Professional Documents
Culture Documents
'\
i; ungo ii
Haein-sa :.e-fp 3-
haeo
Ho-tse ;,'
Ho-tse Shen-hui ,.
Hua-yen ching hsing-yuan p'in shu t fa f::f fj .. J:"W
Hung-chou ;t +\-1
hwadu
Hwaomt;fz.
-v- '(I,f. - 4 11
Hwaom-Son
hyonjung-hyi5n if t,
i 'fL.
lj'i r:!- I(;s
, /..(. ,;\7 /,'1..
zmun-su 1:r..J..:t. \
irhaeng sammae - fr:3- 19f
k'an-hua Ch'an J@
Kanhwa I.tr J-h \;:!
'./ III .,.,lz.W_;t..J.!::...,. .... 1'lt)
ka h
1:. 'I.: -'"
n wa on "f:t
kongan W,
Koryo ,; -10-
Y-l
44 JIABS VOL. 12 NO.2
Kuei-feng Tsung .. mi :$ t'? '*'
u 1'" _I .J...i
. kuJung-hyon g '1' 4:,
Kwon su Ch6nghye kyolsa mun 'Z
Lin-chi . .
Ma-tsu Tao-i i.B- L-
-tr
Mi-an 1i:; .
.. F-'-
munyom-rE ,--:s-
, ;1;. )I''J..
P ansa-su 1}Jif Y
Popchip pyorhaengnok choryo pyongip sagi ;'.1$..l.,'J ft-A..f-L,c,
pyorhaeng ;?,.J fJ
saJi 1-1f.J .
samhyon-mun "'- i F'l
S6ngch' a1 ,tit-
- .!---t
Songgwang-sa 4
Sonmun ch6ngnok H )): ..
Sasan Hyujang i5
suo ilsi l,t ,t% - 8 f
Suryan Kusan t,j fL J-..
Susim kyol 1
'
r ,c. H':..
Susan-sa i.:t. (:#)
T'aego Pou
taesim chungsaeng r_ ,..5"\ 'f:Z L
taesim pombu 7-'-..,6' fl
Ta-hui Tsung-kao f::... ;1--
ton H
tonG H.
tonG ch6msu
tonG tonsu H: ifHr
tonsu iftH
tonsu chOmo l'r '\\
tonsu tonG
ts'an-ch'an f:,. ;,..a
,'f" I{ . o ..... ...... /.,k.
Wan-shan t ung-kuez chz ;Jb .Kl' l},
Yang-ch'i .. L
Yit ..
*- Ii" *
Yi Chongik '-"'-
Yung-ming Yen-shou k<..t1f-lJ.tt
controversy over Dharmakaya
in India and Tibet: A Reappraisal
of its Basis, Abhisamayalamkiira Chapter 8
by John]. Makransky
1. Introduction
Approximately 1200 years ago a disagreement developed in India
over the description of complete enlightenment in Mahayana Bud-
dhism. The disagreement focused on the Abhisamayalarrtkara (AA,
c. 4th-5th century C.E.), a commentary on the Prajnaparamitasutros
ascribed by late Indian scholars to Maitreya.! The AA's eighth and
last chapter explained the final result of the Mahayana path, com-
plete enlightenment (referred to as ''phaladharmakaya'') , in terms
of multiple buddha kayos (buddha "bodies").2 But its verses, dense
with possible meaning, were very ambiguous. Arya Vimuktisena
(c. early 6th century) understood it to be teaching three kayos,
while Haribhadra (late 8th century) thought it taught four. Ratna-
karasanti (c. 1000) believed that their disagreement concerned not
just the wording of the AA, but the nature of dharmakaya in non-
Tantric Mahayana Buddhism as a whole. He sided with Arya
Vimuktisena, as did Abhayakaragupta (early 12th century). Later
in Tibet, the Sa-skya scholar Go-ram-pa bsod-nams seng-ge sup-
ported Arya Vimuktisena in asserting three kayos, while
pa scholars backed Haribhadra's assertion of four. Thus, if we
take Haribhadra as its initiator, the debate over the number of
kayas has continued from the late 8th century to the present day;
having progressed from the Indian to the Tibetan branch of the
Indo-Tibetan Buddhist tradition. Yet most modern scholars, bas-
ing themselves on Haribhadra and his Tibetan followers, have
reported simply that the AA teaches four kayos, as if they were
unaware of the controversy.3
45
46 JIABS VOL. 12 NO.2
Any attempt to analyze the debate is complicated by the
fact that it has been a diachronic discussion rather than a syn-
chronic one. It is not a discussion between two contemporaneous
scholars. Rather it has been an ongoing'interchange conducted
over centuries, in which scholars of each period, attempting to
address the philosophical and religious problems of their OWn
time and place, have written responses to scholars of earlier
periods. Developments in religious thought since the last re-
sponse forced reconsideration of old questions in the light of
new viewpoints. What was important to say about enlighten_
ment, and what methods were used to analyze or describe it,
changed somewhat from age to age and culture to culture.
If the historical perspective is lost, it becomes impossible to
sort out what the whole debate has been about. One complicating
factor is the abhorrence of orthodox scholiasts to give the ap-
pearance of personal innovation. From the perspective of Bud-
dhist traditionalists, the truths of Buddhism were realized by
buddhas and saints (such as Maitreya) and then revealed by them
in sacred scriptures. The commentator's job in explaining those
scriptures was not to innovate, but to explain the meanings
intended by their authors, since those meanings were truths
realized by those authors. At some stage within the Buddhist
tradition the AA was taken to be such a sacred scripture (hence
its ascription, by Haribhadra, to Maitreya). Each commentator
obeyed the unwritten rules of orthodoxy according to which the
only way to reformulate the tradition they received was to read
their reformulation into the texts they inherited. Scholars, like
other people, do not work in a vacuum. They are conditioned
by their historical and cultural context. Although commentators
. made interpretations of the AA appropriate to their own times
and places, they always did so within the context of explicating
the original intentions of its author. Because of this, the debate
over the number of kiiyas took on the appearance of a trivial
disagreement over the meaning of a few verses of one abstruse
text. Although Haribhadra's reinterpretation of AA 8 was in-
novative, it was within the rules of orthodoxy, because he read
his meaning into the received text.
4
Therefore, the debate over the number of kiiyas, examined
diachronically, resolves into a number of different stages of
-discussion in which the issues at stake partially changed over
CONTROVERSY OVER DHARMAKA.YA 47
time. Broadly speaking, I would describe those stages as follows:
the AA's 8th chapter represents an attempt, for the first time,
to homologize two semi-autonomous Mahayana descriptions of
enlightenment: a Prajiiaparamita (PP) sidra description and a
three-kaya Yogacara sastra description (this will be the subject
lllatter proper of this paper). Arya Vimuktisena's task was to
explicate the very dense verses of AA 8 in a form which exposed
its author's intention, while reiterating the Yogacara under-
standing of enlightenment as, in essence (svabhavika1;), an experi-
ence of the highest yogic realization, inconceivable to those who
have not realized it, beyond discursiveness, unconditioned, and
supramundane. Haribhadra, writing several centuries later, felt
compelled in his reading of AA 8 to conduct an analysis of the
kiiyas along clear Madhyamaka lines, in a way which addressed
the conceptions of enlightenment which had developed since
Nagarjuna's time, clarifying the new categories of multiple kayas
. in a way consistent with the earlier Madhyamaka dialectic. View-
. ing enlightenment not only as an inconceivable yogic realization
but also as an object of logical analysis like any other object, he
used his interpretation of the AA to separate out contradictory
elements and assign them to their appropriate domains, result-
ing in four buddha kayas. Later in Tibet, Sa-skya and dGe-Iugs
scholars chose either Arya Vimuktisena's or Haribhadra's view,
depending on what implications for buddhology they saw in
their project of developing an all-inclusive systematic philosophy
out of the thousands of sutras and sastras. they had received from
India. Within that systematic project, the Tibetans perceived a
number of problems as inter-related: problems concerning the
two truths, the perfect knowledge of them (which is enlighten-
ment), and the description of that knowledge as "embodied" in
buddha kayas. To analyze this 1200-year-old controversy, then,
requires that we study it in each of its historical stages.
