvastivadin account of mistaken cogmtIOn, the
(akara), is used in a restricted sense as identical to
(prajiiakara) , and not. in the general. sense in which all
and thought concomItants may ?e SaId to an aspect
). and thought concomItants are. saId to
only In the general sense that the object-support IS
. 0:vn. activity.92 in the re-
sense IS IdentIfied wIth Inslght because It represents the
of the characteristics of the object-field in a par-
way as carried out by insight. 93 Insight characterized by
aspect may be the result of faulty sense organs,
iMfilernents, ignorance, or past action. However, this mistaken
is not associated with the initial moment of externally
;dlrected perceptual consciousness in which insight, though pres-
is not acute. Instead, it occurs only in the subsequent mo-
ijDent of mental perceptual consciousness in which there is dis-
.{criminative conceptual thought (abhinirupaty,avikalpa) , or dis-
:!trimination of the characteristics of the object-field.
r2.Meditative Objects
\:. The Sarvastivadins explain objects perceived in certain
iweditative states also as resulting from the application of a spe-
bfic aspect (akara) to an existent object-support. The meditative
objects in question are those perceived in such states as mindful-
with regard to breathing (anapanasmrti), meditation on the
repulsive (asubhii) , the four immeasurables (apramaty,a), the eight
Fberations. (vimok$a), the eight spheres of .mastery (abhibhva-
yatana), and the ten spheres of totality (krtsnayatana).95 All of
states occur as a result of attention through resolution
(adhimuktimanashara),96 by which practitioners intentionally per-
ceive the object in a certain way, or with certain aspects, in
accord with their resolve. For example, in the ,sphere of totality
with regard to the color blue (nilakrtsnayatana), perceptual con-
sciousness is concentrated on the color blue, and perceives every-
thing, everywhere, exclusively and totally as blue.
These as-
pects (ahara) of totality and exclusiveness are the product of the
practitioner's attention through resolution: that is to say, atten-
is directed in accordance with the practitioner's intention
to perceive the object-field, "blue," as total and exclusive. For
the 98 this perceivedblueness does not
. l' 1 f h d" . eXIst:
l_t ts sImp y t e me Itator.s resolution. 'l'ilz{;l
SarvastlVadms, however, as m the of mIstaken cognifl .. ... ;.
. h h h h . . . IOn
. IstmgUls t e aspects t. at c cognItIOn from
Ject-field that supports It. The practItIOner produces
. h h . l' d I' h .. IOn . ,
. WIt.t e aspects, to.ta lty an rough the powei%\'
of. hIS own. resolutIOn, but thIS cognItlOn IS supported by ari!;!
eXIstent a small patch of blue color.
,,<,: ',:
3. Dream Images .. . . ;.
The claim that dream images are nonexisten:Df
because the dreamer discovers when awakened that events ex.:;.:
perienced in a dream did not actually occur. For example,0l"le'.
eats and satisfies the senses when asleep and nevertheless wakes
up hun!p"y and The defines
as the SImple operatIOn of thought and thought concomitants:;
with regard to an object-support during sleep. Since the fi"d;
externally directed types of perceptual consciousness do not:'
arise in a dream, these object-supports, whether external matet.?;.
ial form, or internal mental factors (dharma), are
only by mental perceptual consciousness. 102 For Sanghabhadra,:
dreaming is the recollection of past object-fields that haveal-.. '
ready been experienced,103 but this recollection is influencedbyi:.
the mind's sluggishness during sleep. For example, in the case;!.
of dream images that have never been experienced as such,
the horn of a hare, the dreamer combines in one place separate;.;
waking memories of a horn and a hare. However, the object-sup";'
port for the dream image is not nonexistent; it is precisely those
past factors that support the various parts of the recollection
The dream images themselves result from several causes; the summarizes as follows: 1) they are
stimulated by other beings, for example, sages, spirits, gods,
and so on; 2) they result from previous experiences, or habitual
activity; 3) they presage a future event, that is to say, the dreamer
first perceives the indicative mark of an auspicious or inauspi-
cious future event in a dream; 105 4) they result from conceptual
thought, specifically, discriminative consideration that occurs in
the waking state when one is about to fall asleep; 5) they result
illness, that is to say; due to a conflict _or among
elements (dhatu, mahabhuta): the
sees a dream Image that conforms to the predommant
106 .

Illusions Magical
%0:',. For the reflected Images, echoes, IllusIOns and
.like objects of sen.sory error or
gt"ognition, medItatIve or dream Images, do not eXIst as
i;:perceived and have .no eXIstent support. For the
reflected Images, and so on, are themselves vaneties
0'bf existent material form. As the Mahiivibha5ii
explains, the
claim that reflected images do not actually exist
"because the object reflected does not itself enter the reflecting
hhrface. Similarly, echoes do not actually exist because all sound
"is momentary, and one moment of sound cannot travel to pro-
)aUCe a distant echo. The Mahavibha5ii responds that these re-
(fleeted images and echoes do indeed exist because they act as
supporting the arising of perceptual consciousness,
:and because they are grasped by the sense organs and, hence,
bn be included within the twelve sense spheres (iiyatana), which
ttbe Buddha declared to Even though the reflected image
and echo are not themselves the original visual material form
or sound, they still consist of material form derived from the
Original object. Indeed, material form can result from a variety
.of causes and conditions: for example, liquid may be produced
from moonlight on a moonstone (candrakiinta) , heat from cow
dung or from sunlight on a sun-crystal (silryakiinta), and sound
from hitting together the lips, teeth, tongue, and so on. These
. varieties of liquid, heat, and sound, though perhaps not pro-
duced in the conventional way, can be said to exist precisely
because they exert the activity ofliquid, heat, or sound. Similarly,
the material form of which a reflected image or echo is composed
actually exists because it has the function of producing cognition.
also argues at length for the actual exis-
tence of reflected images and echoes as varieties of material
form. The reflected image as such, like all composite entities,
exists provisionally, but also like all provisionally existing en-
tities, it has an actually existing basis. In the case of reflected
54 JIABS VOL. 11 NO.1
images and echoes, this basis is the fundamental materi
ments (mahiibhuta) and derivative material form (bhautika).;
of the. fundamental elemen.ts, which are
the ongmal object, reach the reflectIng surface to produc
. 1 fl d . 109 e .. e"
matena re ecte Image. ':'ii*).
In Sanghabhadra's argument for the existence of refie
are that
whIch the SarvastlVadms establIsh an entIty s eXIstence. Firs{i;
notes that his opponents allow only
. b' h fl d' b . am
nonexistent 0 as re ecte Ima?es, to
by perceptual conSCIousness. However, SInce no dIstInctions ca,i/;
be drawn among nonexistents, they should admit thatati';
nonexistent objects are apprehended. Further, '1 ...
the apprehension of an object as correct or incorrect, Which'"
result from the clarity of the sense organ, the distance
object, and so on, are only possible with regard to an
entity. Second, Sanghabhadra criticizes Vasubandu'slll assertion!;.
that the reflected image in no way exists, but is simply a particul{i;
efficacy of a collocation of conditions such that one sees thJ;:
reflection. Sanghabhadra asserts that a collocation (siimagrZ)
not exist as a real entity (dravya) , and therefore cannot be said:
to have its own particular efficacy. Further, he demands
Vasubandhu will not allow this collocation of conditions, thaI
is, the original object and the reflecting surface, to produceii:
existing reflected image. It is the nature of all sepa"::
rately existing conditioned factors to arise from a given colloca7
tion of conditions; similarly, a reflected image that arises from
such a collocation should be allowed to exist as a separate entity::'.
Third, Sanghabhadra
offers several reasons in support of the >,
existence of the reflected image: 1) most importantly, a reflected;
image satisfies the criterion for existence, that is, it serves as the
object-support condition for the arising of perceptual conscious-<,
ness; 2) like all actually existing conditioned factors, a reflected ' ..
image is apprehended only when that reflection is present, and'
the presence of the reflection is dependent upon the collocation
of its requisite conditions; 3) the reflected image is the objeCt-
support of visual perceptual consciousness, which, as an exter':.
nally directed type of perceptual consciousness, is without con-
ceptual thought, and therefore, must be supported by an actually
existing object-field; 4) like all material form, a reflected image
obstruct the arising of other material form (i.e., another
in the place; and 5) a image,
lilT" 'all eXIstent factors, IS produced from vanous separately
Jll':e d' .
:;;{:+'iting can ItlOns.
e'l{lS . - d' f h h '11 . d
SarvastlVa 10S urt er argue t at 1 uSlOns (maya) an
c:eations. (nirma1J,a), reflected .images
of eXIstent matenal form. MagIcal creatlOns conSIst
by magically creative
Itself resul:s from power (abhz-
',-)', developed 10 trance (dhyana).1l3 SImIlarly, 10 the case of
;In '
;illiisions (maya),114 the source of the illusion exists as actual ma-
:,t6rial form and results from techniques in illusion.
and Expressions Referring to a Nonexistent Object
Of all the examples raised by the to prove the
of a of percep.tual con-
sciousness, negatlOns and expresslOns hav10g a nonexIstent ob-
receive the greatest attention from both
;Sapghabhadra and the Abhidharmadipa. 115 Sanghabhadra focuses
treatment of the topic on an examination of the
ihiture and force of negating expressions. First, he cites a
objection that the scriptural passage, 116 "one knows
(asat) as nonexistent," indicates that knowledge
depend upon a nonexistent object-field. Sanghabhadra re-
i;*ponds: 117
What does this cognition take as its object-support? It is produced
supported by a specification (abhidhana)118 that negates existence;
it does not take nonexistence as the object-field by which it is
supported. That is to say, the specifying expression that negates
existence is precisely a particular specification that asserts
nonexistence. As a result, when cognition is produced with regard
to the expression specifying nonexistence, it forms the under-
standing of nonexistence. Therefore, this cognition is not pro-
duced supported by nonexistence. [Objection:] Isn't this specifi-
cation that asserts nonexistence [itself] existent; how can cogni-
tion deny it as nonexistent? [Response:] Cognition is not pro-
duced denying the expression itself; it is only able to cognize
[the object] specified by that [specification] as nonexistent. That
is to say, cognition is produced supported by an object-field that
negates existence, but it is not produced taking nonex:is' "'<f,
its object-field. [Objection:] Whitis this object-field
to negate [Response:] It is
that has ansen WIth regard to nonexIstence. Smce this
is supported by the specification as its object-field ..
. J " ' on d.'"
mands that one should not claIm that thIS [cogmtlOn] is
supported by a nonexistent object-field.
For Sanghabhadra, these negating expressions are oft!);
1) those that have an existent specified Ob.,t. .o ... :l..
. Ject,
(abhidheya) , as in the case of expressIOns such as
(abrahmar}a) , or impermanence (anitya); and 2) those
specified object does not exist, as in the case of expressions
as nonexistence (asat), or absence (abhava). In the first case,the
expressions non-brahman and impermanence implicitly refe?!'
to an existent object: a k:;atriya and conditioned factors,
tively. The specifying expressions, non-brahman or imperma
nence, negate the particular quality, brahman or
within the existent- specified object. This first type of negatiiji
expression produces knowledge in two stages: knowledge first':
depends upon the specification, non-brahman or imperm'<l2
nence, and cognizes that the negated quality does not
Next, it depends upon the specified object, the k:;atriya orcor(
ditioned factors, and cognizes that the negated quality doesn6;:
exist within the specified object. In the second case, the
sions nonexistence, absence, and so on, do not refer implicitly'
to an existent specified object. The resulting knowledge proc'
duced by these expressions depends only upon the specification
itself; it is aware of the nonexistence of that which is negated
in that particular context.
For Sanghabhadra, the existence of a specificatio
(abhidhana) does not demand the corresponding existence of a:
specified object (abhidheya).I2I As in this second type of negation, 1
the specifying expression, "nonexistence," itself exists and call
serve as the object-support for the arising of cognition. However;
the specification does not correspond to an existent specified,
object; that is to say, there exists no specified object, "nonexis-
tence," to which the specifying expression, "nonexistence,"
fers. If all specifying expressions required existent specified ob-
jects, then such expressions as the horn of a hare, the thirteenth.
sense sphere, and the son of a barren woman would also be,
'ired to have an existent specified object. 122 Such specifying
arise in accordance with one's own intentions as a
of both immediately preceding thought concomitants,
", volition (cetana) , and simultaneous thought concomit-
oS. The former acts as the causal arouser (hettlsamtltthiina), and
as in that moment .123
causes of the. speCIfyIng expressIOn are
prior and arousIng .thought.
i/,;ia not the speCIfied object Itself, the object specified by a given
need ?-ot exist. a can
a nonexIstent specIfied object, thls does notJustIfy the
that, in a similar fashion, cognition can be supported
object-field .. the bet,:een .the
.and the the object-field
as a condItion for the ansIng of cogmtIOn, and as such,
this explanation, Sanghabhadra implicitly responds to
that nonexistent object-fields must themselves serve
support for perceptual consciousness because they serve
object-referent in speech. As the Tattvasiddhisiistra
124 "there should be perceptual consciousness that de-
upon the hom of a hare, and so on. If there were not,
would one be able to speak of them?" According to
l'9a,tghabhadra's explanation, one can indeed speak of such
*6onexistent objects, but the specifying expression does not de-
upon a nonexistent specified object, but rather upon pre-
and simultaneous thought concomitants. Similarly, the
of this nonexistent object depends only upon the exis-
ltent specifying expression, and not upon any nonexistent
The and Abhidharmadipa next raise
:the case of the denial of putatively impossible or logically con-
'iradictory objects, as in statements such as "there is no thirteenth
fsehse sphere," or "there is no son of a barren woman." Such'
are to be explained in a way similar to Sanghabhadra's
of the second type of negation, illustrated by expressions
such as "nonexistence," or "absence." Just as in the affirmative
statement, "thirteenth sense sphere" (tray odasayatana) , so in its
idenial, "there is no thirteenth sense sphere," the object-field for
the arising of one's cognition of the expression is not a nonexis-
tent object, "no thirteenth sense sphere," but rather is si
the speech event itself (vaguastuiiuitra).125
The AbhidharmadZpa,126 explains in more detail the pro"F
.' d h b" h Cess.
by whIch negatIOn occurs, an teo t at conditi'fA;
. f h . . f . 1 . Ons",
the arismg 0 t e cogmtIOn 0 a partlcu ar negatIOn. A dei",';
. . h . () . nIaL:
cannot. negat; er an ,or (asatYINi
and of Itself. II thIS were possIble, a king S enemIes would beco<':
nonexist:nt simply as a result of declaring to be so,
a nonexIstent should, through double negatIOn, become ex!:,!
tent. Using the example of negating the horn of a hare
, e.
Abhidharmad'ipa concludes:
.. : 1:1

Neither the horn o.f a bull: nor the of a hare is
through that negative partICle. How IS It then? In dependentl
upon the cognition of a relation between the hare and the element:
of space, cognitions of a lack of relation between real
such as [that relation between] a bull and a horn, and so on,
indicated [in case of the hare and the horn]. " .
Therefore, in denying the horn of a hare, one does not
either an existent (i.e., the horn of a bull), or a nonexistent
the horn of a hare). Instead, one merely denies the relation'
between a bull and its horn perceived previously as it pertains'
to a hare's head, in which only a relation with space is perceived:.
The Abhidharmakosabha$ya raises one final example oLa.
negative expression: "there is the prior nonexistence of
Sanghabhadra 128 explains our cognition of this
accordance with his treatment of the first type of negation. He.
refers to a prior disagreement with Vasubandhu concerning the;
meaning of the phrase, "there is the prior nonexistence Of
sound; there is the subsequent nonexistence [of sound)" (asti
sabdasya prag abhavo 'sti pascad abhava ity ucyate).129 Sanghabhadra'
inquires whether the phrase, "there is the nonexistence," is used
. with regard to an absolutely nonexistent object, or with regard
to an existent object in which something else is negated. Only
the second option, Sanghabhadra claims, is possible. In tha.t
case, the phrase, "there is the prior nonexistence of sound,':
indicates that there is no sound within another existent entity.
The cognition of this prior nonexistence of sound then depends
upon that other existent entity in which sound is not found.
Specifically, it is the substratum (adhi$thana) , or the assisting
Ittfi;'TT"\stances in which sound has not yet arisen that serves as
f h .. f h h "h . h
or t e 0 t e p. .ase, t ere IS t. e
nonexIstence of sound. Thus, cogmtlOn of the pnor
of sound does indeed have an existent object-
that is, the substratum or assisting circumstances that

also defends the explanation attributed to
in the 131 that the percep-
of th.e pnor s?und
71'pbn the future sound ItSelf. For the SarvastlVadms, thIS future
does exist, and therefore may serve as the object-support
(foT the arising of a cognition. It does not, however, exist in the
as the present. A present factor exists characterized
ir:trinsic nat,:re activity, whereas past and future
eXIst only as mtnnslC nature. Therefore, even though
future sound exists as intrinsic nature, it is not heard be-
"l:illse, as future, it does not exert its activity. This future sound
;'ffiay be cognized due' to its existence as intrinsic nature, but
as it lacks activity, it is cognized as nonexistent.

:'{Cognition of Past and Future Factors
For the the most common experience of per-
consciousness without an existent object-field is that of
iwemory of the past, and anticipation of the future. The
:#kas claim that in these cases, the object-field does not exist
;'precisely because the past factors recollected and future factors
ianticipated do not actually exist. Nevertheless, no one would
that recollection or anticipation is possible. Therefore, the
conclude one must admit that thought and thought
concomitants can arise with a nonexistent object-support .
. , . For the Sarvastivadins, however, the mental perceptual con-
of past or future factors, like the perceptual con-
sciousness of present factors, must be supported by an existent
()bject-field. The explains recollection as follows:
)'through the power of habitual practice, sentient beings obtain
lmowledge homogeneous with a certain factor, which enables
Jhem to cognize [that factor when past] in the same way in which
!,twas previously experienced." Vasumitra further suggests three
;causes that make recollection possible: 134 1) securely grasping
60 JIABS VOL. 11 NO.1
the characteristic of the objectpreviously experienced;. 2)';
present oCcurrence of a series homogeneous with that p " ... \;\ .....
. . . rello""
expenence; and 3) not losmg mmdfulness. Therefore on' '.
. , ceoil"
and duly notes an object, one can recollect
later tIme when homogeneous knowledge, or knowledge S;.'.'l'.
' 1 . dk Id" Illliar
to t .at prevIOUS yexpenence nowe ge IS stImulated by., . .. : .."'.,;.',.
. b "1 b' b" .prac-,
tIce, ya SImI ar 0 or y CIrcumstances condu',;!J
to recollection. This recollection then takes the original
. b' ..
now past, as Its 0
The thought concomitant, mindfulness (smrti), which OCC/:}li
associ.ated all moments of th?ught, plays an
role m thIS process of recollectIOn. Whereas the
135 defines mindfulness simply as the
of the object-support,
mindfulness as the cause of the notation (abhiZapana) and
loss of the object-support.
The reasonfot:
Sanghabhadra's inclusion of notation in the definition of mindl;'
fulness becomes in a subsequent argument withtW
master Srllata concerning the existence of
ness as a separate thought concomitant occurring in
ment of thought. 137 Sanghabhadra asserts that notation occlig
in each moment of perceptual consciousness whenever thought
is aware of an object-field.
Therefore, the thought concomifr
ant, mindfulness, functions with regard to present as welLas
past factors. Indeed, as Sanghabhadra suggests, if there werg:
no present mindfulness in the sense of noting the
the recollection of previously experienced objects would be
possible. Mindfulness as the noting of present factors becomes
the cause of their non-loss; this notation, in turn, enables the
arising of subsequent recollection, which takes that past
as its object-support.
... '.
The uses several models to explain knowV
edge of future factors. First, one can infer a future event on;
the basis of the past and present. That is to say, one obseryes
the causal relation between past and present factors and infers,
that a given present factor will produce a certain future factor.'
Or, one anticipates a future effect on the basis of one's observa-
tion of a certain present characteristic or indicative
(phalacihna), which exists in the psycho-physical series as a con;-.
ditioned factor dissociated from thought (cittaviprayuk
Finally, future .(or factors may be perceived
dy as III the case of certam speCial types of knowledge, such
resulting from one's vow As
explains, there are two types of cognition of
"san d f Th fi 'dl .
an uture. e ITSt, unpure worl y cogllltlOn, can
f6nly recollect objects that have been Since
future .has not been expenenced, worldly. can
fcc"ticipate It only dimly. The second, pure cogllltlOn, observes
clarity past and future objects that have never been
In all these cases, however, the direct perception
r1nd resulting cognition of past and future factors demands an

(,vi ." Conclusion
Sarvastivadins counter all such examples of seemingly
;aonexistent objects of cognition by finding, in each case, some
to serve as the object-support. To summarize their ar-
Iglifilent, all perceptual cor:sciousness or kno",,:ledg.e arises only
':iI1ciependence upon an object-support, and thIS object-support,
a condition for the arising of that perceptual consciousness
'bi-knowledge, must actually exist. Since the cognition of such
;thirigs as illusions, dream images, past and future factors, and
's8 on, does occur, it also must have some existent object-support
.as its condition. The Sarvastivadin explanation of these cases
'further implies that the object-support need not exist exactly in
the manner in which it is cognized; hence, there may be a dis-
parity between the content of cognition and the character of
,the object-field in itself.
. Two principles are central to this Sarvastivadin position: 1)
fonditions or causes must actually exist, and therefore, the ob-
ject-support condition (alambanapratyaya) , as one of two condi-
.Hons required for the arising of perceptual consciousness, must
actually, in some manner, exist; and 2) the object-field may exist
in a way other than that in which it is cognized, and therefore,
,cognition or insight may apprehend the object-support with an
aspect (akara) that is not found in the object-field itself. The
however, dispute both these points. The Tattvasid-
dhiSastra, Vasubandhu, and Srnata clearly suggest that 1) even
62 JIABS VOL. 11 NO.1
thouoO"h the obiect-support may, in some sense, be consl'd ..:t!;.:'.j.:f! ..
... . ered'"
condition, it is not the generative cause for the arising of
tual consciousness, and 2) the object-support is the
" . COQ':
tent of cognition.
. Concerning the first point of disagreement, the
asserts that precisely because there is
1 .' WIt ..
out an 0 0ect-support, perceptua conSCIousness is not in e
. . ' . very
case, produced by two causes and condItIOns.
takes. more position: while stil!
cor:dltlo.ns for th: ansmg of perceptual as
scnbed m the scnpture, he remterprets the functIOn of the5h!}
ject-support condition. He distinguishes the object-suPPortc9i'};
dition (iilambanapratyaya) from ?enerative
and dalms that object, though. an object-suPPOrt;
condItIOn, cannot be consIdered a generative condition. Foi'
example, in the case of mental perceptual consciousness
ovijiiiina), the generative cause is that prior moment of mirld"
(manas) within the same mental series. The object of mental)
perceptual consciousness (dharma) is not a generative
rather a mere object-support. Vasubandhu notes that
object-support condition were also the generative
unconditioned factor, nirviir.ta, which cannot functionascj"
generative cause, could not become the object-support
ceptual consciousness.
Since the object-support is not the
generative cause, it need not exist. Therefore, Vasubandhu
dudes that such nonexistent objects as past and future factors
can still be .considered the object-support of perceptual con{
, 1""
The master, SriJata, presents a similar
Mental perceptual consciousness that depends upon past factors;
and so on, is not without an object-support, [but] it doesn6i
depend only upon an existent [object-support]. For what reasow
is this so? [It is so because] mental perceptual consciousness that.
is produced taking the five externally directed types of perceptual
consciousness as. its immediate contiguous condition (samanan,;:
tarapratyaya) is able to experience the object-field apprehended
by the prior [moment] of mind. Such mental perceptual con,.'
sciousness takes this [previous moment of mind] as its cause;if

object-support condition is the object of [that previous momenf;
of] the five externally directed types of perceptual consciousness.
[This previous moment of mind can be said to be its cause be-
cause] this [mental perceptual consciousness] is able to be pro-
duced only when preceded by that [moment of mind], and there-
fore "this [mental perceptual consciousness] exists or does not
exist in accordance with whether that [moment of mind] exists
or does not exist."148 However, this mental perceptual conscious-
ness does not depend only on an existent [object-support] because
at the time [of its arising] that object-field has already passed
away. It is not without an object-support because this mental
perceptualconsciousr:-ess exists does not exist in. accordance
with whether that [obJect-field] eXIsts or does not eXISt. Further,
when one recollects an object-field long past, [the recollection]
is produced in the present time taking the prior [moment of]
perceptual consciousness of that object as its condition because
this recollection falls into the same series [as the prior moment
of perceptual consciousness] and is produced through a mediated
sequence. Even though there are other conditions that give rise
to recollecting perceptual consciousness, it is produced only in
dependence upon that previous object.
'rhus for Srilata, a given moment of mental perceptual con-
;iciousness takes as its object-support that object apprehended
previous moment of perceptual consciousness. Though
previous object has passed away and hence, in SrHata's
::bpinion, is nonexistent, it can still be designated the object-sup-
;lJort condition because it satisfies the traditional formula defin-
a conditioning relation: "wher; this exists, that exists," and
It is important to note that Srilata interprets this formula
indicating a relation among successive conditions; he claims
a condition cannot be simultaneous with its effect, but rather
;tlust precede it. In this case, the existence or nonexistence of
'present perceptual consciousness depends upon the prior exis-
;!tence or nonexistence of this object-field. However, the genera-
tive cause of a present moment of perceptual consciousness is
previous moment of perceptual consciousness within its own
and not the nonexistent object-support.
i In another passage,149 Srllata clarifies the process by which
present mental perceptual consciousness apprehends nonexis-
tent past and future objects. Past and future objects are known
through a mediated process of successive causation; that is to
one infers the nature of past or future objects after having
apprehended the present. As Srllata states: 150
64 JIABS VOL. 11 NO.1

One is able to infer that a given:present effect is produced
a certain type of past cause, and this [past] cause in turn,
. d' h d' allSeS '
from a certaIn cause, an so on, Into t e, lstance past in '," s;,:.!
. h [d h '] . aWay
appropnate to eae case, an t us one attams [past ob";"
through inference just as one would present

This inferential knowledge of various past objects is
from causes are fou?d within .the series of
perceptual conSCIousness Itself. PrevIOus knowledge of a parti/:
lar type functions in a :nediated causal process to produce
ent knowledge, and thIS present knowledge can be said tOt<ik}'
the object-field of this particular previous knowledge as its oWR"
object-support. Thus, the cause of present recollecting kno,vi}
edge is a previous moment of knowledge within its own
and not the content of the present recollection.
cause the past object serves as the object-support for the
ous knowledge, it can, by extension, be considered the
also of the present recollection, even though it no longJ'(
exists. Srilata explains knowledge of past objects not yet
enced and future objects in the same way: one applies a proces;
of inference based on the knowledge .of causes and effects theft
one has already
rejects Vasubandhu's
the generative cause and the object-support condition, and his
identification of the generative cause as a prior moment withir'
the series of perceptual consciousness. In Sanghabhadra's
ion, the scriptural passage stating that perceptual consciousness
is produced in dependence upon two conditions, clearly indK
cates that the basis (iiSraya) , and the (iilambana)
are equally generative causes in the production of perceptual
consciousness. Since the object-support acts as a generative cause;
in production of perceptual consciousness, it must actually!
exist.. ,
In his criticism of Sri"lata's model of the arising of mental
perceptual consciousness in dependence upon past nonexistent
object-supports, Sans-habhadra focuses upon three major;
First, since Srllata does not admit that the five exter-
. nally directed types of perceptual consciousness and the object;
support they apprehend are simultaneous, even these five types
of perceptual consciousness. arise only when their object-fielq

away. The following moment of mental
would then be moments removed from its
Therefore, before Srilata discusses the
and future perceptual conSCIOusness,
first explam how It IS for the five
types ,of perceptual to perceIve a past
i,,,:cSn'existent object-field. Second, Snlata states that mental per-
is not without an
IS to an of the
(';;(!1l"ce of that object-support m some form. Srilata's pOSitIon
,te, . 1 h f
tWbuld then be eqmva ent to t at 0 the Sarvastivadins: a past
'[factor, though lacking activity, is not absolutely nonexistent like
flower, and yet it does not exist like the present, wl;ich is
Itharacterized by both activity and intrinsic nature. Third, Srilata
meaningfully appeal to the traditional formula defining
relations, "when this exists, that exists," and so on,
a model of mediated successive causation because he does
l/n6tallow the existence of past or future factors. According to
model, when the obJect-support exists, the perceptual
that apprehends it has not yet arisen, and when
perceptual consciousness arises, its object-support has al-
;"riaClY passed away. Similarly, in the case of mediated successive
Y<:llusation within the series of perceptual consciousness, the prior
moment of perceptual consciousness no longer exists
its subsequent effect arises. By maintaining a causal rela-
u6nbetween these successive moments, Sri:lata is, in effect, ad-
that there can be a causal relation of depe:r;dence be-
:fween existents and nonexistents, which neither Srilata nor
iSanghabhadra would accept.
Thus, Sanghabhadra concludes
object-support conditions act as generative causes coequal
:with the basis (asraya) in the production of cognition, and must,
in some sense, exist.
(' The second major point of disagreement between the Sar-
(vastivadins and the concerns the relation between
'the content of cognition and the object-support. In all of the
!examples of seemingly nonexistent objects of cognition cited
"previously, the assume that the object-support is
"the object as cognized, orthe content of cognition. For example,
:inthe case of the cognition of two moons, the object-support is
precisely the two moons; in the case of meditation on the spheres
66 JIABS VOL. 11 NO.1
of the obj.ect-support oEone's cognition of total"']""!,
exclusIve blueness IS the total expanse of blue. This Dar t"
. . . .
assumptIOn that the object-support IS the content of co ..
leads to. their the .
not eXIst because m these cases thIS content of cognition
actually existing counterpart. The Sarvastivadins,
sume that one's cognition in mental perceptual
may diverge fr?m the actual character of the
that serves as ItS support. In the case of the cogllltion of't'l'i,;
moons, one's cognition is supported by the single existent
and so on. {tJ;;i!
. assumpti.ans apparent
dIscussIOn of delIberatIve reflectIOn (vzmarsa) on the
characteristics of a perceived object. As Vasubandhu
"When all cognition has an existent object-support,
there deliberative reflec:ion with rei?ard to that [?bject-su'p2i:';
port]? He smce IS the VryJ!
c:ontent of cogmtIO?, If all obJect-.supports eXIst, no
tlon may be questIOned or Judged mIstaken. In other wor(!t\l
deliberative reflection and doubt are possible only so
nonexistent objects are allowed; mistaken cognition wouldtheif:
be cognition based on such a nonexistent object-support,>i.t'
For Sanghabhadra, such deliberative reflection or douDt:if}
only possible with regard to an existent object. The possibility\,
of investigating whether one's cognition of a particular
is accurate or mistaken (viparUa) does not demand that the
ject-support be nonexistent. On the contrary, distinctions, suchi
as that between accurate and mistaken cognition, are
only with regard to or among existents; existence and nonexis'':'
tenee share no characteristic by which they may be compare4P
Accordingly, it is only possible to distinguish accurate from
taken cognition when those cognitions have an existent objed-
support. Therefore, Sanghabhadra assumes that mistaken cog-
nition is not the product of a nonexistent object-support, but,
rather is a function of the accuracy of cognition. The factof
mistaken cognition demands not only an existent object,but
also the possibility that the object in itself and our cognition of
it differ.
Vasubandhu explicitly asserts this identity of the object-
support with the content of cognition in a discussion of
;. of existence of objects of memory and anticipation. 155
,lleasked how past and future factors that do not exist can
object-supports for perceptual consciousness,
responds that an object can be said to exist in the
ner in which it becomes an object-support. That is to say,
tfactors are recollected as "having existed," and, therefore,
described as "having existed;" future factors are antici-
<, as "coming to exist," and, therefore, may be described as
';ihi:ng to exist." Since objects are not recollected or anticipated
';xisting," one cannot claim that they "exist." Further, Vasu-
that .past factors are as they
expenenced III the present; that IS to say, the partICular
of a recollected object are not different from those
object when it was experienced in the present. If, like
one claimed that these past factors "exist,"
be forced into the contradictory position that past
are present, because they are cognized with the charac-
of a presently experienced object. Since, for Vasuba-
the object-support of cognition is the very content of
the object-support of the recollection of a past object
object in its form as presently experienced. But, since
?these factors are not present when recollected, we must conclude
fl '
i:tliatthe object-support has no existent counterpart.
responds by sharply distinguishing the
object-support from the cognition of that object-
For example, when one perceives a pillar as a human
the object-support, the pillar, does not exist as cognized,
\:ihat is, as a human being. Likewise, in the case of past and
Y" .
::Juture factors, though they are cognized as they were or will be
experienced in the present, they exist as past or future.
:;Therefore, precisely because the object-support need not be
:identical to the content of cognition, an existent object-support
1Jlaycondition the arising of an instance of mental perceptual
f6nsciousness whose cognitive content has no existing counter-
;; Thus, underlying these specific controversies between the
,Sarvastivada and schools on the existence or
,nonexistence of the object-support of perceptual consciousness
'are two fundamental points of disagreement: first, concerning
,the causal nature of the object-support, and second, concerning
68 JIABS VOL. 11 NO.1
, h
the relation between the object-support and the Content
nition. The controversies precipitated by these
would provide the background for the; extensive
of the Specifically, their
troverSles concernmg the locatIOn of the perceived ob" "i
. . I' h fd' .
eXlstentla status, t e 0 lrect perceptIOn (pratyaks
nature of knowledge as havmg aspects (sakara), or as
. -k-) d h d'" gWlih'
out aspects nzra ara ,an t e con ItIOmng relations
h' 11 .. d . rougH
w lC perceptIOn occurs were a antlClpate In these earl'&i
cussions. Y.

ADV Padmanabh S. Jaini, ed., Abhidharmadzpa with
., earr
Sanskrit Works Series, Vol. 4, (Patna: Kashi Prasad Jayaswal
. 77 .ro .
Institute, 19 )'".::?
AKB P. Pradhan, ed., ofVasubandhu, Tibetan
Works Series, Vol. 8, (Patna: Kashi Prasad Jayaswal Research
1967) (reprinted ed., 1975)...;; .. '.;
AKV Unrai Wogihara, ed., Sphutarthii Abhidharmakosavyakhya: The
Yas6mitra, (Tokyo: The Publishing Association of the
makosavyfikhya, 1932).;';::
A VB T28.1546. Tr. Buddhavarman.
Fa-pao Chii-she-Iun-shu, T41.1822. ; 'I:::
HTAKBA-p'i-ta-mo-chu-she-lun, T29.1558. By Vasubandhu; tr. HSiian-tsangO-;>
PAKB A-p'i-ta-mo-chu-she-lun, T29.1559. By Vasubandhu, tr. Paramartha.L:;
P'u-kuang Chu-she-lun-chi, T41.1821. 1:"
MA Madhyamagama, T1.26. Tr. Sanghadeva. '"
MN Robert Chalmers, ed., The MaJjhima-Nikaya, 3 Vols., (London: The Patr
Text Society, Henry Frowde, 1896-1899) (reprinted ed., London: The
Pali Text Society, Luzac and Company, Ltd., 1960).'
MVB T27.1545. Tr. Hsiian-tsang.
NAS Nyayanusara, T29.1562. By Sanghabhadra, tr. Hsuan-tsang.
SA Sarrtyuktagama, T2.99. Tr. GUlJ.abhadra. .
SN The Sarrtyutta-Nikiiya, 5 Vols., (London: The Pali Text Society, Henty
Frowde, 1884-1898) (reprinted ed., London: The Pali Text Society,'
Luzac and Company, Ltd., 1960).
T Junjiro Takakusu, Kaikyoku Watanabe, and Gemmyo Ono, eds.,
shinshfl Daizokyo, (Tokyo, 1924-1932). . .
TS TattvasiddhiSiistra, T32.1646. By Harivarman, tr. Kumarajlva.
VB T28.1547. Tr. Sanghabhadra.
VK Vijiianakiiya, T26.1539. Tr. Hsiian-tsang.
For discussions in early non-Buddhist texts see the Nyayasutra
Nyaya-Tarkatirtha, Amarendramohan Tarkatirtha, and Heman-
. eds., Nyayadarsnam, 2 Vols., The Calcutta Sanskrit Series,
(Calcutta: Metropolitan Printing, 1936-44) (reprinted ed., Kyoto:
1982; for perception and the time period of the object perceived
Vol. 1, p. 523; for the composition of objects of perception see
Vol. 2, pp. 1043-1058; for the existential status of the object of
, and false apprehension see 4.2.26-37, Vol. 2, pp. 1072-1089; for
to objects of questionable existential status including the whirling
3.2.58, Vol. 2, p. 897, eye disorders 4.2.13, Vol. 2, p. 1054, dream
4.2.34ff, Vol. 2, pp. 1083ff, a pillar seen as a human being 4.2.35
Vol. 2, p. 1087 .
. 2. Of particular interest to the topic of this paper is Dignaga's Alamba-
See Susumu Yamaguchi (in collaboration with Henriette Meyer),
Examen de l'objet de la connaissance," Journal Asiatique, Vol. 214,
pp. 1-65; Erich Frauwallner, "Dignaga's Text, Uber-
und ErHiuterungen," Wiener Zeitschrift fur die Kunde des Morgenlandes,
930), pp. 174-194 (reprinted in Gerhard Oberhammer and Ernst
(eds.), Erich Frauwallner Kleine Schriften (Wiesbaden: Franz Steiner,
pp. 340-360). For the perception of an object and error (bhranti) in
's Prama'l}asamuccaya see Masaaki Hattori, Dignaga, On Perception, Har-
Series, Vol. 47, (Cambridge: Harvard University Press, 1968),
(1.7 cd-8 ab), pp. 95-97 (notes #1.53-54), pp. 32-35 (II), pp. 116-120
#2.11-28). For the nature of the object of perception and the four
of in the section of Dharmakirti's
see Yusho Miyasaka, ed., Prama'l}avarttikakarika, in Acta In-
Vol. 2, (Narita: Naritasan shinshOji, 1971-72), Ch. 2, vss. 194-238,
72; Ch. 2, vss. 288-300, p. 80. For the commentary ofPrajnakaragupta
the Prama'l}avarttika see Rahula Sankrityayana (ed.), Prama'l}avartikabhashyam
. of Prajfiakaragupta, Tibetan Sanskrit Works Series, Vol. 1,
. Kashi Prasad Jayaswal Research Institute, 1953), Ch. 2, vss. 194-239,
279-303; Ch. 2, vss. 289-301, pp. 331-338. See also Hiromasa Tosaki,
cBukkyo ninshikiron no kenkyu (jokan) , (Tokyo: Daito shuppansha, 1979), pp.
"37-43); vss. 194-238, pp. 294-336; vss. 288-300, pp. 382-393. For a diseus-
,',sion of these four categories of in Jinendrabuddhi's commen-
"lary, Visalamalavati-nama-prama'l}asamuccayatika, on Dignaga's Prama'l}asa-
,'muccaya, see Kensho Hasuba, "Shoshukaku ni yoru Shuryoron no jigenryo
kaishaku ni tsuite," in Indogaku Bukkyogaku ronso, Yamaguchi Hakushi kanreki
kinen, (Kyoto: Hozokan, 1955), pp. 205-212. For the perceptual process in
Tattvasa7[!graha and KamalasIla's Pafijika see S.D. Sastri Tattva-
saf(lgraha, 2 Vols., (Varanasi: Bauddha Bharati, 1968)
Yo!. 1, pp. 448-493; Vol. 2, pp. 670-711; for error and
illusion see Vol. 1, vss. 1311-1328, pp. 479-484; for the two requisite condi-
,tions for perceptual consciousness and the possibility of perceptual conscious-
,ness without an object-support see Vol. 2 vs. 1787 a-b, pp. 614-616, and Vol.
2, vss. 1846-1848, pp. 630-631.
70 JIABS VOL. 11 NO.1
. . 3. There are a number groups within th: Sarvastivada sChd
tmgmshed by geographlCallocatIOn or textual and mstructionall' .?
. . meage
dence of these groups can be found m the vanous Sarvastivad':
quoted in and in the
early SarvastIvadm texts. See BaIYu Watanabe, Ubu Abzdatsuma ron nci'
(Tokyo: Heibonsha, 1954), pp. 111-155; Masao Shizutani, ShojoBuk{-
kenkyu, 1?78), pp. 137-140; Giyu Nishi, "Uhu
okeru hocchIkeI hI-hocchIkeI no shoshu no hakusetsu oyobi gakut6 Ii" ,
kyu," ShUkyo kenkyu (shin), Vol. 11, (1934-4), pp. 564-579, (1934-5) /:.7'/
789. For doctrinal among the ofthe
tary see Watanabe, op. CIt., pp. 253-494; Kosho Kawamura,
no shiryoteki kenkyu, (Kyoto: Dohosha, 1974), pp. 53-206. For doctrinali!'
ences among Sarvastivadin texts see .Kawamura, I
39-52; Taiken Kimura, Abidatsuma ron no kenkyu, Vol. 4, Kimura Taikenz
6 Vols., (Tokyo: Meiji shoin, 1937), pp. 271-324; Ryujo Yamada, Da&"oll
seiritsuron josetsu, (Kyoto: Keirakuji shoten, 1959), pp. 11 0-124.
4. In the case of the Sarvastivada and
"school" not in?icate dist!nct lineages or
affIlIatIOn, but rather SImply dIfferences m doctrmal mterpretation,
tional or textual lineage. See Shizutani, op. cit., p. 256. The history
and Sautrantikas are closely intertwined, with the
as the probable predecessor of the Sautrantikas. See Shizutani, op. cit., p.l?6'1:
pp. 140-147. Though the commentaries cite Sautrantika
tika views separately, references to the DarHantikas are far more
See Yamada, op. cit., p. 84. The later literature, however, refers
sively to the Sautrantikas. Note Yasomitra: "The are a
the Sautrantikas." ity artha!;. AKV p. 40Q.l!;';
Therefore, the correct identification of early masters as orSaUE'
trantikas, if such a distinction was justified in the early period, is exceedinglf;
difficult. See Junsho Kato, "Ibushurinron no tsutaeru Kyoryobu ni tsuite," Daif?;
Bukkyo kara mikkyo e, Katsumata Shunk yo Hakushi koki kinen ronshil, (TokY9:>
Shunjusha, 1976), pp. 175-198'
5. Vasumitra's Samayabhedoparacanacakra records the
sition that thought and thought concomitants must have an
(iilambana). T49.2031 p. 16.b.21-22; T49.2032 p. 19.a.16; T49.2033p3
2l.b.27-28. It does not refer to a difference of opinion between the Sarvas;:
tivada and Sautrantika schools on this issue. T49.2032 p. 17.b.2ff; T49.2032",
p.19.c.llff; T49.2033 p.22.b.20ff. Compare the Kathavatthu. (Arnold C;;;
Taylor, ed., Kathiivatthu, 2 Vols., Pali Text Society, Text series Nos. 48,49,
(London: The Pali Text Society, 1894, 1897) (reprinted ed., 1979), 9.4-7 pp:
6. The later Sarvastivadin tradition includes seven texts in their early.
Abhidharma canon. Following the dating of Hajime Sakurabe (see Hajime
Sakurabe, Kusharon no kenkyu, (Kyoto: Hozokan, 1971), pp. 41ff), to the earliest
period belong the Sang'itiparyiiya by (Ch. Sariputra, tr.
tsang, and the Dharmaskandha by Sariputra (Ch.
tr. Hsiian-tsang, T26.1537). To the next period belong the Vijiiiinakiiyaby
, (tr. Hsuan-tsang, T26.1539), the by PUrI).a (Ch. Va-
.. tr.Hsuan-tsang, T26.1540); and the PraJnaptlSiistra by Maudgalyayana
tr. Dharmapala ?, T26.1538), followed by the Prakara1Japiida
gU!Ilitra; (tr. and T26.1542). The
ecent of the seven texts IS the jnanaprasthana by Katyayaniputra, (tr.
Jadeva, and T26.1544). See AKV p. l1.26ff.
discuSSlOns of the datmg of these texts see Watanabe, op. cit., pp.
: Erich Frauwallner, "Abhidharma-Studien, II," Wiener Zeitschriftfur die
'Siid- und Ostasiem, Archiv fur Indische Philosophie, Bd. 8, (1964), pp. 59-99.
The three Chinese translations, listed in the order of translation,
he (tr. Sanghabhadra ?, T28.1547), the Abhidhar-
(''1rEZihhffi.iiSiistra (tr. Buddhavarman, T28.1546), and the (tr.
.. a summary of the controversy the
and doctnnal dlstmctlOns. among these three translatlOns see Kawa-
cit., pp. 53-206, pp. 118-120, 206. .
8. P. Pradhan (ed.), of Vasubandhu, TIbetan
Works Series, Vol. 8, (Patna: Kashi Prasad J ayaswal Research Institute,
!i1975), the by Vasubandhu (tr. Paramartha, T29.1559,
IJdHsiian-tsang T29.1558), and U. Wogihara (ed.), SPhu(iirthii Abhi-
the Work ofYaiomitra, 2 Vols., (Tokyo: The Publishing As-
:Jqthition of Abhidharmakosavyakhya, 1932). The Nyiiyiinusiira by
!s.inghabhadra (tr. Hsuan-tsang, T29.1562), and a partially extant commentary
Nyiiyiinusiira, the Shun-cheng-li-lun shu-wen-chi, by Yuan-yu (Dai Nippon
$kuz6ky6, 1.83.3). Padmanabh S. Jaini (ed.), Abhidharmad'ipa with
Sanskrit Works Series, Vol. 4, (Patna: Kashi Prasad Jaya-
Research Institute, 1977).
9. In the Nyiiyiinusiira, views are most often represented
illy the teacher, Sthavira. Later sources identify Sthavira as the Sautrantika
Srilata. See P'u-kuang 9 p. 172.a.8-1O; Fa-pao 9 p. 604.a.5-6; K'uei-
chi, Ch'eng-wei-shih-lun shu-chi, T43.1830 4 p. 358.a.9ff. However, from numer-
lbus references in the Nyiiyiinusiira it is clear that Sanghabhadra considers
;Sthavira to be a Damantika. See NAS 3 p. 347.b.6-7; 11 p. 390.c.20ff; 14 p.
'i12.c.9ff; 18 p.442.a.25ff; 19 p. 445.c.3-4; 25 p.482.a.5ff, b.I-2, b.20ff, c.1-3.
See also JunshO Kat6, "Kyoryobu Shurirata (ichi)," Bukkyogaku, VoL 1, (1976),
cpp: 45-65. J unsho Kato, "Notes sur les deux maitres bouddhiques Kumaralata
etSrilata," in Indianisme et Bouddhisme: Melanges offerts a Mgr Etienne Lamotte,
Publications de l'Institut Orientaliste de Louvain, 23, (Louvain-Ia-neuve: In-
stitut Orientaliste, 1980), pp. 197-213.
k 10. For textual references to the dating of the TattvasiddhiSastra, and
tei Harivarman as the author of the Tattvasiddhisiistra and as a student of the
DarHantika-Sautrantika master, Kumaralata, see Kato, "Notes sur Ie deux
martres," pp. 199-200. Paramartha identifies the TattvasiddhiSiistra as repre-
senting the Bahusrutiya school. See Chugan Chozen's Sanrongengi kennyitshu
(1'70.23005 p. 460.c.8ff, especially c.21), which cites autocom-
mentary on his translation of Vasumitra's Samayabhedoparacanacakra
(T49.2033), See Paul Demieville, "L'origine des sectes bouddhiques d'apres
Pararnartha," Melanges chinois et bouddiques, VoL 1, (1931-32), pp. 16ff. How-
ever, there are frequent points of doctrinal similarity between Darst-
Sautrantika positions and those of the Tattvasiddhisiistra. Chi-tsang i
lun hsuan-i (T45.1852 1 p. 3.b.16ff, especially b.24ff) cites various

as to the school of th.e and notes the
between DaqtantIka or SautrantIka views and those of the
See also Shoson Miyamoto, Dai:jo to ShOjo, (Tokyo: Yakumo Shoten,
152-168. Though the exact date of the Tattvasiddhisiistra is not kn C PP;.}
historical references agree that Harivarman precedes Vasubandhu.
11.. N? attention will be later .and
of the early Sarvastl:adm and or
posItIons. For example, among Buddhist sources, see y,'i;l
Embar Krshnamacharya, of
No. 94, (Baroda: ?riental Institute: 1942), p. 34.27ff, p.
24; YUlchl KaJlyama, An Intraductwn ta BuddhISt Ph.zlasaphy, An Annotated Traist?
latzan of the of the M M emOlres ofthe F acuIty of Lette "i
Kyoto University, No. 10, (Kyoto: 1966), p. 62, note #148, pp. 139-140 t\
144. See also Yiiichi Kajiyama, "Sonzai to chishiki: Bukkyo tetsugaku
no ronso," Tetsugaku kenkyu, Vol. 43, (1966 #6), pp. 207-236, Vol. 43, (196;'
#11),.pp .. 1-28; Gadjin "Shoengyosomon no ichi mondai," in
to YUlShzhz, (Tokyo: Iwanaml shoten, 1978), pp. 373-388. Among noil;
sources, see Madhava's V.S. Abhyankar (ed.);:;
Sarvadarsanasa'r[Lgraha, Government Onental Senes, Class A, No. 1
Bhandarkar Oriental Research Institute, 1924) (reprinted ed. 1978); for
characterization of the Sautrantikas as maintaining biihyarthanumeya
Sarvastivadins as maintaining see 2.41-44, p. 19; for
Sautrantika theory of perception see 2.220-263 pp. 33-37, 2.268-371 p. 46j:
See also Ensho Kanakura, "Gekyo no bunken ni mieru Kyoryobu
Indogaku Bukkyogaku ronso, Yamaguchi Hakushi kanreki kinen, (Kyoto:
1955), pp. 55-68; Yiiichi Kajiyama, "Setsuissaiubu no shiso 0 megutte,"
kyogaku semina, Vol. 25, (1977), pp.
12. See AKB 3.85 c p.176,12-13, 4.2 c p.193.2; MVB 3 p.12.b.4ff, 39'
p.200.a.29ff, 76 p.393.c.14ff, 93 p.480.a.26-27; NAS 13 p.407.c.19ff, 14
p.409.b.2ff, 14 p.410.a.4ff, 15 p.417.b.29ff, 15 p.419.c.2ff, 15 p.421.b.22ff,i
18 p.437.c.3ff, 19 p.447.a.l0ff, 52 p.631.c.5ff, 52 p.633.b.27ff, 52 p.634.a.26.\
13. See AKB 3.32 b p.146.4ff; NAS 6 p.365.a.27ff, 15 p.417c.12-":
p.421.c.24, 18 p.440.a.23-24, 20 p.452.a.16ff, 22 p.467.a.22ff, 25,
14. SA 13 #306 p.87.c.26ff; SN 12.43 Dukkhasutta, 44 Lokasutta, 45.,
Natikasutta, Vol. 2, pp. 72-75. cakkhu'r[L ca paticca rupe ca uppajjati ca:'
kkhuviiiiiiir;a'r[L. See also MA 54 #201 p. 767.a.24ff; MN 1.38 Mahiitar;hasa-
nkhayasutta Vol. 1, p. 259. For references in Abhidharma texts see San:
gZtiparyaya T26.1536 15 p.429.a.15ff; Dharmaskandha 10 p's01.b.9ff,
10 p.502.c.20-21, 11 p.507.c.25; VK 3 p.545.b.24; Dhatukaya T26.1540 (shang)<
p.615.c.4; Prakarar;apada T26.1542 2 p.699.a.4; MVB 16 p.79.b.20; AKB 5.25
b p.295.16; NAS 2 p.338.c.22, 51 p.627.c.17, 57 p.658.c.8.
15. MA 7 #30 p.467.a.3ff; MN 1.28 Mahiihatthipadopamasutta Vol. 1,
p. 191.
1'6. MA 47 #181 p. 723.b.16ff, p. 723.c.14ff; MN 3.115 Bahudhatuka-
yol. 3, pp. 62-63.
,17. A distinction between homogeneous (sabhaga) and partially
(tatsabhaga) sense organs and object-fields was developed in
distinguish those that have functioned, are functioning, or will func-
a rnorne'nt of perception (i.e., homogeneous), from those that do not
, but are nevertheless of the same nature as those that do (i.e.,
Jly homogeneous), This category of the partially homogeneous includes
b' fi Id h . d . .
sense organs or? Ie s. t at an pass away WIthout performmg
;1,1,",'. particular functlOn of graspmg or bemg grasped, as well as those future
"I theIr b' f' Id h '11 . h
,:j:!": e organs or 0 Ie stat WI never anse. T e dharma dement, as the
t-field of mental perceptual consciousness, is exclusively homogeneous
ce .it is considered unreasonable that a mental factor will never be ap-
or arises and passes away without being apprehended. AKB 1.39
NAS 6 p.362.a.7ff; AKB 1.42 b p.30.5-7; NAS 6 p.364.a.26ff;
'!M\TB 71 p.368.a.10ff, p.371.a.8ff.
*i:i 18. Perceptual consciousness (vijiiiina) is identified with thought (citta),
mind (manas), and is then described as occurring simultaneously with
concomitants (caitta), each of which carries out its own specific mental
i0Jiullction. Thought and thought concomitants are said to be associated (sa'l'{!pra-
because they to basis. (asraya):
aspect (aMra), tlme penod (kala), and the smgular mstance of theIr
(dravya). AKB 2.34 a-d p.61.22ff; AKV p.141.8.ff; NAS 11
.c.14ff; MVB 16 p.80.b.25ff. For an enumeration of the 46 thought
with which thought may be associated according to the Sarva-
school see AKB 2.23 p.54.3-2.33 p.61.19; NAS 10 p.384.a.8-11
iii;' 19. See MVB 13 p;61.c.7ff and AVB 8 p.51.b.24ff where four views
the proper locus of grasping the object-field are presented: 1)
.;ipharmatrata claims that visual perceptual consciousness, and not the eye, sees
color-form; 2) claims that the insight (prajiia) associated with
perceptual consciousness sees; 3) the claim that the com-
;"plete collocation (samagrz) of causes, including the sense organ, and so on,
and 4) the Vatslputrlyas claim that only one eye sees in each successive
":(ll0ment. The MahavibhaJa replies that the sense organ, speCifically both eyes
tfunctioning together, sees form. See also AKB 1.42 p.30.3-43 b p.31.25;
';.AKV p.80.10ff; NAS 6 p.364.a.23ff; and ADV p.31.1ff; MVB 95 p.489.b.28ff.
tThe AbhidharmakosabhaJya (AKB 1.42 c-d p.31.12) identifies this
'view as that of the Sautrantikas.
20. See AKB l.l6 a p.l1.6ff. vijiiiina'l'{! prativijiiaptil; ... viJaya'l'{! viJaya'l'{!
,'prativijiiaptir upalabdhir vijiianaskandha ity ucyate. AKV p.38.22ff; NAS 3
;p.342.a.15ff. See also NAS 11 p.396.b.6ff, 25 p.484.b.17ff; MVB 72
p.371.b.22ff. Sanghabhadra (NAS 3.p.342.a.17ff; Samayaprad'ipika, T29.1563
:2p.783.b.26ff) clearly delimits the functioning of perceptual consciousness
:to that of apprehending the generic characteristic of the object-field, thereby
:distinguishing the activity of perceptual consciousness from that of its as-
thought concomitants (caitta), which apprehend the specific charac-
74 JIABS VOL. 11 NO.1
teristics of the object-field. See also NAS'il p.390.c.9-11, 11
P'u-kuang 1 p.26.a.3ff; Fa-pao 1 yii p.486.c.7ff; ADV #120 b
Kyokuga Saeki, Kanda abzdatsumakusharon, Vol. 1 (1886) (reprinted ed K 13.; '.
Hozokan, 1978), p. 29. in translating this section of the
akofabhiya, perhaps under the mfluence of the Nyayanusara, modifies "
labdhi" with "tsung" meaning grasps in general, or grasps the generic ch
teristic of the object-field. (HTAKB 1 p.4.a.21; contrast with
21. NAS 11 p.395.a.28ff; NAS 3 p.342.a.18ff; VK 11 p.582.c.20ff F)
the distinction between those thought concomitants associated with m' .or
dh . d 'hh ental'
perceptua conSCiOusness an t ose aSSOCIate WIt t e other five type. ,::.;).
perceptual conSCiOusness see NAS 29 p.506.c.7ff; VK 6 p.559.b.27ff.:>:>
22. AKB 1.10 d p.7.18ff; AKV p.2!.29ff;. MVB 13 p.65.a.12ff;127'
p.665.b.lff. As these passages suggest, thIS partIcular characteristic of ih)'>
as a generic not to be confuseJ:
wIth the common charactenstlc whlCh IS apprehendedoriI;
by mental perceptual consciousness.:,
23. See NAS 28 p.501.b.24-25, 4 p.352.a.20-21. Sarighabhadra
4 argues at length against the
master, Srllata, who claims that the five externally directed types of perceptuarF
consciousness depend upon object-fields that do not exist as real entites. Sriiata'
claims that single atoms are not the object-support of perceptual
they do not constitute the .content of perception. The five externallY\.
dIrected types of perceptual conSCiOusness rely only upon composites
of atoms, and these composites, as such, do not exist as real entities. Therefore,,:i
the five externally directed types of perceptual consciousness do
prehend actually existing object-fields. Sarighabhadra responds by
guishing term "composite" (ho-ho, samagrz, sa'Y(tghiita, samnipata, sa7[!hata ?),
used by Srllata, from aggregation (ho-chi, sa'Y(tcita ?), Sarighabhadra clairns,
that atoms form an aggregation, not a composite, and this aggregation the'n'
allows direct perception to occur. (See also NAS 32 p. 522.a.5-10.) The actuallY.'
existing object-field that causes perception is still, however, the indiVidual.
atom. (See NAS 4 p.352.a.18-19.) This composite (ho-ho), as proposed by
Srllata, exists only provisionally, and hence is apprehended only by mental'
perceptual consciousness. Sarighabhadra's attempt to salvage the Sarvastivadin,'
theory that atoms in aggregation are the object-field of the five externally
directed types of perceptual consciousness by distinguishing ho-ho from ho-chi
constitutes an innovation not found in the commentaries. See MVB.)
13 p.63.c.22-25, 121 p.632.a.24-26; ADV #317 p.277.15ff. See also Sylvain'
Levi, Vijiiaptimatratasiddhi (Vi7[liatika), Bibliotheque de l'ecole des hautes etudes, ,.
VoL 245, (Paris: Librairie Ancienne Honore Champion, 1925), vs. 11 p. 6-7;
Louis de la Vallee Poussin, Vijiiaptimatratasiddhi, Vol. 1, (Paris: Librairie
taliste Paul Geuthner, 1928), p. 44 (notes), p. 45, #1; Junsh6 Kato, "WajulQ
wag6-Ubu to Kyobu no busshitsu no toraekata," Buzan kyagaku taikai kiyo,.
'Vol. 1 (not available to me).
24. For the simultaneity of the sense organ and object-field see AKB,
1.23 a p.15.24ff; AKV p.50.22ff; NAS 3 p.345.c.9ff. For the simultaneity of
'1 ': ',: -
.. obiect-field and perceptual consciousness see AKB 1.44 c p.34.3ff; NAS
e J 7
.374.a.2lff, 8 p.3 4.b.9ff, 4 p.351.b.29ff. .
:;):t'P, 25. AKB 1.29 c p.19.16ff; AKV p.59.4ff; NAS 4p.348.b.5ff. The obJect-
is defined. as that with to :vhich a carries out its
(kantra); the object-support \alambana) IS that IS apprehended
.:tbthought and the thought concomItants. For a companson of the usage of
artha, viJaya, gocara, and alambana in Abhidharma texts see Akira
H;Birakawa, "Setsuissaiubu no ninshikiron," Bungakubu kiyo, Hokkaido daigaku,
;11VoL2, (1953), pp. 7-8; Kyodo Yamada, "AbidatsumaBukkyo niokeru ninshiki
Indogaku Bukkyogaku kenkyu, Vol. 5, (1957-1), pp. 184--187. Un-
Hsuan-tsang does not always distinguish iilambana from viJaya in
translations, making the clarification of Sanghabhadra's understanding of
(thedistinction exceedingly difficult.
26. NAS 15 p.420.c.21-p.421.a.11; AKB 3.32 b p.145.15ff; MVB 16
1. According to the Sarvastivadin system of six causes. (hetu) and
:0/four condItIOnS (pratyaya), the co-present cause (sahabhuhetu), assoCIated cause
the efficient variety of the general cause (kiira1J.ahetu) , the
condition (iilambanapratyaya) , and the sovereign condition
j:l;Zddhipatipratyaya) may be simultaneous with their effects. Though there is some
\ijidifference of opinion (see MVB 16 p.79.a.28ff), generally, according to the
of six causes, the sense organ and object-support are both designated
,:t:fficient general causes, while according to the system of four conditions, the
,I;bbject-support is the object-support condition and the sense organ is the
condition. See MVB 20 p.104.a.4ff; NAS 15 p.417.a.15ff, 18
i::p.438.a.13ff, 20 p.449.c.16ff; TS 2 #17 p.251.a.20-23 .
. :i. 27. NAS 8 p.374.c.2ff. Sanghabhadra (NAS 73 p.736.a.9ff) admits
,'t"three types of direct perception: 1) that through the sense organs (i-ken-hsien-
i/Ziang, ?), which grasps the five external object-fields through
:J'the five sense organs; 2) that through experience (ling-na-hsien-liang,
?), which is the present occurrence of thought and the
:0Jhought concomitants of feelings, concepts, and so on; 3) that through cogni-
(chueh-hui-hsien-liang, ?), which attains the particular and
::coxnmon characteristic appropriate to each factor. This third type of direct
,',;perception arises in dependence upon the first two. The first among these,
'iF direct perception through the sense organs, demands the simultaneity of the
7.:sense organ, object-field, and perceptual consciousness.
28. AKB 1.17 a-b p.11.2lff; NAS 3 p.342.b.1lff. This mental organ
.. also serves as the basis (iifraya) of each of the five externally directed types of
perceptual consciousness, which then have two bases: the past mental organ
their respective present sense organ. See MVB 71 p.369.c.14ff; AKB 1.44
p.34.6ff; NAS 8 p.374.a.24ff.
29. NAS 7 p.366.c.4ff; MVB 71 p.369.c.27-29.
30. NAS 6 p.365.c.2ff. The mental organ, as the immediately preceding
>.moment of perceptual consciousness has as its object-support the object ap-
;:prehended in the preceding moment.
o 31. AKB 3.30 c-d p.143.25ff; AKV p.305.19ff; NAS 29 p.506.c.3ff;
Shun-cheng-li-lun shu-wen-chi, (Dai Nippon zokuzokyo, 1.83.3), 29
p.262.d.6ff; VK 6 In these contact
mental perceptual conSCIOusness (manalJsaT!!SpaTsa) IS explained. This WIty';:
contact with the object-support is called designation (adhivacana) bIUental" .
ecau "
names are the primary object-support of mental: perceptual conscious . se:
. . ness "
because mental perceptual consCIousness operates on Its obiect th "or.;2
speech. . ,;
32. AKBi.33 a-d p.22.19ff; AKV p.64.22ff; NAS 4 p.350.b.5ff;
42 p.219.b.7ff; AVB 23 p.169.b.5. . "':,
_ 4 p.349.a.23-24, 4 p.350.b.llff; MVB 42 See
(MVB 42 p.219.b.7) where conceptual thought In its intri' i.
is with both initial inquiry (vitarka)
InveStlgation (vzciira). For the dIstInctIOn between vztarka and viciira see NAS:
11 p.393.c.29ff. , .
34. MVB 72 p.374.b.5ff; NAS 4 p.349.a.2lff, 4 p.350.b.8; AKB l.3F
a-b p.22.20-23; AKV p.64.29ff.
35. NAS 4 p.350.b.17ff; MVB 42 p.219.b.l0ff. .
36. NAS 4 p.349.a.16ff.
37. AKB 1.48 a p.36.2lff; NAS 8 p.377.a.lff; AKB 2.2 a-b p.39.7fk.
NAS 9 p.378.a.12ff; MVB 9 p.44.b.3ff. .
38. AKB 1.23 a p.15.25ff; AKV p.50.26ff; NAS 3 p.345.c.12; MVB 9
p.44.b.llff. .
39. It is important to note that this ability to apprehend all factors
restricted. Mental perceptual consciousness may not apprehend itself, thought.)
concomitants that are associated with it, and those factors that are its co-present "
causes (sahabhuhetu). These factors may only be apprehended by a subsequent:.
moment of mental perceptual consciousness. See MVB 13 p.65.b.3ff,71'
p.370.c.9ff; NAS 7 p.370.b.22. For these restrictions on knowledge seethe
discussion of the process by which one knows all factors as non-self: MVB9
p.42.c.9ff; AKB 7.18 c-d p.404.22ff; AKV p.630.3lff; NAS 74 p.742.a.27ff..'
TS 15 # 191 p.364.a.4ff.
40. NAS 4 p.350.c.20ff, 4 p.35l.a.23-29; MVB 21 p.l09.b.25.
Sanghabhadra (NAS 58 p.666.a.7ff) identifies entities that exist conventionally
(sarrtvrtisat) as composite entities (ha-ha). There are two such types of compo:,
sites: 1) those like ajar that can be broken into finer pieces by another object
with the result that the conventionally existing jar is destroyed, and 2) those
like water that retain their original conventional nature even when divided,
into smaller amounts; this kind of conventionally existing entity can, neverthe-',
less, be analyzed by insight !prajiiii), which resolves it into its constituent factors.,
When these two types of composite entities are thus broken or analyzed, the
cognition of their composite nature no longer arises. However, these composite
entities are still said to exist conventionally because they have provisional
existence as designated by worldly or conventional names. See AKB 6.4 a-d
p.333.23ff; AKV p.524.8ff; P'u-kuang 22 p.337.b.13; Fa-pao 22 p.728.a.4.
Louis de la Vallee Poussin, "Documents d'abhidharma: les deux, les quatre,
les trois verites," Milangeschinais et bouddhiques, Vol. 5, (1936-1937), pp.169ff. "
4l. MVB 13 p.6l.c.l0-ll. The Mahiivibhiii (MVB 13 p.6l.c.16ff)re-
sponds that this position is not reasonable. If, for example, in the case of
1 p' erception, the collocation had the power of sight, it should see at all
:1 'es, since there IS no tIme when these three are not assembled. The exact
of this collocation as in the is unclear
sense organ, perceptual or all
('llditions), but the purpose of the pOSItIOn IS to refuse to deSIgnate
isolated factor as having prominent causal capability in perception. See
Ji;ijJv p.31.6ff. '" _ . ..
ft.:> 42. NAS 7 p.367.b.24ff. The Abhzdharmakosaohayya attnbutes thIS theory
Sautrantikas (AKB 1.42 c-d p.31.12ff; AKV p.82.27ff). See also ADV
?W44p.33.7ff. ,
.J 43. See NAS 25 p.484.b.19ff where the master, Srilata
the Sa.rvastivadir:- thesis perceptual. is accord-
!:iIlg to its umque of aware (v7Janatz). HIS IS to
:. that perceptual consCIousness eXIsts as an agent, or as a dIstmct factor havmg
;its own unique activity.
. . 44. See NAS 26 p.486.c.18ff.
'1 'i. 45. AKB p.4 73.25 ff. yat tarhi vijiianaT[l vijanati 'ti sutra uktaT[l kiT[l tatra
karoti I na kiT[lcit karoti I yatha tu karyaT[l kara1'}am anuvidMyata ity u.cyate
i/siidrsyena 'tmalabhad akurvad api kiT[lcit / evaT[l vijiianam api vijanatf 'ty ucyate /
:Sadrsyena 'tmalabhad akurvad api ki'f[!cit I kiT[l punar asya sMrsyam / tadakarata I
;.ata eva tad indriyad apy utpannaT[l viJaya'T{! vijanatf 'ty ucyate ne 'ndriya'T{! / athava
(tatha 'tra 'Pi vijiianasa'T{!tiinasya vijiiane karar;..abhavad vijiianaT[l vijanatf 'ti vacanan
kara1'}e kartrsabdanirdesat. AKV p.712.3lff. See also P'u-kuang 30
!,p,448.b.19ff; Fa-pao 30 p.810.a.Iff.
:;;,' 46. NAS 7 p.367.c.Iff.
. 47. AKB 3.32 b p.145.5ff; AKV p.306.27ff. Vasubandhudoes notiden-
"tify this argument as that of the but such identification is justified
.. from references in the Nyayanusara. P'u-kuang (P'u-kuang 10 p. 176.c.4-6)
,and Fa-pao (Fa-pao 10 p.608.a.15-16) attribute this view to the Sautrantikas.
.... 48. NAS 10 p.385.b.15ff; AKB 3.32 p.145.20ff; AKV p.307.17ff. See
.also NAS 10 p.386.b.16ff; 29 p.504.a.29ff. The context for the discussion of
:;:this process model of perception is SrIlata's acceptance of only three thought
.concomitants-feelings (vedana) , concepts (sa'T{!jiia) , and volition (cetana)-
',:rather than the ten thought concomitants (mahabhumikadharmah which are
'. claimed by the Sarvastivadins to be associated with each moment of thought.
See NAS 10 p.384.b.12ff.
49. The (MVB 197 p.984.a.I-3) accepts two types of col-
. location: 1) that among simultaneous factors; and 2) that among factors that
act together to produce a single effect. In the case of the five externally
directed types of perceptual consciousness, the sense organ, object-field, and
perceptual consciousness function as a collocation in both ways. However,
because mental perceptual consciousness, the mental organ, and the object-
. field are not simultaneous, they function as a collocation only in the second
way, that is as producing a single effect. See also AKB 3.30 b p.143.2ff. The
Sarvastivadins claim that this second type of collocation holds only if the
existence of past and future factors is accepted. Since the reject
the existence of past and future factors, they cannot appeal to a collocation
of causes over time-the sense organ, object-field, and perceptual
ness-in explaining the process of perception. See NAS 10
p.421.a.12ff. ..
50. AKB 2.34 b-d.r.62.3ff; NAS 11 16
51. The are characterized as reJectmg both the
between thought and thought concomitants, and the claim that various 7.1?%
functions arise simultaneously. See MVB 16 p. 79.c.7ff, 52
p.463.a.20ff, 95 p.493.c.24ff; NAS Saeki, op. cit., Vol. 1 p.}f';
See also TS 5 #60 p. 274.c.19-67 p.2/8.b.4. For example, Buddhadeva (My':
2 p.8.c.!-9; 127 p .. 66.1.c.17ff; ADV #116 p.!?7ff) identifies the .as varIetIes of thought,. provlslOnally recognizes threes!c!i':l
varIetIes: feel:ngs, c.oncepts, and vohtlOn. Buddhadeva as a
op. CIt., p. 140ff. Th:re !S, variety
Daqtantlka posItIon. For example, The master SrIlata (AKB 3,32'
p.145.20ff; AKV p.307.17ff; NAS 10 p.384.b.12ff) accepts the three-feelin";
concepts, and volition-as thought concomitants, but maintains that
three do not occur simultaneously. See also AKB 3.32 p.146.14ff;
p.309.20ff; NAS 10 p.385.b.15ff, 11 p.390.c.20ff, 29 p.503.b.llff,29:
p.504.a.29ff, 29 p.504.b.15ff. See also JunshO Kato, "Kyoryobu Shurlra.ia'
(III)," Buzan kyogaku taikai kiyo, Vol. 6, (1978), pp. 109-135. J .t;
52. According to the Sarvastivadins and two instances6f;
thought (citta) or perceptual consciousness (vijiiiina) cannot occur simultane;;
ously. See VK 1 p.531.b.6ff, passim. NAS 17 p.435.b.8ff, 19 p.443.b.9ff:;;
p.447.a.22ff; MVB 10 p.47.b.l-p.50.a.19, 140 p.720.a.10ff. .,
53. NAS 8 p.374.b.12ff.;
54. NAS 10 p. 384.c.2ff, 15 p.420.c.18ff, 19 p.447.b.16ff.
55. NAS 15 p. 421.c.5ff.
56. NAS 8 p.374.c.2ff; ADV #77 c-d p.47.13ff. "For the DarHantikas,
nothing is directly perceived. This is due to the fact that the five groupsof
perceptual consciousness have past object-fields; indeed, when the eye and
visual material form are found, perceptual consciousness does not exist, and
when perceptual consciousness exists, the eye and visual material form do not'
exist. Further, this is due to the fact that the apprehension of their own object
is impossible given the absence of the continuation [of the object] in the
moment of perceptual consciousness." hi sarvam /
. paiiciiniir[! vijiiiinakiiyiiniim yadii khalu vidyete tadii vij'iiii
nam asat / yadii vijiiiinar[! sac tadii 'sati
palabdhyanupapattes ca.
57. Siiriputriibhidharmasiistra T28.1548 9 p.590.a.7-8, p. 593.c.16-18.
For the possibility of states of concentration without an existent object-field
see also Siiriputriibhidharmafiistra T28.1548 28 p.701.c.l0, 30 p.717.a.29-b.2:
Though there is some agreement that the Siiriputriibhidharmafiistra represents
the view of the Dharmaguptaka school, the dating of the text is, as yet, disputed.
See Andre Bareau, "Les origines du Siiriputriibhidharmasiistra," Le Musion 63
(1950 #1,2), pp. 69-95; Yamada, op.cit., pp. 79-80; Erich Frauwallner,
"Abhidharma-Studien, IV (Fortsetzung)," Wiener Zeitschrift fur die Kunde Sild-
asiens, Archiv fur Indische Philosophie, Bd. 16, (1972), pp. 133-152; Kimura, op.
fjcil( .,140-160, especially pp. Kogen Mizuno, "Sharikotsuabidon-
tsuite," Indogaku Bukkyogaku ronshil, Kanakura Hakushi koki kin en, (Kyoto:

:kuji shoten, 1966), pp. 109- .
xtffra 58. VK 1 p.531.a.26ff. See also Sariputrabhidharmasastra T28.1548 9
Louis de Vallee Poussin, "La controverse du temps etdu pudgala
:51': sle Vij'iianakiiya/' Etudes Asiatiques, publiees a I'occasion du vingt-cinquieme
de I'Ecole d'Extreme-Orient, (Paris: Publications de
1925): pp. 343-376. .
59. For the possIble IdentIty of thIs Maudgalyayana as the acknowl-
patriarch of the Dharmaguptaka school see Vasumitra's
p.15.b.16-1!; T49.203.2
p.20.b.15-17; Balyu Watanabe, tr., m
issaikyo, Indo senjutsubu, Bidonbu, Vol. 4, (Tokyo: Daito shuppansha,
11(1931), p. 12 note #22; Shizutani, op,cit., pp. 173-181.
:'ii.i 60. For the attribution of this view to the Dharmaguptaka school, see
T49.2031 p.16.c.26-27; T49.2032 p.19.b.12-13;
'i[r49.2033 p.22.a.16-17. Compare the (MVB 76 p.393.a.18ff),
::'thich cites the following contested view: "Further, there are fools who, with
to the intrinsic nature of [factors in] the time periods, deny as
[those of the] past and future and mamtain that [those of the]
Y:presentare unconditioned." See also MVB 13 p.65.b.26-27, 37 p.l90.a.lO-l1.
:;,'i 61. VK 1 p.535.a.8ff.
\::;,f 62. MVB 105 p.554.c.15-17, 136 p.704.a.7-9, 146 p.747.b.15-17, 195
"p.975.a.3-5, 197 p.983.a.23-25. AVB 55 p.393.b.1O-12.
i:, 63. MVB 16 p.79.a.19-21, 55 p.283.a.22-24, 131 p.680.b.26-27, 136
i'p.702.b.13-15. See also AVB 30 p.218.c.14ff.
1',' 64. MVB 44 p.228.b.20ff. See also MVB 108 p.558.a.7ff.
}; 65. MVB 8 p.36.a.16ff. See also Johannes Rahder, "La
8," Melanges chinois et bouddhiques, Vol. 1, (1931-1932), pp.
,. 66. MVB 75 p.390.c.34ff. See, also A VB 6 p.455.c.8ff.
67. MVB 37 p.193.b.2ff.
68. MVB 135 p.696.b.24ff, 44 p.228.b.22ff.
69. MVB 195 p.975.a.2ff.
70. MVB 9 p.42.a.20ff.
7l. See also the Mahaprajiiaparamitasastra T25.1509 26 p.255.a.15ff. A
text that does not include the need for an existent object-support among the
reasons for the existence of past and future factors is the Samyuktiibhidhar-
mahrdayasastra T28.1552 11 p.963.b.2ff. This reason is also omitted from the
two most recent translations of the commentary (MVB 76 p.393.a. 9ff;
AVB 40 p.293.c.18ff), butis found in the oldesttranslation (VB 7 p.464.b.26ff).
72. TS 2 #21 p.255.b.12ff.
73. See also TS 15 #191 p.364.a.7ff.
74. AKB 5.25 p.295.8ff; AKV p.468.28ff.
75. AKB 5.25 b p.295.16 dvayarrt pratztya vijiiiinasyo 'tpiida ity uktam. See
SN 35.93 Dutiyadvayasutta Vol. 4 p. 67; SA 8 #214 p.54.a.22ff. See also SN
12.43-45 Vol. 2 pp. 72-75; ADV #306 a-b p.269.2ff.
lIABS VOL. 11 NO.1
76. TS 2 #19 p.253.c.27ff.
77. NAS 50 p.621.c.I4ff.
78. NAS 50 See. also 17 p.430.a.l0_1I
p.450.c.24-25. For the necessIty of an object-field m the arisinO" of pr ...
NAS 17 p.432.a.7ff. '"
79. ADV #304 a p.262.1. buddhyii yasye '10yate cihnarrt . . . .:>{
80. ADV #304 p.262.3ff. yasya khalv arthavastunal; svabhiivasidd ,>:j
svarupasyii 'viparZtiikiirayii dharmopala10a'IJayii paricchinnarrt lak;;a'IJam
tatsaddravyam zty ucyate .. See ADV #305 c-d p.264.2. "Those [past'anle'i
future factors] have eXIstence lrke present [factors] due to their nature asth<l ..
range of thought and name." dh'iniimagocaratviic ca tat sattvarrt vaTtamiina'" C?,
See also ADV #305 p.268.22-24. "That object whose particular and com Vat
characteristic is determined by cognition having the aspect of that i
and which is referred to ?y the group of names and of factors
by the Buddha, that eXIsts from the absolute standpomt." tadiikiirayii Mal"
buddhyii yasyii 'rthasya svasiimiinyalak;;a'IJarrt paricchidyate yaf ca
ktaniimakiiyadharmakiiyiibhyiim abhidyotyate sa paramiirthato vidyate..}'
81. NAS 50 p.621.c.2lff. See also NAS 15 p.421.b.28ff, 19
Unlike absolut.e nonexistence be according to types i'
because It lacks any partICular charactenstic by whIch It can be distinguished,"
and thereby compared or contrasted. See NAS 17 p.431.c.8ff. .
82. The Abhidharmadzpa (ADV #304 p.262.2ff) adds two types of exis- ..
tence to those mentioned by Sanghabhadra: 1) existence through both (dvaya,','
ubhayathii), referring to entities that can be understood as either real or provi-';
sional depending upon the context; for example, earth (Prthivi), when under-
stood as one of the four fundamental elements (mahiibhuta), exists in an ..
lute sense, and when understood as ordinary dirt, exists only in a conventionaL
sense; 2) relative existence (sattviipek;;ii), which refers to such correlative .
as father/son, teacher/student, or agent/action. The (MVB' 9.
p.42.a.24ff) includes three different classifications of types of existence. The
first includes two types: 1) existence as a real entity (dravya) , such as the.
aggregates (skandha) , or elements (dhiitu) , and 2) existence as a provisional
entity (prajitapti) , such as male or female. The second classification includes
three types: I) relative existence (hsiang-tai, ape10ii ?), as when something exists
relative to one thing and not relative to another; 2) existence as a composite
(ho-ho, siimagrz ?) as when something exists in one place and not in another;
and 3) existence in accord with temporal state (shihjen, avastha ?), as when
something exists at one time and not at another. The third classification
includes five types: I) nominal existence (niima) , such as hair on a tortoise, ,.
the horn of a hare, and so on; 2) existence as a real entity (dravya) , such as ..
all factors (dharma), each of which is defined by intrinsic nature; 3) existence
as a provisional entity (prajitapti), such as a pot, cloth, a chariot, and so on;
4) existence as a composite (ho-ho, siimagrz ?), such as the personality (pudgala),
which is a provisional designation based on a collocation of the aggregates;
and 5) relative existence (hsiang-tai, apek;;ii ?), such as this and that shore, or
long and short.
83. NAS 52 p.636.a.22-24.
84. See NAS 50 p.624.c.6ff. Sanghabhadra uses this point to suggest
'if":<' ast and future factors cannot be said to exist only provisionally. If this
case, they would lack a real basis and could not produce cognition.
261 1 Off
#;\if'ADV #303 p. . .
50 p.622.a.16ff. See TS 2 #19 p.254.a.3ff: "Knowledge also
i},l'.: ates with regard to a nonexistent range." See also ADV #305 p.268.27.
if th_e;e. were even having a nonexistent object-support?"
filiaJalambana eet.
ii';f. 86. TS 2 #19 p.254.a 4ff; NAS 50 p.622.a.19ff; ADV #306 c-d
i1)271.3ff. In the and Nyayanusara, these examples are attributed
.,7P;:ihe I have grouped the various examples in similar categories
of exposition. For discussion of these examples see Yukio Sakamoto,
:.:Ubidatsuma no kenkyii, Sakamoto Yukio ronbunshii, Vol. 1, (Tokyo: Daita shup-
, 1981), pp. 135-156; Shingya Yoshimoto, Abidaruma shiso, (Kyoto:
1982), pp. 146-156; Louis de la Vallee Poussin, "Documents
Yicl;abhidharma: la controverse du temps," Milanges ehinois et bouddhiques, Vol.
{(1936-1937), pp. 25-128.
ji;,' 87. NAS 50 p.623.c.18ff. See TS 2 #19 p.254.b.8ff.
88. NAS 50 p.623.b.8ff.
, 89. See also MVB 8 AKB 6.58 b p.374.26ff; AKV
For an extensive discussion of see MVB 49 p.255.a.21 ff;
iAKB 5.7 p.281.19ff; NAS 47 p.605.c.29ff.
90. NAS 50 p.623.b.17ff. See also MVB 8 p.36.a.2lff.
91. NAS 50 p.623.b.19. See also NAS 4 p.351.b.19ff.
iii':. 92. The meaning of the term akara and the sense in which all thought
thought concomitants are said to have ahara became a controversial issue
;i':forthe Sarvastivada and schools with significant im-
for later Buddhist epistemological theory. For the Sarvastivada-Vai-
;bhasikas and Sanghabhadra, ahara means the discriminative function of in-
Thought and thought concomitants are also said to have an aspect
"(sakara), but only by extension from association with insight, or in the sense
<that they perform their own activity in apprehending the object-support. This
';interpretation stands in sharp contrast to Vasubandhu's concept of ahara as
thought and thought concomitants taking shape or taking on an aspect con-
;sistent with the type or character of the object-support. Contrast AKB 2.34
c-d p.62.6; AKV p.141.29ff; to NAS 11 p.394.c.25-26; Samayapradipika
T29.1563 6 p.803.a.17-18; ADV #482 p.376.3-4. See also NAS 74
p.741.a.2lff; P'u-kuang 1 mo p.26.b.26ff, 4 p.83.b.26ff, 26 p.394.a.2lff; Fa-pao
p.534.c.4ff, 26 p.770.b.2ff; Saeki, op. cit., Vol. 3, p.llOlff.
93. MVB 7 p.409.a.10-11; NAS 74 p.741.b.12ff. For the definition of
iikara as insight see AKB 7.13 b p.401.18ff; NAS 74 p.741.a.19ff; ADV #482
c-d p.375.16ff.
94. MVB 126 p.658.b.27ff. For a discussion of the difference between
error (luan-tao, vibhrama, bhranti ?) and mistaken views (tien-tao, viparyasa), and
their relation to defilements and conceptual thought (vikalpa) see MVB 166
p.841.b.2ff; NAS 47 p.608.c.17ff. For a discussion of the relation between the
production of defilements and conceptual thought see MVB 61 p.315.b.6ff.
82 ]IABS VOL. 11 NO.1
For a discussion of the character of insight when associated with me
ceptual. consciousness as distinguished that associated with the
nally dIrected types of perceptual see MVB 95
95. AKB 6.9 p.337.8-6.13 p.341.6, NA", 59 p.671.a.1-60 p.674 ",k,1
AKB 8.29 pA52.4--S.36 pA5S.10: NAS 79 p.768.c.20-S0 p.774.c 5.
_ . . . In VB s,
Sl p.420.b.S-S5 See also op. Clt., 5.3 pp. 305":'30}'I;
96. For a dlscusslOn of adhlmuktlmanaskara as one of three ty
attention see AKB 2.72 d p.10S.1Iff; AKV p.246.32ff; NAS 20 P,454.
MVB S2 pA22.c.27. For the various meanings of the term ff;,;
Hajime Sakurabe "Sh6ge 'adhimukti' ni tsuite," in Bukkyo go no kenk"u (K
:J' yoto''''
Buneid6, 1975), pp.
97. MVB S5 p.440.b.1Iff, pA41.a.25ff.
98. NAS 50 p.622.a.19; TS 2 #19 p.254.a.4. .
. 99. NAS 50 The (TS 2 # 19
cItes explananon: smce quahty or of the color blue exis't{'
even m thmgs that are not perceIved as blue, thIS blue nature in all thin ,I;
can serve as the object-field for the cognitionof total and exclusive blueness :g:"
100. MVB 37 p.193.bAff; TS 2 #19 p.254.a.7-S, p.254.c.25ff. '"
101. See MVB 37 p.193.b.23ff; AVB 2S p.145.c.11ff; NAS
102. NAS 3 p.346.a.17ff.
103. NAS 50 p.623.c.13ff. The MahiivibhilJa (MVB 3S p.194.a.2Sff)
sents several opinions as to whether or not all dream images must be
result of past experience. Though no explicit judgment is offered,
Mahiivibhi4a clearly favors the opinion that all dream images result
object-supports that have been experienced. See also NAS 3 p.346.a.17ff.
104. MVB 37 p.193.c.24ff.
105. The Mahiivibhi4a (MVB 37 p.194.b.27ff) explains that in the case
of oneiromancy, one knows future events in a dream through inference; one"
infers that a certain event will occur in the future on the basis of an experienced i
cause and effect relation between the past and present. " '
106. The causes for dreams offered by Sanghabhadra (NAS 50 i
p.623.c.9ff) and the TattvasiddhiSiistra (TS 2 #19 p.254.b.13ff) are generally
consistent with those in the Mahiivibhi4a with a few exceptions: both,'
Sanghabhadra and the TattvasiddhiSiistra omit dream images based on future
events, and the Tattvasiddhisastra adds past actions (karma) as a possible cause.
107. MVB 75 p.390.c.3ff. See also VB 6 pA55.c.Sff. ,
lOS. NAS 23 pA70.a.6-474.a.5; AKB 3.11 c-d p.120.20ff; AKV
p.267.29ff. In this section, an opponent offers the example of a reflected
image to disprove the existence of the intermediate state (antarabhava) between
death and rebirth. That is to say, just as there is an interruption between the
reflected image and the original object, so there is an interruption between
death and rebirth and no intermediate state is required. Vasubandhu claims
that since the reflected image does not exist, it should not be compared to .
the aggregates at rebirth. Sanghabhadra, on the other hand, argues strongly,
in defense of the existence of the reflected image, and claims that there is a
connection between the reflected image and the original object.
NAS 23 p.473.a.8ff.
110. NAS 23 p.4 71.b.12ff.
111. NAS 23 p.472.a.22; AKB 3.12 a p.121.5-6. ato na 'sty eva tat kiT[lcit
gryiis tu sa tasyas tadrsal} prabkavo yat tatka darsanaT[l bhavati. See also AKV
, "'112. NAS 23 p.472.b.23ff.
>;,:.113. MVB 135 p.696.b.24ff. For a discussion of the arising of magical
lions from the supernormal power that actualizes the knowledge of objects
duced by magical power abh&'iia) see AKB 7.42
;: .. ii42L6ff; NAS 76 p.752.c.17ff. (especially AKB 7.44 d p.423.5-6; NAS 76
AKB 7.48 p.425.5-7.53 p.429.3; NAS 76 p.754.b.29-76

NAS 50 p.623.b.27ff.
NAS 50 p.623.c.28ff; ADV #306 c-d p.271.lff.
ll\;.; 116. NAS 50 p.622.a.24-25; AKB 5.27 c p.300.18-21; SA 26 #703
117. NAS 50 p.623.c.29-p.624.a.8.
Sanghabhadra clearly distinguishes the sound of speech, which is
Yinaterial form, from name (nama), which is classified as an independent con-
;;aitioned factor dissociated from both thought and material form (cittaviprayuk-
;"?;;siJ'T[1Skara). The specification here would be synonomous with name. See NAS
l4pAI3.a.17ff; p.413.b.16ff; p.414.a.16ff; p.414.a.29ff; p.414,b.22ff;
I#:?;' 119. NAS 50 p.624.a.8ff.
i.;' " 120. Though Sanghabhadra does not identify these two types of nega-
;:tlims, they appear to correspond to the implicative or exclusionary negation
and simple or prohibitive negation
.;which were used extensively in Indian grammatical, ritualistic and philosophi-
See George Cardona, "Negations in Pal)inian Rules," Language, Vol.
",43,(1967-1), pp. 34-56; J.F. Staal, "Negation and the Law of Contradiction:
::A Comparative Study," Bulletin of the School of Oriental and African Studies, Vol.
(1962-1), pp. 52-71. For references in later Buddhist texts see Yuichi
::kajiyama, "Three Kinds of Affirmation and Two Kinds of Negation in Bud-
i'ghist Philosophy," Wiener Zeitschrift filr die Kunde Sildasiens, Bd. 17, (1973), pp.
, 121. NAS 50 p. 624.a.lSff.
122. Sanghqbhadra here uses the argument that expressions can lack a
specified object because otherwise there would be no worldly speech that lacks
meaning, Sanghabhadra then cites another opinion that all expressions must
"have a specified object because these expressions are specifications. In the
:',case of expressions such as "nonexistence" or "thirteenth sense sphere" the
'specified object would be the name or concept and not some objective "nonexis-
tence" or "thirteenth sense sphere." See also AKB 5.27 c p.300.7ff; AKV
p.475.1lff. The (MVB 15 p. 72.c.2-5) similarly explains that all
names are able to manifest meaning and that even names such as the "thir-
teenth sense sphere" manifest the concept, "thirteenth sense sphere."
123. The terms hetusamuttkana and are used to explain
84 JIABS VOL. 11 NO.1
the immediate causes by which manifest verbal or corporeal
tirupa) arises. See AKE 4.10 p.203.13ff; AKV p.364.17ff; NAS 36 P.S47 '17ia/f;;;
MVB 117 p.610.a.5ff."
124. TS 15 #191 p.364.b.9-10',;.l:fti1
125. ADV #306 c-d p.271.16-17. "If one claimed that there is
ha:ing a due to existence [of a
object-field of cogmtlOn IS the demal of the thirteenth sense sphere [w
. . , e would'<;
reply] no, because It has been demonstrated by the Lord that this [cogn' "'Ai
is merely based upon speech."
asadiilambanii buddhir astf 'ti eet / na / Bhagavatai 'va viigvastumatram
nirr;,ftatviit. In the Abhidharmakosabh[4ya, the Sarvastivada- res ,/
to a similar objection claiming that the name (niima), "thirteenth sense sp!?n""
serves as the object-supyort of one's of of the
sense sphere. S.2/ c p.300.8-9 Then IS th.e object-support of the';
perceptual conscIOusness of the statement, there IS no thirteenth sen'r!
sphere?" That has only name as its object-support." atha trayodasam
nii 'stf 'ty asya vijiiiinasya kim iilambanam / etad eva niimiilambanam. See also AKV.<
p.4 7 5 . 14ff. Since the Sarvastivadins claim that names exist as real entitiesE
classified as factors dissociated from thought and material form,
serve as the existent object-support for the arising of cognition. .,
126. ADV #306 c-d p.271.1-1S, p.272.3-1S.:>
127. ADV#306 c-d p.272.13-15. tasmiin naiio na
/ ki'r{! tarhi / saSiikiiSadhiitusa'r{!bandhabuddhyapek:;e'YJa,'[(
'vadyotyante. . .... ',';;
128. NAS 50 p.624.b.4ff. See also AKB 5.27 c p.300.9ff; AKV p.47S.
129. NAS 17 p.431.b.12ff; AKB 2.55 d p.93.7ff. See also Louis de. la,
Vallee Poussin, "Documents d'abhidharma: textes relatifs au nirvaI,la et aUll:
asarpskrtas en general, II" Bulletin de nieole Franqaise d'Extreme-Orient, Vo1.30,:;
(1930), pp.277ff. In this section, Vasubandhu cites the Sautrantika opinion,.
that the unconditioned factors (asa'r{!skrtadharma)-space (iikiiSa), cessatio,n,'
through application (pratisa'r{!khyiinirodha), and cessation not through
tion (apratisa'r{!khyiinirodha)-do not exist as real entities, but rather are mere'
absences (abhiiva). (AKB 2.55 d p.92.4) Nevertheless, the Sautrantikas assert'
that unconditioned "factors can be said to exist in the same way in
can be said that there is the prior or subsequent nonexistence of souncl:
However, this mere statement that they exist does not mean that absences'
(abhiiva) themselves exist as entities (bhiiva).
130. NAS 50 p.624.b.22ff.
131. AKB 5.27 d p.300.10ff.
132. Sanghabhadra distinguishes absolute nonexistence, like the horn
of a hare, from the nonexistence of that which has not yet been produced
(i.e., a future factor), or has already passed away (i.e., a past factor). These.
last two are nonexistent only in the sense that they lack activity. Even though
they do exist as entities having intrinsic nature, they are recognized to be
nonexistent in comparison to the present, which is characterized by both
activity and intrinsic nature. NAS 15 p.419.c.5ff. '
133. MVB 12 p.55.c.29ff; AVE 6 p.42.b.16ff. For a discussion of how
ction occurs without a personality (pudgala) or a continuous substratum
11 p.55.a.I6-12 p.58.c.18.
MVB 12 p.57.c.24-26. .
135. AKB p.54.22-23. smr

r:,'anslauon (PAKB 3 p.178.b.I4-15) corresponds to thIS defiDl-
Hsuan-tsang (HTAKB 4 p.19.a.20-21) in his translation adds
. ") . 'bl d . f
fii"_chi(abhdapana, or notatIOn, POSSI y un er the mfluence 0
explanation. Yasomitra (AKV p.127.32ff) comments: "Mind-
is that by.connection with the mind not forget the object-
I;ii.::' 'p'" ort, and, as It were, notes that [object-support]. yadyogiid iilambanaT(l. na
"sliP. . - 'bh'l - , - 'h
tac ca z apatz va sa smrtz ..
NAS 10 p.384.b.7-8. See also ADV #112 p.69.6-7 "Mindfulness
form the functioning of thought. It is the notation of the object of
and has the characteristic of not losing action that has been, will be,
G)'fis being performed." cittavyiipiirarupii smrtil}, / cittasyii 'rthiibhilapanii krtakarta-
Compare Abhidharmiivatiirasiistra
'1(28)554 shang p.982.a.18-19: See also ADV #446 p.360.14-16. For defini-
of smrti in terms of abhilapana in the early Sarvastivadin Abhidharma
;}rhts see Dharmaskandha T26.1537 7 pA85.a.7; Dhiitukiiya T26.1540 (shang)
14.c.20ff; SarigftiParyiiya T26.1536 16 pA33.b.6ff, 17 pA37.a.13ff; Pra-
r;apiida T26.1541 2 p.699.c.17ff. .
:1'1' 137. NAS 10 p.389.b.12ff. .
);f':.;. 138. This statement is also significant because it indicates that for the
'siivastivadins, mindfulness as notation (abhilapana) operates not only in mo-
of mental perceptual consciousness, but also in all moments of the five
directed types of perceptual consciousness. Howe"er, since mindful-
associated with the five externally directed types of perceptual conscious-
.pess is weak, it is not considered to be conceptual thought through recollection
(NAS 4 p.350.b.17ff). This view is to be contrasted with
;thatof Yasomitra (AKV p.65.10-11) who claims that mindfulness does not
operate as notation in moments of the five externally directed types of percep-
i;t,llalconsciousness: "Because mindfulness associated with the five types of
'perceptual consciousness does not operate through the notation of the experi-
;enced object, it is not considered to be conceptual thought through recollec-
'(ion." paiicavijiiiinakiiyasaT(l.prayuktii tu nii abhiliipa)
pravrtte 'ti nii 'nusmara1Javikalpa itf
. 139. See P'u-kuang4 p.74.b.2lff; Fa-pao 4 p.527.c.13ff.
140. MVB 11 p.5l.b.14ff. Here examines the problem
.of how the Buddha knows the sequence in which future factors arise, since
they, as yet, lack sequence and are disordered. (vyiikula). Compare MVB 179
p.897.b.24ff; AKB 2.62 a-b p.98.29ff; AKV p.233.30ff; NAS 19 pA44.b.9ff;
P'u-kuang 7 p.135.a.8ff.
" 14l. Though the particular cittaviprayuktasaT(l.Skiira is not identified in
,this passage in the it can only refer to possession (priipti). For
frriipti described as cihna see Yasomitra (AKV p.148.22-23) who quotes
Sanghabhadra (NAS 12 p.397.bA-6): "Possession is the indicative mark of
the knowledge that 'this belongs to that,' and is the cause of the non-disappear-
ance of factors that have been obtained." idam asye 'ti jiiiinacihna '
dhadharmiivipraniiSakiira'I'Ja'T{l ca priiptir ity iiciiryasarighabhadrah. 'Y[l,frr
142. MVB 178 p.895.a.26-179 p.898.a.12. See also MVB 76 P.395
AVB 40 p.295.c.lff; VB 7 p.466.a.14-15; 7.37 a-b PA17.19fi
NAS 75 p.750.b.18. For the vanous mterpretations of
see MVB 178 p.896.a.13ff. Two of the Buddha's
also. here: power of the knowledge of previous
whIch knows factors, the power of the
of death and rebIrth (cyutyupapatt1Jniinabala), whIch knows future facto ."
(MVB 100 p.517 .a.3ff) discusses the complex issue of the
" I h ' se POW",
ers, contrastmg them wIth the supernorma power t at actualizes the
recollection of previous birthstates
and the the knowledge
and See also AKB 7.29 c p
NAS 15 p.746.a.18ff; AKB 7.42 p.421.6ff; NAS p.752.c.17ff. ',,:,\i
143. NAS 51
144. TS 15 # 191
145. AKB 5.27 c p.299.20ff; AKV p.474.9ff; NAS 51
146. For a discussion of whether or not unconditioned factoi;':;:
(asa'T[iSkrtadharma) may serve as causes, and if so, as what type of
AKB 2.55 d p.91.18ff; AKV p.218.18ff; NAS 17 p.429.a.3ff. ,j;:ft
147. NAS 19 p.447.b.29-p.447.c.9.
148. For various interpretations of the general definition of causal
tion-"when this exists, that exists; from the production of this, that is
ed" (asmin sati: 'dam bhavati asyo 'tpiidiid idam utpadyate)-see AKB
p.138.28ff; NAS 15 p.419.a.7ff, 25 p.482.a.ff: AKV p.297.9ff. For Stilata's';:
interpretation of the nature of this causal relation see NAS 15 p.419.a.7f{'i;i
149. NAS 51
150. NAS 51 p.628.c.6-8. "1
151. NAS 51 p.628.a.4ff. See also Fa-pao 7 p.578.b.2ff.'f
152. NAS 19 p.447.c.9ff; 51 p.628.c.27ff.,'
153. Sanghabhadra (NAS 19 p.448.a.8ff) also SrIlara's theory,::;
of the secondary or subsidiary element (sui-chieh = anudhiitu ?, or chiu-sui-chieh.J
= purviinudhiitu ?), which SrI!;ita uses to account for all types of causal relations:"
Sanghabhadra identifies this secondary element with the seeds (b'ija) proposed
by Vasubandhu. Since both the secondary element and seeds function causally:.
only through a successive relation within the psycho-physical series, their "
proper operation requires the existence of past and future factors. Since
neither SrIlata nor Vasubandhu admits the existence of past and future factors,
their models are, in Sanghabhadra's opinion, untenable. See NAS '18
154. AKB 5.27 c p.300.16-17 . ... sarvabuddMnii'T{l sadiilambanatve kuta 'sya
syiit... NAS 50 p.622.c.13ff. Yasomitra (AKV p.4 76.7-10) glosses
arsa with investigation (viciira), or doubt, (sa'T{ldeha). He comments: "When
there are existent and nonexistent object-supports of cognition, this deliberac.
tive reflection is possible; not otherwise." sadasadiilambane tu buddhZniim ayar{!
sambhavati nii 'nyathii.
:155. AKB 5.27 c p.299.24ff; AKV p.474.15ff; NAS 51 p.628.a.27ff.
156. NAS51 p.628.b.llff.
The Pali carum is OUT oldest source joT the word of the Bv.t:iJiha. We are a charity publishing Pali lex'
trarutations, dictionaries and primers to increase public awartmess oj Buddhist literature. If you lulu'
dijJiculty in finding our books, please write Jor our list oj publications or order your book direct af1.i
Better still, fUrther Ollr work by bcwming a member, All members receive 10% dis ;OIJnt on purcJlilSi
and Sponsoring members can also clwose a free book every year.
Pali Text Society
73 Lime
OxJord 0)(3 7AD
Sponsoring Membership:
1 year..,,16
5 years. ... 70
OJ $

Ordinary Membership:
1 year. ... 7
5 years. ... 30

We invite applications Jor Research Fellowships from suitably qualified persons, working in tll(fl(/doh
Pali Studies. Applicants will usually be in tAe Jourth year oj a course oj graduate research, or its equivalent'
The course oj reseaTch will be expected to lead to a publishable book, on the publication rights 0/ which tilt
Pali Text Society will have first option. .
Felwwships will be tenable Jor one year in the first instance, with a possibility of renewal Theirvalue
will be fixed at the appropriate level Jor a comparable research worker in the applicant's country of domicilt "
Letters oj application and requests Jar information should be sent to Mr KR Norman, Faculty a/Oriental
Studies, Sidgwick Avenue, Cambridge CB3 9DA.

"',,,)." ,
The Vimalak'irtinirdeia-sutra (VNS) has long been admired
its felicitous blend of literary and philosophical qualities. Its
tstrong and ambitious narrative-as well as its delight in the
only remind us at times of the jatakas and avadanas,
:iwhere the abstractions of Buddhist metaphysics are made ap-
for a lay audience through engaging moral fables.
o/,In the VNS we find very few of the scholia which distend and
i,break the narrative tension of other large sutras; the action keeps
building momentum toward a final and climactic en-
between the Buddha and the householder from Vaisali:.
;;qur dramatic expectations are skillfully lifted, and, in the end,
;!;\Vell satisfied.
Yet the VNS is not a "novel" or an "epic" in any sense we
!!Would recognize: just as the moral-phantasmagorical works of
a Blake, or a Dante elude any single genre, so one
do justice to the VNS by analyzing it purely as a tale.
in what follows is my view that the VNS is a Buddhist
,work sui generis which merits and demands its own critical ap-
,proach. For this text more than for most, the success of a doc-
!trinal argument is hinged on the success of a poetics, and vice
The plot line is far more than just an armature for a
scholastic discourse; it both vivifies and takes its direction from
jhe issues being debated by Vimalaklrti and his guests. Because
it so admirably merges the visionary with the conceptual, the
:,aesthetic with the scholastic, the VNS stands as perhaps the
'purest example of philosophical drama the surviving Mahayana
jmaterials provide.
90 JIABS VOL. 11 NO.1
I. Salvific Magic
This essay focuses on an especially sensitive point ofc
. - Ont ',,'
bet:-veen Mahay.ana and Buddhist literary
bolIsm.: the of. the pheno:r:nenal world bif
an enlIghtened bemg. The VNS IS sImply burstmg with
tricks, usually justified (when justified at all) as
b dh
' ." kill' ,,- k 'I A . of .
o S m means, ?r upaya- ya. ccording
doctnne of upiiya, an enlIghtened bemg-a bUddha or)';;
bodhisattva-possesses a special ability and prerogative
method is best suited to character
karmIC dISposItIOn of the student. For a bookish student
upayic approach might involve a doctrinal sermon or a
debate; for a merchant, a parable involving gold pieces;
superstitious man, a magic spectacle intended to dazzleaniil
beguile. Upiiya takes as many forms as the many dispositions ii
sentient beings, and only the superior)nsight of the enlightened:
being would seem to guarantee its appropriateness to the situ)
ation at hand.j:
Even in the Pali literature, the Buddha can
found resorting to magic tricks as a teaching device; but only;
in the Mahayana does the practice seem to have a true efflorev:
cence, becoming in certain respects a literary setpiece. It seemi'
plausible that the VNS enjoyed an early and important role in'
this vogue, for magic is central to its texture as a work of imag:
ination. It is certain that the VNS's colorful imagery inspired a"
host of poetic and visual artworks, from the poetry of Po
to the sculptures of Lung-men and the large stele now housed
at the Metropolitan Museum of Art in New York. . ... ,
At the same time, the VNS is a deeply philosophical text-yet
what role can magic possibly play in a sober Madhyamaka
tic? Certainly, the liberal use of illusion as a plot catalyst raises.
a host of difficult philosophical questions. But perhaps this is
exactly its function. When we witness one of the Buddha's ,
lous displays of legerdemain we are prompted to ask: What is
the nature of the phenomenal world such that it can be magically
manipulated in this way? And what is the moral stance of the
being who does so? How can we be sure of his good intentions?
By examining several instances of upayic magic I hope to
show how the VNS's two primary rhetorics, the visual and the
work in tandem to address these questions. In par-
I will examine the tension between the Buddha and
as rival magicians-for it is .here, in the magical
rl'butest between a otherworldly hlerophant and an ex-
worldly that the m?ral and philosophical
rt?lhIlenslOns of upaya come most sharply lUto focus. We wIll see,
156, that the VNS's careful exploration of upaya is really some-
of a Trojan Horse concealing a much larger question:
of the true status, both moral and ontological, of the

fir The UPayic Magician
i::t I have suggested that the VNS is eminently concerned with
'the notion of "skill in means" (Skt. upaya-kausalya, Tib. thabs la
'!fnkhas pa). The entire second chapter of the Tibetan text is
dedicated to the topic, and the first concludes with a dramatic
'demonstration of upaya in action as the Buddha transforms the
into a panorama of jewels (Th 18).2 The chapters on
:the humiliation of the sriivakas and bodhisattvas are essentially
'illustrations of Vimalakirti's upayic mastery, and as the text
tmoves toward its dramatic denouement we witness ever more
examples of his power to conjure.
:';i At the outset, the VNS appears to exhibit a certain structural
'similarity to the SaddharmapurJ(larika, which also debates upaya
(in its second chapter while furnishing examples of upaya at work.
the similarity between these two important Mahayana texts
.is only superficial: with a closer look we can discern a basic
'disagreement over the proper agency of upaya. This disagree-
ment has wide implications, not only for the emergent Mahayana
doctrine of upaya , but also for the early Mahayana view of the
bodhisattva. While both texts ultimately use the upaya issue as a
means of drawing ontological and ethical distinctions between
buddhas and bodhisattvas, they proceed in very different ways and
reach quite different conclusions.
In the Saddharmapu17darika or Lotus Siitra, the Buddha has
recourse to the upaya doctrine in explaining how a buddha can
communicate the inscrutable nature of reality to a non-buddha. 3
A sharp distinction is drawn between the understanding of a
buddha and that of a bodhisattva: in a dialogue between
of m.etaphysical truths takes place
tlflCe, while m the dlalogue of a buddha an. d a bodhisattv .... h. '1

. d '. at e
employ upaya m his insights.
bodh2Sattva s lesser WIsdom places a barner between him and En!
metaphysical enlightenment, and all of a buddha's efforts Y
directed toward its removal. The episten::ololgical gap
a buddha and a sravaka, or a mere prthagjana, of course, is ev
wider, and the buddha must adjust his upaya accordingly, moldi eB
it to the disposition of his listener. This flexibility allows
wide range of upayic acts, including, as the famous parable6t
the burning house illustrates, a type of deception.
condoning unethical behavior, however, the Saddharmapur}{la_
r'ika recognizes that the supreme wisdom of a buddha allows him'
to act in ways which, for a non-buddha, might be moq.lly
The buddha's special status as an enlightened being removes
his behavior from the formulaic rubric of the Vinaya and
it in a realm of upayic compassion. "And it is precisely because
a buddha stands outside of "the burning house of the
world" that he has a responsibility to employ his upaya to
others. He alone is capable of fully discerning the true nature
of reality, including the inward karmic dispositions of
trapped in the vagaries of rebirth. . ... , .....
So argues the Saddharmapun,gar'ika. By contrast, the upiiyd
chapter of the VNS opens by extolling Vimalakirti's long service
to the past buddhas and his dedication to the bodhisattva caust
in the present time (Th 20). It is understood, and will later be
stated explicitly (Th 43), that he is a bodhisattva living as a layma!l
for the sake of saving sentient beings. Ye.t like a buddha, Vim-.
alakirti is said to have "assimilated an understanding of skill in
means" (thabs mkhas pa rtogs par khong du chud pa,' Osh 155). The
remainder of the text will systematically outline Vimalaklrti's
upaya, even going so far as to establish parallels with the
Buddha's own acts of upaya.
Evidently, then, the VNS does not accept the
mapuryjar'ika's limitation of upaya to the buddhas: here even a
bodhisattva may practice upaya, and in doing so he does not.
appear to raise the moral question addressed by the parable of
the burning house. This observation raises a trio of questions
which orient us for the present study: (1) What is the ontological
tuS "of magical upaya as it is presented in the VNS? (2) What
:i;:fits rnoral_ status, if any? And (3) Are there differences in the
Jla,gical upaya of a bodhzsattva and that of a buddha?
51ir The Magical Parasol
{{ As with so many Mahayana sutras, the VNS opens with a
spectacle which the.plot a supramun-
status and sets the phIlosophICal dIalogue 111 gear. A huge
:;ssernbly of disciples and bodhisattvas has gathered in Amrapali
,;iohear the Buddha. From the crowd appears a group ofLicchavi
Nouths, led by the Bodhisattva Ratnakara and bearing jewelled
i'parasols as offerings for the Buddha. When the last parasol is
iIaid at the Buddha's feet, the offerings are miraculously trans-
fformed into a single giant parasol which appears to cover the
"rriany universes (Osh 147-8). The particular features of all the
seem to be reflected on the underside of the apparitional
;.parasol, complete with the buddhasof the ten directions preach-
in their respective domains. This cosmic vision astonishes
ithe multitude and prompts Ratnakara to kneel and address a
ilaudatory hymn to the Buddha (Th 13-15).
i. What is the ontological nature of this incident? The text's
,'choice of language provides a number of interesting clues. As
for the transformation itself, we are told that it is "by the power
of the Buddha" (sangs rgyas kyi mthus; Osh 147)-by his
anubhiiva-that the giant parasol "appears to cover" (khebs par
snang ngo; Osh 148) the many universes. The use of the Tibetan
verb snang ba ("to appear"), which Thurman's translation ig-
nores, is significant here, for it points to the linked doctrines of
illusion and emptiness which figure importantly in later parts
of the text. The magical demonstration reveals the entire galaxy,
. and most especially, as Ratnakara sings, the fields of the sugatas
or buddhas. It is a sort of beatific vision, a glimpse of the beyond
not unlike that which Beatrice grants Dante; yet it is manifestly
derivative, being only a magically-induced reflection, a second-
order appearance.
The distinction between the cosmos itself and its mirror
image might seem trivial in light of the tremendous grandeur
of what is being shown, but it is important to keep in mind the
special symbolism of such imagery in the Mahayana. The th';};i
. fl' d '11 d bl .
of vIsual re ectlOns an 1 usory ou es IS a potent on .;)/
Mahayana rhetoric; it almost always functions as a didacti/i{H
go.ry for the empty nature of reality.6 Though the text In ..
. l' I h' 1 a es
no exp lClt onto oglCa statement at t IS ear y stage, by introd';j
ing the theme of the magical mirror it begins to lay
groundwork for a rhetoric of emptiness to appear in full
ater on. ... 1
The incident of the parasols has an equally subtle functi;;
. h' h d' f h' . on;
WIt m t e ramatlC structure 0 t e text: It establishes ale
axis in the narrative's geography, a point d'appui against whiCh
the other side of the plot's action-that centered on
now steadily begin to pull. Seated amid his ret:.
inue in Amrapall, the Buddha uses his power (anubhava) toi
project the entire universe over his head, thereby renderini
himself its spatial and symbolic epicenter. This is none othef
than the axis mundi motif as articulated by Eliade and others'1
but in our text it serves to throw the scene at Amrapali
high relief, placing the assembly in a privileged position froni
which all cosmic events and destinies (gati) may be';
Yet almost immediately the action will leave this high
ground and develop a second dramatic axis at Vimalaklrti's
house. Just as the VNS allows a bodhisattva the power of
upaya, so, in dramatic terms, does it allow his presence and
actions to parallel and in some ways compete with those of the;
Buddha. This peculiar tension between buddha and bodhisattva
hints, perhaps, at the historical ascendancy of the bodhisattva as;
an alternative model of perfection-and perhaps, too, at a slow
struggle to accomodate both buddhic and bodhisattvic
paradigms into Buddhist dogma without subtracting from the
prestige of either. The geographical distance separating buddha
from bodhisattva in the early chapters of the text-a distance
which seems more and more treacherous as the disciples tell
their stories and thus embellish our picture ofVimalaklrti-may
mirror a very real ambivalence about the precise relationship
between a buddha and a bodhisattva, a need to show that these
two enlightened beings are in some ways similar but in other
ways quite distinct. To test such a supposition we must take our
cues from the plotting of the text, for it is here that the symbolic
geography so carefully established at the outset is put through
_ -

:W.Enthronement of the Dharma-king
.' .
. iN' The spectacle of the giant parasol is followed immediately
laudatory hymn (Th 13-15). The young Lic-
verses intensify the emphasis on the Buddha as cosmic
while preparing the way for a direct comparison with

:;,;i:i; .The Buddha emerges from Ratnakara's hymn as the
of an Il1,dian pure, virtuous, rich in
\tleeds, he IS an ascetIC followmg the path of peace (dge sbyon zhz
lam brten; Osh 148). Yet at the same time he is a bull of
elrien, a leader, a Dharma-king, a Lord of Dharma (Th 13).
ZEpllowing the early Buddhist paradigm, he teaches the law of
co-origination in order to liberate beings. Doctrine
:fNerses 4, 7, 8) interweaves with sacred or hagiographical history
6) to apotheosize the Buddha and elaborate his special marks
(9, 10, 11). Besides elevating the Buddha to a position
0:oL supreme respect, Ratnakara's paean serves as a brief
catechism, deftly laying out the philosophical stance
text and validating it by fusing it with a portrait of the
The Buddha is a teacher of the doctrine and a sacred
}!ong, poised at the center of all destinies (gati) and able, through
.his special powers, to appear individually to each disciple. As
':we will see, this homology with kingship will ultimately help to
"distinguish the Buddha from Vimalakirti even as it raises the
':i:ritical issue of kingly power and legitimacy.
, Having established the sacred identity of the vision's creator,
'the text goes on to ponder its cosmological significance. Ratnak-
ara, on behalf of his Licchavi cohorts, asks the/Buddha to explain
the purification of the buddha-fields (buddhak$etra) (Th 15), as
he has already (in verse 2) identified the vision as revealing "the
superb and radiant fields of the Sugatas" (Th 13).
But what exactly is meant by "purification" here? The
Buddha's discussion details the ways in which a bodhisattva, by
the merits of his own practice, draws karmically-ripened beings
,into his buddha-field when he attains enlightenment. A
,bodhisattva perfects his generosity, and thereby causes generous
96 JIABS VOL. 11 NO, 1
beings to be reborn in his field; he perfects his morality
' b ' d h' DI ' , alJ.d.
mora y emgs rawn to tlmately,
a causal cham of several lmks, the punty of the buddha-fi1cii.
becomes a reflection of the purity of the bodhisattva's own
(Th 18). Since has become a this
cannot be anythmg less than absolute. Why IS It, then,
"sup.erb and radiant fiel?s ofthe Sugatas"
magIC parasol replete all the and features6l1
the natural cosmos? How IS such a kaleIdoscopIC VISIOn reflectiy'!i
of the Buddha's inner state of
. Sariputra finds himself wondering these very things.BSPf:'.'
cifically, he questions whether the apparent impurity of thfi:
magically revealed buddha-field is somehow a reflex of the implll'J;:
ity of the Buddha's mind when he was still a mere bodhisattvd/
(Th 18). And here we encounter the text's first real
argument concerning the nature of both upayically-controllec(;
reality and the cosmos as it appears in everyday life. Ler.i.W1
examine it closely. ,.i;n
The Buddha addresses Sariputra's doubt by
appearance of the buddha-field in terms of the perceiver's mentaLJ
state: "Sariputra, the buddha-field of the Tathagata is pure, bllr,lj
you do not see it" (Th 18). In explaining the vision this way;;];
the Buddha implies a twofold reality: the first aspect
genetic or creative aspect-is pure because the buddhqi:;
at the center of the buddha-field is pure. But the second
its apparent or phenomenological aspect-conforms to the
position of the mind which perceives it. As the Buddha putsiti;
"What do you think, Sariputra? Is it because the sun
moon are impure that those blind from birth do not see the
(Th 18). .
Thus, in ontological terms, the appearance of the
tells an ordinary man nothing about its intrinsic nature.
radical statement is only emphasized, of course, by the
the vision under discussion is a spectacle magically
. the Buddha's upayic powers. But in what sense is this
upaya? The Buddha's upayic emanations seem to be
I;>y the narrowness of Sariputra's mind. As Brahma
monishes him, "Lord Sariputra, because there are
lows in your mind, in reflecting on the Buddha-knowledge
are convinced that it is not wholly pure" (btsun pa shi'i ri

: ; .. i"i!;::
( ..
,-,; - .
;;KtS la rnthon dman yod cing sangs rgyas kyi ye shes La bsam pa yongs
dag par nges so;. Osh 154). Has failed
for Or, more provocatively, IS It I?osslble that a
enlIghtens one sort of man can sImultaneously
and mIslead another? .