5
It is logical to begin such a study by analyzing the received
text upon which the debate explicitly centered at every stage,
i.e. the AA. This will require a fresh look at AA's 8's place in the
history of Mahayana thought. But even apart from the debate,
it is well worth a fresh reexamination, because at the time that
it was written, there was a tremendous diversity in the descrip-
tions of buddhahood in Mahayana sutras and sastras, reflecting
a diverse set of views which had developed in different milieus
48 JIABS VOL. 12 NO.2
and textual traditions. When the AA's eighth chapter is examined
in relation to its textual antecedents and historical context, it
can light on the by early
Mahayana masters to homologIze these dIverse VIews. I believe
that the reason AA 8 has always been so difficult to interpret is
that its presentation of buddhahood is neither an independent
creation, nor a restatement of what was said in other treatises
of its time. Rather, it represents a synthesis of two different
ways to describe buddhahood: a Yogacara siistra way and a PP
sidra way. It functions like a grid to map a Yogacara model of
enlightenment onto the Prajiiiipiiramitii sutras.
6
II. The Heart of the Controversy: Abhisamayalamkara Chapter 8,
vss. 1-6
The controversy over the number of kiiyas centers on the
first six verses of AA, chapter 8. Verse one describes a buddha's
svabhavikakaya, Intrinsic Body:
sarviikiiriir{L visuddhir{L ye dharmii/:t priiptii niriisraviiJ;J
sviibhiiviko munel; kiiyas te$iir{L prakrti-Iak$arJa/:tl1 AA 8.1
The undefiled dharmas which have obtained purity in all
respects,
The Intrinsic Body of the Muni has their innate nature as
its characteristic.
Whatever this first verse means, all commentators agreed
that it teaches the first kiiya of a buddha, the sViibhiivikakaya,
understood in some sense to be the innate nature of the "unde-
filed dharmas." The undefiled dharmas (niriisrava- or anasrava-
dharmii/:t) are a buddha's pure mental qualities, his gnoses (jiiana),
obtained through the complete realization of the Mahayana
path. Verses two through six list these undefiled dharmas, divided
into twenty-one types, and then relate them to the word "dharma-
kiiya": '
bodhipak$iipramiirJiini vimok$ii anupurvasa/:tl
naviitmikii samiipatti/:t krtsnar{L daiavidhiitmakamll 8.2
abhibhviiyataniiny prakiiriirJi prabhedata/:tl
CONTROVERSY OVER DHARMA.KA.YA
. arar;ii prar;idhijftiinam abhijftii/:t pratisamvidal:t/ I 8.3
sarviikiiriis catasro 'tha suddhayo vaJitii daJal
baliini daJa catviiri vaiJiiradyiiny arakJar;amll 8.4
trividha7[l s7[lrtyupasthiina7[l
vasanayii/:t samudghiito mahatI karur;ii janell 8.S
iiver;ikii muner eva dharmii ye
sarviikiirajftatii ceti dharmakiiyo 'bhidhZyatell 8.6
"The factors which foster enlightenment, the measureless
thoughts, the liberations, the nine meditative attainments,
the ten meditative totalities, the bases of overpowering di-
vided into eight kinds, the meditative power blocking others'
passions, the knowledge resulting from resolve, the super-
natural knowledges, the analytical knowledges, the four
total purities, the ten sovereignties, the ten powers, the four
forms of fearlessness, the three ways in which [a buddha]
has nothing to hide, the threefold mindful equanimity, the
nature of never forgetting, the complete destruction of
[negative] propensities, the great compassion for living
beings, the eighteen qualities unique to the Muni, and om-
niscient wisdom": thus is the dharmakiiya denominated.
49
Arya Vimuktisena, author of the earliest commentary ex-
tant, understood all six verses to be teaching one kaya of the
buddha, which is first called "svabhavikakaya," and later "dharma-
kaya." He read "dharmakaya" of verse 6 as a synonym for "svabhii-
vikakaya" of verse 1.7 Over two centuries later, Haribhadra re-
interpreted the verses, arguing that Arya Vimuktisena had been
mistaken in his understanding of these two key terms.
"svabhiivikakaya" of verse 1 and "dharmakaya" of verse 6 were
not synonyms, he said. They referred to two different aspects
of buddhahood: the first being the emptiness of the undefiled
dharmas, and the latter being the collection of those dharmas
themselves.
8
All commentators agreed that after the sixth verse, AA 8
taught two more kayas: sambhogikakaya (Enjoyment Body) and
nairmar;ikakaya (Emanation Body). Therefore, the debate over
whether it teaches three or four kayas, actually resolves into a
debate over whether its first six verses teach one kaya or two.
Three concepts found in these verses are at the very heart of
the controversy: 1. svabhiivikakaya, 2. dharmakaya, and 3. the
undefiled buddha dharmas (anasravadharmas). Any attempt to
50 JIABS VOL. 12 NO.2
resolve th.is controversy us to go back history
of BuddhIst thought and examme these concepts withm textual
traditions antecendent to the AA.
III. Sarviistiviida Abhidharma
Traditionally, the mark of a Buddhist has always been his
or her going for refuge to the Three Jewels (triratna): the
Buddha, the Dharma and the Sangha. Sarvastivada scholars
posed the question: Precisely what is the Buddha refuge? When'
one takes refuge in the Buddha what is one taking refuge in?
The gives a reply:
One who goes to the Buddha for refuge goes for refuge to
the afaik.$a dharmas which make him a buddha; [the dharmas]
because of which the person is called "buddha"; [the dharmas]
by obtaining which he understands all, thereby becoming
a buddha. What are those dharmas? KJayajiiana, etc., together
with their attendants.
9
It goes on to say that one goes for refuge not to the Buddha's
physical body, referred to as his "rupakiiya," but to the
dharmas comprising his mind. The reason is that these dharmas
are undefiled (aniisrava) , while his body remains defiled even
after enlightenment.
On the same issue, the SarvastivadaMahiivibhii$iiSiistra says:
Some say that to take refuge in the Buddha is to take refuge
in the body constituted by the Tathagata's head, neck,
stomach, back, hands, and feet. It is explained, then, that
the body, born of the father and the mother, is [composed
of] defiled dharmas, and therefore not a source of refuge.
The refuge is the Buddha's afaik.$a dharmas which comprise
enlightenment (bodhi), i.e. the dharmakaya.
1o
In these formulations, the Sarvastivadins identified the qual-
ities which made a buddha a buddha, that is, his essence. They;
identified this essence to be the undefiled qualities of his mind:
his aniisrava (aSaik$a) dharmas. And they called it the
kiiya," which could be translated in this context as the "Body of
CONTROVERSY OVER DHARMAKAYA
51
lfUndefiled] Dharmas." It was dharmakiiya,.a unde-
IDled essence, as opposed to hIS rupakiiya, hIS physIcal body,
iWhich constituted the Buddha refuge.
It appears there were different traditions within Sarvasti-
as to .the identity of the dharmakiiya's undefiled dharmas.
notes that some scholars identify the Buddha ref-
primarily with the eighteen dharmas exclusive to a buddha,
!the so-called "iive7Jika dharmas," which coexist with his kJaya-
Jiiana. II explained at length in t?e Kosabhii$ya, where
,:they are IdentIfIed as the ten powers (dasabala) , four fearless-
!fiesses (vaiSiiradya), three mindful equanimities (smrtyupasthiina),
!and the great compassion (mahiikarunii).12 With these, other men-
qualities, possessed by both buddhas and non-buddhas, are
:i:lescribed. 13 Together this collection constitutes close to the same
riist of undefiled dharmas which is presented throughout the PP
and in AA 8 vss. 2-6 quoted earlier.
t Later, the Kosabhii$ya uses the term "dharmakiiya" in a new
It describes buddhahood as the ''phalasampad,'' the "attain-
of the fruit." In this context the term "dharmakiiya" refers
!io buddhahood in its entirety, not just to its undefiled mental
ualities. Vasubandhu explains that dharmakiiya, meaning
phalasampad, includes four attainments: jfiiinasampad (gnosis at-
:trunment), prahii7Jasampad (riddance attainment), prabhiiva-
(power attainment), and rupakiiyasampad (physical body
:attainment).14 This "dharmakiiya phalasampad" of Sarvastivada
.inay be a precursor of the AA's "phala dharmakiiya," which also
!refers to buddhahood as a whole.