Jewelled Cosmos

The miracle of the parasols almost immediately undergoes
mutation which seems to remove any doubt as to the
complete control over it. With a of his to
the Buddha transforms the cosmos mto a matnx of
wherein each beholder perceives himself to be seated on
throne (Th 18-19). The text intimates that this trans-
ifiguration applies not (only?) to the cosmos reflected in the giant
!parasol, but also to the cosmos in its everyday form. If this is
illiecase, the Buddha has shifted his ontological demonstration
:Iiom the domain of the overtly magical apparition (the giant
to that of the "natural" world. This is an ambi-
,Mus step, for it dramatically equates the common phenomenal
the illusion built of maya.
tee. Such a step is consistent, of course, with the philosophical
the text: when the Buddha later "blunts" (brtuL ba) his
vision, eighty-four thousand beings who had been "de-
W:.()ted to the grandeur of the buddha-field" suddenly see that "all
\things are by nature but magical creations" (Th 19). As is often
:.t4ecase with Mahayana pericopes of magical upaya,9 it is not so
much the vision itself but its withdrawal which precipitates an
in the beholder. Here the visual rhetoric of the
;tbemirrored duplicate cosmos, the successive transfigurations
and home a specific doctrinal point,
with such force that each witness to the spectacle has a
'.(eligious awakening. As we would expect, these awakenings vary
the karmic disposition of the being who is awakened: while
!the sravakas, for example, extract a lesson in impermanence
trom the incident, the bhi/0us are released from their asravas.
later, in the eleventh chapter, the Buddha will further
this principle of dispositionaLity by asserting that each
emphasizes a particular type of disposition and thus

a particular brand,of upaya (Th
But what of Sariputra's concern about the pUrity rl"'1:
buddha:field? The Buddha answered his disciple by statinO . r:.: .. ........
. S- . ,. I" d . d h' h d g t at
It was anputra sown Imlte mm w lC cause various "h'Ii:s
and lows" to infest the original vision of the cosmos. ThrJ
his magical upaya the Buddha now makes it possible for Sarip
to see the field as pure .symbolized by the jewelled
Yet what has changed-Sanputra, the cosmos, or the illusi \\
The ontological question is confounded further by a
ous one: If the Buddha is indeed the creator of his own
field, why would he choose, as a compassionate being, tOfil&"Y
with impurities? Or, in thaumaturgic if the
the power to transform the cosmos by hIS upaya, why doesh'
transform it toward greater impurity?<;1;.,
This is, in essence, a Buddhist wording of the problemdf
theodicy, or the explanation of the profane's perdurance,ih
default ofthe sacred's power. 10 As Wendy O'Flaherty has showl}
the problem of impure creation found its classical Hindu
tion in Vedic and Pural)ic myths, where negative aspects
universe are created from the more baneful parts of
own body. II For Sariputra, however, the cosmos is homologizeg";
with the sacred being's mind, not his body; Sariputra's moment}
of doubt reflects a psychologization of the Vedic
god" concept, and as such poses a new series of dogmatic
mas. A mapping of the variegated, sacred/profane cosmos:
against the Buddha's presumably pure mind could hardlybe
supported in Sariputra's naive, devotional view. . S(;.
The specific vocabulary of the 'Jewelled cosmos"
reveals a curious stance toward these ontological and religions;2
questions. Rather than speaking of the Buddha "creating" tht
series of apparitions, the text prefers to say that he simply!
"shows" them (stan). Is the Buddha the actual creator ofhif'
buddha-field, or merely its presiding genius? There is evidence.'
in the text to support both positions; side-by-side with the con(
sistent use of stan to imply a mere manipulation of what already'
exists, we find a curious passage in which the Buddha speC\ks,
of a bodhisattva "making" (bya ba,' Osh 151) a buddha-field to.!
ripen beings. But the use of stan in adverbial phrases ("showas\
X") is common and consistent enough to suggest that the text
is generally more comfortable in describing the mode of an .
ran"ce than in assessing its ontological value. Rhetorically,
uses stan in a ontol?g-
.fa11y its use seems to mdicate a certam mdeCIslOn regardmg
Ie' f _. D h h h
,\ . statuS o upaylC magIc. oes t e text mean to say t at t e
of the jewelled cosmos is a revelation of the "true" nature
rthe buddha-field-and thus of the ,Buddha's mental state?
The Buddha puts it this way: "Sariputra, the buddha-field
like this; but for the sake of ripening lower beings, the
shows it as contaminated by many faults" (Osh 154).
'ithe notion of "like this" is conveyed in Tibetan by 'di 'dra,
$rqbably a tad iva or a similar
$llrase whose ImpreCIslOn m the source languages IS much lIke
fits imprecision in English. We have no clue as to whether the
of the jewelled vision to the buddha-field is one of simile,
Groimesis, metaphor, or identity. In short, even in this pivotal
the ontological status of the upayic vision is left open

07:" We also detect here a subtle shift in the Buddha's rhetoric:
at first he insisted to Sariplftra that the impurities in
:ihe cosmic vision were the result of Sariputra's own deficiency,
;;Howhe suggests that these impurities are his own upayic crea-
::tion. Yet these two positions are not necessarily contradictory.
;ltis, after all, the translation of ordinary reality through the first
and then the second which enlightens Sariputra-the
stages taken together form a soteriological unity, and thus
credibly under the rubric of upaya. While the text sidesteps
firm ontological analysis of the various levels of visions,
it clearly exposes the two key aspects of the upayic process:
the importance of the beholder's karmic disposition; and
:;second, the process of progressive insight as steered by an en-
;lightened being. The incessant use of visual language (ston,
roots the Buddha's upayic magic firmly in the
;phenomenological realm; though upayic conjurings may pro-
;yide their beholders with a realization of ontological emptiness,
own ontological status remains indefinite. As the case of
demonstrates, it is the magical-upayic process, as rep-
in the three stages of the jewelled cosmos vision, which
ers most-for it is only by considering a juxtaposition of
that the disciple is able to see ontological questions in a
,new light. No single vision provides the necessary point and

counterpoint for this characteristically Mahayanist
By the same token, the problem of the Buddha's own rno9:1\
. h' 1 .. Dh ki . ental'
punty--or IS egltlmacy as a arm a- ng, to VOIce the oh';\i,
implications-is deftly skirted by an invocation of skill in rn
. By transforming the variegated cosmos a jewelled
the B.uddha shows. not of hiS own butili,sj,
of. observes m the worlcf";;
These impunties have their ongm m upaya, as the rnetarnoii?,
phosis of upaya proves. .;,5{,'
With this established, the theodical pro?lerrr. is also
the Buddha causes the buddha-field to mamfest Impurities soler);
in order to lead lower beings to liberation. With the
possible economy, then, the VNS is able to transmute a challeng6;
to the Buddha's legitimacy and good intentions into a vivid
proof of his compassion. By the close of the first chapter
text has apotheosized the Buddha as a supreme manipulatorbr
phenomenological realities, a compassionate magician Whose'
intentions are always salvational and consonant with
Dharma. But what of Vimalakirti, that magician of equally irn::;
pressive powers? 0":)

. ;',:
" f:-
VI. Illness as Metaphor
With the introduction of the Bodhisattva Vimalakirti in the'
VNS's second chapter, a second dramatic epicenter beginst6.
emerge, one which presents a rhetorical counterpoint to the
Buddha's actions and words. Having elevated the Dharma-king
to a stature of religious supremacy and established his seataF
Amrapal1 as the center of the cosmic array, the text shifts quicklY
to Vimalakirti's house within the city of Vaisali, outlining the.
Bodhisattva's character through an elegiac recitation of his quaP.
ities and attainments. From the start. he is described through
contrasts, oppositions which have the effect of casting his life"
as a continuous play between illusion and reality: he wears
layman's clothes, yet lives as a sramar;a; he has a wife, a son and
harem, yet practices purity; he appears to be surrounded
retinue, yet practices solitude; he appears to eat, yetis
by the virtues of his meditation (Osh 155-6). Once again,
we have a rhetoric of appearance: the spiritual being
one thing while actually doing its opposite. Nor is there
as to duplicity. We to.ld
that Vimalakirtl lIves III Valsah for the sake of npemng sentient
beings throl)gh upaya (Osh 155). If he to games and
t!ambling houses, It IS for the sake of convertmg garners and
farnb1ers; if he into it for the sake of
tbe evils of deSire (Th 21). HIS actIOns are not capnCIous but
are, on the contrary, carefully crafted as salvific strategies.
:f'D We have here an upaya which, while grounded in the manip-
of .also itself in worldly life. The Buddha
hls upaya m essentially two modes, the spectacular and
the verbal-homilitic; this is appropriate to his dual role as a
cosmic or semi-docetic saint, on the one hand, and a teacher of
iDen on the other. By contrast, Vimalakirti, as a bodhisattva,
kmploys his upaya initially through his worldly actions, prefer-
Hng to teach by example. This aspect of his spiritual praxis is
the main theme of our introduction to him as a protagonist,
'lIrid by its position in the_ text stands in sharp contrast to the
:[potheosized Buddha at Amrapali.
Immediately, however, the relationship between the
'Buddha and Vimalakirti is complicated. We learn that Vim-
hlaklrti, through his "skillin liberative technique," "shows" (bstan
himself "as if' (lta bur) ill (Osh 157). This illusion-the
hint of what will soon emerge as a formidable power of
illusion-has the effect of drawing visitors to his house to inquire
'after his health. Their visits give the malingering Bodhisattva
pretext for a discourse on the emptiness and imperma-
of the human body (Th 22-3). To undermine the human
:trust in the body's reality Vimalakirti characterizes
;it as prone to disease, ownerless, inert, selfless, lifeless, unreal,
.yoid, insensate, filthy and false; in a series of similes he likens
itto a mirage, a machine, an illusion, a dream, a reflection, an
a cloud, and an old welL This didactic interlude, besides
opening the text to a common universe of Mahayana imagery
concerning emptiness, 12 deepens its preoccupation with illusion.
In clear contrast with the action of the first chapter, in which
the Buddha displays both verbal and magical upaya in grand
Vimalakirti's first exercises of upaya are quite modest and
only subtly magical. They serve in all cases merely to set up or
introduce a doctrinal discourse, and no claim is made for their
--------- -----------------
102 jIABSVOL.llNO.l
own power to enlighten. Whereas the beholders of the
. 1 . 1 b h . h b arasolt
an Jewemrrac es are roug t to eplp any y the
of the. visual thos: :vho
sory sIckness conceIve the spmt of enlIghtenment only fro ..
b 1 d
' h' h h ". k" d l' III tIle'l
ver a Iscourse w
lC .c t SlC man e Ivedrs. illness
an elegant metap or lor Impermanence an Splfltual SI'ckn"Jii",,;ti
its meaning comes out the
Moreover,.though It IS COnjured,'!;"
there IS nothmg outwardly magICal about the
position; far from being a bold celestial spectacle of the
type, it is only a homely affliction. The contrast of upayic
though it will seem to vanish as the text moves forward,
scores an important tension between the buddhic and}:}
bodhisattvic paradigms of the VNS. As we will see, this tensioJ:;:
drives the story forward, not only dramatically, but philosophi{{
cally as well. ".:<.;;

VII. The Gadfly of Vaisiil'i

The two long chapters following that in which
is introduced are dedicated to an exploration of his verbal,
chiefly non-magical, upiiya. As the disciples and then the1
bodhisattvas in the Buddha's retinue relate their tales of
ation at Vimalakirti's hand we begin to get a clear
the gentleman from Vaisali: as something of a
relentless debunker who makes it his business to confront
Buddhists and challenge their conduct. In this respect V{j
malakirti must sometimes remind us of Socrates, who loiters aU
the fringes of Athenian society to unmask hypocrites and attac(:
the cant of the Sophists, One by one the Buddha's followers)
beg to be released from their assignment to call on the invalid;'
and as their stories unfold and interweave, the figure of Vi<
malakirti grows to larger and larger dimensions, challenging.;
that of the Buddha himself. Moreover, the very length of the.,'
"reluctance" chapters (twenty-nine folios in the Peking edition):
builds an element of suspense into the sidra, pointing ever mor,e"
strongly to the absent Vimalakirti in his sickroom within in the,
city. By withholding him from us the text intensifies his mystique}:
and effectively builds anticipation toward a confrontation
later point.
dramatic technique reinforces the general tension be-
buddha and bodhisattva which we have been highlighting.
<lW h d . 1 f
that t. e ISCIP es use a request of the
all, thelr Dharma-kmg-cannot but create a dIrect rIvalry
" 'een him and the Vimalakirti they dread. The implication oj
anecdotes is that even a mere bodhisattva's power can be such that
ddha's wishes must be lift unanswered-a profound implication
, especially as the reluctant disciples must refuse the
to his face:
the dISCIples temponze, of course, they repeat fragments
challenges and sermons, thereby illustrating the
i';i&tiety of his upayic devices. Yet for the most part, the incidents
i:lia've nothing particularly magical about them; the Bodhisattva's
is of the dialectical or confined. to ,:ell-
*chosen words and retorts.
The mam pomt of the sectIOn IS to
Vimalaklrti's way life: attitude as. he moves
kHlrough the lay world, and to bmld him mto a compellIng charac-
with a formidable personal mythos.
must assume further, though, that the passages are
to describe Vimalakirti in soteriological or hagiographical
If he is not an overt hypocrite-if he practices what he
admonitions provide us with a clue to his own
of attainment and place within the Buddhist spiritual
$ierarchy. Can we assume otherwise when he advises Sariputra
yourself in contemplation in such a way that you
lmimifest all ordinary behavior without forsaking cessation" or
fito"manifest the nature of an ordinary person without abandon-
fririgyour cultivated spiritual nature" (Th 24)? These admonitions
perfectly with the initial descriptions of Vimalakirti in
second chapter. In remonstrating with Mahama-
in fact, the roving Bodhisattva gives us a succinct
of his own present activities: "You should be adept in
to the spiritual faculties of living beings ... you should
the Dharma in order that the continuity of the Three
1Jewels may never be interrupted" (Th 26). He flourishes his
. ability to intuit "the spiritual faculties of all beings" in the pres-
of Puma, causing some young monks to recall their prior
i;llves and recognize their hodhic natures (Th 28-9); PUfl).a con-
',fludes that Vimalaklrti surely must be something more than a
sravaka, for the psychic feat he has just witnessed is onf'
traditionally belongs only to a tathagata. This
. is boldly. by Upali, ,;ho warns his
lOllS: Do not entertam the notIon that he IS a mere
Why? With the exception of the Tathagata himself there is no disc'
capable o.f cOI?petir;,g with his Or
the oEhiS wIsdom (Th 31,. ItalIcs AIld.[;
mdeed, the senes of tales we hear about Vlmalaklrtl's litigio;':'
encounters with the disciples would seem to confirm
assessment. '.' .. ,} .
I noted above that the upayic mode for the sravaka chapte;
is almost exclusively verbal. But as we move
bodhisattva chapter we begin to notice the intrusion of small
magical motives, minor upayic spectacles which prepare
SrI, The fmal two pencopes of VimalakIrti s humIlIatIOn of
bodhisattvas-those involving J agatirpdhara and Sudatta-offe,{,i
pointed parallels to two of the Buddha's powerful feats, leaving:;)
no doubt that the man from Vaisall is in some sense a rival of';;
the Blessed One. In the J agatirpdhara incident,
able to unveil a magic trick of Mara's, exposing the BUddha;s',
old foe and leaving him hoist with his own petard. Not
does the Bodhisattva unmask "Indra" as the evil tempter, but '.
he takes the opportunity to convert the twelve thousand maidens':
with whom Mara had meant to compromise Jagatlrpdhara (Th',
37-8). This is certainly a symbolic homology of the Buddha's;
famous defeat of Mara during the night under the tree of en-"
lightenment; it implies that Vimalakirti shares with the Buddha:
an. ability to identify the personification' of the adharma andO
uproot it.;.
To complete the homology and thread it once more into,'
the present-time storyline, Vimalak"irti is described by Sudatta ....
as having performed a magic trick very much like the Buddha's
conversion of the parasols in the first chapter. The miracle . ,'.
reveals a universe called MarIci and shows a pearl necklace;
transformed into a pearl-inlaid pavilion (Th 41). Though the (
specifics are different from those of the parasol miracle, the ....
formal structure of the spectacle is much the same: like the:
parasols, the pearl necklace is a devotional gift; like the cosmic:,
canopy, another universe is displayed magically; and as with
of the parasols a giant one., . the
;li;!lnaIllent of pearls IS transformed mto a smgle large pavIllOn.
implication of the incident is that Vimalak'irti's
, when employed, are comparable to
of the Bud a. a l?ere can, like
Buddha, freely mampulate realIty to serve hIS upaylC ends.

Emptiness and Upaya
With Mafijusri's call on the invalid Vimalakirti, the two foci
narrative finally begin to converge. Mafijusrl, the only
bold enough to take on Vimalaklrti, does not hesitate
';;tqengage him in a lively philosophical conversation. The topic
the entry point, Vimalakirti's illusory illness. Once
the Bodhisattva exploits his metaphor to the fullest, argu-
;';i#g that his illness is merely a reflection of the more basic ill-
i:i1iess-the spiritual cupidity-of sentient beings. Here at last he
calls himself a bodhisattva, and admits that his affliction
;;;1s'1 device of maya. These clarifications, following rapidly upon
another (Th 42-3), point inevitably to the central doctrine
But they also provide an opportunity for Vimalakirti
emptiness with the bodhisattva path, the path of compas-
J;sionate upaya. Specifically, emptiness acts as a corrective to "sen-
compassion" or a "compassion of attachment" (phan
iyon'du Zta bar . .. snying rje; Osh 183; Th 46). In this sense,
}:emptiness guides the bodhisattva, allowing him to be compassion-
and energetic in his dealings with others without clinging
;!'t() these relations.
r'; In this discourse we see a new relationship between upaya
sunyata. In previous instances of upayic magic, the alterna-
.<tion of "natural" phenomenal appearances and plainly spectacu-
liar appearances functioned to reveal ordinary phenomenal re-
;;ality as essentially empty, analogous to the coruured images or
in a mirror. Now, however, Vimalakirti argues that
lthe understanding of emptiness can guide upaya itself, maintain-
,jng its appropriateness in a philosophical sense. Because the
bodhisattva regards all beings as empty, there is no danger that
he will become attached to them in the course of discharging
,:his upayic duties. True compassion purifies the bodhisattva to
the point where his voluntary incarnation resembles the
liberation, or nirvary,a (Th 46); cloying compassion, on the
hand, sends him through a cycle of reincarnations fired
karma which spurious atta.chments UPaya,
not only mampulates and dIsplays emptmess, as it
takes its cues from emptiness and is circumscribed by it. '
ness acts as a moral force for the bodhisattva as he moves
the world on his mission of liberation, protecting him
backsliding and helping to keep his vision pure.
IX. The Magical Thrones
With this key point established, Vimalaklrti is freed to LAr.r.,;,
cise his magic to its fullest extent. The Bodhisattva's
now grounded both morally and metaphysically; it remains
to distinguish it from that of a buddha. As the head-on
between Vimalaklrti and the Buddha draws closer, the VNS
explore Vimalaklrti's bodhisattvic magic with closer
integrating it with the bodhisattvic paradigm laid out in
"reluctance" chapters. The most important locus for this
nation is a miracle. performed in the middle chapters: the
portation of the lion thrones into the Bodhisattva's
Sariputra, the hapless Pierrot of the sidra literature,
himself the butt of several gentle drolleries once he
malaklrti's house. One of these serves as the trigger of the
spectacle: looking about him, the practical-minded
wonders where the crowd of visitors will all find seats.
alaklrti reads his mind and teases him, "Reverend
did you come here for the sake of the Dharma? Or did
come here for the sake of a chair?" (Th 50). The two
engage in a brief exchange which interprets the question
chairs as one of attachment to the physical body. In the
however, Vimalaklrti honors Sariputra's wish by
transporting 32,000 thrones from another universe
room. The miracle exhibits a number of "impossible"
the thrones, which are of large dimensions, arrange
in the house without any crowding; the house enlarges
accommodate them; the city is not obscured by the
51':"'2). The more advanced bodhisattvas transform their phys-
bodies so as to mount the lofty thrones; Vimalakirti gives
special teaching which enables :hem to
SUlt. respectful bow to the Tathagata of
the sravakas to? are to seat then:selves.
,,';! this IS accomplIshed through Vlmalahrtl s legerdemam, and
world is left just as it was before.
Sariputra revels in this achievement, prompting the
,i, clhisattva to dilate upon the nature of his powers over phe-
reality. One who has attained the "inconceivable liber-
(acintyavimoksa) has the ability to alter the appearance of
and time at will; he can put Mount Sumeru, the largest
li:;frnountains, into a mustard seed, or make the passing of a
seem like the passing of an aeon (Th 52-3).
Critically, however, the efficacy of these miracles depends
the karmic receptivity of those who are meant to experi-
them. Thus, even the most ambitious upayic magic is sub-
the of 'Yhile the text
that thIS upaylC dlSpositlOnahty IS the responsIbIlIty of
himself, the present passage implies that disposi-
}r,.Honality is rooted in the karmic matrix of the world around-
in the subtle interplay between the magician'S power
maya and the karmic context of the beholder.
Such a necessity begins, in fact, with the very notion of the
miracles appear as miracles only to those who are
to be disciplined by miracles" (Th 52). Not only is the
setting and form of the apparition a function of the
karmic disposition; its very existence, its
lj!'phenomenological possibility, depends on the presence of a re-
mind. The present passage suggests, then, that upayic
is more than merely a positive quality added to
by its creator: it is a root condition without which the
J;tlIJiracle cannot come to be. The dogmatic implication of this
?:/principle is that the miracle of an enlightened being cannot
from anything other than upaya. Gratuitous magic-the
sport or lila of the classic Indian gods-is not the magic
:;Ipfa buddha, nor that ofa bodhisattva. Magic has a purpose, and
U.:this purpose is rigidly governed by the force of the karmic situ-
in which it is unfurled. 14

X. The Goddess
, The deservedly famous chapter on the dialogue
Sariputra and the Goddess contains,' besides their notew
d d b f
' 1 d h' h OllllY""
pas, e a er 0 mter u es w IC press forwardke'<t{l
phllo_sophICal Issues, Foremost am?ng these
Mahayana paradox of empty compaSSIOn: If abodhzsattva
all beings essentially empty: how why
exerCIse compaSSIOn toward them? VImalakIrtl's answer is t<\!:<,
d h
' b dh' h Id 11 h'
an to t e pomt: a 0 lsattva s ou reca IS own
knowing his own emptiness, and help others to attain
while knowing theirs, In short, the bodhisattva's empathy
world beings allows
hIm-to a lazy m!uhsm to theIr aid,
or not, suffenng IS an experzenced reahty whICh can be mitigated
with proper insight and application, .. !/.;::,:
This by :-eturning t? the
and acceptmg ItS lIved value If not ItS vendI cal value, dovetails;.,'
perfectly with the metaphysical position the text has
toward upayic by e,mphasizing t.he cognitive
of both suffenng and magIC, the VNS IS able to propound.a:J
rigorous philosophy of emptiness without falling into pessimisni'\
or passive nihilism. Here, too, we find the basis for the
moral view of upiiya: the bodhisattva's magical means are not:
arbitrary, though they may be founded on emptiness; on the
contrary, the karmic matrix which underpins the'
phenomenological world guarantees that they will never turn,
pernicious. Because the dispositionality of upiiya is in some sense
beyond the magician's control, arising as it 'does from the karma
of the beholder, magic is always experienced in such a to
serve the soteriological interests of the experiencer.
We might say, then, that upiiya finds its justification on the
ontic plane.
Importantly, however, it succeeds soteriologically
by shifting its beneficiary from an ontic and conventional view
of reality into an ontological mode: that is, it makes him aware
of the essence of his ontic experience, which is only emptiness,
This play between ontic and ontological, besides allowing us to
grasp the rhetorical structure of the text, provides the basic
paradigm of the Mahayana Buddhist epiphany. After each in;
stance of magicalmiiyii, some or all of the witnesses have a
went of awakening in which they see, in a pregnant sense,
underlying structure of reality. Magic creates a disruption,
in the unbr?ken continuity of ontic awareness,
it is through thIS dIsruptIOn that the nature of the phenom-
world is disclosed as emptiness. ,
episode of the Goddess and Sariputra attacks one of
wost types of immer,sion o.r blind.ness: the satkii-
l;;;1]d,f$ti, or VIew of the body s realIty. As m S.addhar-
a sex change the of
'the body's mamfestatIOn; and also as m the Lotus, It IS Sanputra
triggers the spectacle.
By the Goddess's magical power
,J6yin gyi brlab pa; Osh 1 99) is made to appear (snang) as the
herself (lha mo de cz dra ba; Osh The Goddess's
'prankish conjuring is intended as a reply to Sariputra's inquiry
to why she does not transform herself (bsgyur) out of her
state;, when the miracle has taken effect she taunts him,
i:IReverend Sariputra, what prevents you from transforming
from.rour female (Th 62).. . .
The ensumg exchange brmgs the rhetonc of lllusIOn back
:"t6center stage, arguing that one's physical appearance is mere
f#yiya (Th 62). Philosophically, Sariputra and the Goddess con-
Mude that something unmade cannot be changed-and that all
{things, like Sariputra's female form, are neither made nor
changed. This conclusion is foreshadowed in the very opening
;Sieps of their dialogue, as the explicitly introduces the
i:theme of magical maya: "Reverend Sariputra, if a magician were
incarnate a woman by magic, would you ask her, 'What pre-
iyents you from transforming yourself out of your female state?'"
'(Th 61). The word for "magician'" here is sgyu ma mkhan-liter-
;klly, "maya-adept." That Sariputra is himself subject to the va-
garies of illusion, even in his physical being, is a break in his
qntic continuum which he can hardly ignore. The philosophical
'thread of the scene is emptiness; rhetorically, it serves to create
.a transition from the still dominantly verbal upaya of the middle
.chapters to the ambitious visual upaya to come. Whereas the
'miracle of the transported thrones focused attention on the role
of the individual's karma in governing magic upaya, Sariputra's
transformation radically undercuts any tendency to identify this
individual karma with the physical body. Scene by scene; a clear
;theory of upaya emerges, preparing the way for the magical
finale of the closing chapters.
XI. Two Emptinesses
. Chapte:-s Eight and Nine, coming in the wake of
mterlude wIth the Goddess, further pursue the lesson of
tiness she has given him. In contrast to the disciple
plodding Chapter Eight gives .us a
bodhzsattva whose behavIOr IS a perfect expreSSIOn
metaphysical understanding:':?:
Of the true bodhisattvas,
The mother is the transcendence of wisdom,
The father is the skill in liberative technique;
The Leaders are born of such parents. (Th 67)
While dilating upon. the bodhisattva's :profound
of the nature of realIty, the verses whIch compnse the bodyof:I';
the chapter also stress his worldliness, his self-chosen
ness in the sarp.saric order. The bodhisattva metamorphoses
self into a hell-being, an animal, a courtesan, a head of state,";[i;%

holy man-whatever guise will serve him in developing liViN(l:ip
beings. His upaya expresses itself in a willingness to
shapes, to melt into a series of magical masks, to utterly
his personal identity. In this sense it becomes, of
object lesson in nairatmya, or the unfoundedness of
self. The bodhisattva is emptiness in action. Though always
ent in times of need, though always energetic in easing the,;':j
suffering of those around him, the true bodhisattva is
empty-and this ineffable quality is, perhaps, his
metaphysical teaching: ':I,,!;
They manifest birth voluntarily,
Yet they are not born, nor do they originate.
They shine in all the fields of the Buddhas,
Just like the rising sun. (Th 68)
Chapter Eight, then, establishes both the bodhisattva
and its metaphysical emptiness. In Chapter Nine, the
of emptiness receives a more abstract dialectical treatment,';

shifting from an ontic account of the bodhisattva's path
t"pOIO . 1 d b h h" h Dh d
ontologlCa e ate over ow e enters t e arma- oor
(Th 73). While Chapter Eight clearly demon-
the emptiness of the person, Chapter Nine concerns itself
?:,stL . f Th . fhd .
emptmess 0 concepts. ese two wmgs 0 t e octnne
shadow the text's dichotomy of visual and
;v:g'ihal upaya-lay the groundwork for the meeting of the
famous.silence, far from
all possIbIhtIes for verbal dIscourse m favor of purely
i5fihive upaya, actually unites these two emptinesses: part of verbal
after all, is knowing silence, or inscrutable
, can be eloquent. The skillful alternatIOn of dIscourse and
liaction is precisely what defines Vimalakirti's bodhisattva practice.

lfi'jtrA Meeting of Magics

The final three chapters of the VNS witness a long-awaited
between the Bodhisattva Vimalakirti-empty though
be-and the Buddha. Here at last we find a dramatic
of the tension which has been rigorously maintained
the text's two chief spiritual figures: as the invalid trans-
himself ane!. his visitors (by magic, of course) to the
seat at Amrapall, we begin to see the intent of the
established homology between the two characters. Two
of sainthood are at play here, and the figure of Vim-
i[;lJ-!akirti, by a subtle refraction, illuminates for us an important
or historical motion in the Mahayana concept of bud-
The interplay of the two saints, while plainly meant
;'lovalorize the bodhisattva as a character model, also reveals some-
significant about the Buddha's changing status.
it;; Chapters Ten and Eleven, those leading up to the final
serve to revive the theme of upaya, fitting it into the
;fosmological scheme with a chain of references to the multitude
systems and their respective ways of life. Vimalakirti's
power is so vast that he "sends magical incarnations to
the buddha-fields of the ten directions, and all these incarna-
accomplish the buddha-work for all the living beings in all
buddha-fields" (Th 80); but by incarnating voluntarily in
,:::;the lowly Saha world, full of the "wild and uncivilized," the
has the precious opportunity to
tues whlCh_ befit only such a rung of eXistence (Th
In the Saha the Buddha fmds unadorned verbal
to be the most sUItable mode of upaya, for the most part
identifying should be done and what (Th
Bo?hlsattva,. too, .finds great salvatlo?-al power in
pr.actlCe of simple ethICal of a .non-magl:al
thIS passage, a m.ore conventIOnal View of upaya
magICal upaya WhiCh permeates the rest of the text; it serve:'!
. d h h h .. f s to.
remm us, per t . at t e ,:,anetles 0 are as
the buddha-fields ,:,hICh bodhzsattvas act upaYICally. Once agciiriii
we have a steady mSIstence on what I have called the
ality of skill in means: the karmic context of its exercise
mines the specific character of upaya, guaranteeing its liberativ "
efficacy. Ie
Despite the Buddha's stated on verbal teachinf.'
however, the present text abounds In mIracles. Why? As .we:
have seen, the various miraculous happenings which
narrative forward serve primarily to catalyze or provoke a
response: a question, a dharma-lecture, a dialogue. It is
that, as the saying has it, "miracles quickly convert the commori, ..
. man" (asu prthagjanasya rddhiravarjanakarZ)18; rather, it is
of miracles provoking the beholder into metaphysical
The interplay of visual and verbal rhetorics is most pronounced.
when Vimalakirti at last betakes himself and his company to the:
Buddha-and it is here, too, that the subtle differences
Vimalak'irti's magical upaya and that of the Buddha become
most apparent. J"l,:
Chapter Eleven begins with a small but portentous miracle: "
the Buddha's surroundings at Amrapali: begin, of their
accord, to expand, while the assembled disciples begin to 1
on a golden hue (Th 84). The Buddha interprets these signs as
an indication that Vimalaklrti is coming to visit, but the text{
leaves the source of the miracle unnamed. Is the omen the resulL!
of Vimalaklrti's magic, the Buddha's, or neither? We are riot
informed. But Vimalaklrti soon confirms the Buddha's surmise
by magically transporting himself and his company to the
den. Upon his arrival the Bodhisattva makes the customary
gestures of reverence to the Tathagata, prostrating himself be-.
fore the Dharma-king and assuming an uncharacteristic and
silence. The vastness of his own upayic powers does
eprevent him from rendering proper homage to il miglior
>this setpiece is followed by a brief dialogue between
'uda, the Buddha and Vimalaklrti concerning the ambrosial
which bodhisattvas have taken lunch which now
!.',.".'.'..',p..sfrom theIr pores to perfume the aIr. Doctnnally, the core
see ., h "fi f h'
diSCUSSIOn IS t e slgm lCance 0 t IS amrta as a metaphor
upayic like a
i:1l1enectar stays m the dnnker s body untIl he has attamed the
appropriate to his spiritual stage (Th 85). For beings
have not yet formulated the thought of enlightenment, the
will be digested only when they do so; for those who have
conceived the bodhicitta, the nectar will persist until they
so on.
remains for the Buddha to expand Vimalaklrti's explana-
to describe the general principle of dispositionality through-
the buddha-fields, as well as the unity of all buddhas despite
iiQeir outward difference (Th 86-7). This discourse prompts the
bodhisattvas from the ambrosial buddha-field to ask the
for a final teaching to take home with them. In what
jJfollows, the Buddha brings together the ontic and the ontological
of the text, arguing that, ontically, the bodhisattva must
['plunge into the world of compounded things ('dus byas; Osh
while ontologically he must not rest in the uncompounded
';['dus ma byas la mi gnas; Osh 225). This double mandate throws
',!rIto high relief the career of the bodhisattva as an engaged being,
who comprehends the ultimate nature of things yet
applies himself to salvational work in the world. The
,entiretext, in fact, when considered from the rhetorical perspec-
has exemplified this position, skirting a full-blown ontology
'iufavor of a persistent concern with praxis. Though the VNS
;iscertainly not without its philosophical interludes, in compari-
with other Mahayana works of its time it is rather uniquely
::cohcerned with playing out the actual practices of the
.bodhisattva's life, deferring many scholastic problems which
might blur its pragmatic vision. Needless to say, this vision is
,largely one of upaya .
... ' ." . It is in the final chapter of the VNS that the career of the
ybodhisattva-and his soteriological value-are given their ulti-
114 JIABS VOL. 11 NO.1
mate vindication. The Bodhisattva Vifualakirti stands
Blessed One, surrounded by a multitude of admiring
As if in challenge, the Buddha asks: "Noble son,
wou!d the you (Th 91).
alakIrtl s answer IS hIghly sIgmficant-it IS, m fact, the
his own dramatic and philosophical function within the t
. to the Saint the is
mg but an mterplay of OpposItes, an mexpressIbl: being
eludes human concepts: h.e does not dwell
n?r anse the past, nor mto the future;. he
wIth the realIty of form, yet IS not form; he abIdes m
reality (de bzhin nyid; Osh 227), yet between it and him there:,1'i\
no connection (Th 91). Not devoid of characteristics, he
possessed of them either; he is neither name nor sign; he Js':s
neither good not evil. On the whole, he is simply
(don du 'ang brjod pa ma lags pa; Osh 228). No verbal
can express him (Th 92)'/i':
The Buddha as painted here is ineffable; even a
as advanced a Vimalakirti can barely grapple with his
being. As we will see presently, this view of the
so far removed, perhaps, from the docetic view of
creates the need for a more manageable character model,;'!,;'
personification of enlightenment whose (apparent)
manifest and approachable. The final miracle of the VNS
to establish the Bodhisattva as the man for the job:
between the buddhic order and the samsaric, the bodhisattva acts .
as a religious interpreter, a sort of link who makd
enlightenment comprehensible to those in the lower worlds.
Moreover, this pivotal role for Vimalakirti explains much"
about the construction of the text itself: in particular, it helps
to account for the subtle and tireless parallelism which is main-
tained between Vimalakirti and the Buddha, on the one hand,'
and Vimalakirti and the common householder on the other.
The Buddha himself plainly realizes this important function .'
of the Bodhisattva, for he overtly countermands Vimalakirti in
order to emphasize bodhisattvic embeddedness in the sarp.saric
order. Sariputra asks Vimalakirti where he might have died in.
order to reincarnate in the lowly Saha world. In reply, and',
perhaps with a bit of dry wit at Sariputra's effrontery, Vimalakirti.
promptly deconstructs the question, arguing that since all things<
magical creations, he cannot have been rein car-
;fLi d (Th 92-3). Surprisingly, though, the raises his
"nate V 1 1- ., d . S .
to sidestep Ima a nrtl s non-answer an gIve anputra a
I . I 1 V 1 k- . h . f h
, ogICa . rep. y: Ima a lrtl, . says, IS. rom t e
Ulllverse, whICh IS under the spintual gUIdance of
. _.,. .
VimalakIrtl s dlssemblmg, then, the Blessed One
him sq,uarely spectrum,
Sariputra s questlon on Sanputra s own level. ThIS leads
to relent and offer a more ontically comprehensible
his own_ reb.irth, admitting that he has i?- fact
;"teincarnated m the Saha umverse, whether he and the umverse
mere maya or not (Th 93).
The assembly then wishes to see the Abhirati universe. The
reading the thoughts of the disciples, orders Vim-
to reveal it: or, in more interpretive terms, the Dharma-
commands his proxy to perform some magic for the good
iOLthe multitude. Vimalakirti executes the miracle as ordered,
;}iuid the company is rightly astounded. This accomplished, a
ghly significant exchange occurs between Sariputra and the
(Buddha, the final dialogue of the text proper: the Buddha asks
if he has seen the conjured universe, whereupon
';Sariputra extolls Vimalakirti in terms which make the
rtodhisattva's religious function clear at last:
I saw it, Lord! May all living beings come to live in a buddha-field
as splendid as that! May all living beings come to have miraculous
powers just like those of the noble Licchavi Vimalaklrti!
We have gained great benefit from having seen a holy man such
as he. We have gained a great benefit from having heard such
teaching of the Dharma, whether the Tathagata himself still ac-
tually exists or whether he has already attained ultimate libera-
tion. (Th 95)
Sariputra's speech reveals a certain metaphysical doubt with
regard to the Buddha: though standing before him, and indeed
being addressed by him, the ordinary disciple is simply uncertain
whether what he sees as the Buddha is real or merely a trick of
magic. Vimalakirti, however, despite his own Madhyamaka
rhetoric to the contrary, is real enough-or seems so-to have
116 JIABS VOL. 11 NO.1
a profoundly Hberative effect upoll his
Here, I believe, we have the essential division
Buddha as religi?us. teachers. The
portrayed In the VNS IS a cosmIC fI'gure, well on his w ..
docetic apotheosis; Vimalakirti, as his proxy in the
h h h Id . f h ld pace,.!
an In t e ouse 0 ,IS a man 0 t e wor
magic but engaged in .the external affairs
mon men. We mIght even wonder, In fact, whether Vimalakr:,l
himself is a creation of the Buddha's, an upayic vision
conjured in order to infiltrate the everyday world. But
ical possibility aside, the Buddha's use of
spectful praise for him, his insistence on his worldly
ness, his deferral to Vimalakirti's magic in the final scene--':'is;
plainly upayic in the wider sense. The Buddha seems to
Vimalakirti is a more approachable figure for the avera.gei"
dISCIple, a character model one can actually hope to emulaie.Zj
In this sense, Vimalakirti secures the worldly foundation()f;
religious life while the Buddha himself is raised to a more aha:
more docetic or metaphysical status.

Conclusion .
The VNS presents us with two of the most compellings
characters in the early Mahayana literature, each drawn witH;
remarkable finesse and considerable philosophical rigor. Byt
playing them against one another through mechanisms of plot.
and rhetoric, it maintains a sharp tension between them, high-
lighting certain parallelisms which unite' them while acknowl
edging their profound differences. The Buddha is no
the humble teacher of the Pali texts; he is rapidly
supramundane, divinized figure, whose earthly presence is built
on the shaky edifice of maya. Against this apothetic upwarc!
pressure the VNS installs the Bodhisattva as anchor, establishing;
him resolutely in the sar:p.saric world as a constant exemplar of
the dharmic teaching. Vimalakjrti stands at the critical point of
juncture between a 'cosmic order and a conventional one, acting
as a living embodiment of Nagarjuna's two truths, ultimate and
relative. His life, though immediately accessible to a layman of
Vaisali, constantly intimates the greater ontological order, like}
slowly wearing away to reveal an older metaphysical
(c1 tf1.lth. .
The chief rhetorical mode of the VNS is, of course, verbal.
alongside t?e runr:ing all
like a' phIlos0.r:>hIcal IS a VIsual rhetOriC
force, of spectacle. LIk.e ?f upaya, thIS
rhetOriC exhIbIts a finely-tuned dlSposltlOnahty, conform-
to the contours of its karmic
;i'lhg differently to beholder. thIS diSposltlOnahty
to be a functlOn of the pedagogICal sItuation or encounter
not merely a skill the upayic teacher has learned. In this
other ways, the miracles of the text display a structure and
%ecessity of their own, hinting that maya, even as illusion, must
course have a certain logic.
:it;. Philosophically, however, this possibility is never addressed.
::iinptiness becomes a way of describing the mode of appearance,
'J'Aot its ontological status. And this emphasis on modality is per-
;Jectly consonant with the upayic methods of the bodhisattva, for
f$hilosophical constructs cannot, we are told, capture reality.
r';6' The VNS is eminently a teaching in motion, a text which
as much through its plot events as through the dialogues
its protagonists. To grasp its philosophical orientation cor-
F;ectly it is essential to give due weight to the narrative and
structures which underrun it. And it is this quality-an
quality, to put it most simply-which sets the Vimalakir-
;'tinirdesa in a class of its own, making it one of the most enjoyable
'Mahayana texts to read and reflect upon. To the very extent
',.that it can draw us in with its beauty and humanity, its own
:upayic power stands undiminished by time.
1. In the use of these terms I mean to suggest that the predominantly
yisual materials of the text-particularly, of course, the moments of magical
upaya-are governed by a logic of persuasion much as are the verbal dialogues.
This position is inevitable if one accepts the upayic intentionality of the spec-
tacle-a hardly radical position in light of the fact that each major spectacle
is followed by a report of religious epiphanies experienced by its beholders.
In referring to the "discursive" rhetoric of the text, I mean the text's
'verbal elements taken in the widest sense: philosophical discourses, laudatory
118 JIABS VOL. 11 NO.1
verses, tales, and even the Goddess's playful jibes at Sariputra. ",';\1
2 The VMS does not survive in Sanskrit; it is generally Delt th ;eG.'.'
.'. .. at
canomcal TIbetan verSIOn gIVes the most coherent and fluent reading a ?\'r
the surviving editions. In what follow .. s I have adopted two systems for
. I r I h R Th ' I 'd ntl -,i
mg textua relerences. n cases were. urman s UC! translationr,
Holy Teaching of Vimalakirti, seems to me to capture the sense of the Tib3e;;i'
"11 . . h r (Th ) . h etan: '\
accurate y, WI CIte passages m t e lorm xx ,WIt xx representin: :"ii;';
page number from his translation. In cases where I disagree with Thurmg}"'i;
or highlight a grammatical point, I will either the
text m romamzatIOn followed by a page reference to ]Isshu Oshika's critiI,I;
edition (see Bibliography), or simply an Oshika reference with no text.
ences to other canonical and secondary works are given in endnotes. cr,,"
3. L. Hurvitz, tr. Scripture of the Lotus Blossom of the Fine Dharma (New"
York: 1976) 22-25 et passim. "
4. Ibid., 58-83."
5. Ibid., 60-61.
6. For a more detailed discussion of visual illusion in Mahayana
see my paper, "Discourse in the Larikavatara-sutra,"Journal
(11: 1983), section one. For a representative Hindu parallel of the "doubled;
cosmos" motif, see W. O'Flaherty, Dreams, Illusion and Other Realilies
1984) 103-109. . .... , ... '
7. Cf. Eliade, Myths, Dreams, and Mysteries (New York: 1960) 60. Eliaae',
stressed throughout his work the connection between the axis mundi concept':,'
and the myth of paradise, a resonance which is perhaps not out of place in;'
the present text. ... '.
8. The text tells us that Sariputra's skepticism regarding the purity ()f;
the magically-revealed buddha-field arises "by the power (anubhiiva) of the'
Buddha" (sang rgyas kyi mthus; Osh 153). This is the exact phrase used earlie{
to account for the creation of the spectacle in the first place. It cannot be'
determined from the text whether the repetition of this phrase is meant to
imply that the Buddha, in a new act of magical manipulation, has put the,
thought into Sariputra's mind, or whether the reference simply points back;'
ward to the initial creation of the spectacle. Thurman opts for the former
reading; but if an act of overt thought-control is suggested here it is difficult'
to explain why the Buddha must then read Sariputra's thought telepathically
'(yongs su rtog pa thugs kyis mkhyen; Osh 153). '
9. Compare the vanishing Buddha in the RavaI)a chapter of the Larikii-"
vatara-sutra; cf. my 1983 paper, op. cit., pp. 270-275, for an analysis.
10. See O'Flaherty, The Origins of Evil in Hindu Mythology (Berkeley:
1976) 1-3.
11. Ibid., 139-140.
12. Hamlin, op. cit., 277-8.
13. The reluctance stories effectively convey Vimalaklrti's ability to ad-
just his upaya to the disposition of his auditor. Thus, his challenges to the,
sravakas are geared to vinayic issues, right views and conduct in sarpsaric lik,
By contrast, his lectures to the bodhisattvas chiefly concern nirvaI)ic ' .
Verbally, then, Vimalaklrti's upaya reflects the same "dispositionality" we found
l:?;i'the Buddha s magIC.
Thurman's and Lamotte's endnotes to these vignettes detail the subtle
f.\i;.,.: .. . . ys .. in which the sermons and challenges are adjusted to legendary attributes
. ldh h'b
characters mvo ve : t us, to c oose ut one example, the well-known
Mahakasyapa is by Vimalaklrti explicitly on grounds of
of austentles 26-:, .116). [Yet, see followmg note.]
14. But If the power ofupaYlc magIc IS regulated in some sense by karma,
is its actual origin? The close of Chapter Six gives a surprising answer,
a partial one. ':imalaklrti asserts that all the Maras and beggars who
ii(''ccost abodhlSattva are m fact other bodhzsattvas who, wearing an upayic disguise,
to test the resolve of their peers. Their power is said to arise from their
austerities" (Th 55). Common people (phal po ehe) never have the
(or authority; mthu) to trouble a bodhisattva.
:.: This seems a rather strange passage for a Mahayana text devoted to
the career of a worldly bodhisattva. It resonates deeply, however, with
(the timeless Indian image of the ascetic, the tapasvin whose austerities heat
throne when become too great. As King in Kalidasa's
puts It,
Ascetics devoted to peace
possess a fiery hidden power,
like smooth crystal sunstones
that reflect the sun's scorching rays.
[B.S. Miller, tr. Theater of Memory:
The Plays of Kiilidiisa (New York: 1984) 105]
t;Viinalakirti is no forest-dwelling yogi, but we should remember that he is, in
literal sense, chaste. The bodhisattva's austerities not only grant him the
fpbwer to manipulate maya but also protect him from the corruptive force of
!{.tlie multitude-a statement with which apologists for many religions might
('concur. At this formative stage of the Mahayana, then, we see that insight
is not felt to be a sufficient explanation for the bodhisattva's extraordinary
there still lingers the idea of the hermit who buys his attainments
a pound of flesh, even if the rigors of such a course are only sketched
lin passing.
'" 15. The ontic (ontisch) / ontological (ontologisch) distinction is borrowed
;berefrom Heidegger. In general terms, the ontic is that which pertains to the
ilived phenomenal world as lived; it implicitly assumes the experiential immedi-
lacy of entities encountered "in the world." The ontological, on the other hand,
,pertains to any attempt to articulate the conditions for the possibility of this
'Jnl,nediacy of being-namely, the essential structure of the "rea!." For
'lI1adhyamaka metaphysics, of course, the key ontological truth is emptiness.
fo 'say, "I see a miracle!" is to make an ontic statement; to say "Miracles are
founded on emptiness" is to invoke the ontological. Ontic awareness-Heideg-
. ger's "everydayness" (Alltiiglichkeit)-assumes without question the reality of
things: this is the error (bhranta) of the common man, according
itO Buddhism. For Heidegger, the philosopher describes a circular movement.
from this naIve awareness to a awareness or ontology, the
. h . Th b dh' d . f h VNS . h" n back'"
agam to t e ontlc. e 0 lsattva octnne 0 t e ,wIt Its lilsistence
the bodhisattva's return to the world, makes a similar circular movem ,on.k
though rather than founding its ontology on the doctrine of Being (S .en)t::,;;
Heidegger does, its reposes itself on emptiness (Sunyatii.). '>!:;
For a discussion of the ontic and the ontological, see John Macqu
and Edward Robinson, tr. Being and Time (Harper and Row, 1962) pp.
+ ff.

16. Hurvitz, op. cit., 200-l.
17. Thurman's "concealing his miraculous powers" is absent from
Tibetan recension at my disposal (Osh 218).
18. This stock aside occurs in several texts. See, for example,
adiinasatakam (Suprabha story, 4.4) and Divyiivadiina (133.9 and 192.8).
Selected Bibliography
The Vimalak'irtinirdeia-sutra
Xylograph and critical editions:
1. Tibetan Xylograph text
Suzuki, T., ed., The Tibetan Tripitaka: Peking Edition (Tokyo and Kyoto: 1957)Y:
bKah-hGyur, mDo sNa Tshogs VIII(34): 74-101. Tibetan title: 'Phags-pa/,
dri-ma med-par grags-pas bstan-pa zhes bya-ba theg-pa chen-po'i mdo. [Tibetan ',.' :
translation of Ch.os-nyid Tshul-khrims.] "
2. Chinese texts
Takakusu, J., and Watanabe, K., et. al., ed" Taisho Issaikyo (Tokyo: 1924-35). '
Texts: T. 474, T. 475, T. 476. [Chinese translations of Kumar,ylva,
Hsuan-tsang, Chih Ch'ien.] ,
3. Critical editions and romanized texts'
Bailey, H.W., ed., Khotanese Buddhist Texts (Cambridge: 1951; rev. 1981) pp.'
104-/113. [Khotanese fragments mentioning Vimalaklrti.]
Kara, G. Le sutra de Vimalak'irti en Mongol: texts de Ergilu-a Rincin: Manuscript
de Leningrad (Monumenta linguae mongolicae collecta, no. 9) (Budapest: ,:
1982). [Romanization of a Mongol VNS manuscript.] ,
Osshika, J., ed., "The Tibetan Text of Vimalak'irtinirdeia." Acta Indologica I
(1970) pp. 137-240. [Romanization of Tibetan text.]
Prasadika, and Joshi, L. M., Bhotzya saT(lskara1'}a, saT(lSkrta udviira eva1/!
hindz anuviida (English title: Vimalakzrtinirdeia-sutra: Tibetan Version,
Sanskrit Restoration and Hindi Translation) (Bibliotheca Indo-Tibetica, vol.
V) (Sarnath: 1981). [Critically edited, typeset Tibetan text, Sanskrit resto-
ration, Hindi translation].
Modern translations:
Boin, S., trans. The Teaching ofVimalakirti (Vimalak'irti-nirdesa): From the French
Translation (of Lamotte), with Introduction and Notes (Sacred Books of the
Buddhists, vol. 32) (London: 1976). [English rendering of Lamotte's
;,:' French translation (below); certain material has been omitted.]
J., and Tazeko, Y., Das Sutra Vimatak'irti (Das Sutra tiber die Erlosung)
t. (Tokyo: '1944; rev. 1969). [Cerman translation from Kumaraji:va's
t: Chinese.]
, H., "The Vimalak'irti-sutra," serialized in Eastern Buddhist III-IV. [En-
glish translation from the Chinese.]
, Vimalak'irti (Louvain: 1962). [French translation
;.. from TIbetan and Chmese.]
:tuk, c. (Lu Kuan Yu), The Vimalakirti Nirdesa Sutra (Wei Mo Chieh So Shuo
:, Ching), (Clear Light Series) (Berkeley and London: 1972). [English trans-
"'" lation of Kumarajlva's Chinese.]
iThurrnan, R.A.F., The Holy Teaching of Vimalak'irti: A Mahayana Scripture (In-
i, stitute for the Advanced Study of World Religions series) (University
J. Park and London: 1976). [English translation from the Tibetan.]
Critical apparatus:
,j{itarnura, S., "Variant Narrative Texts of the Vimalak'irti-nirdesa sutra, " Journal
of Indian and Buddhist Studies 24:2 (1976).
bshika, j., "Appendices to the Tibetan Translation of the Vimalak'irtinirdesa
.... (Acta Indologica I): 1. Concordance of the Four Translations; II. Cor-
.':. rigenda," Acta Indologica III (1975).
Oshika, J., "An Index to the Tibetan Translation of the Vimalak'irtinirdeia,"
,. Acta Indologica III (1975).
Weller, F., "Bemerkungen zum Sogdischen Vimalak'irtinirdeia-siltra," Asia Major
X (1935).
NAME e:-):-(=-M--id--:-d:-:l-e-n-a-m-e-) ----2:;

MEMBERSHIP TYPE (circle one) Full, Student, Life, Institutional SubsCriber';;
Year(s) for which application is being made. ___ _______ _

o Please check this box if you do not want your name and address
in our mailing list when it is given out.
o Payments may be made in either US Dollars or Japanese Yen. lABS membe;s':
and subscribers who wish to pay in US dollars (no Yen to this address please!)',
should send their payments to:,,;'
Lewis Lancaster, Treasurer,
c/o Dept. of Oriental Languages STUDENT MEMBER
The University of California INSTITUTIONAL SUBSCRIBER
Berkeley, CA 94720 USA LIFE MEMBER


lABS members and subscribers who wish to pay in Japanese Yen (no
to this address please!) should send their payments to the lABS accotiht':
supervised by Dr. Akira Yuyama, Regional Secretary for Asia: )
No. 041-4502280
Kokusai B ukkyo Gakkai
(Ordinary Account)
Toranomon Branch of the
Mitsubishi Bank
Back issues are available to individual members for $17 per issue ( 3,740)l,
or $25 per volume ( 5,500). Life members may purchase back issues atj;'
discounted rate of$12.75 per issue ( 2,805) or $18.75 per volume (
Prospective individual members from economically developing countries
contact the Treasurer for possible subsidized rates.
NOTE: Journal issues are sent in September and December of each
a special fourth-class rate. If members and subscribers do not notify the
secretariat in Berkeley of address changes sufficiently in advance, we can
be responsible for lost

Sanskrit In the
I\}{alacakra Tan tra
"WJohn Newman
Vajrayana Buddhist texts were composed in at least three
:ifndian languages: Sanskrit, Apabhrarpsa, and an East Indian
;dia1ect usually referred to as "Old Bengali." By far the greatest
of Vajrayana .literature was in the
of. pan-IndIC. culture. Thls IS not VaJ-
like Buddhism as a whole, developed among dIverse
"Uhguistic communities, and it is only to be expected that Indian
Buddhists used the common language of educated
ldlscourse to communicate their ideas.
ii' The Sanskrit of the Vajrayana literature, however, is not
fthe Sanskrit of Pa1).ini. According to M. Winternitz: "The
in which the [Buddhist] Tantras are written, is, as a
!'Hlle, just as barbarous as their contents" (Winternitz 1933 :40 1).
'Tp.e question remains as to the exact nature of this linguistic
:;i'parbarism"-is it due simply to incompetence on the part of
;Yajrayana Buddhist authors, or does some other factor, such
'as Buddhist Hybrid Sanskrit, come into play here?l
", Perhaps the earliest Western comment on the Sanskrit of
faVajrayana text is L. de la Vallee Poussin's description of the
ofNagarjuna'sPancakrama: "11 est ... redige en sanscrit,
;#ans une langue riche de termes techniques et de particularites
ou prosodiques, mais qui, du point de vue
phonetique, est correcte et n'exige pas les vastes connaissances
surete de main que suppose l'edition du Lotus ou celle du
lvIahavastu" (de la Vallee Poussin 1896:VII). In other words,
Sanskrit of the Pancakrama does not always follow pa1).inian
'!,lorms, but it is not what we today would call Buddhist Hybrid
or Buddhist Arsa.
scholars reject the i?fluence of
Sansknt on the language of VaJrayana texts. D.L. Snell
'b' h S k' f h H . 7' groVer.
escn mg t e ans nt 0 t e evaJra .Lantra, says: "The li'{r,:
guage need not be graced by the term Buddhist
just. bad Sanskrit" (Snellgrove 19?9:xi). C.S. George,
metICulous study of the portIOn of the Carj,{;lamaharo$aiiil!
Tantra, appears to concur wIth. "The lang
the Tant:aJ. . . IS Sansknt. Alth?ugh the
ulary IS ofter: techmcal, thIS seems
confuse the Issue of language by descnbmg It as 'Buddhist It'.)
brid Sanskrit' or even 'Buddhist Sanskrit.' The subject
is Buddhist, but the language is Sanskrit, close indeed
Sanskrit of the Epics" (George 1974:14)",f
Other scholars, on the contrary, discern a relationshipbt'
tween the Sanskrit of the Vajrayana literature and BUddhist':
Hybrid Sanskrit. B. Bhattacharya says: "The Sadhanamala, J()t'
all intents and purposes, is written in Sanskrit, but the Sanskrit
used here is far from what we usually understand by the
It is the Sanskrit of the Buddhists,-similar to t,hat employea'
in the Mahiivastu Avadiina, theLalitavistara, the
the Kiirar;r;lavyuha, the Saddharma-Pur;r;larika, and similar works"]
(Bhattacharya 1925:viii). T. Skorupski, commenting on the lari0;
guage of the SarvadurgatipariSodhana Tantra, says: "The text9f
this Tantra, like many other works of this kind, has many',
peculiarities of Buddhist Hybrid Sanskrit. To a Sanskrit
who is not acquainted with this kind of literary work and whtJ
has no sympathy for Buddhist Hybrid Sanskrit many grammaC
ical endings will appear simply as wrong" (Skorupski 1983: 118).'
Most pertinent to the present essay is H. Hoffmann's de: .
scription of the Sanskrit of the Paramiidibuddha-the Kalacakra "
mulatantra: "[The Kalacakra mulatantra] is not only written ill
Buddhist Hybrid Sanskrit but in a very corrupt barbariail
Sanskrit of a semi-Indian region in the far north-west",.
(Hoffmann 1973: 136). A characterization such as this, however,
must be evaluated in light of what the Kalacakra tradition itself
has to say about the question of "correct" language.
. In what follows we edit and translate the Sanskrit and Tibe
tan of a yassage from the Vimalaprabhii, the great commentary
on the Sri Kiilacakra (the Kalacakra laghutantra). This passage
describes the language of the Paramiidibuddha, the Sri Kalacakra,.
g Vimalaprabhii and, by extension, the entire early Kalacakra
The Sanskrit text of this passage was first published
;,ieventy years ago,s but Western scholars have not given it the
:;&'ttention it deserves.
ca susabdavadinarp susabdagrahavinasayarthasa-
ra1).atam asritya kvacid vrtte 'pasabda1:;/ kvacid vrtte
yatibhanga1:;/ kvacid avibhaktikarp padam/ kvacid var1).a-
svaralopal)7/ kvacid vrtte dlrgho hrasva1:; hrasvo 'piS
dirghal)/ kvacit pancamyarthe saptaml caturthyarthe
kutracit parasmaipadini
dhatav atmanepadam at-
manepadini. parasmaipadam/ kvacid ekavacane
bahuvacanarp bahuvacana ekavacanam/ purplinge
napurpsakalingarplo napurpsake purplingam/ kvacit
talavyasakare dantyamurdhanyau
/ kvacin murdhanye
/ kvacid dantye talavyamurdhanyau/ evam
anye 'py anusartavyas tantradesakopadeseneti 12/ tatha
mulatantre bhagavan aha/
sucandra sarvabuddhanarp deyarp
ca guru1).arp ca bharyaduhitrputrakam//(l)
gandho bhavati medinyarp toye ruparp raso 'nale/
vayau sparso sabda dharmadhatur
gandhadhu padidi pe bhi1:; khana panadivasasai1:;/
pujayitva sada mudrarpl4 guror dadati satsuta1:;//(3)
ity evam adayo .'nye 'py apasabda
agamapathadl6 iti/ evarp tikayam api su-
likhitavyarp mayarthasara1).atam
srityeti/ atha yena yena prakare1).a kulavidyasu sab-
bhavatp9 tena tena prakare1).arthasa-
ra1).atam asritya buddhanarp bodhisattvanarp dharma-