W. Dharmakayaand Buddhadharmas in the Prajiiaparamita Sutras
.. The full enlightenment of a buddha, samyaksarrtbodhi, is not
treated at any length as a separate topic or chapter within the
PP sutras (except in the revised PP to be discussed later). In
fact, reference to "dharmakiiya" and "rupakiiya" in the PP sutras
!s only very occasional. However, these sutras do refer to buddha-
hood indirectly, and often, when they present formulaic lists of
dharmas" (sarvadharmii/:t). The "all dharmas" are understood
comprise all phenomena in the universe, as described in
Abhidharma.
15
Included among all phenomena, of course, are a
52 JIABS VOL. 12 NO.2
buddha's undefiled dharmas (aniisravadharmas), as they are listed'
in AA 8. They are presented in extensive or abbreviated forn{
throughout the PP sutras. As the collection of buddha's mental'
qualities, his aSaik$adharmas, they constitute what the
stivada Abhidharma referred to as his "dharmakiiya." It is impoi-_!
tant to note;- however, that unlike in theAbhidharma, nowhere
in the PP sutras is the collection of buddha's undefiled dharmas
in itself identified as being the "dharmakiiya." The reason fof
this probably lies in the difference between the ontologies of
the Abhidharma and the P P sutras .Ji'
The purpose of the Abhidharma was to negate the apparent.
permanence, etc., of things by analytically finding the dharmas
which were their ultimate constituents. In contrast to this, the
purpose of the PP sutras was to negate the ultimacy of
dharmas themselves, to deny their self-existence (svabhava). rd
formulaic repetition of the dharma lists, which are drawn
from Abhidharma, was done in order to deny the self-existence'
of everyone of the dharmas listed. The PP's analysis
salvific insight (prajna) does not find dharmas. It finds only their
emptiness of self-existence (svabhavasunyata).16 This
is known as the "prajnaparamita," the perfection of wisdom. It!
conjoined with the mind seeking enlightenment for the
of all other beings (bodhicitta), is the very heart of the Mahayana'
path, which when completed, issues in buddhahood.l
Like the Abhidharma, the PP sutras identify the dharmaka:jd;;
not the rupakaya (physical form), as that which really constitutel
a buddha, his essence. But they differ as to what that
is.17 In the Abhidharma it was the buddha's undefiled
these were his dharmakaya.But nowhere in the PP sutras is
dharmakiiya directly identified with the undefiled dharmas.
is because the highest attainment in the P P is not a
of dharmas, no matter how exalted, but rather the perfect
tion of the emptiness of all dharmas. Since "dharmakaya" is
of the words used to describe that highest attainment, its
ing in the PP sutras is quite different from its meaning in
Abhidharma. In this regard, two observations should be
1. From the perspective of prajnaparamita, the
along with all other dharmas, are not perceived. What is not;
perceived by perfect wisdom cannot be taken as the very essenct-
of a buddha. IS 2. This means that unlike the Abhidharma, the
,,-,
CONTROVERSY OVER DHARMAKAYA
53
"'utras-do not identify the buddha's dharmakaya with the collection
per se. They identify it instead with sunyata,
'the emptiness of all dharmas, and with prajiiaparamita, the reali-
ia
tion
of that emptiness. 19
Three Kayas in the Emerging Yogacara
In a number of early Mahayana sutras, along with references
)0 the formless dharmakaya of the buddha, there are physical
of buddhas which go far beyond what is found in
the Pali canon.
20
Attempts have been made by scholars to trace
(the historical development of these ideas in Buddhism prior to
the full-blown advent of the Mahayana.
21
Here I will just note
:t}lat certain sastras seminal to a newly emerging Yogacara school
'reformulated earlier two-kaya descriptions in order to accomo-
;aate these new forms. They presented a new theory of three
',Was: the svabhiivikakaya, the sambhogikakaya, and the nairma-
,/t;ikakaya. Here "sambhogikakaya" was the term for the exalted
'io/thiigatas of the Mahayana sutras, while "nairmalJikakaya" re-
to a buddha's infinite emanations into the realms of living
22 Both of these kayas were to be considered sub-categories
?fthe earlier, wider category: rupakaya. The svabhiivikakaya cor-
responded broadly to what the Mahayana sutras called the
It will be the focus of whatfollows.
-" The earliest text known to formally introduce and explain
Jdistinct terminology of three kayas was the MSA, in its ninth
chapter, on enlightenment (bodhi). The MSA served as the basis
,for extensive discussion of the three kayas in the Msg, which
yften quotes it. These two texts with their commentaries seem
:.t(}(onstitute a core Yogacara literature upon which was based
'piscussion of three kayos in numerous other texts: the Kaya-
trayasutra and Kayatrayastotra, Kayatrayavatarasastra, Ratnagotra-
,vibhiiga, Buddhabhumivyakhyana, etc. Brief mention is also made
of three kayas in the DhDh V and in the MA Vbhiisya (the AA, as
case, will be discussed below).23 The MSA and Msg were
f.uthored in the formative period of the Y ogacara school, the
f?rmer perhaps in the 3rd to 4th century, the latter in the late
1
th
century C.E.24 Together, these texts give us a good picture
of the intellectual milieu in which the three-kaya theory first
54 JIABS VOL. 12 NO.2
appeared. They explain the three-kaya theory by demonstratin
its relation to other Y ogacara models of enlightenment: a s r a y a ~ .
paravrtti, dharmadhatuvisuddhi, vimalatathata and nirvikalpajiiiina
and dharmakaya. Here, I will only make a few points particularl;
relevant to AA 8. .
To begin with, the MSA and its commentaries agree With
the PP siltras that while the undefiled dharmas are acknowledged
to b ~ qualities ofa buddha, they ar? not taken as his defining
qualIty or essence. MSA 9.4 says: [Buddhahood] consists of
excellent qualities, but it is not defined by them."25 Sthiramati's
commentary explains that buddhahood is obtained by a c ~
complishing the various undefiled dharmas, etc., and when ob-
tained, can be said to possess those qualities. But it is not defined
by them, because those qualities, as understood through concep_
tual construction (parikalpita) , are not the nature of a buddha.
Buddhahood involves no such conceptual tonstruction.
26
We are also reminded of the PP siltras when MSA 9.79 says
about enlightenment: "Those who see no attainment have the
supreme attainment."27 Sthiramati comments: "At the buddha
stage there is the highest attainment. That is the not seeing of
the attainment of a sambhogikakaya, a nairmar.tikakaya, the ten
powers, the four fearlessnesses, i.e., the not seeing of any of the
[buddha] dharmas. Why is that? Because it is the supreme attain-
ment, the highest of all dharmas, the dharmakaya."28 The passages
which precede and follow this make clear that buddhahood,
although associated with a collection of undefiled dharmas and
form bodies, is not to be identified with them. It is identified
with the dharmakaya, explained here as non-conceptual gnosis
(nirvikalpajiiana). In other places, it is explained as purified such
7
.
ness (vimalatathata). 29
We saw above that the Abhidharmakosa, in one verse, used
the term "dharmakaya" in a special sense, to designate the state
of buddhahood in its entirety. The term also carries this sense
in Yogacara texts, notably the Msg and the commentaries on
the MSA, where the dharmakaya is identified with a buddha's
asrayaparavrtti (the transformation of the basis).30 The concept
of "asrayaparavrtti" in Y ogacara texts is a model for full enlighten-
ment in which the basis of ordinary existence is transformed
into the full enlightenment of a buddha, through a process of
yogic realization. Different models of asrayaparavrtti compete
CONTROVERSY OVER DHARMAKAYA 55
PVith each other in early Y ogacara. Different models for the basis
samaliitathata, sarrtklesabhiigaparatantrasvabhiiva, etc.)
said to be transformed through yogic practice into different
fbtreSponding models of enlightenment (dharmakaya, nirvikalpa-
and nirmalatathata, dharmadhiituviSuddhi, etc.).31 But at the
of the literature at which the three kiiyas appear, all such
are considered equivalent to each other. When the dhar-
f!taMya is identified as a buddha's iiSrayaparavrtti in the MSA, Msg
related texts, it refers to the yogic attainment of full en-
as a In this usage it carries
i!Jte same n:eamng as the term phala-dharmakaya" of the AA
.
Now, when the earliest siistras known to teach three kayas
MSA, Msg, RGV, AA) list them, the first is called "svabhavi-
'lJ),kiiya." MSA 9.60 bhii$ya and Msg 10.1, lO.3 present what may
ibe the earliest Y ogacara definition of svabhiivikakaya. They define
being the dharmakaya, whose character is iiSrayaparavrtti. In
ljilier words, they equate svabhavikakaya with dharmakaya in its
of buddhahood as a whole. But why, one might ask, do
another term for all w.e already have
many of these terms. The answer IS that there IS buddhahood
actually exists, i.e., as a buddha has realized it (svabhavi-
t/iakiiya); buddhahood as arya bodhisattvas perceive it (sambhogika-
and buddhahood as others perceive it (nairma7),ikakaya).
t!buddha has achieved only one buddhahood, the dharmakaya.
II'hat kiiya as it actually exists, as it is in its own nature (svabha- .
'f&JkalJ,) is the "own-nature body," svabhiivikakiiyaly,; as experienced
iirya bodhisattvas, causing them to enjoy the dharma (sambho-
rgikab,) , is sambhogikakaya; and as experienced by others in its
reinanated forms (nairma7),ikaly,), is nairma7),ikakaya. This is the ex-
Ilanation of the three kaya names given in MSA 9.60 bhii$ya.
32
is consistent with the way the terms are used throughout
literature.
33
The first kaya is the real one. It is what a
tbuddha actually is, formless, and known only to a buddha. The
toiher two kayas are how that kiiya manifests in physical forms
1jio the unenlightened.
34
t Because within the early three kaya theory the svabhiivikakaya
;;is understood to be buddhahood, i.e., to be the dharmakaya, as
actually exists, the early commentaries began to substitute the
fterm "dharmakaya" for the term "svabhiivikakaya" in the list of
;>,',>,
;4-:.
(f'e
::;-,'
56 ]IABSVOL.12NO.2
three kayas. In later literature, it gradually becomes the norm
to name the three kayas: "dharmakaya, sambhogakaya, nirmii'I'Ja_
the term possessed two related
meanmgs m Yogacara lIterature: dharmakaya meamng full en-
lightenment as a whole, and dharmakaya meaning the first o'r
three kayas. And these two meanings were mediated by the term:
"svabhiivikakaya. "
Some of the points made here are particularly relevant to
the AA, and should be summarized: 1. In early Yogacara siistras
although buddhahood is conceptually understood to possess
undefiled dharmas, it is not to be identified with them. It is more
properly identified as being the culmination of a process ()f
yogic realization which goes beyond conceptual construction
understood as asrayaparavrtti, and referred to as "dharmakiiya';
(also referred to as "dharmadhatuvisuddhi," "anasravadhatu," etc.).
2. The first of the three kayas, svabhiivikakaya, is identified as
being the dharmakaya, buddhahood, as it actually exists; as itis
known only to a buddha. 3. It is therefore typical in
literature to use the word "dharmakaya" with two closely related
meanings: dharmakaya meaning buddhahood as a whole, and
dharmakaya meaning the first of three buddha kayas.
VI. Abhisamayalarp.kara Chapter 8's Relation to the Large Prajiia-
paramita Sutra
We must now look at AA chapter 8's relation to the PP
sutras. It is the Large PP sutra, especially in its 25,OOOsloka version.,
which served as the textual basis for the AA.36 Near the end of
one version of this sutra, the version referred to in modern
scholarship as the "revised Paiicavir(tsatisahasrika Prajiiapara-
mita,"37 there is a section which centers on the state of buddha-
hood, describing it in terms of more than two buddha kayas.
Some important late Indian scholars, and all Tibetan scholars
I am aware of, quoted this section and understood it to be the
primary textual basis for the AA's teaching on the buddha kiiyas
(AA 8 verses 1-33).38 Its passages are numbered "VIlLI,':
"VIII.2," and "VIII.3" in Conze's editions of the sutra. In
translation they read as follows
39
:
CONTROVERSY OVER DHARMAKAYA
VIII.l svabMvika/:t kaya/:t
Again, Subhuti, of those all-dharmas, which are like a dream,
which are nonentities, which have nonexistence for own-
being, which are empty of own-hlarks, which are perfectly
pure through the knowledge of all modes, which are unde-
filed, the essential original nature, which has one mark only,
i.e. no mark, should be known as the Tathagata, the Arhat,
the fully Enlightened One. It is thus that the bodhisattva,
the great being, should train in perfect wisdom.
rJnanatmako dharmakaya/:t]
Subhuti: What again, 0 Lord, are those undefiled all-dharmas?
The Lord: The 37 wings of enlightenment, the holy unlim-
ited, the eight emancipations, ... the four perfect purities,
the ten perfections, the ten powers, the four grounds of
self-confidence, the three ways in which (the Tathagata)
has nothing to hide, the threefold mindful equanimity, the
nature which is never bewildered, the knowledge of all
modes, the knowledge of the paths, all-knowledge-these,
Subhuti, are the undefiled all-dharmas. It is thus, Subhuti,
that the bodhisattva, the great being, should train in the
perfection of wisdom.
VIII.2 sambhogika/:t kaya/:t
Moreover, Subhuti, when he has trained in perfect wisdom,
when by the full attainment of just these dharmas he has
known full enlightenment, his body always and everywhere
adorned with the 32 marks of the superman and his 80
accessory characteristics, the Tathagata, the Arhat, the fully
Enlightened One, demonstrates to the bodhisattvas, the great
beings, the supreme dharma of the Mahayana which brings
them unsurpassed delight and joy, happiness and ease. It
is thus that the bodhisattva, the great being, should train in
perfect wisdom.
VIII.3 nairma1'!ika/:t kaya/:t
Moreover, having trained in perfect wisdom, having,
through the full attainment of just these dharmas, known
full enlightenment, the Tathagata, Arhat, the fully en-
lightened Buddha, in the ten directions, in endless and
boundless world systems, during the whole of time, works
the weal of all beings by means of a multiform cloud of
transformation bodies. It is thus that the bodhisattva, the
great being, should train in perfect wisdom.
57
Late Indian scholars (at least from the time of Ratnakara-
.58 JIABS VOL. 12 NO.2
santi) and Tibetan scholars up to the present day have aSSUIlle;!,;;"
that AA 8 was commenting directly on these siltra passages. For!!]
that reason they all understood it to be a straightforward
tion of the multiple kayas (either three or four) as they were)
taught directly in the PP siltras.
4
0)/i
The bad news, for those who have relied on Tibetan corre!
mentaries for their understanding of the AA, is that this section
1
of the siltra did not exist at the time theAA was written.
41
is very strong evidence that passages VlII.1-VIII.3 were an'!
interpolation, added to the PP siltra long after the AA's
tion. This means that AA 8 was commenting not on this section;
but on a different section of the siltra, a section which, wheif
properly identified, can give us a better. picture of the
w'
What is the evidence of a late interpolation? Firstly,
were three Chinese translations of the entire
kii PP siltra: (291 C.E.), Kumarajlva's (403 CE.), anall
Hsiian tsang's (659-663 C.E.). Passages VIII.1-VIII.3 are
found in any of them.42 To my knowledge they are not
in any Chinese translation of any Large PP siltra. This
that they were probably a late addition to the 25,000 PP
an addition not known to Chinese translators up to the
century. Secondly, passages VIII.1-VIII.3 are not found in
editions of the unrevised Large PP siltra extant in Sanskrit
Tibetan, including the 100,000, 25,000 and 18,000 sloka
sions. They are only found in one special version of the
PP, the revised edition, found in the Tibetan canon but nevef)
translated into Chinese.