Vimalaprabhii 1.3
"In order to destroy the attachment to correct language of
those (brahman sages
who) advocated correct language,
(Kalki Yasas) relied on the meaning.
In some verses (of
the Sri Kiilacakra) there are ungrammatical words. In some
verses the caesuras are lacking. Some have words without
case endings, In some, letters and vowels are elided (BHSG
2.3,2.17,2.72,2.84 ff., 2.90, 3.106, 3.122, 3.118). In SOIDe
verses long vowels are short, and short vowels are long
(BHSG 1.10, 3.1-46). In some the locative case is used for
the ablative case (BHSG 7.82), and the genitive case is Used
for the dative case (BHSG 7.63; Whitney 294b, 297a'
Holtzmann 297). In some a middle voice is attributed to
root that possesses an active voice, and an active voice is
attributed to one that possesses a middle voice (BHSG 37.22
ff. & 37.10 ff.; Whitney 529a, 774; Holtzmann 530, 774).
In some the pluralnumber is used for the singular number
and the singular number is used for the plural numbe;
(BHSG l.10, l.101, 25.4). The neuter gender is used for
the masculine gender,and the masculine gender is used
for the neuter (BHSG l.10, 6.1; cf. Boltzmann 263). In
some the dental (sa) and the cerebral (5a) are used for the
palatal letter sa; in some the dental and the palatal are used
for the cerebral; in some the palatal and the cerebral are
used for the dental (BBSG 2.56 ff.; Boltzmann 63). There
are also other such things that must be understood in con-
formity with the instructions of the tantra teacher. Likewise,
the Bhagavan (Buddha) said in the basic tantra (the
Paramadibuddha) :
Sucandra, disciples should constantly offer desired things-
wives, daughters, and beloved sons-to all and
Odor arises from earth, form from water, taste from fire,
tactility from wind, sound from the unchanging, the sphere
of phenomena from the great sky'//(2)
Constantly worshipping the mudrii with perfumes, incense,
lamps, and so forth, and food, drink, clothing, and so forth,
the noble son gives her to the guru.//(3)
A yogi should understand ungrammatical words like these,
and others too, by reading the sacred texts. Likewise, I
(Kalki PUl)<;larika) must write the (Vimalaprabha) commen-
tary relying on the meaning, in order to destroy conceit in
correct language. Thus, Buddhas and bodhisattvas teach
the Dharma for the sake ofliberation-relying on the mean-
ing, they use the different vernaculars and the different
la"nguages of the grammatical treatises, whichever eliminate
conceit in family, learning, and correct language."
It should be noted that I have translated the Sanskrit of
tiWispassage .accordance the Tibetan translation. An ex-
ti{ption to thIS IS the verses CIted from the mulatantra, where the
t;1jbetan faithfully reproduces the grammatical solecisms of the
As Bu ston points out, the genitives sarvabuddhiinam in verse"
in lc, and gurol:t in 3d must all be glossed as datives
ston In the first ca is .syntactically redun-
and IS a VedlC form of the mstrumental plural
"'orthe a-stems (BHSG 8.110; Whitney 329d).
the locatives in the second verse-medinyam, toye, anale,
so forth-are to be taken as ablatives (Bu ston 1324:61114-
2c sabda is lacking a case ending.
:.f,".. In 3a & b adi should properly come at the end of the com-
In 3a -dzpebhil; is, again, a Vedic form of the instrumen-
for the a-stems. In 3b-vasasail;, vasas, neuter, is treated
a-stem (BHSG 16.26; Whitney 1315). In 3d the plural
'Ha,aati should properly be dadati, singular (BHSG 28.11).
is important to note that the langu_age described above
ilsnot Buddhist Hybrid Sanskrit (Buddhist AsF. Edgerton
Buddhist Hybrid Sanskrit "is a blend of Middle Indic
iWith Sanskrit, but its basis, or substratum, is Middle Indic; the
features are secondarily and superficially laid on"
((Edgerton 1956:134). Edgerton stresses that Buddhist Hybrid
3$anskrit "originally was, or was based on, an ancient, pre-Chris-
Atian, Middle Indic vernacular. That is, it is not, and never was,
'$anskrit" (Edgerton 1954:2). In particular, "[BHS] vocabulary
'isto a very large extent not Sanskrit, but Middle Indic" (ibid);
'.and "[Middle Indic words] stamp the language of the [BHS]
;w6rks containing them as based upon another dialect than
anskrit" (BHSG 1.37). "Even the latest Buddhist Hybrid
;Sanskrit texts still retain numerous words, lexical items, which
'show their vernacular origin" (Edgerton 1956: 134).
"."... The language of the Kalacakra literature, on the other hand,
)s Sanskrit. As the passages cited in this essay exemplify, it is
based on a Middle Indic dialect; it contains very few Middle
!Indic words.
128 JIABS VOL. 11 NO. I
The grammatical "rules" gl'ven in the Vimalaprabha c:;:
deed be applied to the language depicted by Edgerton
Buddhist Hybrid Sanskrit Grammar, but many of these feat IS:
are also common to the Sanskrit of the Mahabharata, the Pura
and the Hindu tantras (cf. Goudriaan 1981 :27-28).,:
irregularities are the of a
ImgUlstlc from an appear to
have been deliberately mtroduced mto the Kalacakra literatur
Early Buddhist traditions record that the Buddha
thorized his followers to recite the buddhavacanam in their in:_
vidual dialects in order to convert the unconverted (cf. BHst
l.6-1.13; Edgerton 1954:5-6; 1956:130-133). In the same}
spirit, concern for the meaning rather than the words of the
Dharma is often exhibited in the early Kalacakra literature.
example, in the Paramadibuddha the Buddha says:
yena yena prakareI)a sattvanarp paripacanam/
tena tena prakareI)a kuryad dharmasya ddanam//(5)
yogi sabdapasabdena dharmarp. grhI)ati yatnatab/
desasabdena labde 'rthe sastrasabdena tatra kim//(6)26
One should teach the Dharma in whatever fashion completely
matures sentient beings.//(5) . ..
A yogi zealously grasps the Dharma through grammatical and
ungrammatical words. When one gets the meaning from the
local words, what is the use of technical terms?//(6)
Likewise, the verse introduction to the Vimalaprabha says:
sabdasabdaviciraI)a na mahati sarvajiiamargarthinam
mahatarp marge pravrttib sada/
sattvanam adhimukticittavasatab para
any a vyakaraI)e surahiracita sabdadivadarthinam//(37)
apasabdad artham api yogi grhI)ati
toye payo pibanti harps as tad uddhrtya//(38)
na vyaiijanasaraI)ata
sada mahata
desasarpjiiabhir arthe jiiate kirp sastrasabdena//(39)
jiianarp tad eva na bhavati udite yasyapasabdasabdab
sarvajiiasya na ya sa pradesiki jagati//(40Yo
Those intent on the path to omniscience do not greatly discrimi-
nate between grammatical and ungrammatical words. They al-
ways enter the path of the great ones, even through the lowly
languages of various countries. The Omniscient One's language
is Other,3! in accordance with the dispositions of sentient beings.
Quite different is the language of those intent on arguing about
the words and so forth the gods and nagas arranged in the gram-
A yogi grasps the meaning even from vernaculars and ungram-
matical words. Swans draw out and drink the milk mixed in the
In the realm of ultimate reality great ones never rely on the
letters. What is the use of technical terms when one understands
the meaning through the local expressions?//(39)
That which grammatical and ungrammatical words can express
is not gnosis. That which is parochial to the world is not the
language of the Omniscient One'//(40)
. PUl)Q.arika, the author of the Vimalaprabhii, tells us that his
'father Yasas introduced "ungrammatical words" and so forth
in the text of the Sri Kalacakra "in order to destroy the attach-
'ment to correct language of those (brahman sages) who advo-
correct language." Likewise, the irregular Sanskrit of the
;Vimalaprabhii is intended to "destroy conceit in correct lan-
guage." Excessive esteem for Sanskrit appears to have had a
'deleterious effect on some members of the Buddhist community
'during PUl).qarika's time:
anena pradesikasarpskrtaikavacanena buddh0
pi pradesiko
bhavati sarvasattvarutasvabhavinya vinal iha
sabdavadinan tirthikanarp pal)c;iitanam abhimanarp
balamatlnarp bauddhanam abhiprayab/ yatha
brahmahariharadayab sarpskrtavaktaro
vasaivadlnam iHadevatab tatha.smakarp ya bud-
dhabodhisattvab sarpskrtavaktaro bhavantitil iha na ca te anena
praddikasarpskrtaikavacanena sarvasattvarutair dharmaddakab
sarpgHikaraka . bhavanti buddhabodhisattvab
vinal ato devajatipratibaddha praddika buddhabodhisatt-
vanarp na syad iti nanasattvarutadharmadesakatvat/34
"If he did not use the omniscient language that has the
nature of the utterances of all sentient beings, if he
this parochial Sanskrit speech, then the Buddha would be only
chial as well. Here in the land of the Aryans,35 foolish
, lsts
see the arrogance of the scholarly heterodox proponents of
rect language, and corne to believe: 'Just as the chosen dei./o
. ' . uesof
the brahmans, Val;iI).aVaS, Salvas, so forth-Brahma, Rari,'
Hara, and so forth-speak Sansknt, so too our chosen deities'
the Buddhas and bodhisattvas, speak Sanskrit.' But here th.\
, ese ,.
Buddhas and bodhisattvas do not use the utterances of all sentient.
beings to teach and redact
the Dharma solely by means afth'\
parochial Sanskrit speech, without using the omniscient
guage. Therefore, Buddhas and bodhisattvas do not use a para
chiallanguage that is with birth as a because they'
teach the Dharma usmg the utterances of vanous sentient be-
We are not dealing with simple bad Sanskrit in the Kalacakra;
literature because the irregular grammar employed in the sri
Kalacakra and the Vimalaprabha is not the result of
The authors of these texts, Yasas and PUl)<;iarlka, demonstrate .
considerable knowledge of the full range of brahmanicallear
ing, including grammar, prosody, and poetics. They
correct, even elegant, Sanskrit when they so desire. Moreover,>
in the Vimalaprabha PUl)<;iarika sometimes even points out the'
irregular forms appearing in the Sri Kalacakra, and explains>
how they deviate from standard usage. For exam:ele: ...., \,
(l) Comment on the declension of kalayoge in Sri Kiilacakra
I.26d: kalayoga iti pancamyarthe saptami (Vimalaprabha (S) B 31 bi3; .. '
U 77.19); the locative is used for the ablative.
(2) Comment on the declension of jnanadhatau in Sri Kiilacakra
II.24a: iha sarire apanavayur jnanadhator bhavati atrapi
camyarthe saptamz (Vimalaprabha (S) B 57bl7; U 168.26); the loca-'
tive is used for ablative. .
(3) Comment on the number of $atsandhil; in Sri Kiilacakra
II.25d: $atsandhir iti bahuvacane ekavacanarrt (Vimalaprabhii (S).8.
58a/2; U 169.12-13); the singular is used for the plural. .. ,:.
(4) Comment on mahi in Sri Kalacakra II.25d: mahiti hrasvo
bhilparyayal; (Vimalaprabha (S) B 58a/3; 169.14); short vowel fof
long. . .. :.
(5) Comment on antrameghal; in Sri Kalacakra n.34b: antrameghii
'ntra ity avibhaktikarrt padarrt antra1J,i megha bhavantiti (Vimalaprabhii
60a/l; U 174.27-28); a.ntra is ending.
(6) Comment on the declensIOn of karrJe m Srz Kalacakra II. 79c:
ity agamapathat pancamyarthe saptami (Vimalaprabha (S) B
U 213.11); the locativ: is used for the ablative .
,;. The fact that the Sansknt of the early Kalacakra literature
strewn with irregular grammatical forms presents special prob-
for the editor and translator. If these forms were used
"consistently, it would be simple enough to learn to recognize
.,Iand understand them. Unfortunately, this is not the case. The
clqegular forms appear in verses and prose passages that are
;otherwise written in standard Sanskrit,37 and sometimes an ir-
form is disguised in such a way that it can be interpreted
:meaningfully, but wrongly, as though it were standard usage.
;' An example of this is kalac chunye$u in Sri Kalacakra 1.43.,
;lwhich the Tibetans consistently translate as dus kyis stong pa rnams
;las. Kalat as an ablative of instrumental use is not extraordinary,
but the locative sunye$u as an ablative would be extremely prob-
lematic if we did not have the Tibetan translation (cf. Holtzmann
; Another example is PUl).Q.arlka's usage of the words vivarta
(,and sar(l,varta in his comment on Sri Kalacakra 1.4. These terms
;usually mean "evolution" and "devolution," respectively (cf.
3.90), but the Vimalaprabha exactly inverts their
'meaning: lokadhatutpado nirodho veditavyal; samvarto vivartakalas
,ceti; 'jig rten gyi khams 'byung ba dang 'gag pa chags pa dang 'jig
pa'i dus kyang rig par bya'o (Vimalaprabha (S) B 22a/6; U 54.18-19;
(T) 425/4). On first glance the Tibetans seem to have blundered
'in translating sar(l,varta as chags pa and vivarta as 'jig pa, but this
idiosyncratic usage is confirmed by PUl).Q.arika's use of sar(l,varta
in apposition to utpada and utpatti: atal; samvartad utpadakalavasat
Sunye$v iti; des na chags pa ni 'byung ba'i dus kyi dbang gis stong pa
rnams las shes pa (Vimalaprabha (S) B 22a/6; U 54.19; (T) 425/4);
and kalayogat prajatar(l, samvartotpattikalavasat; dus kyi sbyor ba dag
. las rab tu skyes shes pa chags pa skye ba'i dus kyi dbang gis (Vimala-
prabha (S) B 28a/1; U 68.12; (T) 45717 [comment on Sri Kalacakra
The linguistic peculiarity of the Kalacakra literature once
again underscores our dependence on Tibetan translations and
Indian and Tibetan commentaries for correct editing and trans-
lation of the Vajrayana literature in Sanskrit. Without their aid
]IABSVOL. ll-NO. 1
we would be at a loss to establish the Sanskrit texts, not to sp 'k";
. h' . 38 ea
of understandmg t elr meamng. '''''-.
The language of the early literature is not Bud.
dhist Hybrid Sanskrit (Buddhist nor is it simply substa.
dard Sanskrit. It is Sanskrit into which various types of nonsta n.
dard forms have been intentionally introduced. Most of then
irregularities are common to Buddhist Hybrid Sanskrit and
classical Sanskrit. While some of these solecisms have e
pedagogic purpose, for the most part they are designed to
teract pedantic arrogance. How far these findings can be applied
to other Vajrayana texts will be an important subject for future
research. '
APPENDIX: Rare Vocabulary from the Kiilacakra Literature I
Although the Kalacakra literature contains negligible Middle Indicvocab_
ulary, it does present a number of Sanskrit words that are either rare or
unattested elsewhere. (Here we are not concerned with technical terminology;
which is notorious for being poorly represented in current dictionaries.) Unless'
otherwise noted, the following words are not found (with these meanings, at
least) in the lexicons of Bohtlingk and Roth, Bohtlingk, Schmidt, Monier"
Williams, Apte, or Edgerton.
(1) ekalolZbhuta m. (Tib. gcig tu 'dres par gyur pal "become blended into one":
niriivarar;,atii samarasatvam ekalolZbhutatva1!l fun yam ity ucyate; 'di mams sgrib
pa dang bral ba nyid dang ro mnyam pa nyid dang gcigtu 'dres par gyur pa nyid la
stong pa zhes brjod do (Vimalaprabhii (S) B 19a12; U 47.22-23; (T) 407/3-4).
"Their quality of being free from obscuration, of having a single taste,of
being blended into one, is called 'void'." (The things "blended" here are various
aggregates, elements, faculties, and so forth that are components in the
Kalacakra abhidharma.) -loli- appears to be derived from Vlur,l or Vlul; cf.
lolita. Schmidt lists ekalolzbhiiva in the sense of "Begehren" [desire] (Schmidt
(2) kalka m. (Tib. rigs) "clan": atal}, kalasaguhya-prajiiiijiiiiniibhi{ekatal}, sar-
vavarr;,iiniim ekakalko bhavatil sa kalko 'syastiti kalkZl tasya gotra1!l kalkZgotraT(! vaj-
rakuliibhi{ekatal}, sakalamantrir;,iim iti nztiirthal},; bum pa dang gsang ba dang shes
rab ye shes kyi dbang bskur ba 'di las rigs thams cad rigs gcig tu 'gyur rol rigs de 'di
la yod pa'i phyir rigs ldan nol de'i rigs ni rigs ldan gyi rigs te sngags pa mtha' dag
rdo rje'i rigs kyis dbang bskur ba'i phyir ro zhes bya ba nges pa'i don tal (Vimalaprabhfi
(S) B 8b/3-4; U 22.8-10; (T) 345/2-3). "The vase, secret, and wisdom-gnosis
initiations make all the castes into a single clan. Because he possesses that,
clan, he is Kalki. The definitive meaning of this is: 'His lineage is the lineage
Kalkl because all mantra adepts are initiated into the vajra family.''' Simi-
'lady, Srz Ka:acakr:: "
:X< . so 'yar(L srzmanJuvayra!; suravaranamzto vaJragotrer;a kalkZ
. dattva sakalamunikulany ekakalkar(L
';'lIe (i.e., Manjusri Yasas) will be Sri Manjuvajra, saluted by the best of gods,
;I{alkl by means of the vajra lineage. Having given them the vajra initiation,
::he will make all the families of sages into a single clan."
.,The usual meanings of kalka-"dregs", "feces", "sin", and s6 forth-are
clearly inapplicable here. This usage of kalka is uniquE to the Kalacakra liter-
ature (cf. Newman 1985:64 & n. 4; 1987a:94). F6rdiscussion of other
;etymolo.gies of and kalki/-in (193,:)' .'
(3) pratzsena f. (Tlb. pra phab pa) prognostIC Image : pratyak:;a!; svaClttapratz-
'bhiisO yogina1[t gagane pratibhiate kumarikaya adarsadau pratisenavad iti; gzhon nu
ina mams kyis me long la sogs pa la pra phab pa bzhin du mal 'byor pa mams kyi
rang gi sems kyi 'od gsal mngon sum du nam mkha' la snang ba (Vimalaprabha (S)
B 16b/6; U 42.23-24; (T) 394/3-4). "Like a maiden's prognostic image in a
Ihirror and so forth, the clear light of the yogis' own minds appears manifest
1n the sky." The "sky" here refers to the void. (N ote: I take kumarikaya!; ;is a
genitive singular against the Tibetan plural)
L1.35ab (U 4.21-22) notes that mahamudra IS SimIlar to the eight prognostic
.images In his Paramarthasar(Lgraha-nama-sekoddeSatzka Naro
refers to the Pratisenavataratantra on the eight kinds of prognostication:
'pratisimavataratantre kila
(read: pratisenavatara ukta!; (Carelli 1941 :49.6-7). "Indeed,
ithe Pratisenavataratantra speaks of the manifestation of prognostic images in
the.eight: mirror, sword, thumb, lamp, moon, sun, water well, and eye."
Pratisena appears to be a Sanskritized form of (Prakrit?) prasena (m. or n.),
'"na (f.): "eine Art Gauklerei" [a kind of conjuring] (Bi:ihtlingk 176.2). Edgerton,
following Bi:ihtlingk, lists prasena with a query, and also cites Mahavyutpatti
#4268: prasenam; gsal snang (BHSD 389.1). Edgerton translates gsal snang as
"bright light, or bright clear," but it is most likely the old orthography for dag
'snang, "a mystic vision." It is interesting to note that prasenam appears in the
Mahiivyutpatti in the section on tantric terminology (Mahavyutpatti #4234-
:(4) lz (Tib. Ii) "Khotan": bora lZ ca cfnadidesesu . . . bod dang
Ii dang rgya nag la sogspa'i yul mams suo . . shambha la'i yul la thug pa'i bar du
(Vimalaprabha (S) B 40al2; U 10l.8; (T) 521/4). "In the countries of Tibet,
!{hotan, and China and so forth ... through to the land of Sambhala ... " The
context here is a discussion of the duration of daylight in various regions
north of India. This passage of the Vimalaprabha is the source for the only
other known occurance of lZ in Sanskrit, Abhayakaragupta's Kalacakravatara:
bhota l"i ca sambhalavi;;ayantar(L (Asiatic Society of Bengal MS G.4732
f.6b/5). L'i is an example of a very rare phenomenon, a Tibetan loanword in
134 JIABS VOL. 11 NO.1
1. For previous study of the language of Sanskrit Vajrayana texts'
Bhattacharya (1925:viii-x); Snellgrove (1959:viii-xi); George
Tsuda (1974:6-27); Skorupski (1983:117-118).<'
2. D.S. Ruegg notes: "Ar,1'a is indeed in several respects a more con
nient (and a less linguistically questionable) term than Buddhist
to designate the basically Middle Indo-Aryan language of much of the Can "}t.,
and could therefore be used instead by modern scholars, at least for cert
stages of BHS" (Ruegg 1986:597). Weadd the qualification
distinguish this language from the J aina and brahmanical ar}a-s (cf.
1933:430; Goudriaan 1981:27). ..,
3. As we will see, the Kalacakra literature is not written in BUddhist
Hybrid Sanskrit, and its language is not simply "very corrupt" or "barbarian;"
Sanskrit. Hoffmann has not presented any evidence to support his view that
the Kalacakra originated in "a semi-Indian region in the far north-west." The'
Vimalaprabha, in any case, was by its own account written in India (see note
35). The earliest historically identified proponent of the Kalacakra, AtiSa's
guru Pil).<;io, was born in Java (Newman 1985:71-75; I987b:96-106).
4. For discussion of this literature see: Newman (1985:52-54, 58, 63-
65, 73; I987a; 1987b). I believe all of these texts were composed in India,
during the first few decades of the 11 th century. . ,
5. By H.P. Shastri (1917:78-79), under the heading "The
did not care for Correct Sanskrit." f\
6. Indian scholars have noted its significance: B. Bhattacharyyr
(1924:iv) quotes Shastri's Catalogue. P.e. Bagchi (1934:v) cites the Catalogue,
and offers an English rendering and an interpretation (cf. Goudriaan 1981:27,
n. 130). B. Banerjee gives a brief synopsis of this passage in English (Ban':;:
dyopadhyaya 1952:73), and alludes to it in the introduction to his recel1S
edition of the Sri: Kalacakra (Banerjee 1985 :xxii). J. U padh yaya refers to it in
the introduction to his recent edition of the first two patalas of the Sri: Kiilacakra ".
and the Vimalaprabhii (Vimalaprabha (S) U xv, xxiv).
Tibetan scholars were well aware of the grammatical anomalies of
early Kalacakra literature. See: Bu ston (1324:610-612); mKhas grilb<
( 1434:444-448).
7. Vimalaprabhii (S) N: -svarolopal;.
8. Vimalaprabhii (S) N: krasval; hrasvapi.
9. Vimalaprabhii (S) D: paraspai-.
10. Vimalaprabhii (S) D: napu7[lSakarrt.
11. Vimalaprabhii (S) N: danta-.
12. Vimalaprabhii (S) D: -desakeneti.
13. Vimalaprabhii (S) N: -dhiitum.
14. Vimalaprabhii (S) D: mahiimudrarrt.
15. Vimalaprabha (S) N: iidayo py apaSabdiis tadanye pi.
16. Vimalaprabhii (S) N: agarigama- (emendation deletes -riga-).
17. Vimalaprabha (S) D: suiabdiibhimanarrt nasa.
18. Vimalaprabhii (S) N: -sar;,atarrt (emendation adds -ra-).
19. Vimalaprabhii (S) N: bhavaci.
20. Vimalaprabhii (S) N: (emended to desa-).
921. Vimalaprabhii (S) N 21b/4-22a/4; D 17bll-9; cf. U 29.21-30.6. I have
"o;tandardizedthe sandhi, the orthography with respect to use of the avagraha
ih;tid doubling of consonants after -T-, and supplied the darp;ias. Unfortu-
the leaf containing this passage is missing from Vimalaprabhii (S) B, by
;far the best of the three MSS available to me. The Tibetan for the passage
Vimalaprabha (T) 36115-36217: /sgra bzang po smra ba de dag mams kyi sgra
bzang po la zhen pa spang ba'i slad du don la rton pa nyid la brten tel tshigs su
bead pa kha eig tu zur ehag gi sgra dang/ tshigs su bead pa kha eig tu geod
mtshams nyams pa dang/ kha eig tu mam dbye med. i..a'i tshig dang/ kha eig tu
dbyangs dang gsal byed [better: yig 'bTU dang dbyangs] phyis pa dang/ tshigs bead
kha eig tu ring po la thung ngu dang thung ngu la ring po dang! kha eig tu lnga
pa'i don la bdun pa dang/ bzhi pa'i don la drug pa dang/ kha cig tu gzhan gyi
tshig can gyi byings la bdag nyid kyi tshig dang! bdag nyid kyi tshig can la gzhan
gyi tshig dang/ kha eig tu gcig gi tshig la mang po'i tshig dang/ mang po'i tshig
la geig gi tshig dang/ kha eig tu pho'i rtags la rna ning gi rtags dang rna ning gi
rtags la pho'i rtags dang/ kha eig tu rkan las byung ba sha yig la so dang spyi bo
las byung ba dang! kha eig tu spyi bo las byung ba la so dang rkan las byung ba
dang/ kha eig tu so las byung ba la rkan dang spyi bo las byung ba ste/ de Ita bu
gzhan yang rgyud ston pa po'i man ngag gi rjes su 'brang bar bya'o/ de bzhin
du rtsa ba'i rgyud las beom ldan 'das kyis/
Izla bzang sangs rgyas thams cad dang/
Ibla rna rnams kyi slob rna yis/
/ehung rna bu mo bu sdug dang/
/rtag tu 'dod pa'i dngos po dbul/(l)
/dri ni sa la 'byung ba ste/
jehu la gzugs dang me la ro/
/rlung reg 'gyur ba med la sgral
/nam mkha' ehe la ehos kyi dbyings/(2)
/dri dang spos sogs mar me dang/
Ibza' dang btung sogs gos rnams kyis/
Irtag tu phyag rgya mehod byas tel
/dam pa'i bu yis bla ma'i dbul/(3)
Izhes gsungs te de Ita bu la sogs pa'i sgra zur ehaggzhan yang mal 'byor
pas lung bklags pa las rtogs par bya'o/ de bzhin du 'gre! bshad las kyang
sgra bzang po'i mngon pa'i nga rgyal nyams par bya ba'i slad du bdag gis
don la rton pa nyid la brten te bri bar bya ste/ mam pa gang dang gang
gis rigs dang rigs pa [read: rig pa] dang sgra bzang po'i mngon pa'i nga
rgyal zad par 'gyur ba'i mam pa de dang des don la rton pa nyid la brten
te yul gyi skad gzhan dang sgra'i bstan beos kyi skad gzhan gyis sangs
rgyas dang byang chub sems dpa' mams thar pa'i don du ehos ston to/
22. The "brahman sages" referred to here are the thirty-five million
brahman sages of Sambhala. The story told in the Vimalaprabhii of how Kalki:
Yasas converted them to the Vajrayfma is translated in Newman (1985:59-63;
1987b:304-314). The passage on grammar edited and translated here im-
mediately follows the conversion story, and makes up the end of the thO
uddesa of the first pata1a of the Vimalaprabhii. lrd
23. This refers to the first of the four pratisara?,}as; d. Mahavyutpat';
#1545-1549. '
24. Vimalaprabhii (T) usua,lly translates the Sanskrit according to the,
meaning rather than the words: It does not attempt to reproduce grammatic I
irregularities, but provides the sense. In this instance, however, it "mecha
cally" reproduces the irregular grammar because the irregular grammar
is part of the meaning.'
25. In editing a passage of the Kalacakra mulatantra, Hoffmann
"Middle lndic" out of correct Sanskrit forms metri causa (Hoffmann 1973: 13i
n. 3 & 7). This is difficult to justify given the fact that the verses of
Paramadibuddha are often hypercatalectic or catalectic: cf. verses lIa, 13c, 15b
in Reigle (1986:5-6, 9).
26. Sanskrit after Reigle (1986:5); cf. Vimalaprabhii (S) U 24.27-30.
verses are part of a twenty-one and one-half verse quotation from the
Paramadibuddha that appears towards the middle of the third udde.a of the
Vimalaprabhii. Vimalaprabhii (T) 35117-35212:
Imam pa gang dang gang dag gisl
Isems can rnams ni yongs smin byed!
Imam pa de dang de dag gisl
Ichos ni bstan par bya ba yin/(5)
Isgra dang sgra nyams dag gi [read: gisJ chosl
Imal 'byor pa ni 'bad pas'dzinl
Iyul gyi sgra yis don myed pal
Ide la bstan bcos sgra yis ci/(6)
27, Vimalaprabhii(S) D: -sara?'}atii .
. 28. Vimalaprabhii (S) N: mahatiirrt.
29. Vimalaprabhii (S) D: -sabda{t sabdii{t.
30. Vimalaprabhii (S) N 3a/6-3b/2; D 3al2-5; d. U 5.3-12; leaf missing
in B. These are verses 37-40 of the first uddesa of the Vimalaprabhii.Verse 37
is flawless sardulavikr'irjita; 38-40 are iiryii. Pur:H;iarlka employs a wide variety
of metres in the Vimalaprabhii, and his Sri Paramiirthasevii is composed of
various kinds of Vimalaprabhii (T) 307
Ikun mkhyen lam don gnyer ba rnams la sgra dang sgra min
mam dpyad chen po medl
Isna tshogs yul skad ngan pa yis kyang chen po mams kyi lam
la rtag tu Jugl
Isems can mams kyi mas pa'i sems kyi dbang gis thams cad
mkhyen pa'i gsung gzhan lal
Ilung stan dag la [better: brda sprod dag laJlha klus bkod pa'i' sgra
sags rtsod pa don gnyer mams kyi gzhan/(37)
Iyul gyi skad dang zur chag sgra las kyangl
Imal 'byor ldan pas don ni 'dzin byed del
Ichu la '0 rna nges par zhugs pa del
In gang pas rab tu phyung nas 'thung bar byed/(38)
Idon dam de nyid yul la rtag tu nil
/chen po mams dag tshig la rton pa mini
Iyul gyi mi [read: ming] mams kyis ni don shes Ia/
Ibstan bcos sgra dag gis ni ci zhig bya/(39)
Igang zhig ZUT chag sgra dang sgrar gyur pasl
Ibrjod pa de nyid ye shes rna yin zhingl
Igang zhig 'gro ba dag la nyi tshe bal
Ide ni thams cad mkhyen pa'i gsung rna yin/(40)
31. Here "Other" refers to the transcendent aspect of the Kalacakra
. triad: eva7[L sarvatra vajrayogo bahye adhyatmani pare yoginii 'vagantavya iti; de
h.zhin du rdo rje rnal 'byor yang phyi dang nang dang gzhan thams cad la rnal 'byor
pasrtogs par bya'o (Vimalaprabhii (S) B 17ai7-17b/l; cf. U 44.10-11; (T) 398/2-3).
~ ' A yogi should realize the vajrayoga everywhere in the Outer, Inner, and
Other." .
32. Vimalaprabhii (S) B: sarvabuddho (emendation deletes sarva).
33. Vimalaprabhii (S) B: omit b h i i ~ i i (emendation adds).
34. Vimalaprabhii (S) B 13ai7-13b/l; cf. U 34.11-18. MS B spells sattva
as satva, and I have supplied the darpjas, but otherwise I have retained its
orthography in this and all other quotations from it. Vimalaprabhii (T) 372/6-
Isems can thams cad kyi skad kyi rang bzhin can thams cad mkhyen pa'i skad
med par legs par sbyar ba'i skad nyi tshe ba gcig pu 'di yis yin na sangs rgyas
kyang nyi tshe bar 'gyur TOI 'phags pa'i yul 'dir sgra smra ba po mu stegs pa
mkhas pa'i mngon pa'i nga rgyal dang Idan pa mams [sic!] mthong nasi ji !tar
bram ze dang khyab Jug pa dang zhi ba pa la sogs pa mams kyi 'dod pa'i Iha
tshangs pa dang khyab Jug dang drag po la sogs pa mams legs par sbyar ba
smra ba po yin pa de bzhin du bdag cag gi 'dod pa'i lha sangs rgyas dang byang
chub sems dpa' mams legs par sbyar ba smra ba po yin no zhes pa nil sangs
rgyas pa byis pa'i blo can mams kyi bsam pa stel 'dir sangs rgyas dang byang
chub sems dpa' de dag thams cad mkhyen pa'i skad med par legs par sbyar ba'i
skad nyi tshe ba [add: gcig bu] 'di yis rna yin te sems can thams cad kyi skad kyis
chos ston par byed pa po dang yang dag par sdud par byed pa po yin nol de'i
phyir sangs rgyas dang byang chub sems dpa' mams ni Iha'i skye ba dang 'brei
ba'i [add: skad] nyi tshe bas rna yin te sems can sna tshogs kyi skad kyis chos ston
pa po yin pa'i phyirl
35. This is one of several instances in the Vimalaprabhii in which
PUJ.l.<.iarIka writes "here in the land of the Aryans/' demonstrating that this
text was written in India. (Elsewhere in the Vimalaprabhii "the land of the
Aryans" is dearly defined as India [cf. Newman 1985:61; 1987b:309-310].)
36. sa7[Lgitikiiraka; yang dag par sdud par byed pa po. The use of this term
to indicate the "redactor" of a text is not quite clear at BHSD 548, s.v. sa7[Lgiti
(3). It is often used in this sense in the Vimalaprabhii: e.g., King Sucandra, an
emanation of VajrapaJ.l.i, redacted the Paramiidibuddha, and KalkI Yasas, an
emanation of MaiijusrI, condensed the Paramiidibuddha and redacted it in the
form of the Sri Kiilacakra (Newman 1985:54,63; 1987a:93-94).
37. Likewise Skorupski: "It must be said that one does get frustrated by
138 JIABS VOL. 11 NO.1
the fact that the forms. are used side with the
hybrid forms. It is practically Impossible to discern the prinCIple of using h
correct grammatical forms instead of the hybrid ones or vice versa" (Skoru t J
1983: 118). The can be said about the .ear!?
again, I do not thmk we are confronted with hybrid forms m the strict sense
of forms that developed directly from Prakrit. ...
38. I agree completely with Snellgrove'S remark that we must be
with "a text that accords with the required sense as it may be ascertained front
the Tibetan translation and the several commentaries" (Snellgrove
cf. Tsuda (1974:6-16).
Apte: Vaman Shivaram Apte, The Practical Sanskrit-English Dictionary: Revised
& Enlarged Edition (Kyoto: Rinsen Book Company reprint [1957 ed.] .,
1986). ,.
Bagchi (1934): Prabodh Chandra Bagchi (ed.), Kaulafiiananir1Jaya and Some.
Minor Texts of the School of Matsyendranatha (Calcutta: Metropolitan
ing & Publishing House, 1934) [Calcutta Sanskrit Series no. 3]. .
Bandyopadhyaya (1952): Biswanath Bandyopadhyaya [Biswanath BaneIjee],:
"A Note on the Kalacakratantra and Its Commentary"Journal of the Asiatic
Society: Letters 18 (1952) pp. 71-76.
Banerjee (1985): Biswanath Banerjee (ed.), A Critical Edition of Sri Kalaca-
kratantra-raja (Collated with the Tibetan version) (Calcutta: The Asiatic Soc.
ciety, 1985) [Bibliotheca Indica Series no. 311].
Bhattacharyya (1924): Benoytosh Bhattacharyya, The Indian Buddhist Iconog-
raphy (London: Oxford University Press, 1924).
Bhattacharya (1925): Benoytosh Bhattacharya (ed.), Sadhanamala vol. 1
(Baroda: Oriental Institute reprint, 1968) [Gaekwad's Oriental Series "
no. 26].
BHSG/BHSD: Franklin Edgerton, Buddhist Hybrid Sanskrit Grammar and Dic-
tionary (Delhi: Motilal Banarsidass reprint, 1972) vol. 1: Grammar, vol. 2:
Bohtlingk: Otto Bohtlingk, Sanskrit-Worterbuch in kurzerer Fassung (Graz:
Akademische Druck-U. Verlagsanstalt reprint, 1959).
Bohtlingk and Roth: Otto Bohtlingk and Rudolph Roth, Sanskrit-Worterbuch
(St. Petersburg/Leipzig: Kaiserlichen Akademie der Wissenschaften,
1855 ff.).
Bu ston (1324): Bu ston Rin chen grub, dPal dus kyi 'khor lo'i bshad thabs sgra
rig mkhas pa'i rgyan ces bya ba; Lokesh Chandra (ed.), The Collected Works
of Bu-ston: Part 4 (NGA) (New Delhi: International Academy of Indian
Culture, 1965) pp. 599-614 [Sata-pitaka Series vol. 44].
Carelli (1941): Mario E. Carelli (ed.), Sekoddesatika of Naapada (Niiropii)
(Baroda: Oriental Institute, 1941) [Gaekwad's Oriental Series no. 90].
de la Vallee Poussin (1896): L. de la Vallee Poussin (ed.), Etudes et textes
tantriques: Paiicakrama (Gand: H. Engelcke, 1896) [Recueil de travaux
publies par la Faculte de Philosophie et Lettres de l'Universite de Gand,
16me fascicule].
Edgerton (1954): Franklin Edgerto?, H?brid Sanskrit Language and
1,: Literature .(Banaras: Banaras Hmdu Umverslty, 1954).
:tdgerton (1956): Franklin Edgerton, "The Buddha and Language" Indian
... Historical Quarterly 32 (1956) pp. 129-135.
(George (1974): Christopher S. George (ed. and trans.), The
. Tantra: A Critical Edition and English Translation, Chapters I-VIII (New
Haven: American Oriental Society, 1974) [American Oriental Series
'It vol. 56].
Gdudriaan (1981): Teun Goudriaan and Sanjukta Gupta, Hindu Tantric and
. Siikta Literature (Weisbaden: Otto Harrassowitz, 1981) [A History of In-
" dian Literature 2.2].
Hoffmann (1973): Helmut Hoffmann, "Buddha's Preaching of the Kalacakra
Tantra at the Stupa of Dhanyakataka" German Scholars on India vol. 1
. (Varanasi: The Chowkhamba Sanskrit Series Office, 1973) pp. 136-140.
Boltzmann: Adolf Holtzmann, Grammatisches aus derll Mahabharata (Hil-
;; desheim/New York: Georg Olms Verlag reprint [1884 ed.], 1981) .
. mKhas grub (1434): mKhas grub dGe legs dpal bzang, rGyud thams cad kyi
rgyal po beam ldan 'das dpal dus kyi 'khor lo mchog gi dang po'i sangs rgyas kyi
rtsa ba'i rgyud las phyung ba bsdus ba'i rgyud kyi 'grel chen rtsa ba'i rgyud kyi
rjes su 'jug pa stong phrag bcu gnyis pa dri ma med pa'i 'ad kyi rgya cher bshad
pa de kho na nyid snang bar byed pa zhes bya ba; Yab sras gsung 'bum: mKhas
grub KHA (Dharamsala: Tibetan Cultural Printing Press [Shes rig bar
Mahiivyutpatti: Ryozaburo Sakaki et al (ed.), Mahiivyutpatti (Tokyo: Suzuki Re-
search Foundation, 1970).
Monier-Williams: Monier Monier-Williams, A Sanskrit-English Dictionary
. (Delhi: Motilal Banarsidass reprint [1899 ed.], 1981).
Newman (1985): John Newman, "A Brief History of the Kalachakra"; Geshe
Lhundub Sopa (ed.), The Wheel of Time: The Kalachakra in Context (Madi-
son, WI: Deer Park Books, 1985) pp. 51-90.
Newman (l987a): John Newman, "The Paramiidibuddha' (the Kalacakra
mulatantra) and Its Relation to the Early Kalacakra Literature" Indo-
Iranian Journal 30 (1987) pp. 93-102.
Newman (1987b): John Ronald Newman, The Outer Wheel of Time: Vajrayiina
Buddhist Cosmology in the Kiilacakra Tantra (Ann Arbor, MI: University
Microfilms International, 1987) [U niversity of Wisconsin - Madison Ph.D.
Reigle (1986): David Reigle, "The Lost Kalacakra Mula Tantra on the Kings
of Sambhala" [Kiilacakra Research Publications no. 1] (Talent, OR: Eastern
School, 1986).
Ruegg (1986): D. Seyfort Ruegg, "Review of Georg von Simson, Sanskrit- Wor-
terbuch der buddhistischen Texte aus den Turfan-Funden 3" Journal of the
American Oriental Society 106 (1986) pp. 596-597.
Schmidt: Richard Schmidt, Nachtrage zum Sanskrit-Worterbuch in kurzerer Fas-
140 JIABS VOL. 11 NO.1
sung von Otto Bohtlingk (Leipzig: Otto Harrassowitz, 1928).
Schrader (1937): F. Otto Schrader, "The Name Kalki(n)" The Adyar Libra.'
Bulletin 1 (1937) pp. 17-25. ry
Shastri (1917): Hara Prasad Shastri, A Descriptive Catalogue of Sanscrit Ma
scripts in the Government Collection Under the Care of the Asiatic Societ n
Bengal vol. 1, Buddhist Manuscripts (Calcutta: Asiatic Society of B e n ~ : ~
1917). . ...... .
Skorupski (1983): Tadeusz Skorupski (ed. and trans.), The Sarvadur a-v
tiparifodhana Tantra (Delhi: Motilal Banarsidass, 1983). go'
Snellgrove (1959): D.L. Snellgrove (ed. and trans.), The Hevajra Tantra Part.
2 (London: Oxford University Press reprint, 1980) [London Oriental
Series vol. 6].
Sri Kalacakra: text as given in Vimalaprabhii (S) U.
Tsuda (1974): Shinichi Tsuda (ed. and trans.), The Samvarodaya Tantra: Selected
Chapters (Tokyo: The Hokuseido Press, 1974).
Vimalaprabhii (S) B: Asiatic Society of Bengal MS G.l 0766; palm leaf; old
Bengali script; dated 39th regnal year of Harivarman of Bengal (11th-
12th cent.); described in Shastri (1917:79-82).
Vimalaprabhii (S) D: The Institute for Advanced Studies of World Religions
film-strip no. MBB-1971-24-25; paper; devanagari; (Stony Brook, NY:
The Institute for Advanced Studies of World Religions, 1971).
Vimalaprabhii (S) N: Asiatic Society of Bengal MS G.4727; palm leaf; old
Newarl script; description and extracts in Shastri (1917:73-79). . ............. .
Vimalaprabhii (S) U: J agannatha U padhyaya (ed.), Vimalaprabhiitikii of Kalki Sri.
PU1Jr}arika on Sri Laghukalacakratantraraja by Sri ManjuSriya.fa vol. 1 (Sar- ...
nath: Central Institute of Higher Tibetan Studies, 1986) [Bibliotheca
Indo-Tibetica Series no. 9].
Vimalaprabhii (T): Lokesh Chandra (ed.), The Collected Works of Bu-ston: Parti
(KA) (New Delhi: International Academy of Indian Culture, 1965) pp.
301-603 [Sata-pitaka Series vol. 41].
Whitney: William Dwight Whitney, Sanskrit Grammar (Camht:idge: Harvard
University Press reprint of the 2nd [1889] edition, 1967). .
Winternitz (1933): Maurice Winternitz, A History of Indian Literature vo1.2
(New Delhi: Oriental Books Reprint Corporation reprint, 1977). .
,Two New Fragments of Buddhist Sanskrit
'Manuscripts from Central Asia
By Richard Salomon and Collett Cox
1. Introduction
The two manuscript fragments presented below were re-
ported in Huang 1983, p. 51 and illustrated there in plates
;XXXVI and XXXVII. They are said there to have been found
at the site at Ruoqiang (Charkhlik, a), southeast of the Tarim
Basin in the Xinjiang-Uighur Autonomous Region of the
People's Republic of China. Paleographically these fragments
resemble most closely Sander's Gupta alphabet type A,f (Sander
1968, Tafeln 9-20), dating from the 3rd-4th centuries A.D.;
'note for example the forms of a and sa in fragment 2, and ka,
without a curved tail at the bottom, in both fragments. The
script of no. 1 may be slightly later that that of no. 2, and in
some respects resembles Sander's Gupta B,h (4th-5th centuries
A.D.), for instance in the shape of a and sa. Both fragments
may thus be dated around the 4th century A.D.; possibly a little
earlier in the case of no. 2, a little later in no. 1.
Both fragments are clearly written and reasonably accurate,
though there are some scribal errors such as omission of vowel
signs and visarga, deletion of one element of a consonant con-
junct, confusion of dentals and retroflexes, etc. All of these are
more or less obvious and typical of central Asian Sanksrit man-
uscripts, and have been emended in square brackets with as-
. terisks.
II. Fragment of a Stotra Text
Fig. 1: Fragment of a stotm
Fragment no. 1 (Fig. 1) consists of a single leaf of a paper
manuscript, measuring 10 X 2.5cm. It is mostly intact except
for the left edge and a portion broken away at the upper right
(recto), so that a few akJaras at the beginning of each line and
in the last quarter of the first and last verses are missing. (Missing
akJaras are indicated by X; portions of missing akJaras by-.
Where possible, conjectural restorations of the missing portions
are indicated in the translation.) There are five lines of writing
on each side, each line corresponding to a single verse in anU$tubh
(sloka) meter, with a space in the middle between hemistichs.
The verses, lOin all, are not numbered.
1) X-ac(i)nt(ya?)dhhutagul)e tvayi kalyal)acetasi /
vikkriyarp naspadarh lebhe yatha dh(i?)XX-mmat(a) /
2) XXX-eHate kascit tvayy asadhu tam eva tu /
krpayasedhikatararp matevatmajam aturam /
3) svadubkaik:aturarp jagat /
tvarp svadub.khany anadrtya paradub.khaturab. sada /
4) XXXya dubkhani svapral).ativyayair <"pi /
ya te pritir abhut sadho sa api na t[*a]vat /
5) XXXXX(l)okoyarp. na tatha p10yate yatha /
krpapesalasarp.tanab pioyase tvarp. parartibhib /
1) XXXXX(bh)uc chatrau putre vanugraha(nt)aram /
iva /
2) XXXXr vahen murdhna muner api /
yasya te /
3) XX-b khedyamanopi cchidyamanopi casakrt /
nayasid vikkriyarp. dhira grahakkranta ivoourat [sic; read -rat]/
4) XXXvyasanavarte karul).a satvavatsala /
na tatyaja api tvarp. dharman iva dharm(m)ata /
5) XXnityanubaddharp. ca tvarp. /
notsehire taXXXXXXX /