43
Thirdly, and most importantly,
Vimuktisena (early 6th century), who wrote the first AA
mentary extant, tells us that the PP siltra of his time did no.t]
contain the passages in
Within the eight chapters of the AA, there are seventy topics;:l
!he last four topics are the subject of chapter 8. According
Arya Vimuktisena, they are: sviibhiivikakiiya, siimbhogikakaya!:
nairmii1J,ikakiiya, and nairmii1J,ikakiiyasya karma (the
kiiya's activity in the world).44 The 'Primary purpose of
Vimuktisena's commentary is to align each topic of the AA.
its corresponding passage in the 25,000 PP siltra. He does
by identifying the AA topic and then quoting or
corresponding siltra passage. We know when he is quotiIlS
:,J,ll
CONTROVERSY OVER DHARM.AK..AYA
59
r.'fparaphrasing the sutra, rather than giving his own explana-
by his use of one or more of the standard. expressions
frllployed in Sanskrit to signal quotations: "yad aha . .. iti" ("as
{the siitra] said"), "yathii" ("as [said in sutra]"), "sutre" ("[as] in
We sutra"), or "iti" (indicating a direct quote).45 For all 66 topics
!{the AA's first seven chapters, he invariably follows this proce-
ytre and methodically marks his references to the sutra. It is
f{gnificax:t, then, that he sud?enly stops quotin.? when
:lli:tfoducmg the first two topICS of chapter 8, svabhavzkakaya and
fainbhogikakiiya. There is no mark of reference to the sutra by
or paraphrase. He just presents his own explanations.
fhen, upon introducing the third topic and fourth topics,
'ftdirmiir;,ikakiiya and karma (activity), he resumes quoting the PP
ffttra. However, his quotes are drawn not from passages VIII.1-
Win.3 presented above, but from the passages in the sutra which
follow them (VIII.4-VIII.5 in Conze's numbering
_
If: What does this mean? At the point where Arya Vimuktisena .
explanations of the sViibhiivikakiiya and siimbho-
he tells us. He says: "As for the teaching of these two
[iJiyas] , they are taught in the section of the [PP] sutra which
fe'aches the nairmiir;,ikakiiya's activity, [in the section on] the
(flieans of collecting disciples which is the giving of supramun-
'llihe they were not taught earlier."46 He is
that t e PP sutra does not contain any distinct sections
iiya and siimbhogikakiiya. He finds their textual basis
/ h h fi d . .c
the same place were ems a textual basIs lor the
ftiirmiir;,ikakiiya's activity. And that is in a later portion of the
(VIII.5), quite different from the passages we quoted above
[XIII.I-VIII.3) of which he was completely unaware.
47
This
means that VIII.1-VIII.3 were added to the PP sutra
Ome time after Arya Vimuktisena, which was obviously a signif-
time after the AA was composed.. .
Where did these interpolated sutra passages come from? If
Vimuktisena's own introductory remarks on each of the
kiiyas are compared to PP passages VIII.I-VIII.3, it is
clear that these passages were composed and inserted into
sutra using Arya Vimuktisena's re:narks as their basis. For
reasons given above, we know that Arya Vimuktisena's intro-
comments on sViibhiivikakiiya and siimbhogikakiiya are his
..
,.
____
60 JIABS VOL. 12 NO.2
own, and are not paraphrases of PP VIlLI-VIII.3. We know:1J
in fact, that he had never heard of those passages. He also
prior
Its textual basIs III PP VIII.4. Arya VUl1uktIsena's own
tory remarks on sViibhiivikakiiya, siimbhogikakiiya, and
kiiya are very' similar in wording to PP VIII. 1, VIIL2, and VIII.3
i
j
respectively. Late Indian and Tibetan scholars, seeing the
ness between Arya Vimuktisena's remarks and PP VIII.l_!?
VIII.3, naturally assumed that he was paraphrasing the
But in fact the reverse was true. PP VIII.I-VIII.3 were inserted;
into the siltra as a paraphrase of Arya Vimuktisena!48>&
A careful reading of Arya Vimuktisena's commentary, then";:;
tells us three things: 1. Contrary to what late Indian and Tibetan"
traditions believed, AA 8 was not based on PP passages VIII.l1
5
VIII.3. 2. These Yf passages were a late interpolation. They:
were written taking Arya Vimuktisena's AA 8 commentary as theifi
. basis. 3. AA 8 was probably based on the section of the 25,000
PP sidra identified by Arya Vimuktisena, consisting of
VIII.4 and VIIL5 (which immediately follow the interpolateg'
passages VIILI-VIII.3 in the revised
What do PP passages VIII.4 and VIII.5 teach, upon which;;
AA 8 was actually based? Surprisingiy, they do not center ort"
buddhahood, not even mentioning the buddha kiiyas or a
activity. Instead, their teaching concerns the four ways in which'
bodhisattvas gather disciples (catviiri sa1{Lgraha vastilni). The first
of these ways is the giving of gifts, which includes the giving o(
material gifts and the gift of dharma. Within the gift of dharma.
are all the practices and realizations of Buddhists and non-Bud-
dhists, all the dharmas of the three vehicles, including the achieve:'
ment of the undefiled buddha dharmas, 32 marks and 80 signs;
etc. 50 Thus, the qualities of buddhahood are not even the focus'
of the passage. They are merely included within 4 large
tory of realizations imparted by bodhisattvas. N umerous
to help beings are also mentioned in the passage. But they are
carried out not by a buddha, but by bodhisattvas, the buddha merely
observing them.
51
Consistent with much of the rest of the PP
siltra, the passage focuses on the activities of bodhisattvas, wh()
work for living beings by engaging in the practice of prajiiiipara;.
mitii conjoined with skill in means. Its mention of buddha
is ancillary. . .
CONTROVERSY OVER DHARMAKAYA
61
If AA chapter 8 were based on the P P passages quoted
we might follow Tibetan scholars in concluding that AA
8 whatever number of kayas those passages
But It was not based on them. It was based on passages
f1which the same sort of list of buddha dharmas which
to be found scattered throughout the P P sutras. In explicat-
iiirig them, AA. 8 was just explicating the PP sutras'
ofreferrmg to buddhahood. But at the same tIme, WIthout
ft"rty clear basis in the PP sutra, it used the specific terms "svabhii-
'iiJikakaya," "sambhogikakaya," and "nairma1'}ikakaya," which it
;'plainly drew from Yogacara sources.
52
it
:YIlI. Conclusion
The author of AA, then, by explicating the PP's lists of
ilfiud,dha qualities, was explaining the way the P P sutras generally
lireferred to buddhahood. And, at the same time, he was relating
to the way Yogacara texts generally talked about buddha- .
What he sought to explain was not just the meaning of
short PP passages, but the relationship between the differ-
ways buddhahood was generally described in two of the main
textual traditions of his time, the PP sutras and the
sastras.
r:t Up to his time, nobody had explicitly related the PP's bud-
Iridhology to the increasingly popular Yogacara descriptions ..
the PP and the Yogacara talking about the same state of .
i;enlightenment, or not? Surely the author of the AA would want
()o say that they were. But this would mean that what the PP
{'referred to in terms of "undefiled dharmas," "marks and signs,"
etc., must be the same thing that the Yogacaras
cireferred to in terms of "the three kayas". The obvious question
;would then be: how do the two descriptions correspond? Which
'items in the PP descriptions correspond to each of the three
')kayas of Y ogacara? This is the question the author of the AA
have wanted to address. And this would mean that the
;AA was indeed a three-kaya text, mapping Yogacara concepts
;:onto the PP sutra.
Given this background, let us pretend for a moment that
we were the AA's first commentator (with no commentaries to
62 JIABS VOL. 12 NO.2
refer to). We analyze its 8th chapter, cognizant of the Yoga.ca.r
and PP traditions which were prevalent at the time of its
position.
53
It is highly likely we wOllld interpret it as follows.
The first of the three kiiyas ofYogacaraois called both
kiiya" and "dharmakaya". This is to be eqllated with the
of the PP sutras. It is often designated in the PP by listing the
names of the llndefiled buddha dharmas, but it is not to be iden_
tified with them, since it is beyond such designations. The
sambhogikakaya of Y ogacara corresponds to the buddha in the PP
sutra who is said to possess the 32 marks and 80 signs. And since
the nairmii1Jikakiiya of Yogacara must have some correspondencf!
in the PP, the limitless forms emanated by bodhisattvas in
PP (section VIII.5) will have to be understood as emanations
of the buddha himself, their activity then, being his activity. As
for the title of the chapter, it is also called "dharmakaya," where
the term carries its second Yogacaran sense, meaning buddha-
hood as a whole. We should not be surprised, then, that this is
precisely the interpretation of AA 8 which was made by its first
great commentator, Arya Vimuktisena.
54
And it continued to
be the standard interpretation for several hundred years after
him.55
We conclude, then, that AA 8 was not newly presenting a
theory of four kayas, as many have claimed, but was instead
performing a task which was far more pressing at the time it
was written: to show, for the first time, the relation between PP
descriptions of enlightenment and Yogacara descriptions.
cording to this theory, the AA is teaching three kayas. But it does
so idiosyncratically, because rather than explaining them within
a strict Yogacara context (as in MSA, Msg, etc.), it tries to show
how they are tacitly expressed in PP passages which make no
explicit mention of them. Here, I have formulated this theory
based upon AA chapter 8, its textual antecedents, its first com-
mentators, and historical considerations. There is much more
evidence to support it in the Sanskrit of the AA and in other
texts of its period, but that will be the subject of another paper.
56
NOTES
1. Haribhadra (c. 770-810 C.E.), to my knowledge, was the first to
ascribe the AA to Maitreya. He did so in his Aloka and Sphutartha (Amano, p. 2).
CONTROVERSY OVER DHARNI.AK:AYA 63
that AsaIiga and Vasubandhu wrote commentaries on the AA,
these have m:verbeen found. If true, the AA was composed by the
C.E. The first commentary extant in any language is Arya Vimukti-
(c. early 6th century). If this was the first commentary, it would put
's terminus ad quem in the 5th century.
2. The last verse of the AA names its final topic: "dharmakayaphalam,"
dharmakaya," meaning the state of buddhahood. In Arya Vi-
AA Vrtti (Peking 5185, fol. 100-3-7) the AA's final chapter is
;'chos kyi sku'i skabs bslab pa'i 'bras bu'i leu," "The Dharmakiiya Section,
rh.,ntpr on the Result of the Trainings" ('''bras bu" = ''jJhalam''). In Hari-
Sphutartha (Amano, p. 262) the AA's final chapter is called "Dharma-
"Complete Enlightenment: the Dharmakaya."
In Indian commentaries on theAA (those by Arya Vimuktisena, Bhadanta
Dharmamitra, Ratnakarasanti, Abhayakaragupta), the word
in "dharmakiiya" is etymologized in one or more of three ways: kaya =
support, basis ("dharmakaya" = the support of all excellent qualities,
; kiiya = sarira: body ("dharmakaya" = body of dharmata); or kaya =
collection or accumulation ("dharmakaya" = collection of excellent
dharmas). The term "kaya" in "rupakiiya," in both pre-Mahayana and
' .. "lva.uu texts, has generally meant "sar'ira," "body."
3. e.g. Conze, PP Lit., p. 103; Dutt,MahiiyanaBuddhism, p. 155; Poussin,
. pp. 790-791; Obermiller, Analysis of the Abhisamayiilamkara, pp. 11-12.
4. As far as we know, Haribhadra was the first to claim that the AA
four kayas.
5. The statements of this paragraph were made for the purpose of
i!i,'hT'OV1,OlI' 111: the broader context into which the subject matter of this paper fits.
aware that these statements require a great deal of supporting eviden<::e, .
purpose of this paper is to begin providing that evidence, starting first
'an analysis of the AA's textual antecedents. Evidence for: my description
of the controversy's other historical stages will be taken up in future
6. One point should be made at the outset. Although the participants
debate always read their views into the AA, if our analysis finally decides
the AA taught one kaya theory rather than the other, it does not comprise
i!3.refutatjion of the other theory. It only establishes which theory the AA taught.
apart from the AA, both the three-kaya and four-kaya theories of en-
Il1)111lC:11lJIlClllin non-tantric Mahayana Buddhism are quite supportable within
tradition; based on other scriptures and on reason. '
7. Abhisainayiilamkaravrtti, Pk 5185, Vol. 88, fols. 92-4-6-92-5-7.
for the first chapter of this text has been edited by C. Pensa. Up to
present time, the rest of the text is available only in the Tibetan canon.
8. Sphutiirtha, Amano, pp: 268-270. Aloka, Wogihara, pp. 914-916.
9. 4.32. '
10. 34.7. Po us sin, L'Abhidharmakosa, ch. 6, p. 76.
11. 4.32, 7.28.
12. Ibid. 7.28.
13. Ibid. 7.28: araT}iisamiidhi, prar;.idhijiiii1ia, the four pratisamvid, the six
64 JIABS VOL. 12 NO.2
abhijitas, the four dhyanas, the four arilpyasamapattis, the four aprama7].as, th';
eight eight abhibhvayatanas, and the ten krtsnayatanas. The 37
are described in detail in chapter 6. h
14. Ibid. 7.34.
15.' For the list of saivadharmal;" see Conze, Mahaprajiiaparamita Sutra
fols. P 165-169, section I.5; Conze, Large Sutra, pp. 120-123.),;
16, e.g, Conze, The Gilgit Manuscript of the
prajiiaparamita, fo1. 276b, p. 35. Paiicavim: Conze, Mahaprajiiaparamitii, fol.:
P.524:'
"tad bodhisattvo mahiisattval;, prajiiaparamitayaT{! caran dvayo sunyatayo sthitva at yaH;:
ntaSunyatayaT{! anavaragraSunyatayaT{! ca sattvanaT{! dharmam desayati / sunyaT{!
dhatukam eta[nJ niistyatra rilpaT{! va vedanii va samjiia va samskarii va vijiiiinaiT/,J
va skandha va dhatavo va ayataniini va api tu khalu sarva ete dharma avastukii',;
abhavasvabMviis . .. " "Thus the Bodhisattva, Mahasattva, engaging in the
fection of insight, having stood in the two emptinesses: the boundless empti_i1
ness and the emptiness without beginning or end, teaches the dharma fcir'
living beings. [He tells them:] 'Everything in the three realms is empty. Here
there is no form, feeling, perception, mental formations, or consciousness;;'
There are no skandhas, no elements, no sense fields .... Rather, all these'
dharmas are unreal. Their self-existence is non-existent .... '" Note here atil
in the notes which follow that where I do not name the translator into English;;
the translation is mine. .;;,
17. By the word "essence" here, I mean the most important quality
qualities of a thing, those qualities without which, it would cease to be that:
thing: The discussion here concerns the PP, which denies the self-existence"
of all phenomena. Therefore, when I say "essence" I do not mean any kina'
of self-existent or independent nature. I mean that which makes a buddha
buddha,:;
18. e.g. PaitcaviT{tSati: Conze, Mahaprajiiaparamita, fo!' P 78: d
"yena samanviigato bodhisattvo mahiisattvo na kaT{!cid dharmti,i;t:
prajanati saT{!Skritam vii asaT{!skritam va kusalam va akufalam va savadyamvii'
anavadyam va siisravam va aniisravam vii samklesam va niJ;,klesam va laukikam va,
lokottaram va va vyavadanam va. yena bodhisattvena
sattvena Weid dharmo na na sruto na mato na vijiiatal;,. idam bodhisattvasya
Z
mahiisattvasya pariSuddham prajiia Translated in Conze, Large Sutra;'
p. 77: "A bodhisattva who is endowed with that wisdom eye does not know any
dharma-be it conditioned or unconditioned, wholesome or unwholesome;
faulty or faultless, with or without outflows, defiled or undefiled, worldly or
supramundane. With that wisdom eye he does not see any dharma, or hear;'
know, or discern one. This is the perfectly pure wisdom eye of a bodhisattva. ,,'
19. fo1. 94:
"saeet kaufika ayaT{! te jambudv'ipal;, paripufIJ-aS cUijika baddhas tathagataior
r'ira7].aT{! d'iyeta/ iyaT{! ca prajitaparamita likhitvopanamyeta/ tata ekatere7].a bhagena,
pravaryamano 'nayor dvayor bhagayol;, sthapitayol;, katamaT{! tvaT{!' kaufika bhagam;
grhr}'iyal;,// sakra aha/ saeen me bhagavann ayaT{! jambudv'ipal;, paripuT7].aS e#iM;'
baddhas tathagataSar'ira7].am diyeta/ iyam ea prajiiaparamita likhitvopanamyeta/
ekatarer}a bhagena pravaryamano 'nayor dvayor bhagayol;, sthapitayor imam evaha1jl,
bhagavan prajiiaparamitaT{! parigrh7].'iyaT{!// tat kasya hetoJ;,! yatha 'Pi nama tathag
atd
CONTROVERSY OVER DHARMAKAYA 65
,C 'etff citrfkiire1Ja I etad dhi tathagatanaT{! bhiltarthikaT{! sarfram /1 tat kasya heto!t/
likltam hy etad bhagavata dharmakiiya buddha bhagavanta!t/ ma khalu punar imaT{!