1) In you, with your holy mind and inconceivably (?; [*
a]c(i)nt)ya)) wondrous virtues, perturbation (read vikkriya fot
vikkriyarrt?) found no place, like ...
2) [*Even if] someone does ([yadyapi *cehfate) ill to you, you pity
him all the more, as a mother does her sick child.
3) This world has no care for the sorrows'[*of others] ([* para-
and is afflicted by its own; you disregard. your
own sorrows and are always afflicted by those of others.
4) The joy which you (felt), Holy One, in [*dispelling] the sor-
rows [*of others] vyaslya du/:tkhiini] even at the cost
of your own life; even they (the others) did not (feel) so great
(a joy).
5) This world is not pained [*by its own afflictions] ([*svartibhir
eva] laka 'yam) as much as you, with your tenderly compassion-
ate heart are pained by the afflictions of others.
1) [*For you] whose mind is filled with compassion, there was
no ([*na te kimcid a](bh)uc chatrau) difference in the kindness
(shown) to an enemy or to a son, any more than to your left
or right eye.
2) ... would carryon the head even the footdust of the Sage, of
you whose every action bespeaks [? udgariJ the touchsto ..... ......
compaSSIOn. .;
3) Even when tormented and cut to pieces again and a ;.
. gaIn
[*your body?] ([*tvaddehajlJ?) dId not undergo perturbaf',
[vikkriyanzJ, 0 Steadfast One, like the moon when
by an eclipse. ......
4) In the whirlpool of evil [*of samsara] ([*samsarajvyasaniivart)!
(your) compassion, tender to (all) beings, never once
doned you, any more than the nature of things could abandon
the things themselves."
5) Faults such as impatience could not affect you, who were .
and always dedicated. . . . .
Although we have not succeeded in identifying the text with
any previously published stotra, it closely resembles in both style'
and content several other Sanskrit poems of this class, such as
the Vany,iirhavarrj,a and Satapancasatka or AdhyardhaSataka attri-
buted to Matrceta, which have been published from manuscript
fragments found in central Asia (see Schlingloff 1955 and 1968;
Shackleton Bailey 1951). The following instances are charac-
With R2cd, krpayasedhikataranz matevatmajam aturam and V4bc,
karur}a satvavatsalal I na tatyaja k;ar}am api, compare verse lOcd
of the Rahulastava (Schlingloff 1955 p. 90), karur}a tva na tatyaja
mata sutam ivaurasam.
With R3ab, [*paradul;}khiinapek;'idanz svadul;kaikiituranz jagat, com-
pare Rahulastava 6ab (Schlingloff 1955 p. 89), parathe niravek!;asya
janasyatmanzbharer iha.
With R4, [*pare$am vyasjya du/:tkhiini svaprar}ativyayair api I ya te
pr'itir abhut sadho sa te$am api na t[*ajvat, compare Satapaiicasatka
17 (Shackleton Bailey 1951 p. 46), pararthe tyajata/:t prar}an ya pr'itir
abhavat taval I na sa na$topalabdhe$u prar}i$u prar}inanz bhavet II.
With Vlab, [*na te kimcid a}(bh)uc chatrau putre vanugrahantaram,
compare Munayastava 7ab (Schlingloff 1955 p. 86), vad-
hakapatyayor yasya manas tulyanz pravartate.
With V3b, cchidyamanopi casakrt, compare Satapaiicasatka ISb
(Shackleton Bailey 1951 p. 47), cchidyamanasya te 'sakrt.
With V5a, .. . nityanubaddhanz ca tvanz, cf. Satapaiicasatka 22c
(Shackleton Bailey 1951 p. 50), na te nityanubaddhasya.
Further parallels could be cited, but these examples should suE-
j;'6ce to demonstrate the cl?se. similarity of the new fragment to
;:the known stotras. In partICular, the closest resemblances seem
be .with the anonymous Riihulastava and with Mat:rceta's
!;;Satapancasatka (especially the Hetustava portion, verses 10-26).
fragment is thus almost certainly from a stotra, previously
;:llhdiscovered as far as we have been able to determine, of the
of Mat:rceta; whether by Mat:rceta himself or by an im-
Jitator we cannot say for sure, although if the latter it is a good
'tirnitation of the master's style.
,flIl. Fragment of an Abhidharma Text
Fig. 2: Fragment of an Abhidharma Text
Fragment no. 2 (Fig. 2), part of an Abhidharma text, is on
palm leaf and measures 8.7 by 1.3 em., with three lines on each
side. It is complete except for a break at the upper left (of the
recto), with four or five ak;;aras missing from the beginning of
line R1 = V3. In the left margin of the recto is a numerical
sign, apparently 40, indicating the number of the folio.
1). X X X X rta(?)vatIti kecit tavad ahul,1 dvividharp kusalarp
sasravarp. canasravarp ca tatra ya[t*]
2). sasravarp tad upadayatlti athava dvividharp kusalamm [. :
' hi' . SlC]
upa lp a arp vlsarpyoga- ,
3). phalarp ca tatropadhiphalarp upa[da*]nakam iti atroc ...
. -k h h - yate
ne a vlpa a etu. pan-
1). k:;;yate upapattihetur iha sa ca [a*]kusalab
yady upapattihetur akusa- .' .....
2). lasyana -lab. syan kakid up::
apadyeta Itl tad dhl tatra kusalam astltl ...
3). -o-e X X -astiti uktarp hi bhagavata viviktab iti
rocyate na vayarp .
1) .... Now, some say, "The virtuous is twofold, with contamina_
tion and without contamination. Of those, that which.:.:
2). is with contamination furthers attachment." Or, "The virtu:\'
ous is twofold, resulting in a substratum, and resulting
3). in disconnection. Of these, the one resulting in a substratum
is the basis of attachment." To this it is said, "It is not the
cause of maturation that'is being examined here;
1). it is the cause of rebirth that is being examined here; and
that is [un]virtuous." To this one [might] say, "If the cause
of rebirth were unvirtuous, .
2). no realm of form nor formless realm would arise at all; for
that is virtuous there,
3). " .is ... For the Lord has said, 'One is free from desires.'"
To this it is said, "We do not. ...
Both the style and content of this fragment suggest thatit \.
is from a Buddhist Abhidharma commentarial treatise. Though
efforts to locate the passage in any extant Sanskrit text or frag-
ment, or any Chinese translation were unsuccessful, there is a
marked similarity in style to two fragments discovered at Kucha, ....
which were published by E. Waldschmidt. (Waldschmidt 1965
#15, #18, pp. 9-12) All three fragments are written in the form
rof a dialogue in which the views of the proponent are indicated
:by the phrase atrocyate, those of the opponent(s) kecid
atraha, etc. Both partles employ argument and scnptural
':Citations as authorities in support of their positions. Further, all
fragments treat a relatively sophisticated point of doctrinal
'controversy. The dialogical expository style and complexity of
doctrinal investigations suggest an Abhidharma text ofthe mid-
dle or later period, that is, contemporaneous with or following
;the initial compilation of the Vibha5a compendia (c. 2nd century
(Lamotte 1958 p. 648; see Kimura 1937 pp. 207ff).
. . Although neither the specific topic under discussion in this
;fragment nor the sectarian affiliation of either party is explicitly
:identified, the following doctrinal issues suggest that the topic
is probably karma, or possibly citta, and at least one statement
.bythe opponent (V. 1-3) is fully consistent with Kashmiri Sar-
.vastivada- doctrinal positions:
'!.(,' >
:1). Rl The two categories of contaminated, sasrava, and uncon-
;laminated, anasrava, are used to classify all dharmas in an early
!;Abhidharma text, the Sariputrabhidharmasa,stra (SAS 1 p.
,527.b.23ff), and become common in Abhidharma texts from
tthe middle period on. (PP 5 p. 711.b.9; JP T.26.1544 2 p.
926.a.llff; MVB 76 p. 391.c.2lff, VB 7 p. 463.a.19ff; AVB 40
293.b.8ff; MVB 95 p. 490.a.26, AVB 47 p. 360.b.22) The
classification of virtuous dharmas according to these two
categories is also frequent. (MVB 67 p. 346.a.28, A VB 35 p.
2). R2 The second classification of "the virtuous" according to
the two categories of that having substratum as its effect (upa-
dhiphala) and that having disconnection as its effect (visarrtyoga-
phala) does not appear in any extant Abhidharma text. It is, in
part, clarified by a passage from the Jnanaprasthiinasastra OP
T.26.1543 7 p. 851.b.19ff; T.26.1544 12 p. 979.b.23ff; MVB
123 p. 640.b.24ff) describing the effects of sasrava and anasra-
vakarma. Here, karma as a whole, including both sasrava and
anasrava, is said to have three possible effects: 1) the effect of
uniform outflow (ni5yandaphala) and 2) the effect of maturation
(vipakaphala), which are themselves sasrava and are produced
by sasravakarma, and 3) the effect of disconnection
(visar(iyogaphala), which is anasrava and may be produced b'
. either sasrava or anasravakarma. Dharmasrl's Abhidharmahrd . Y
(T.28.1550 1 p. 815.a.7ff) and
(T.28.1551 2 p. 843.b.5ff) contam simllar descnptions of the
threefold effects of karma, but the later Sar(iyuktabhidharma_
hrdayasastra (T.28.1552 3 p. 897.b.3ff), after presenting the
theory of the threefold effect, adds the two effects_th
purwakaraphala and the adhipatiphala-which all together cons/
tute the set of five effects characteristic of Kashmiri
theory. The two, ,?f the jnanapra_
sthanasastra, and both Dharmasn sand Upasanta's Abhidharma_
hrdaya present a theory of three possible effects that predates
or rivals the theory of five effects typical of the later
position (MVB 21 p. 108.c.3ff, 121 p. 629.c.4ff).
The "effect of disconnection" mentioned in these passages
and in this fragment dearly correspond. However, the identity
of the "effect resulting in substratum" (upadhiphala) mentioned'
in the fragment with the "effect of uniform outflow" (ni$yan-
daphala) and the "effect of maturation" (vipakaphala) is uncertain.
Virtually the only occurrence of the term upadhi in the
Abhidharma texts is in the terms "nirvar;,a
with a remainder of upadhi," and "nirvar;,a
without a remainder of up ad hi." (MVB 32 p. 167. 14ff, AVB17
p. 126.a.8ff. See also Schmithausen 1969 pp. 79-81 #2.) The
character yib , can be used t6 translate upadhi, (YBS 50 p.
576.c.27ff) and yiguo
, as in the Jnanaprasthiina T.26.1543 (tr.
Sarighadeva), the Abhidharmahrdaya T.28.1550 (tr. Sarighadeva),
T.28.1551 (tr. Narendrayasas), and the Sar(iyuktabhidharmahrdaya
T.28.1552 (tr. Sarighavarman), could then conceivably be the
equivalent of upadhiphala. However, we find Sarighadeva in the
jnanaprasthiina (T.26.1543 17 p.851.b.20) usingyiguo to translate
a term for which Xuanzang's translation (T.26.1544 12
p.979.b.25) dearly suggests Similarly, Buddha-
varman in the T.28.1546 (AVB 10 p. 74.c.27)
uses the term yiguo, where Xuanzang (MVB 18 p. 90.c.1) has
dearly translated The term upadhi appears fre-
quently in Pali suttas and Chinese translations of the agamas: 1)
as that which provides the basis for suffering (MN #26 vol. 1
p. 162, MA 56 #204 p. 776.a.12), or as one link in a succession
of factors that give rise to suffering, old age, and death (SN vol. 2
;108, SA 12 #291 p. 82.b.lOff(upadhitransliterated); Norman
334-336); 2) as that from which one becomes free in attain-
(MN. #16 vol. 1 p. 454, MA 50 #192 p. 74?a.8ff,
and 3) as Juxtaposed to attachment, upadiina as 1ll upa-
dkyupadanavinibaddho . .. (Tripathi 1962 pp. 45,168; Pali: upa-
'Jupadana . .. SN vol. 2 p. 17,. 12 'p'
One reference to upadhz 1ll the MaJJhzmanzkaya (MN # 11 7
:vol. 3 p. 72) describing right views (sammaditthi) parallels its use
:in this fragment: a contaminated right view (sammaditthi sasavii)
connected with meritorious action (puiiiiabhagiya) has its result
!In upadhi (upadhivepakka), whereas an uncontaminated noble
right view (sammaditthi ariya anasavii) is the member of the noble
!path Upadhi, .as the basis for is associated
;with suffenng and functIOns as a component III the causal proc-
'ess leading to birth and death. In this sense, upadhi and visar(l,yoga,
. or disconnection, represent mutually exclusive categories. How-
ever, no passage was found that clearly juxtaposes upadhiphala
>and visar(l,yogaphala.
3). R3-Vl No explicit reference to the pair, upapattihetu and
;vipakahetu, or to these two as causes producing upadhi was found
in any Abhidharma text. However, the canonical use of upadhi
indicates that it is clearly associated with the process of rebirth.
Further, in an explanation of the meaning of upadhi in the
terms, and the Maha-
vibhasa (MVB 32 p. 168.a.lff, AVB 17 p. 126.a.28-29) distin-
guishes between the upadhi of defilements (kleSa) and the upadhi
. of rebirth (upapatti). Although the term vipakahetu, appears fre-
quently in Abhidharma causal systems of all periods, upapattihetu
,has a much more constricted use. Upapattihetu appears paired
with abhinirvrttihetu, the cause of proceeding; the former refers
to the cause of specific rebirth states, and the latter, to the cause
that leads to rebirth in general. (AKB 6.3 p. 333.5ff; YBS 5 p.
301.c.7ff, YB Bhattacharya p. 108). (For the pair abhinirvrtti-
sarrtyojana and upapattisar(l,yojana and their relation to the inter-
mediate state and the future rebirth state according to Sarvas-
theory see AKB 3.41 p. 153.16ff.)
Sanghabhadra in the Nyayanusara (NAS 49 p. 618.a.13ff) lists
abhinirvrttihetu and upapattihetu with vipakahetu in a group of
three causes: abhinirvrttihetu is the cause of not abandoning, or
150 JIABS VOL. 1) NO.1
not surpassing realms and stages; upapattihetu is the cause tl
. makes one be reborn; and vipakahetu is the cause that
?ne the maturatio? ?f after one is Accoral
mg to Sanghabhadra, abhznzrvrttzhetu and upapattzhetu differ from'
vipakahetu in. that they are. causes for the process of rebirth.
Though logIcally one might assume that Sanghabhadr'
d.ivision of causes into abhinir:rttihetu, upapattihetu
vzpakahetu IS a refinement of an earlIer twofold division int .,
upapattihetu and vipakahetu attested in this fragment, no
basis for this hypothesis has been found.
4). Vl-2 Given the objection of the opponent in Vl-2,
cause of rebirth were unvirtuous," the proponent's staternent in
VI, sa ca kuialal;, should probably be emended to read so,
cakusalal;, "and that is un-virtuous." With this emendation, the
proponent suggests that the cause of rebirth under discussion.
is unvirtuous. The opponent's subsequent objection (Vl-3) pro-
vides an important clue concerning both the function of the
prior distinction between upapattihetu and vipakahetu, and the:
opponent's identity. The opponent's first point-if this cause'
of rebirth were unvirtuous, the realm of form and the formleSs
realm would not arise-implies that this cause of rebirth,if
unvirtuous, would produce an unvirtuous effect. In other words,
the cause of rebirth functions through a causal relation of
larity producing an effect similar to it, as opposed to the cause
of maturation, which functions through a relation of difference
(MVB 19 p. 98.b.5ff). An unvirtuous cause of rebirth must then
produce an unvirtuous effect. This unvirtuous effect could nOt
occur within the realm of form or the formless realm because;
the opponent states, "that is virtuous there." This statement is
consistent with the Kashmiri Sarvastivada- position
that unvirtuous dharmas are not found in the two upper realms'
of the realm of form and the formless realm, and therefore,
whatever defilements (klesa) are found there are indeterminate
(avyakrta) (MVB 3 p. 14.b.8ff, 38 p. 196.b.12ff, 50 p. 259.c.9ff,
141 p. 724.c.3ff, 144 p. 741.b.4ff; AKV p. 392.32-33). This
position opposes the and Mahasanghika view that.
all defilements are unvirtuous (MVB 38 p. 196.a.15ff, 50 p.
259.c.9; Masuda 1925 p. 27), and therefore, by implication, that
defilements of the realm of form and the formless realm must.
be unvirtuous .
. r Though the opponent seems to represent the Kashmiri Sar-
perspective, the identification of the pro-
as the frag.ment as part of a
:Jastra IS hIghly problematIc. DespIte Sanghabhadra's ObVIOUS
:rdiance on a written text in his frequent references to the views
;:of the master Sthavira (Srllata), no fragment of an
:Hndependent siistra has yet been found. Instead, the
,'proponent could well represent the view of another branch of
tthe Sarvastivada. Or, this fragment could represent a section
'embedded in a larger text quoting the views of an opponent
.'who presents his position in a dialogue in which he is the pro-
ponent. In that case, the proponent and opponent of the frag-
.ment and the larger text would be reversed .
.. 5). V3 The opponent concludes his reasons for the impossibility
(.of rebirth in the realm of form or the formless realm with a
:'scriptural citation in V3: "one is free from desires (vivikta/:t
Though brief, this citation echoes the common for-
:mulaic description of the process by which one passes from the
realm of desire through the four trance states in the realm of
form: "one traverses, attaining the first trance state ...... that
is free of desires, free of evil and unvirtuous dharmas." ( .. . viv-
ktarrt kiimair vviviktarrt piipakair akusalair dharmmai/:t . . . prathamarrt
dhyiinam upasarrtpadya viharati. Dietz 1984 p. 62; DS 12 p.
512.c.23ff. See also DN #2 vol. 1 p. 73; MN #13 vol. 1 p. 89,
MA 25 #99 p. 586.a.18ff; MVB 80 p. 415.a.23ff, AVB 41 p.
311.b.7ff, VB 10 p. 488.a.2ft). One would normally expect viv-
ikta to be construed with the instrumental, as we find in this
canonical passage, reflecting the common idiom of the instru-
mental with verbs of separation (von Hiniiber 1968 .149 p:
162; UV 30.28c-d p. 399 kamebhir vipramukto ... ). However, in
this fragment, the locative, is probably not an anomaly
. but rather is due either to a confusion of the locative for the
.instrumental (Edgerton 1953 .7.30 p. 44, .7.81 p. 47), or re-
flects another verbal idiom witli in the locative. (Sen 1953
p. 410; UV 2.9c p. 114 tv apratibaddhacitta . .. ; UV 18.15c
p.245 atrptam eva . .. ). Since this scriptural passage states
that in attaining the first trance state in the realm of form, there
152 JIABS VOL) 1 NO.1
is freedom not only from desires, but also from unvirtu '.
'd .c h OUs
dharmas, it would provl e support lor t e opponent's suggesti
that there is nothing unvirtuous in the of form or
formless realm. Remarkably, we find this very scriptural
ence used by Sanghabhadra in the Nyayanusara (NAS 49"
617.a.24ff) in an argument with the master
(Sri:!ata) in a context identical to that of this fragment: thatis
Sanghabhadra attempts to refute Sthavira's suggestion that sine'
all defilements are unvirtuous, there are unvirtuous dharmas
the realm of form. However, the same caution noted at the end
of the previous section concerning the attribution of this frag-
ment to the must be repeated here.
AKB-Pradhan, P., ed., 1975. Abhidharmakosabhf4yamofVasubandhu, (Tibetan
Sanskrit Works Series Vol. 8.) Patna.
AVB-Abhidharmavibhf4iiSiistra. T.28.1546.
Dietz-Dietz, S. 1984. Fragmente des Dharmaskandha. (Abhandlungen der
Akademie der Wissenschaften in Gi:ittingen. Philologisch-Historische
Klasse, Dritte Folge, Nr.142.) Gi:ittingen.
DN-Rhys Davids, T.W. and J. E. Carpenter, eds., 1890. The Digha Nikaya,
Vol. 1 London. .
DS-Dharmaskandha. T.26.1537.
Edgerton-Edgerton, F., 1953. Buddhist Hybrid Sanskrit Grammar and Dictionary.
(Vol. 1, Grammar). New Haven.
von Hiniiber-Hiniiber, O. von, 1968. Studien zur Kasussyntax des Pali, besonders
des Vinaya-Pitaka. Miinchen.
Huang-Huang Wenbi
. 1983. Xinjiang kaogufajue baogao (1957-8ye. (An Ar-
chaeological Tour of Xinjiang (1957-8)). Beijing.
JP-Jiianaprasthana. T.26.1544. (Abhidharmf4taskandhasiistra. T.26.1543.)
Kimura-Kimura, T., 1937. Abidatsumaron no kenkyu. Kimura Taiken zensM,
Vol. 4. Tokyo.
Lamotte, E., 1958. Histoire du bouddhisme Indien. (Bibliotheque du Museon,
Vol. 43). Louvain.
MA-Madhyamagama. T.1.26.
Masuda-Masuda, J., 1925. "Origins and Doctrines of Early Buddhist
Schools." Asia Major. 2:1-78.
MN---,-Chalmers, R., ed., 1896-99. The Majjhima Nikaya, 3 Vols. London.
NAS-Nyayanusara. T.29.1562.
Norman-Norman, K.R., 1970. "Middle Indo-Aryan Studies VIII." Journal
of the Oriental Institute, Baroda. 20:329-336.
:'\SA_Sarrz,yuktiigama. T.2.99.... .,
"Sander-Sander, L., 1968. Palaographzsches zu den Sanskrzthandschriften der Ber-
<" liner Turfansammlung. (Verzeichnis der Orientalischen Handschriften in
it, Deutschland, Supplementband 8.) Wiesbaden.
JAS_Siiriputriibhidharmafiistra. T.28.1548.
:"Schlingloff (1955 )-Schlingloff, D., 1955. Buddhistische Stotras aus ostturkistanis-
, ' chen Sanskrittexten (Deutsche Akademie der Wissenschaften iu Berlin.
iF Institut fUr Orientforschung, Veroffentlichung Nr.22). Berlin.
'sihlingloff (l968)-Schlingloff, D. 1968. Die Buddhastotras des Miitrceta. Fak-
similewiedergabe der HandschriJten. (Abhandlungen der Deutschen
Akademie der Wissenschaften zu Berlin. Klasse fUr Sprachen, Literatur
and Kunst, 1968, Nr.2). Berlin.
iSchmithausen-Schmithausen, L., 1969. Der Nirvii'f}a-Abschnitt in der ViniS-
, cayasarrz,graha'f}t der YogiiciirabhilmiJ;,. (Veroffentlichungen der Kommission
fUr Sprachen und Kulturen Siid- und Ostasiens, 8. Philosophisch-His-
torische Klasse, Sitzungsberichte 264, Band 2.) Wien.
,Sen-Sen, S., 1953. ,"Historical Syntax of Middle Indo-Aryan." Indian Linguis-
, tics. 13 (in Vol. 3, reprint ed. of Vois. 1-15):355-412.
"Shackleton Bailey-Shackleton Bailey, D.R., 1951: The Satapaiicasatka oj
"" Miitrceta. Cambridge.
SN-Feer, M. L., ed., 1884-98. The Sarrz,yutta Nikiiya oJthe Sutta Pitaka. 5 Vois.
,',' London.
T.-Takakusu,]., et al. eds., 1924-32. TaishO shinshil Daizokyo. Tokyo.
: TripathI-TripathI, C. 1962. FiinJundzwanzig Siltras des Nidiinasarrz,yukta.
(Deutsche Akademie der Wissenschaften zu Berlin. Institut fUr
Orientforschung, Veroffentlichung Nr.56. Sanskrittexte aus den Turfan-
funden, 8.) Berlin. '
"UV-Bernhard, F., 1965. Udiinavarga, Band 1. (Abhandlungen der Akademie
der Wissenschaften in Gottingen. Philologisch-Historische Klasse, Dritte
Folge, Nr.54.) Gottingen.
Waldschmidt-Waldschmidt, E., 1965. Sanskrithandschriften aus den TurfanJun-
den, Teil 1. (Verzeichnis der Orientalischen Handschriften in Deutsch-
land, Band 9,1.) Wiesbaden.
VB-Bhattacharya, V., 1957. The Yogiiciirabhilmi oj .Aciirya ASa1iga, Part 1. Cal-
YBS-Yogiiciirabhilmisiistra. T.30.1579.
Chinese terms
International Journal for
Philosophy of Religion.
Bowman l. Clarke; Frank R. Harrison III, The University of
Georgia, USA
The organ of no single institution or sectarian school
philosophical or religious, the International Journal f o ~
Philosophy of Religion provides a medium for the exposition
development, and criticism of important philosophical i n ~ .
sights and theories relevant to religion in any of its varied
forms. It also provides a forum for critical, constructive, and
interpretative consideration of religion from an objective
philosophical point of view.
Articles, symposia, discussions, reviews, notes, and news in
this journal are intended to serve the interests of a wide
range of thoughtful readers, especially teachers and students
of philosophy, philosophical theology and religious thought.
Subscription Information iSSN 0020-7047
1987, Volume 21-22 (6 issues)
Institutional rate: Ofl. 222.00/US$92.00 incl. postage/handl-
Private rate: Ofl. 85.00/US$35.50 inc!. postage/handling
Private subscriptions should be sent direct to the publishers.
Back Volume(s) Available
Volumes 1-20 (1971-1986)
Martinus Nijhoff Publishers
Price per Volume
excl. postage
Ofl. 95.00/US$38.00
Ofl. 75.00/US$28.00
Ofl. 80.00/US$27.50
(Vols. 19-20)
Reflections on R. S. Y. Chi's Buddhist
1tJ:;()rrnal Logic ...
\' .'
Tom J. F. Tillemans
It has been almost twenty years since Richard Chi published
Formal Logic, I a work which attempted virtually every-
',thing at once: it was filled with various interpretations of the
'ihistory of Buddhist thought, particularly that of Dignaga as
in the short synopsis of valid and fallacious reasons known
the Hetucakra; at the same time it sought to present a
;:,philosophical analysis and evaluation of Dignaga's thought by
/abundantly using the techniques of Anglo-American analytic
;'philosophy and formal logic. The book, in spite of its forbidding
is certainly worth the necessary effort on the part
;.of an open-minded reader; no doubt it has its share of question-
:able interpretations, as we shall see below, but Chi's approach
and his portrayal of Dignaga's thought are still of relevance for
:the growing number of philosophers who recognize the impor-
tance and interest of understanding non-western systems of
logic and epistemology.
Two recent events make it fitting to once again examine
.Chi's contribution: one is the sad news of Prof. Chi's death, and
the other is the fortunate fact that his book has been reprinted
in India by Motilal Banarsidass. Given that the book was written
many years ago, it is, of course, inevitable that the perspective
of a present-day writer on these issues will exhibit important
differences from that of Chi. Nevertheless, I think that it is of
interest to take stock of some of the positive and negative sides
of Chi's book and, in so doing, address some aspects of the
problem as to how one might approach the subject of "Buddhist
formal logic."
It should be clear by now that formal logic can be profitably
156 JIABS VOL. 1 rNO. 1.
used elucidating the thought of early logicians, ?e they WesterJ
or ASIan; w.ehave. a number of of now and Chi
?as the of.bemg one the this approach;
m BuddhIst lOgIC. The ObVIOUS pitfall, ?owever, IS basing enor:;
mous on mcomplete or shaky
derstandmgs of the BuddhIst texts: If one does not have a realI'
clear picture of the philosphical notions which' one is seekiri
to translate, or if those notions themselves are fuzzy, the
of using formal techniques will be, to say the least, unilluminating. '.
1. The Trairupya
. good example a fruitless, use oflogic
IS ChI s treatment of Dignaga s theory of the tnply characterized
reason (trairupya). (Here, let me refer to two articles published
by S. Katsura: "On Trairupya Formulae" and "Dignaga 011
Trairupya."2) On pp. 40 et seqq. we find Chi's section 126, "In-.
terpretation of the Trairupya," where the author bases his opinion.'
on what Dignaga's theory was all about on a passage from Hsuan-
tsang's Chinese translation of the Nyayapravesa (NP), a work
which was itself most likely written by Sar'lkarasvamin. Aside
from the obvious incongruity of grounding an interpretation
of the fine details of Dignaga's thought on a text which is not
actually by him, and is at any rate far and away less important
than the Prama1J,asamuccaya, Chi bases his whole understanding
on the Chinese translation of NP, when in fact we have the
Sanskrit original and the Tibetan: Here is what Chi says on p. 41:'
The above way of interpretation is not merely my personal specu-.
lation. The Chinese translation, although usually very poor, is
accurate enough in the rendering of the Trairupya. The word
piena which means "pervade" or "pervasive" is used in the first
and the third clauses; while the word ting
which means "neces-
sary" is used in the second clause. According to the Chinese
rendering, the Trairupya [as found in Hsuan-tsang's trans. ofNP]
should be translated as follows:
The pervasive presence of the hetu in the subject;
The necessary presence of the hetu in some similar instances;
The pervasive absence of the hetu from dissimilar instances.
Now let me reproduce the Chinese, Sanskrit and Tibetan
"ts of the NP passage in question and confront Chi's rendition
some of Katsura's results concerning "only" (eva) the NP:
, -rl -.J-: - ...j-Q 1- 1tt 1-1- - )'Ifl: - E- -:;:
thinese: 1.1 II'l :::- 'l'D.. floJ <;j E.7 .::-. f,F) '/"
C', IA. r-;:) 0 ,-'-'1 -t:. l.,'t- ,'<'1 0 - tJ<-
t or] :iL 11'l 'r.:t. I ',:i:..
hetus trirupab, I ki1'(t punas trairupyam I pak.$adhar.matva1'(t
sattva1'(t vipakJe casattvam iti/3
'Tibetan: gtan tshigs ni tshul gsum mo II tshul gsum po de yang gang
na I phygos kyi chos nyid dang I mthun pa'i phyogs nyid la yod
;par nges pa dang I mi thun pa'i phyogs la med pa nyid du nges pa
;yang ngo II (Peking bstan 'gyur ce 180b5-6),
i'i' The first thing to be noticed, following Katsura, is that the
:Sanskrit does not have "only" (eva), whereas the Tibetan, in the
second and third clauses, has nyid, which here has the sense of
:'eva rather than the abstract tva or tao As for the Chinese, pien
'hnd tinl do not correspond to anything in the Sanskrit, but are
.rather Hsuan-tsang's additions, just as nyid was added by the
1:Tibetan translator. Now, Chi understood ting in a rather normal
'Chinese way4 as meaning "necessary." He explains on p. 42 that
:iing you hsing
"assured presence," "not failing to be present" or "bare presence"
and includes two possible cases, namely: the pervasive presence
and the partial presence (sapakJavyapaka andsapakJaikadesavrtti).
" There is, however, every reason to believe that here the
character ting is being used, perhaps infelicitously, to convey
'the sense of "only" (eva = nyid): in other words Hsuan-tsang
also made the same "addition" which the Tibetan translator
,made. The point then is not (as Chi would have it) that the
,second clause simply asserts that some, 'but not all, instances of
the reason are in the set of similar instances. Rather, the clause
asserts that all instances of the reason are in that set.
equivalence, ting = eva, while fairly rare, is attested in A .
. Hirakawa's Index to the but more signific-
antly, we find that both pien and ting are used to render eva in
other trairupya formulae in Hsuan-tsang's translation of NP,
formulae where eva actually does occur in the Sanskrit.
No doubt, if we want to translate the NP passages strictly
on the basis of what the Chinese says, the result will be ambigu-
158 JIABS VOL. 11 NO.1
ous. But if we bear in mind. that ting and pien are. be,ing Usedto
render eva, then the meanmg of the passage wIll oe that th .
hetuis only present in the subject, it is present in only the simile
instances and it is only absent (i.e., entirely absent) in the dissirti
lar instances. If this does not look different from
formulation in the Nyayabindu, so be it: it seems that both
tsang and the Tibetan translator of NP may very well have hatf
a Dharmakirti-style formulation of the trairupya with eva in mind
when they made their "additions." Anyway, leaving aside Np
we now have fairly good evidence that Dignaga in
PramatJ,asamuccayavrf;ti did use eva in the second characteristiC'
viz., the anvayavyapti, when commenting on tattulye sadbhiiv()'
and moreover, that he may have accepted the equivalence of
the second and third (i.e., the vyatirekavyapti) characteristics.'
Chi seems to have based his stance largely on an articIe()f
K. Potter, "Dignaga and the development ofIndian logic", which
he reproduced in his book and which distinguished three
of the development of the trairupya. Given the results whichwe
have now, I doubt that these phases can be accepted as Potter
described them. So, in sum, I will not present the superstructure
of formulae (see pp. 42-43) which Chi based on his or Potter's
interpretation, for I think it is sufficient to say that the Bud-
dhological problem of what Dignaga, or even Hsiian-tsang, held
with regard to eva was a much more thorny one than Chi or
Potter made it out to be, and until that problem is clearer in
our minds, much of Chi's formal treatment of the trairupya and
his philosophical comparisons with John Stuart Mill are really
beside the point.
II. Is the Hetucakra extensional?
On p. xix Chi writes:
The Hetucakra was intended to be an extensional study of various
kinds of major premises about whether they can yield valid syl-
By "extensionality of logical formulae" we mean that any
two equivalent formulae may, in contexts in a given theory, be
freplaced by another., for example, if the .sentences P
trid Q are eqmvalent, then It can be shown that, III formulae
;20111Posed only of sentential connectives and quantifiers, Q can
substituted for P in any given sentence and the truth-value
Ebf that sentence will remain unchanged. If, however, we have
:;itheory which uses epistemic statements such as "X knows that
.t .. " or modal statements such as "Necessarily ... ," then there
(will be many statements which are not extensional, but rather,
The sentence
:;, (P == Q) --_. (X knows that P == X knows that Q)8
,is false, and thus sentences containing "X know that. .. " are
said to be "intensional" rather than "extensional."9
.. , Now, one of the key problems in deciding to what degree
hhe Hetucakra is extensional is how to interpret the
?asadhiirar;,anaikantikahetu, viz., "the reason which is uncertain
,because it is [too] exclusive," where, according to Dignaga, the
'jreason is supposed to be absent from both the similar instances
.;(sapak$a) and the dissimilar instances (vipak5a).10 The usual in-
'terpretation, which Chi also presents and which I have elsewhere
::termed "the orthodox scenario," is that a reason such as "audi-
bility" (sravar;,atva) , which is co-extensive with the dharmin,
' .. "sound," is literally excluded from both the sapak5a and the
;vipak$a. The sapak5a here would be all things which are imper-
'manent, except for the dharmin, sound. So, audibility cannot occur
ln the set of sapak5a for proving sound's impermanence, for
such a sapak5a simply does not exist: it would have had to be
;something which was audible was not a sound.1i And a
<fortiori, "audibility" does not occur in the dissimilar instances
'(vipaksa) either, for there is nothing which is audible and which
;is also permanent. 12
. This version of the asadharar;,ahetu is not silly or wrong, but
it is worth our while to see that it is not the only version: far
.. from it. Indeed, I have argued that there was a controversy in
Tibetan Buddhism over this question, with the Sa skya pas main-
' . taining something like the orthodox scenario, while the dGe
.Jugs pa maintained a different position, basing themselves on
; definitions of sapak5a and vipak5a such as those found in Ratna-
. karasanti's Antarvyaptisamarthana. But it is, of course, not particu-
160 JIABS VOL. 11 NO. Y
larly I?ersuasive to just. later logicians in <:rder to answer the
questIon as to what Dlgnaga thought; what IS much more
vincing is that Dharmaklrti, in Pramar;,avarttika IV's long discus ..
sion of the asadharar;,ahetu, not .the
scenario, but rather comes up wIth a verSIOn (sImIlar to thedG
lugs) which int:rpret .this as essentially I
problem of an epistemlC and mtensIOnallogic m that it involves
contexts such as "X knows that ... ". I leave it up to the reader
to judge whether DharmakIrti was an accurate exponent of
Dignaga's thought on this matter.
Karikas 207-259 of the pararthanumana chapter of
Pramar;,avarttika form part of a larger section loosely treatingof
Dignaga's Hetucakra, and specifically concern the refutationo[
the Naiyayika's argument that living bodies have selves
because they have breath and other animal functions (pra1J,tidi).14
Although Dharmakirti does not discuss the sound-(im)perrna_
nent-audibility example very much, he does explicitly state in
karika 218 that the asadharar;,anaikantikahetu, "breath, etc.,"js
completely similar logically to the example found in the
Here are some of the key verses along with extracts
from commentaries.
Context: In k. 205 and 206, Dharmakirti has been putting
forth the recurrent theme that the certainty of the reason's being
excluded from the dissimilar instances depends upon there
being a necessary connection (avinabhava) between it and th'
property to be proven. Such a connection will guarantee the
pervasion (vyapti), i.e., the concomitances in similarity (anvaya)
and in difference (vyatireka). Thus, given such a connection, the
reason would be excluded from the dissimilar instances, butifl
the case of the asadhara1J,ahetu, such a connection cannot be
established; hence there is no such exclusion.
Devendrabuddhi's introduction to k. 207: [Objection:] If in this ..
way the Master [Dignaga] did not exclude (ldog pa ma yin na) the'
special case [i.e., the aSiidhiira1J,ahetu] [from the dissimilar in-.
stances], then why is it said that it is excluded from the similar
and dissimilar instances?16
k. 207: [Reply:] It is just from the point of view of merely not
observing [the reason among the dissimilar instances]
spoke of it being excluded. Therefore [i.e., since the
uncertain when it is due to merely not observing the reason],
[the Master said that the reason] is uncertain. Otherwise [if there
were the certainty that it is excluded from the dissimilar intances],
[the reason] would be demonstrative (gamaka).17
k. 220: By saying that [the sadhana] is excluded just from the
contrary of what is to be proven [viz., the dissimilar instances],
it is asserted [by implication] that it is present in what is to be
proven [viz., in the similar instances]. Therefore, it was said that
by means of one [viz., the vyatireka or the anvaya], both will be
demonstrated by implication.
The point of k. 207, then, is that Dharmakirti wants to
interpret "absence in the vipak$a" metaphorically: it does not
mean that breath, etc., are in fact completely absent from what
>does not have a self, but rather that the debaters do not observe
that breath, etc., occurs in things which have no self. But, al-
though the debater might not see something, that does not
necessarily mean that it is not there. In that sense, the debater
does not ascertain absence, for indeed, as k. 220 makes clear,
if breath, etc., were really absent in the dissimilar instances, then
the vyatirekavyapti would hold; hence, the anvaya would hold
'too, and the reason would be valid!
So in brief, "exclusion" or "absence" is to be interpreted
metaphorically as meaning "non-observation." And precisely
because non-observation is not probative, the essential point of
.the asadhilra1J,ahetu, according to Dharmakirti's interpretation of,
Dignaga, is that the debaters do not know or ascertain vyapti,
be it the reason's absence in vipak$a or its presence in sapak$a.
Finally, consider the following important passage from
Dharmakirti's Svavrtti to Prama1J,avarttika I (Svarthilnumana) k.
28, along with KarI:lakagomin's TiM.19 (I have underlined the
Svavrtti passages):
katharrt tarhy asadharar;atvac chravar;atvarrt nityanityayor nastfty ucyata
ity aha I kevalarrt tu ityadi I sravar;,atvasya bhiivanis-
cayabhiivat I sravar;,atvarrt nityiinityayor nastity ucyate I. Now then,
how is it that audibility is said to be absent in both permanent
and impermanent [things] because it is an exclusive [attribute]?
[Dharmaklrti] answers: But it is just. .. etc. [It is just] because
audibility is not ascertained as being in either permanent or
impermanent [things] that audibility is said to be absent from
what is permanent or impermanent.
162 JIABS VOL. 11 NO.1
Our conclusion must be that, at least following Dharmaklrf
the Hetucakra cannot be treated extensionally, but will invol 1,.
epistemic, intensional statements. 20 Ve
III. Quantification and semantics
A number of years ago the Dutch logician, E.W. Beth
sketched out a set of questions which a study of a
logic should approach if it is to be of use to a modern philosopher
of 10gic.
Some of his more general questions (e.g., "Does the.
language of the culture in question give suitable means of ex-
pression for formal logical reasoning?") have, I think, been more
or less satisfactorily answered by now with regard to Buddhist
logic, but others, in particular those which concern the basic
structures of Buddhist logic, such as quantification, negation,
implication, intensionality/extensionality and semantic theory>
are still to a surprising degree unresolved. Whenever one
tempts to explain such structures, one would do well to bearin
mind Beth's caution that "one should be prepared for the pos"
sibility that the treatment of formal logical problems is combined
or mingled with considerations of another, for example, epis-
temological, psychological or grammatical nature."22
Now let us, by way of an example, look at quantification ill
Buddhist logic. It seems to me that one of the recurring problems
in using formal logic to handle pervasion (vyapti) and such no-
tions is that we often disregard the peculiar semantic theories
of the Buddhists, and then we simply go ahead and use first
order predicate calculus (with some modifications here and'
there) plus a semantics which is more or less as we might find
it in an elementary logic textbook. Of course, there can be a
certain utility in deliberately over-simplifying things. But I
would argue that a satisfactory theory of basic formal structures
in Buddhist logic must take into account Buddhist metaphysiCs
and especially the semantic theory of anyapoha ("exclusion of
what is other").
Let us first look at what Chi did with quantification. We
find an account of Dignaga's three "operators," pervasive pre-
sence (vyapaka) , absence (avrtti) and partial presence'
(ekadesavrtti) , where Chi uses standard first order predicate caI7 ..
quantifiers (Ex) Fx (Read: "There is an x, such that x has
f) and (x) Fx (Read: "For all x, x has F). He then argues (cf. pp.
)o{iv and xxv) that Dignaga's version of pervasion between F
arid G was:........., (Ex) (Fx &-, Gx) plus the additional premiss (Ex)
(Fx & ?X!: I that what means that per-
vasion IS a conjUnctIOn of two dIstmct premIsses," IS that the
ekample constitutes a premiss, viz., that there are instances of
:Fand G. At any rate, if these quantifiers are taken normally (as
,in an elementary logic textbook), they should range over a do-
.main composed only of actually existent things. And that leads
difficulties in the case of Buddhist logic: (a) pervasion (and
hence quantification) is not restricted to existent things, but also
concerns inexistent but possible items, such as the rabbit's horn,23
.and even inconsistent, impossible items, such as the barren
woman's son; (b) even in the Hetucakra itself, we find examples,
such as space (iikiisa), which from the point of view of Buddhist
.metaphysics do not really exist. 24
, A fairly good solution to the conundrum is that of A. Mac-
;Dermott, who uses R. Routley's R*.2
We can introduce a univer-
quantifier, (1TX) P, ranging over all possible items, be they
or inexistent. We can then define another quantifier as
Thus "(1TX) Fx" should be read as "For all possible items, x, x
has F". Read as "for some x," i.e., as a quantifier which
does not imply that the value of the variable exists. Pervasion
,then would become:
('Trx) (Fx- Gx) & (2:x) (Fx & Gx), i.e.,
,(2:x) (Fx & -,Gx) & (2:x) (Fx & Gx) .
. (The second conjunct would show the premiss that F and G
must not be uninstantiated.)
This approach does, however, have the distinct disadvan-
tage that it cannot easily handle inconsistent items, a point which
,seems to have given MacDermott herself some qualms.
problem is symptomatic of the fact that the semantics which she
used bears little resemblance to the Buddhists' own semantics
164 JIABS VOL. 11 NO.1
as developed in the theory of apoha. (In fact, even Routley h'
self had some misgivings about the domain of an
ofR* being composed in part Qfinexistent items, and he
a preference for a substitutional interpretation, where the a
main of suggest another
III BuddhIst lOgIC, an which, in spite
OElts pnma faCIe strangeness, has the ment of staying close t
the Buddhist's own semantic theories. 0
. We know that for Dignagaand Dharmaklrti words do not
directly refer to particulars (svalak;a1Ja); instead, what is spoken
about is always a mentally created fiction, a samanyalak;a1J
universal characteristic, which consists in the exclusion of all
. that is other than the object. This fiction has a type
of mind-dependent existence (we could say "subsistence"), or
in Buddhist terminology, it is said to be "conventionally
_ (sar{Lvrtisat); it is, however, only indirectly connected with what
is ultimately real, i.e., svalak;a1Ja; in sum, it is a type of proxy
over which our thoughts .and words range. The details of this
connection between proxies and svalak;a1Ja is beyond the scope
of this paper, but suffice to say that the proxy, largely because
one confuses it with the svalak;a1Ja leads us to understand and
act on various svalak;a1Ja in the world. (Dharmakirti, in
Prama1Javarttika III, k. 57, illustrates the process by an example
of a person who mistakes the light emitted by a jewel through
a keyhole for the jewel itself; such a person will nonetheless
know where the jewel is, and will eventually obtain it.) Equally,
there can be "proxies," such as that of the rabbit's horn or that
of the barren woman's son, which have no corresponding
svalak;a1Ja. The upshot is that the Buddhist adopts what B.K.
Matilal has termed a "pan-fictional" approach: .
The Buddhist, in fact, would like to put all the objects over which
our thoughts and other psychological activities may range at the
. same level; and this will include not only (a) things which do
exist now (i.e., which are assumed to be existent by the common
people or by the realist) but also (b) things which do not exist
now (i.e., past and future things), (c) things which cannot exist
(viz., the rabbit's horn) and also (d) things of which it would be
a logical contradiction to say that they exist (viz., the son of a
barren woman).27
To return to the problem of quantifiers and their interpre-
tation, the essential idea is to let them range over a domain
composed entirely of proxies, i.e., samanyalaksa'f}a. We could use
the quanti.fiers (1TX) and (2:x) and let them range over domains
of proxies, which all exist at least conventionally. While the
barren woman's son is contradictory, his samanyalaksa'f}a proxy
is not, and gives us no special problems. We also introduce a
function which assigns svalaksa'f}a to some of these proxies. This
function will often assign the same svalaksaTJa to several different
samiinyalaksaTJa proxies; the samanyalaksaTJa denoted by the words
!'impermanence" and "producthood" (krtakatva) , for example,
are connected with the same svalaksaTJa.
An interpretation I will consist of the ordered quadruple
[D, D',j, g], where D is a non-empty set of samanyalaksaTJa, D' is
a possibly empty set of svalaksaTJa,f is a function assigning ele-
ments of D to individual variables and constants, and sets of
in D to n-ary predicates. g is a function assigning
.svala10a1Ja in D' to proxies in D. Truth and satisfaction could
proceed more or less normally, except that in the case of an
atomic formula such as Fa (interpreted as "a is impermanent"),
the formula would be true when the proxy, a, which is not itself
impermanent, is conventionally or commonly thought to be im-
permanent. (In the case of the barren woman's son we should
probably have to say that his proxy is not even conventionally
thought to be a son.) I do not wish to pretend that this is
philosophically wholly satisfying as a theory of truth-at any
rate it would need an accompanying account of what conven-
tional truth is for Buddhist logicians.
Suffice to say that this
line of approach to quantification and its semantics in Buddhist
logic is also faithful with regard to the philosophical stance of
.Dignaga and Dharmakirti's thought.
IV. Final Remarks
It seems to me obvious that we cannot reasonably attempt
a philosophical analysis of Buddhist logic as extensive and tech-
nical as that of Chi's study on the Hetucakra until we are much
more familiar with the main Indian and Tibetan texts and have
a clearer idea of the doctrines about which we wish to
166 JIABS VOL. 11 NO.1
philosophize. Now, Chi his work Was ju ..........
. beginning-that should be stressed. However, the next
1 h k
. II . lIne ..
someone tack es t e Hetuca ra we espeCla y need to hay .........
of. the and
?oth of wh.ICh sectIOns onthii'
subject. The Hidlgenous TIbetan rtags rzgs texts, which of tiC
have a section on the Hetucakra and provide useful definitio
of the sorts of valid and fallacious are also die
value here. Fmally, there are many other TIbetan works, such',
as dGe 'dun grub pa's Tshad ma rigs pa'i rgyan and rGyal
rNam 'grel thar lam gsal byed and N gag dbang bstan dar's
tary on the Hetucakra, which have long and valuable sectionsori
the problems at stake and merit serious study. .'
1. R.S.Y. Chi, Buddhist Formal Logic, Part I, A study ofDigniiga's
and K'uei-chi's Great Commentary on the Nyiiyapravesa. Dr. B.C. Law Trust Fund,:
Volume I, first published by the Royal Asiatic Society of Great Britian, 1969::,
Reprinted by Motilal Banarsidass, Delhi, 1984. There are no differences
tween the two printings, except that the errata indicated in the first editidIil
have been corrected in the reprint. Let me specify that in what follows, I have,",.
for typographical reasons, changed Chi's symbolic notation in certain places.,
Specifically, will be represented by "&" (read: ... and ... ),
tion by "v" (read: ... or. .. ), material implication by" -" (read: if. .. theri));
negation by "-," (read: ... not. .. ) and equivalence by"""" (read: .. .if and.:.
only if. .. ). . ....... .
2. "On Trairupya Formulae," in Buddhism and Its Relation to
gions: Essays in Honour of Dr. Shozen Kumoi on His Seventieth Birthday, Kydtoi.
1986, pp. 161-172. "Dignaga on Trairupya, " Journal of Indian and Buddhit'
Studies (IndogakuBukkyogakuKenkyu), XXXII, 1, December, 1983, pp. 544-538: <
3. See p. 1 of The Nyiiyapravesa, Sanskrit Text with Commentaries, ed. A.B,;
Dhruva, Gaekwad's Oriental Series 38, Baroda, 1968. .,
4. Cf. modern Chinese yi tinl.
5. For the use of eva / "only" at stake in the trairupya see Y. KajiyaIIla, ..
"Three kinds of affirmation and two kinds of negation in Buddhisr:
philosophy," Wiener Zeitschrift fur die Kunde Sudasiens 17, 1973, pp. 161-175,.,.:
See also the corrections of B.S. Gillon and R.P. Hayes, "The role of the particlf
eva in (logical) quantification in Sanskrit," WZKS 26, 1982, pp. 195-203.
6. See Katsura (1986) op. cit. p. 162. To take one of Katsura's
NP, p. 1, 13-14: tatra krtakatvaf[t prayatniintariyakatvaf[t vii pa10adharmal;,
;:i-sti vipa4e niisty eva I. Taish6 p. lib:
eva ttt '7 PJf Ii jjJ
'11 .j! ! t yt ;li J
\:Jb 'Ii
7. See Katsura (1983) op. cit.
8. For an explanation of the symbols used, see n. 1.
9. See A. Grzegorczyk, An Outline of Mathematical Logic. (Dordrecht: D.
;:'Reidel Publishing Co., 1974) pp. 222-224. I have discussed the role of these
:intensional epistemic statements in Indian and Tibetan Buddhist logic in my
and referential opacity in Tibetan Buddhist apoha theory," in B.K.
and R.D. Evans (eds.) Buddhist Logic and Epistemology. (Dordrecht: D.
Reidel, 1986), pp. 207-277.
i 10. In another article I have argued against taking Buddhist inferences-
"for-oneself (sviirthanumiina) and inferences-for-others (pariithanumiina) as
::.being some sort of syllogism; I will not go into the details here. See my article
"i';Sur Ie pariirthanumiina en logique bouddhique," Asiatische Studien I Etudes
;.J,siatiques 38, 2, 1984, pp. 73-99.
: 11. See my paper, "On sapaf0a, " forthcoming in Acta Orientalia Hungarica,
"Proceedings of the Csoma de Karas Symposium held in Visegrid in Sept.
;:1984. On p. 363 of his article, "On the theory of intrinsic determination of
::universal concomitance in Buddhist Logic," Journal of Indian and Buddhist
7, 1, pp. 364-360, Y. Kajiyama gives the "orthodox scenario":
If the reason belongs exclusively to the minor term, as in the case of audibility
which is supposed to prove momentariness of sound (minor term), no homolo-
gous cases [i.e. which are audible and momentary are available.
12. In fact, the Hetucakra discusses the reason "audibility" in the context
'of proving that sound is permanent. We frequently find both sorts of proof,
, viz., of sound's impermanence or sound's permanence. See Kajiyama op. cit.
"p. 363, and T. Stcherbatsky:Buddhist Logic. Leningrad 1930, vol 2, p. 208, n.
1. Indeed, logically the problem is the same. If we are proving that sound is
permanent, then there will be no vipaf0a, i.e., impermanent things other than
,.sound, which will also be audible. Note that Chi (p. 17) definesvipak,5a as z
. ( hz), i.e., the set of all things which do not have the property to be proven.
'If he wants to make his account of the asiidhiira7Jahetu work, he would have
to specify vipaf0a as z ( fz & hz), i.e., the set of things which are not the
and which do not have the property to be proven. Cf. n. 33 on p. 135
:pf M. Tachikawa, "A Sixth-century manual of Indian logic," Journal of Indian
>Philosophy 1, 1971, pp. 111-145: "Both the sapakJa and the vipaf0a must be
different from the paf0a. Therefore the mark is present neither in the sapaf0a
nor in the vipaf0a." If we kept Chi's definition,z (-,hz) , then we would be
(forced to say that "audibility" does occur amongst the things, like sound, which
;"are not permanent, and hence that it does occur in the vipaf0a when one is
; proving that sound is permanent!
168 JIABS VOL. 11 NO.1
13. The question turns largely, I would think, on Dharmaklrf' ,,'
. ..' (" . d"" I S add!"
tIOn" of the word nges pa = msczta ascertame ) m hIS renditi .-
Prama'f}aviniscaya II) of Dig nag a's formulation of the trairupya inPraman on (cf.;:.'
h h Th
" k . E S . .asamuc.
caya's Svart anumana c apter. IS IS ta en up m . temkellner's "Re "{
'. . h " h . . F h ifi fi G T" marks'
on msatagra a'f}a, iort commg m estsc n t or . uccz. ThIs term i ",
explicitly present in Dignaga's formulation but the sense, judging from h
f P
- . dd' t er
passages 0 . seems to Its a . Ition. Surprisin r?'
enough, when ChI (d. p. 30) cItes the Nyayabmdu s verSIOn of the trair- g r
he. omits word, which makes.for a big change: I think it is fair to
nzsata I nzscaya makes an extensIOnal treatment of the trmrupya incom r ')
d d
. pete.
an Istortmg..;,
14. Cf 3.2.4 Uddyotakara maintained that the reasou' ..... .
was one where only the contraposition held (kevalavyatirekin). iHg
15. srava'f}atvena tat tulyar[! pra'f}adi vyabhicaratal},.,
16 .. Prama'f}ava;ttikapaiijika, (sDe dge edition, reproduced in sDe dge
tan Tnpztaka, bsTan gyur Tshad ma, Tokyo, 1981ff), 310a3: gal te 'di ltarslob
dpon gyis khyad par ldog pa ma yin na I ji ltar mthun pa'i phyogs dang mi mthun:
pa'i phyogs las de ldog pa yin no zhes bshad ce na I.' ..
17. adaya kevalar[! vyatirekita I
ukta 'naikantikas tasmad anyatha gamako bhavet II.
My additions in pada c and d have been made on the basis of Manorathanan_
din's Prama'f}avarttikavrtti (ed. R. Sari.knyayana, Patna 1938-40):
sanamatre'f}a vyatirekaniscayad anaikantika acarye'f}oktal}, I anyatha
tirekaniScaye gamako hetur bhavet I. Note, however, that with regard to padaa'
and b, this latter commentator strangely glosses vyatirekitokta, whereas
following Devendrabuddhi's line of thought,.as well as the general thread 6f
the argumentation, vyatirekitokta would seem more logical. I
essentially followed Devendrabuddhi here. Cf Paiijika 31 Oa4: gang gi phyir mi.
mthun pa'i phyogs la de mthong ba med pa tsam gyis ldog pa yin la I de'i phyir
nges pa yin no II.
18. asadhyad eva viccheda iti sadhye 'stitocyate I
arthapattya 'ta evoktam ekena dvayadarsanam I
For additions, see Manorathanandin ad. k. 220.
19. P. 84in ed. R. Sari.kl;tyayana, reprinted in Kyoto: Rinsen Books, 1982.
20. A few remarks on the subject of Chi, and indeed most other
modern writers on Buddhist logic, have defined as those items, with'
the exception of the dharmin, which have the property to be proved. It is important
to stress that this view on would involve serious problems of a formal
nature: in particular, the equivalence between anvaya and vyatireka does not
hold. Chi seems to have recognized this fact (cf p. xxxvii). Specifically, he
argued that a formula such as (x) (fx - gx) & (x) (hfx & gx) - hx) .
--(x) (fx - hx) is "completely wrong", but he then used this as an argument
to scrap any equivalence between the dnvaya and vyatireka and to try to show
that Dignaga perhaps had some sort of way out of John Stuart Mill's charge
that the syllogism committed a petitio principii. This last bit is, to say the least,
rather far-fetched.
Now, given our previous discussion concerning Dignaga's trairupya, we
p, to recognize that anvaya and vyatireka were taken to be equivalent
may . f D' - H h h h . . .
at the tIme 0 Igna?"a.. owever, w et er t .at IS so or not, It IS
cl?ar DharmakIrtI held them to be eqUIvalent. But then
;'bsurdity anses: If we must exclude the dharmzn from DharmakirtI
have been making a gross logical blunder in claiming this equivalence.
that J.F. Staal, in an in Logic" E.
P. Suppes and A. ed., L0f5l.c, ':'1ethodol?gy and
i stanford , 1962), obscured thIngs by claImIng that In DharmakirtI's verSIon of
lithe trairiipya in the Nyiiyabindu, the anvayavyiipti and the vyatirekavyiipti are in
'hd equivalent, i.e., the latter is the contraposition of the former. Staal used
T. Bailperin's restricted variables (see "A Theory of Restricted Quantification
h&I1,"]ournal of Symbolic Logic, 22,1. 1957: pp. 19-35 and 22, 2,1957: pp.
where the expression ax F(x) denotes any of the values of x such
i;that x has F. He also introduced a relation A(y,x) meaning that y occurs in a
'locus x, h means the hetu, p means the and s means the siidhya. He then
'defined and as:
., (1) ax h(x = p) & A(s,x)) and
,: (2) ax-,A(s,x)
;:'respectively. The result is that the anvaya condition becomes:
? (3) (x) (A(h,x)-(x = i.e.,
.' (4) (y) (A(h,y) -(y = ax h(x = p) & A(s'x, i.e.,
'. (5) (y) (A(h,y)-h(y = p) & A(s,y).
i''vyatireka becomes:
(6) (ax-,A(s,x-,A(h, ax--,A(s,x)), i.e.,
; (7) (x) hA(s,x)_ ,A(h,x))
However, contrary to what Staal maintains, (6) is not the contraposition of (4): (4) and
) are not equivalent. The formula,
(8) (y) (A(h,y)_ (y =ax h(x = p) & A(s,x == (ax-,A(s,x -,A(h, ax
is false. This becomes clear when we eliminate the restricted variables as in
(5) and (7) .
. (9) (y) (A(h,y)_ (--,(y = p) & A(s,y) == (x) (,A(s,x)_ -,A(h,x is dearly
. false. The problem arises precisely because of the presence of the formula ", (y = p )."
What is perhaps worse, following this interpretation of in Dhar-
. rnakirti's trairiipya, the anvayavyiipti would in most cases be a false statement.
speaking, the anvaya would state that everything which has the
: hetu-property is a member of the set of things which have the siidhya and are
not the dharmin. That's usually false: the dharmin can certainly have the hetit-
property. So leaving aside Dignaga, we would have to say that the logicians
followed him continually made howlers of the most abysmal sort-and
here we are violating the fundamental hermeneutical principle that
.' one should always seek an interpretation which presupposes that the author
,was intelligent and had a consistent position in mind. Now, I would readily
; grant that the term is vague, but it seems that to "recuperate" Dhar-
"rnakirti's (and probably Dignaga's) statements, the at stake in the
!Jrairiipya theory cannot exclude the dharmin. This is more or less the conclusion
,that the Tibetan dGe lugs pa scholastics reached when they formulated two
. sorts of proper (mthun phyogs) and taken etymologically
170 JIABS VOL. 11 NO.1
(sgra bshad du 'jug pa'i mthun phyogs). It is -only in the latter type that h
dharmin is excluded. See my article, "On Note that followin \l
Tibetans' idea of (proper), it would come down to everything (!
has the property to be proved (siidhya), and (proper) Would be thi(:h
things which do not have this property (asiidhya). Not only would this
all the logical problems which plague us, but it is noteworthy that
. Irq
in k. 220, for examp.le, (see above) uses very siidhya and asiidhd'
and Manorathanandm and Devendrabuddhl (cf. Pan;zkii 312b6-7) gloss th
ks d
'ks .. em
as sapa .a an mpa.a respecuvely.,.;:
21. pp. 131-133 of his Aspects of Modem Logic. Dordrecht: D. Reidel
Publishing, 1970.;
22. op. cit. p. 132.,
23. Budd.hists (cf. his Viidanyaya)tci'7
Ratnaklru argue m the followmg way: All thmgs which do not produce their
effects successively or simultaneously are incapable of casual efficacity, likea
rabbit's horn. What is not momentary does not produce its effect successively
or simultaneously. Therefore what is not momentary is incapable of casual
efficacity and hence does not exist." See p. 60 in K. Mimaki, La Rifutation
bouddhique de la permanence des chases (sthirasiddhid0a1Ja) et la preuve de la momen_'
taneite des chases Paris, 1976. The pervasion here between
"not producing effects successively or simultaneously" (F) and "being incapable
of casual efficacity" (G) cannot follow Chi's model, for the point is that there
does not exist anything whatsoever which has F or G.}
24. It could very well be argued that not all cases of pervasion for Dig'-;
nagean logicians must be accompanied by examples. In particular, Dignaga
and his followers also used consequences (prasariga), which behave quite dif_i
ferently from the valid and fallacious reasons which are the concern of the
Hetucakra: they do have pervasions between their terms, but it is not usual to
give any examples at all. To take an illustrative case, Dharmaklrti in"
Pramii1Javiirttika IV, k. 12 (cf. also Pramii1Javiniscaya III) explains Dignaga's;
use of consequences by taking the prasariga that a Naiyayika universal (siimanya)
would have to be many different things because it is present in a multitude
of particulars. There is not discussion of an example at all in the commentaries,
nor in the Pramii'lJ,aviniScaya. Note however that Manoratbanandin in com-
menting on k. 12's parakalpitail,! prasarigo dvayasar{lbandhiid ekiibhiive
glosses dvayasar{lbandha as vyiipyav'yiipakabhava, which is essentially the
Chi is seeking to explain in Dignaga. See my article, "Pramii'lJ,aviirttika IV (1)"
in the Weiner Zeitschriftfur die Kunde Sudasiens 30,1986, pp.143-162.
rathanandin and the other commentators make it clear that the pervasion of
the consequence is logically equivalent to the pervasion in the consequence's
contraposition (prasarigaviparyaya), which yields a valid reason. But it is onlf
with regard to this contraposed form that one would need to present art
25. See A. MacDermott, An Eleventh-Century Buddhist Logic of 'Exists'.
Dordrecht: D. Reidel, 1969. R. RoutIey, "Some things do not exist," Notre,';
Dame journal of Formal Logic 7, 3, 1966, pp. 251-276. ','
26. See notes 24 and 29 in MacDermott op. cit. Routley himself had
.. . me suggestions as to how to handle a thoroughly Meinongian logic with
items, but I doubt that it is worth our while to enter into the
In M'" "AP
Ii tails. For an attempt at a emonglan semantIcs, see T. Parsons, . ro-
to Meinongian semantics," Journal of Philosophy, 61, 1, 1974, pp.
. .
27. p. 103 in B. K. Matilal, "Reference and existence in Nyaya and
Buddhist logic," Journal of Indian Philosophy 1, 1970, pp. 83-110.
28. See S. Katsura, "Dharmakirti's theory of truth," Journal of Indian
Philosophy 12, 1984, pp. 215-235.
Chinese terms
Policy Conflicts in Post-Mao China
A Documentary Survey, with Analysis
"Burns and Rosen have provided us with the most up-to-date,
and accessible collection of materials on post-Mao China now
-Lowell Dittmer, University of California,
An East Gate Book 360 pp. Cloth $39.95 Paper
China's Establishment Intellectuals
" ... one of the most absorbing and penetrating accounts to date of the
tension-filled, often puzzling relationship between China's intellectuals
the Chinese party/state." -Paul A. Cohen, Wellesley
An East Gate Book 250 pp. Cloth $35.00 Paper
State and Society in the Taiwan Miracle
Thomas Gold utilizes the "historical structural approach" to offer the
comprehensive sociological analysis of Taiwan's unquestioned success at
id economic growth with stability and equity.
An East Gate Book 176 PIP. 'Cloth $29.00 Paper
The Marginal World of Oe
In this study of the writings of Japan's most
gious writer, Professor Wilson brings to bear a
range of contemporary critical techniques.
An East Gate Book 168 pp. Cloth
Media and the Chinese Public
This translation makes available an extensive survey
of the Beijing media audience carried out by the
Beijing Journalism Association. The poll is the
large-scale media survey in China. (A special issue of
Chinese Sociology and Anthropology.)
200 pp. Paper $14.95
M. E. Sharpe, Inc.
80 Business Park Drive, Armonk, New York 10504
Divine Revelation in Pali Buddhism, by Peter Masefield. London:
George Allen and Unwin, 1986. xx,187 pages.
Peter Masefield applies his broad knowledge of the Thera-
vada canon and commentaries to two distinct purposes in this
book: first, to clarify a number of difficult or neglected concepts
in Pali, and second, to present a theory about the development
of Buddhism. Unfortunately, the combination of purposes de-
tracts from the overall contribution of Masefield's labors, mainly
because a weak historical thesis seems to restrict the scope of the
word studies which in themselves are serious and valuable.
Masefield begins by saying that, "the sad fact is that much
of the basic terminology and symbolism of the Nikayas is still in
need of detailed investigation" (p. xv). There is no denying this,
and it is to Masefield's credit that he attempts to rectify what
many of us deplore. He is most successful in the first chapter,
"The Spiritual Division of the Buddhist World," in which he
considers the distinctions between the ariyasiivaka and the puthuj-
jana. Masefield has collected a number of passages which suggest
that the puthuJj"ana was indeed considered as apart from the Bud-
dhist path (prthak; see the PTS dictionary, s.v. puthuJj"ana, which
says this meaning of separateness "is not felt at all in the Pali
word"), and, according to his conclusions, eternally so. Masefield's
discussion is quite thought-provoking, but it is regrettable that
he does not consider those passages which speak of the kalyii1Ja
puthuJj"ana (e.g., inPatisambhidiimagga and the Dlgha commentary)
which seem to suggest the opposite.
Chapters Two and Three also make contributions to our
understanding of Theravadin systematic thought, but they are
more grab-bags of suggestion than sustained arguments. Chapter
Two, "The Path," begins with the question of how "right view"
can be the beginning of the eightfold path, and continues with
considerations of the organization of the path according to the
threefold division of slla, samiidhi, panna, the idea of Dhamma as
sound, and the concept parato ghosa. Chapter Three, "The Goal,"
takes up different schemata of the fruits of the path. Masefield,
arguing that these fruits are actually discontinuous goals, ques-
tions whether they could be attained after the parinibbiina of the
Buddha and his great disciples.
174 JIABS VOL. 11 NO.1
Masefield's answer is negative, and his sweeping conclusions
are presented in a short (18 pages) fourth chapter in which his
"historical" thesis finally becomes explicit. For Masefield, Bud-
dhism was a short-lived esoteric movement. The "true Dhamma"
lasted only "a mere seventy years," and thus the subsequent Bud-
dhist tradition is merely the misguided product of puthuyjana
monks who, by definition, were unaware of this cataclysmic fact
(pp. 162, 163).
This might work as "theological" argument, even if a very
nihilistic one. As history, however, this is very poor stuff. Indeed
the requirements of good historiography are not
strong suit. He is inconsistent in his appeals to textual history,
justifying his selection of texts on the grounds that they "may,
for reasons of style, be said to form a literary unit" (p. xvii), while
also invoking the idea that these same texts have suffered a later
redaction which distorted their original message (p. 46).
Masefield organizes his investigation with elastic categories like
"the Nikaya period" and "the Abhidhamma period" which he
never specifies temporally. He has a disconcerting tendency to
make debatable links, such as his identification of Mi-Ia-ras-pa
as "the Tibetan counterpart of Ailgulimala" (p. 92). There are
problems of historical fact, such as his blunt statement that the
Vimuttimagga is a non-Theravadin text. And it is quite surprising
that the translator of commentaries attributed to Dhammapala
could simply state that this author wrote in the late fifth century,
ignoring the discussions about his date and works which have
appeared in the scholarly literature.
Masefield seems to have a low opinion of his co-workers in
the field of Buddhist studies (e.g., p. 54). This in itself is innocu-
ous, but I can't help wondering whether this opinion has led
Masefield to ignore generally the contributions of other scholars.
While there are profuse references to Pali literature, there are
remarkably few to secondary literature. In a book that seeks to
overturn much of scholarly opinion (especially the conventional
idea that the early Buddhist movement had a strong exoteric
orientation), more detailed reaction to individual arguments
would seem to be required. Moreover, Masefield is misleading
about his use of his colleagues' labors. He quotes verbatim existing
translations without acknowledgement (e.g., pp. 71, 73, 130,
153). This may not be uncommon these days, but it becomes a
disservice because Masefield is often critical of individual trans-
lators by name.
In his preface, Masefield seems to step back from his histor-
i ~ a l thesis: "My next reading of the Nikayas will probably cause
me to rethink some of the claims made in the present work but
if ... enough has been said to stir others into realization of the
need fQr a re-examination of the Buddhism portrayed in the
Nikayas, my efforts will have been rewarded" (p. xx.). I expect
that few students of Buddhism will want to adopt the general
claims of Masefield's book, but if my own experience is at all
representative, it will indeed send many back to the texts for a
fresh reading.
Charles Hallisey
Studies in Buddhist Art of South Asia, edited by A.K. Narain (New
Delhi: Kanak Publications, 1985), 139 pp., 54 figs., US $50.00
Studies in Buddhist Art of South Asia publishes seven papers
presented at an international conference on Buddhism held at
the University of Wisconsin in 1976. The essays are arranged
chronologically by topic, with the first three dealing with the
origin of the Buddha image and the remaining four with diverse
subjects. The three essays on the early Buddha image are, I think,
particularly suggestive and deserve careful reading by scholars
interested in early Buddhist doctrine and art.
The first of the early Buddha-image essays, by the volume's
editor A.K. Narain, proposes that the earliest anthropomorphic
Buddha images occur on a coin-type of the Saka king Maues
who reigned ca. 95-75 B.C.E. in the area of the Swat Valley and
Kashmir. Narain argues that the cross-legged figure on Maues'
coins, although long considered but mostly rejected by scholars
as a Buddha image, should be reconsidered as indeed the
Buddha. He suggests that the ideological underpinnings for the
creation of the anthropomorphic image came from the Sarvas-
tivada school of "Hlnayana" Buddhism and its philosophy of
"realism." Narain feels the Sarvastivadins associated themselves
with the Sakas as patrons in a mutually advantageous political
and economic alliance that allowed for the creation of the Buddha
image. The period from the appearance of the Maues coin-type
to the numerous examples of anthropomorphic images during
the reign of the Kw;;aI).a king Kaniska some 200 years later is,
according to Narain, one of experimentation. Extant Buddha
images of the period are few, however, and mostly on coins,
176 JIABS VOL. 11 NO.1
except for those on the Bimaran reliquary (which Narain accepts
as dating to the period of Azes II, ca. 30 B.C.E., due to four of
his coins found inside).
Most of Narain's evidence is not new, but he argues that it
has not been fairly analyzed and spends considerable time review_
ing past arguments. As with the other two writers on the early
image, Narain assumes that the reader has considerable familiar-
ity with the now extensive and increasingly complicated Buddha_
image bibliography.
Such familiarity is particularly helpful in understanding Joe
Cribb's essay on the origin of the Buddha image as revealed by
images on the coins of King Unlike the images on the
coins discussed by Narain, whose identification as the Buddha is
controversial, the Buddha images on coins are clearly
labeled as such. Cribb argues that on all the coins there are only
three different basic image types and three inscriptions, two that
identify the Buddha Sakyamuni and one the Buddha Maitreya.
That Maitrep on the coins is dressed in princely clothes and is
labeled a Buddha points out that the later distinction between a
buddha and bodhisattva was not made in time.
Cribb's coin evidence is helpful, and he proposes to use it
to comment on a vast array of theological and art historical issues,
such as the chronology of Gandharan sculpture and the early
artistic relationship between Marhura arid Gandhara. His basic
methodological assumptions are that the coin images were mod-
eled on existing sculpture; that unlike the sculptural models,
however, the coins are securely dated by inscription to
reign; and that; therefore, the coin images can be used to identify
sculptures of reign. Unfortunately, to properly judge
Cribb's essay would require a paragraph-by-paragraph analysis,
as, I think, there are serious problems with. many of his ideas
and conclusions. His key assumption, for example, that the coin
images were modeled on sculpture gets off to a shaky start when
he repeatedly shows that there are, in fact, very few likely
sculptural models and none that fit the coin images precisely.
Rather than come to the logical conclusion that the coin images
are not modeled on sculpture, but are independent creations
that share with sculpture certain underlying characteristics, Cribb
is forced to find his sculptural models scattered fro.m Gandhara
to Mathura (although the mints are all in Gandhara), each exhibit-
ing only this or that characteristic. The arguments made in this
essay are not well served by either the awkward writing style or
the unfortunate typographical errors for the figure numbers.
(Readers are advised to read Cribb's previously published ;;trticles
where much of the same material is presented:
Buddha Coins-The Official Iconography of Sakyamuni & Mai-
treya,"Journal of the International Association of Buddhist Studies 3.2
[1980]:79-88, and "The origin of the Buddha image-the numis-
matic evidence," in South Asian Archaeology 1981 [Cambridge,
The third article, on the origin of the Buddha image, written
by John Huntington, is an interesting attempt to judge, based
primarily on analyses of literary references, how early the an-
thropomorphic Buddha image was invented. We have seen that
Narain places it at ca. 100 B.C.E. in Kashmir or Swat; Cribb
concludes that there is little evidence for images before
reign and locates their emergence at Mathura; Huntington, on
the other hand, argues that there is a probability that images
were made during Buddha's own lifetime (5th c. B.C.E.), most
likely in Magadha. He suggests that it was not the monks but
the laymen who prompted the first making of images, and their
motivation was to enable themselves to gain merit by viewing the
Buddha (buddhadarsanapur;ya). While one may question certain
of Huntington's propositions in this long essay, the cumulative
evidence he presents does strongly argue for the existence of
Buddha images long before their extant examples (in stone) in
the 1st c. C.E. The question has always been: where is the earlier
archaeological evidence? Huntington tentatively presents a pos-
sible Maurya-period piece (ca. 3rd B.C.E.), a small stone relief
image, which would be the earliest example thus far known; but
one must wonder why there are no examples in the extensive
Buddhist sculptural remains of the 2nd and 1st centuries B. C.E.
The fourth article in the book is a straightforward analysis
of an iconographic formula popular in Gandharan art, the
Buddha flanked by two weapon-holding attendants in narrative
scenes. Its author, Joan Raducha, points out that identification
of the guardians is, however, anything but straightforward. She
shows that textual references do not explain or identify the atten-
dants. Rather, she relies on reconstructing the religious context
for attendant and protective deities in Gandharan sculpture and
in popular beliefs, concluding that the attendants are most likely
VajrapaI).i and Pafichika. Raducha reminds us of how few of the
often very prominent subsidiary figures we can identify in Indian
sculpture, as they are creations of concepts and beliefs not neces-
sarily recorded directly in texts.
The fifth article also deals with iconography. Janice Leoshko
178 JIABS VOL. 11 NO.1
presents a number of PaJa style reliefs of the Bodhisattva
Amoghapasa, a multi-armed form of Avalokitesvara. She finds
that Amoghapasa was popular primarily in the area of Gaya and
only during the 9th and early 10th centuries: She stresses that,
although it is not usually done, the possibility of such geograph_
ical and chronological specificity should be considered when deal-
ing with iconographical questions .
. The sixth article is by Walter Spink, on the internal chronol_
ogy of Cave 7 at Ajanta. Spink argues that Cave 7, and primarily
its major Buddha image, was the result of two separate artistic
campaigns, one that lasted from ca. 462-468 and the second
from ca. 477-479. As with Spink's previous work on Ajanta, one
stands in awe of the careful sifting of evidence that allows the
reconstruction of events that appear to explain what we see.
Spink is attempting to give, in far greater detail than is usually
possible in Indian art due to the lack of historical documentation,
a detailed explanation of the chronology of the making of the
monument. In many ways, he is suggesting what was in the minds
of the cave's makers, what their decisions were, and how these
decisions resulted in what we see. When one recalls that Spink
is working with almost no hard historical data, his results are
Is he correct? The explanations are, to my mind, too plaus-
ible, too helpful, to be "incorrect." They may be reconstructions,
with some pieces out of place, but most of the edifice must be
original and it enables us to understand the art in an unusually
direct and personal fashion.
The final article, by Martha Carter, is on the colossal Buddha
images at Bamiyan in Afghanistan. She makes the interesting
suggestion, based on the Chinese pilgrim Hsiian-tsang's com-
ments on his visit to Bamiyan in 632, that only the smaller (the
127 foot) of the present two colossal rock-cut Buddha images
existed at the time of his visit. HSiian-tsang does mention two
colossal images, but one he describes as made of metal andjoined
together in parts. This description has puzzled scholars (who
assumed he had mistaken stone for metal), but Carter proposes
to accept it at face value-that there was a now lost metal image.
Since HSiian-tsang estimates its height at 100 feet, this would be
a very large metal image indeed. Carter suggests that the larger
rock-cut image (the 175 foot) was not carved until the end of the
7th or beginning of the 8th century.
Most scholars today accept that the two extant monumental
stone relief Buddhas at Bamiyan are not as early as was thought
just a decade or so ago, when they were often dated 2nd/3rd c.
for the smaller and 4th/5th c. for the larger. That they both were
made after 600 C.E. has been shown, for example, by the work
of Deborah Klimburg-Salter and Zemaryalai Tarzi. Carter's sug-
gestion, ,however, that one of the colossi Hsuan-tsang mentions
is metal is difficult to judge. Although she points to examples of
monumental bronzes from both Western and Asian antiquity, a
100-foot standing Buddha in metal appears to me an interesting
but unlikely possibility. As with the other six essays, however,
Carter's article is an important contribution that will be of interest
to all students of Buddhist art and religion.
Robert L. Brown
Traditions of Meditation in Chinese Buddhism. Edited by Peter N.
Gregory. Studies in East Asian Buddhism no. 4. The Kuroda
Institute for the Study of Buddhism and Human Values. Hon-
olulu, University of Hawaii Press, 1987. 266 pp.
Since the publication of Studies in Ch'an and Hua-yen in 1983,
the first volume in the Studies in East Asian Buddhism series edited
by Robert M. Gimello and Peter N. Gregory, a number of signif-
icant contributions to our understanding of East Asian Buddhism
have appeared in this excellent series. The present book, which
is a collection of six lengthy articles on different aspects of medi-
tation in Chinese Buddhism, is the most recent. Despite the fact
that meditation in one of its many forms has always been at the
heart of the Chinese Buddhist tradition, surprisingly little has
been written on this topic from a scholarly point of view. For
this reason the present collection is a very welcome contribution
towards a deepening of our understanding of the contemplative
aspects of Chinese Buddhism.
The book opens with a long, very interesting and perceptive
introduction by Peter N. Gregory, the editor. Recapitulating the
views of previous and current authorities on Zen/Ch'an Bud-
dhism, he points out the need for revising many of our fixed
opinions on Chinese Buddhism meditation, which hitherto has
tended to be identified solely with Ch'an Buddhism. Gregory
presents his views with detailed consideration of the hermeneutics
of the various traditions of meditation within Chinese Buddhism,
i.e., the methods of meditation seen in relation to their underlying
doctrinal structures, "and thus gIves us a complex, but at the same
time useful methodological framework with which to approach
and understand the issue under discussion.
The first paper, "Meditation in Fa-hsiang Buddhism" by
Alan Sponberg, is devoted to a hitherto little studied subject (at
least among Western scholars), namely meditation as practiced
in the Fa-hsiang tradition, which originated with Hsuan-tsang.
The author has chosen to present his topic through a discussion
of two methods of meditation, both of which were designated
by the polyvalent term kuan. The first use of this term relates to
the practice of visualization, here highlighted by an example of
Maitreya visualization found in Hui-li's Ta Tang Ta Tzu-en Ssu
San-ts'ang Fa-shih chilan. In its second usage kuan appears as a
method of "discernment" or "insight," such as is found in the
traditional prajnaparamita literature, but of course based on the
dharmalak:;a1}a perspective. Sponberg gives considerable space to
a general discussion of the importance of defining the meaning
of "meditation." In his concern for clarification of terminology
the author suggests that the Sanskrit term bhavana would be a
better designation for meditation than the more limited dhyana,
since it includes any practice productive of nirva1}a. While much
of Sponberg's argument on the issue of dhyana is certainly valid
and raises important questions, I think that he might have de-
voted a separate paper to a discussion of this important topic.
By so doing, he would have been able to treat a wider range of
Fa-hsiang practices in the present paper, one obvious drawback
of which is precisely its narrow focus on only two aspects of
Fa-hsiang meditation. This limitation is unfortunate when we
consider the copiousness of the existing sources. What is lacking
in particular is a discussion of the important Tantric practices in
Hsuan-tsang's tradition, practices which the author at least ought
to have mentioned when formulating his broad definition of
meditation as bhavana.
N ext follows a very interesting and detailed paper by Daniel
B. Stevenson, "The Four Kinds of Samadhi in Early T'ien-t'ai
Buddhism." Using very wide-ranging source material, the author
gives an in-depth presentation and discussion of the structure
of early T'ien-t'ai ritual and contemplative practices. Relying
mostly on the writings of Chih-i (548-597), Stevenson shows how
the meaning of chih-kuan, the central type ofT'ien-t'ai meditation,
covers a wide range of practices, including rituals of repentance
as well as buddhanusmrti, Lotus Samadhi and i-hsing san-mei, the
so-called "one practice samadhi." The influence of Tantric Bud-
dhism is also noted. The author treats a large complex corpus
of texts expertly and succeeds in showing how structured and
multi-faceted the meditation practices of the early Tien-t'ai tra-
dition were, A minor point of criticism concerns a certain lack
of clarity in the way in-which Chih-i's master Hui-ssu (515-577)
is presented when the author discusses the origin and structure
of the early Tien-t'ai practices. An acknowledgment of Paul Mag-
nin's La Vie et l'Oeuvre de Huisi (515-577): Les Origines de la Secte
Bouddhique Chinoise du Tiantai, Publications de l'Ecole Franc;aise
d'Extreme-Orient, vol. 116 (Paris: Ecole Franc;aise d'Extreme-
Orient, 1979) would have been appropriate in this respect.
Bernard Faure's "The Concept of One-Practice Samadhi in
Early Ch'an" sets out to discuss the main methods of meditation
on so-called Northern Ch'an Buddhism, i.e. the tradition
popularized by Shen-hsiu (605?-706) and his successor P'u-chi
(652-739). As Stevenson shows, I-hsing san-mei constituted an
important part of the higher practices in the Tien-t'ai regimen
and had a fixed meaning as a special type of practice. In the
Ch'an trad'ition one-practice samadhi took on a more general
meaning, and was redefined as a designation for a number of
practices such as shou-i (maintaining an unwavering concentra-
tion), kuan-hsin/kan-hsin (contemplating the mind), and even nien-
fo (Buddha invocation). Faure's paper demonstrates a versatile
and sure grasp of a large range of sources as well as of the general
history of the contemplative traditions in Chinese Buddhism.'
This allows him to establish a plausible as well as clear presenta-
tion of the topic under discussion. However, when he discusses
the respective practices of Northern and Southern Ch'an, I am
rather reluctant to accept his view that the traditional distinction
between gradual and sudden enlightenment is invalid. Clearly
the practices of meditation advocated by Shen-hsiu and the earlier
patriarchs of the East Mountain/Northern School were gradual,
if not in theory, then at least in practice. This can be seen in
Tao-hsin's entry in Leng-ch'ieh shih-tzu chi, in Hung-jen's Hsiu-hsin
yao lun and in Shen-hsiu's Kuan-hsin lun, all of which speak of
preparation and a gradual entry into samadhi in their respective
sections dealing with the actual practice(s) of meditation.
In his paper "Ch'ang-lu Tsung-tse's Tso-ch'an I and the
'Secret' of Zen Meditation" Carl Bielefeldt discusses what he calls
"the earliest and in some ways most influential" (pp. 130) manual
of Ch'an medita,tion and presents a very fine translation of the
work. The text in question is the Tso-ch'an I by the Yiln-men
monk Tsung-tse (d.u.), a relatively short treatise on meditation
182 JIABS VOL. 11 NO.1
encountered as a part of the Ch'an-yuan ch'ing_
kuez, complIed In 1103. I feel, however, that the author in his
conscientious effort at contextualizing the Tso-,ch'an 1 in relation
to earlier Tien-t'ai manuals of meditation (such as the Mo-ho
chih-kuan and especially the T'ien-t'ai hsiao chih-kuan) , may, to
some extent, have looked in the wrong place. It is true that
Tsung-tse mentions Chih-i and his Chih-kuan in the manual, but
this indirect reference is solely concerned with how to check
demonic disturbances while meditating, and does not refer to
the actual practice of chih-kuan. In fact, Bielefeldt himself cor-
rectly concludes that the Tso-ch'an 1 does not have much in com-
mon with the much larger Tien-t'ai texts on meditation. Further-
more, the doctrinal differences between the Ch'an schools of the
late Tang/Five Dynasties Period-with the exception of the Fa-
yen School-and the revived T'ien-t'ai School of early Northern
Sung are fairly obvious and would have made a direct link there
problematic. For these reasons the author might have looked in
other directions for material with which to illuminate the possible
origin(s) of the manual as well as other similar Ch'an texts of a
later date. Personally I would have tried to relate the Tso-ch'an
1 to references to meditation in material from the Yun-men
School (since Tsung-tse belonged to that school), and also to
information to be found in the other lineages. A close scrutiny
of the Tun-huang Ch'an mss. is most likely to yield significant
information in this regard. Slightly later meditation manuals and
related texts such as the Tso-ch'an 1 by Fo-hsin Pen-ts'ai (d.u.)
and the Tso-:ch'an ming by Fo-yen Ch'ing-yuan (1067-1120), both
of the Lin-chi School, would also serve as useful points of refer-
David W. Chappell follows next with his "From Dispute to
Dual Cultivation: Pure Land Response to Ch'an Critics." This
paper examines two late T'ang Pure Land masters' response to,
and counter-attack on, general Ch'an criticism of the relevance
of their practice. The highlights of this paper are Chappell's very
interesting examples of the Pure Land master H ui-jih's (680-748)
anti-Ch'an polemics which are seen in relation to the criticism
directed by the Ch'an master Ling-yu against the excesses of
many Ch'an adherents. Likewise Chappell's discussion of the
Pure Land dialectics of Fei-hsi (d.u.) serves as a useful point of
departure for understanding the later dual cultivation of Ch'an
and Pure Land practices. For unknown reasons the author has
chosen not to mention the importance of Ch'an and Pure Land
practices in the Fa-yen School, a syncretism which is perhaps the
. most outstanding example of dual cultivation in Chinese Bud-
dhism. I am also hesitant to accept Chappell's view of the average
pure Land Buddhist's attitude to practice. He asserts that "at a
fundamental level they were aware that all was empty and there
was nothing to attain" (pp. 189). Although such an argument
was put forth by Tan-Iuan from the point of the two levels of
meaning, i.e., as a philosophical view, such an understanding-if
employed in practice-would simply prevent the desire for re-
birth in Sukhavati, the very goal of the Pure Land devotees. In
order to go to the Pure Land for rebirth one must believe in it!
Another minor point on which I disagree with the author con-
cerns his insistence that "the Southern SchoQI of Ch'an emerged
. partially by defining itself over against various Pure Land
tices" (pp. 174). Southern Ch'an emerged-I believe-as a gen-
eral reaction to intellectualism and relative practices current in
Tang Buddhism as a whole. (This includes the practices em-
ployed by various lineages within the Ch'an tradition itself).
The final paper of the volume is E.. Buswell's
"Chinul's System of Chinese Meditative Techniques in Korean
Son Buddhism." In this paper, which may serve as a supplement
to the author's momumental work, The Korean Approach to Zen:
The Collected Works of Chinul, the focus is on the teachings and
methods of meditation propounded by Chinul (1158-1210), the
reviver of Korean Son Buddhism during the middle of the Koryb
Dynasty (936..;..1392). Buswell presents five aspects of Chinul's
thought on practice, including Amit'a invocation. However,
Chinul's Son developed in two basic stages. In the early stage it
shows strong influem:e from the Plaiform Scripture (Liu-tsu fa-pao
t'an ching) attributed to Hui-neng. The teachings which de-
veloped around the combined practice of samiidhi. and prajiiii
especially left a significant impact on Chinul during his formative
years. The later and more mature stage is that based on Ta-hui
Tsung-kao's k'an-hua or hua-t'ou practice, i.e., the type of medita-
tion which employs kung-an. Buswell's impressive knowledge of
the Korean sources is readily apparent in his demonstrations of
how these Chinese Ch'an techniques influenced and shaped
Chinul's Son in the two basic stages.
Traditions of Meditation in Chinese Buddhism is a collection of
informative and highly interesting papers written by some of the
best American scholars in the field of East Asian Buddhist studies.
The volume is significant because most of the information it
presents is new or presented in a new light, and one of the book's
most marked characteristics is the inner coherence and structure
which connects the individual papers. This places meditation in
a central position in the history of Chinese Buddhism and thus
gives of its importance: My only
real pomt of cntICIsm of thIS otherwIse excellent collectlOn is the
limitation of topics treated. If the book is intended to present
the more important traditions of meditation in Chinese Bud-
dhism, then I would have liked to see articles on meditation in
the Nan-Pei Chao period (preferably on meditation literature),
one on Chen-yen (Tantric) Buddhism in the Tang, one on the
influence of Tibetan practices after the Yuan, and one on Ch'an
practice in the Ming. Such papers would have given the present
compilation more perspective as well as a deeper historical dimen-
sion. However, the book is an important contribution to our
understanding of the role of meditation in Chinese Buddhism
and can be warmly recommended.
Henrik H. Sorensen
IDr, c. Bielefeldt
rofpt. of Religion
CA 94305
(;." '
'1.R.L. Brown