:ihikiava{L satkiiyaT{! kayaT{! dharmakayaparini:;pattito maT{!
'drak!yathai/a ca bhutakofl prabMvito yad prajiiiiparamltai/"
,'iiBhagavan: If, Kauslka, on the one hand you were gIven thIs world filled up
: . the top with relics of the tatMgatas; and if, on the other hand, you could
't
6are
in a written copy of this perfection of wisdom; and if now you had to
between the two, which one would you take? Sakra: I would take just
this perfection of wisdom rprajnaparamita]. Because of my respect for [it as]
the guide of the tatMgatas. Because in actuality it is the body of the tatMgatas.
'As the Bhagavan has said: 'The dharmakayas are the buddhas, the bhagavans.
Bilt, monks, you should not think that this [physical] body is my true body.
Monks, you will see me from the perfection of the dharmakaya. And this
"tdthiigatakiiya should be seen as brought about by the true limit, i.e. by the
:. perfection of ,,:isdom.'"
.," PP, fols. 512-514:
"evam ukte dharmodgato bodhisattva maMsattva{L sadapraruditaT{! bodhisattvaT{!
mahasattvaT{! etad avocatl na khalu kulaputra tatMgata{L kutascid agacchanti va
"'gOa;hanti val acalita h'i tathata ya ca tathata sa tatMgata!t/ na hi kula putranutpada
agl1cchati va gacchati va yai canutpada{L sa tatMgata!t/ na hi kulaputra bhutako.tya
<agamanaT{! va gamanaT{!va prajnayate ya ca bhutako.fi{L sa tatMgata{L I na hi kulaputra
Siinyataya agamanam va gamanam va prajniiyate yii ca sunyata sa tatMgata{L I . .. na
hi kulaputriinyatrebhyo dharmebhyas tatMgata{L yo' ca eva dharmii1JaT{!
lathata yo, ca sarvadharmatathatii yii ca tatMgatatathata tathatiil nasti kula-
'putra tathataya dvaidhZkiira{L ekaivai;;ii tathata kulaputra tathatii na dve na tism
gOr;anavyativrta kulaputra tathatii yad utasattviitl ... dharmodgata ahal evam etat
kulaputraivam etatl evam eva kulaputra ye kecit tatMgataruperJa va va
te tatMgatasy' agamanartl ca gamanaT{! cakalpayantil ... sarve te biila-
1fJtiya iti vaktavya!t/ tadyatM 'Pi nama sa eva yo 'nudake
Udakasamjnam utpadayatil tat kasya heto!t/ na hi tatMgato rupakiiyato
tdharmakayas tatMgata!t/ na ca kulaputra dharmata agacchati va gacchati val evam
eva kulaputra nasti tatMgatanam agamanaT{! va gamanaT{! val . .. kulaputra sar-
vadharma{L svapnopamii ukta Bhagavata I ye kecit kulaputra svapnopamiin sarva
dharmar{lS tatMgatena nirdeiitan yatMbhutaT{! na prajananti te tatMgatan nama-
kiiyena va rupakayena va abhiniviSya tathiigatanam agamanaT{! vii gamanar[l va kalpa-
yantil yatM 'Pi nama dharmatam aprajananto ye ca tatMgatanaT{! iigamanaT{! va
gamanaT{! va kalpayanti sarve te biilajatfya{L prthagjana};, . . . I ye khalu puna};, kula-
putra svapnopamiin sarvadharman svapnopama};, sarvadharma iti tathagatena desitan
yathiibhutar[l prajananti na te kasyacid dharmasy' agamanaT{! va gamanaT{! va kalpa-
yanti ... te dharmataya tathagatar[l prajananti I . . . ye ca tathagatasyedrSir[l dharma-
tii1[! prajananti te asanna anuttaraya};, samyaksambodhes caranti te ca prajnapara-
initiiyar[l caranti/" "Dharmodgata: Son of the family, tathagatas certainly do not
:tome from anywhere, nor do they go anywhere. For, indeed, such ness (tathata)
is unmoving, and the Tathagata is suchness. Nor, son of the family, does
non-arising come or go; the Tathagata is non-arising. Nor is a coming or
?,oing of the true limit (bhutakotiM known; the Tathagata is the true limit. Nor
1sa coming of emptiness (Sunyata) known; the Tathagata is emptiness ... Nor,
son of the family, is the Tathagata other than the dharmas, for that which is
66 JIABS VOL. 12 NO.2
the suchness of these dharmas, that which is the suchness of all dharmas th
which is the suchness of the Tathagata, is just this one suchness. For suchn
has no division. This suchness is just one, son of the family. Suchness is :8,8
two, not three. Suchness is beYQnd enumeration because it is not a bei:
t
(asattvat). .g
[Dharmodgata gives a metaphor of a foolish man who mistakes a mira':
of water for actual water. He asks Sadaprarudita whether the
has come from anywhere or goes anywhere. Sadaprarudita replies that, sine!"
there is no water in the mirage, there is no coming or going of water, and
the man who believes there is water in the mirage is foolish.]
Dharmodgata: Injust the same way, son of the family, those who
adhered to the Tathagata through his form or his voice imagine a coming or
going of the Tathagata. They are to be called foolish and stupid, just like the
person who perceives water where there is no water. [This is] because the
Tathagata is not to be seen from his rupakiiya. The dharmakayas are the tatha_
gatas, and the real nature of things [dharmata] does not come or go. Likewise
there is no coming or going of the Tathagata .... The Bhagavan has said
that all dharmas are like a dream. And those who do not know all dharmasto
be like a dream as explained by the Tathagata, they adhere to the tathagatas
through [their] name body [namakaya] or form body [rupakaya] and imagine
there is a coming or going of the tathagatas . ... But those who know all dhaT1l!l.ls
to be like a dream as they really are, as explained by the Tathagata, they do.
not imagine a coming or going of any dharma, ... they know the Tathagata
by means of the real nature [dharmata] . ... Those who know such a real nature
[dharmata] of the Tathagata, they practice close to full enlightenment; they
practice the perfection of wisdom rprajiiaparamita]." (Portions of this passage
are very close to Vajracchedika PP, vs. 26).
Paiicavimsati: Conze, Mahiiprajiiaparamita, fo1. P 485b:
"subhutir aha bodhir ityucyate kasyaitad adhivacanaml bhagavan aha: bodhir
iti subhute sunyataya etad adhivacanarrt tathataya etad adhivacanam bhUtakoter etad
adhivacanam dharmadhiitor etad adhivacanam. . . . I api tu khalu subhute buddhanai[!
bhagavatarrt bodhis tasmad bodhir ityucyatel api tu khalu subhute buddhair
bhagavadbhir abhisambuddhas tasmad bodhir ityucyate I" "Subhuti: 'Enlighten-
ment' is spoken of, Bhagavan. For what is that a designation? The Bhagavan:
'Enlightenment' is a designation for emptiness. It is a designation for thusness.
It is a designation for the true limit. It is a designation for the dharma-realm .
. . . Moreover, Subhuti, because the buddhas, the bhagavans, have this enlighten-
ment, it is called 'enlightenment.' Moreover, Subhuti, because it is realized by
the buddhas, the bhagavans, it is called 'enlightenment.'''
Note in these passages how enlightenment (bodhi, dharmakaya, tathiigata-.
kaya, etc.) is associated both with thusness, tathata, and with the gnosis that
realizes it, prajiiaparamita.