t. of Art, Design & Art History

300 Dickson Art Center
!lirtiversity of California,
'''','';',',L', ,os Angeles

Ilks Angeles, CA 90024

!Dept. of Asian Languages &
IWLiterature DO-21
IUniversity of Washington
WA 98195
fiir. C. Hallisey
l'Pepartment of Theology

North Sheridan Road
IL 60626 .
!1tt.. Hamlin
North Racine #3E
rChicago, IL 60657
Dr. J. Newman
Dept. of South Asian Studies
1244 Van Hise Hall
University of Wisconsin
Madison, WI 53706
Dr. R. Saloman
Department of Asian Languages
and Literature DO-21
University of Washington
Seattle, WA 98195
H.H. Sorensen
0stasiatiske Institui, Ki<lbenhavns
Kejsergade 2'
1155 Ki<lbenhavn K
Dr. T.J.F. Tillemans
25, rue Winke1ried
2300 La Chaux-de-Fonds
Both the Editors and Association would like to thank Indiana Univep
sity and Fairfield University for their financial support in the produc-
tion of the Journal.
The Editors wish to thank Mr. Kevin Atkins for his invaluable help
in the preparation of this issue.
Copyright The International Association of Buddhist Studies 1988
ISSN: 0193-600X
Indexed in Religion Index One: Periodicals, American Theological Li-
brary Association, Chicago, available online through BRS ( B i b l i o ~
graphic Retrieval Services), Latham, New York, and DIALOG Infor-
mation Services, Palo Alto, California.
Composition by Publications Division, Grote Deutsch & Co., Madison, WI 53704.
Printing by Thomson-Shore, Inc., Dexter, MI 48130.