The bases for this analysis are the descriptions of dharmakaya in versions
of the 8000, 18,000 and 25,000 PP sutras avC'jlable in Sanskrit and Tibetan,
and closely related descriptions inthe Vajracchedikii PP and the SaptaSatika
PP. According to Professor Lewis Lancaster's studies of the development of
the 8000 PP sutras in Chinese translations, the accounts of dharmakaya I have
CONTROVERSY OVER DHARMAKAYA
67
en probably belong to the middle and late stages of the 8000 PP, whose
v. st Chinese translations were made in the early 5th and mid-7th centuries
linson, pp. 16, 30t. The Vajracchedika PP was translated into Chinese at
:beginning of the 5th centUry (Conze, PP Literature, p. 60). Obviously,
se pp sutras had a significant period of development in India prior to their
':slation in China. A very similar account of dharmakaya vs. rupakiiyais found
the Samadhirajasutra, whose terminus ad quem has been put in the 4th
!tury (Regamay, Three Chapters from Samadhirajasutra, pp. 11-12. But see
open, "Notes on the Cult of the Book," pp. 153 ff. and "Sukhavatl,"
204 where he notes that available evidence has pushed back speculative
mations of the dates of the 8000 P P, Vajracchedika PP, and Samadhirajasutra,
: ng the latter two to perhaps the 2nd century C.E.). It is widely held among
ern scholars, including Lancaster ("The Oldest Mahayana Sutra," p. 36),
the two-kaya theory found in the P P sutras was a forerunner of the
adira three-kaya terminology,- a terminology which appears in the AA
-bhavikakaya" "sambhogikakaya" "nairma1Jikakaya"). It is likely, therefore, that
PP conceptions of dharmakaya discussed above developed prior to the
'iod of composition of the AA (ca 4th to 5th century), although expressions
them continued to be added to the PP sutras throughout the following
turies.
. One point should be made parenthetically. Lancaster identified one men-
n_of the word "dharmakiiya" ih a passage of the 8000 PP which seems to
-:ry the meaning "collection of dharma texts," rather then the meanings
entified here: thus ness and prajiiaparamita. Lancaster believes that this pas-
ke is part of the earliest stage of development of the 8000 PP text ("The
Idest Mahayana Sutra," p. 36). What I am focusing on here are the meanings
the word "dharmakaya" in the middle and late texts which became especially
portant to the Yogacaras and, I believe, to the author of the AA.
'. 20. Sutras such as Sukhavatzvyuha, Vimala-
inirdesa, etc. In them, exalted tathagatas are described presiding over pure
hakJetras, e.g., Sakyamuni, of the PP sutras, Vairocana of the
aT{lSaka, Amitabha of the Sukhavatzvyuha, etc. And descriptions are given
buddhas and bodhisattvas emanating infinite arrays of forms to teach living
ngs at the times and places fitted to their qeeds.
'. 21. N. Dutt, Mahiiyana Buddhism, pp. 136-170; Hobogrin, article: "Bus-
in" by P. Demieville; La Vallee Poussin, La Siddhi, pp. 762-813, "Notes sur
Corps du Bouddha." .
I:,!' 22. This summarizes part of the description of these two kiiyas found in
and Asvabhava's :ommentaries on MSA 9.61 ;lOd in Vasubandhu's
Asvabhava's commentanes on Msg 10.30.
23. MAVbhii.yya 4.14. DhDhV, sDe dge phi, fo!' 47b4, 51b6. RGV, chapter
presents a three-kaya theory at some length. Because its focus is so squarely
the theory of tathagatagarbha, it stands apart somewhat from the other
!t:xts mentioned here. However, it relates its basic model of enlightenment,
'l!.!rmala tathata, to the theory of three kayiis in much the same way that the
tSA, Msg, and their commentaries relate dharmadhatuviSuddhi and nirmalata-
to the kayas. It quotes from MSA 9, and in one portion
f:'"
riG":'
1)-\,
68 JIABS VOL. 12 NO.2
of its second chapter it is clearly applying the MSA's buddhology to its theory
of tathagatagarbha. See RGV, Johnson pp. 85-88, Takasaki, p. 4l.
24. A bibliography of modern scholars' speculations on the history of
early Yogacara can be found in Nakamura, Indian Buddhism, p. 263. Summaries
are found in Ruegg, La Theorie, pp. 30-55; Davidson, "Buddhist Systems of
Transformation," pp. 14-49, 126-149. Davidson reexamines the questions of
authorship of all early Yogacara sastras, and concludes that the authorship of
the MAV, DhDh V, and AA is still unknown. I agree. _
25. Levi, MSA 9.4, p. 34.
26. Vrttibhii4ya, sDe-dge mi, fols. 108a2-108b4.
27. Levi, MSA 9.79, p, 48.
28. Vrttibhii4ya, sDe-dge mi, fols. 144a7-144b1.
29. Vrttibhii4ya, sDe-dge mi, fols. 144a2-144b7 (commenting on MSA vss.
79-81). On full enlightenment as vimalatathatii, see MSA 9.56-59 bhii4ya and
vrttibhii4Ya.
30. Msg 10.3 characterizes dharmakiiya as iiSrayapariivrtti, the complete
transformation of the basis which is full enlightenment nirvii7j.a,
Msg 9.1). The precedent for this is Samdhinirmocanasutra, 10.1-10.2, where
the Bhagavan tells Maiijusrl that the dharmakiiya of the tathiigatas is to be
identified with their iiSrayaparivrtti (on the etymologies and general semantic'
equivalence of -pariivrtti and -parivrtti in classical Y ogacara texts, see Davidson,
pp. 152-3). MSA 9.60 bhii4ya makes the same characterization. At 9.77 the
bhii4Ya closely relates dharmakiiya with the anasravadhiitu, a MSA model of full
enlightenment. Sthiramati's vrttibhii4Ya on MSA 9.60 and 66 identifies dharma-
kiiya directly with dharmadhiituviSuddhi, another Y ogacara model of full en-
lightenment.
31. Davidson, pp. 199-259 separates out several different models of
iiSrayapariivrtti in Yogacara. The important point here is that the Yogacara
understood its models of full enlightenment, including the three-kiiya model,
in terms of iiSrayapariivrtti, i.e., as the completion of a process of yogic realiza-
tion, not just an an object of logical analysis.
32. Levi, MSA bhii4ya, p. 45: "trividhaly, kiiyo buddhiiniiT(t/ sviibhiiviko
dharmakiiya asrayapariivrttila!?Ja7j.aly,/ siiT(tbhogiko yena dharma-
saT(tbhogam karotif nairmiiniko yena nirmiinena satviirtham karotif"
33. MSA 9.60-62, bhii4Ya and Msg7.11, 10, 1, 10.3 and
upanibhandhana: RGW ch. 2 preamble and vss. 2.38-2.61; Kiiyatrayastotra;
Kiiyatrayasutra; KiiyatrayiivatiiraSastra; the three kiiya chapter ("sku gsum mam
par 'byed pa") which appears in later editions of the Suvar7].aprabhasasutra.
34. This hearkens back to the PP sutras' opposition of dharmakiiya to
rupiikiiya which we saw earlier (the dharmakiiya being what the Buddha actually
is; the rupakiiya being what fools think he is). Typical in Yogacara literature,
is the description of the sviibhiivikakiiya as "pratyatmavedam," "known only to
himself' (to the Buddha), not to others (Kiiyatrayastotra vs. 1; Kiiyatrayasutra
Pk 949 Vol. 37, fo1. 108-3-2; RGV 2.42). MSA 9.60 describes the sViibhiivikakiiya
as "subtle" (su!?Jma), Sthiramati explaining this to mean that it is not a cognitive
object for sriivakas or pratyekabuddhas (vrttibhii4ya sDe dge mi, fols. 166b5-6).
35. MSA chapter 9 may well be the first presentation of three kiiyas in
CONTROVERSY OVER DHARMAKAYA
69
'yogacara literature. It labels the first 'of the three kayas :"svabhavikakaya" (not
::;dharmakaya"). The next earliest texts to teach three kayas are probably Msg
:chapter 10. (based on MSA 9), AA chapter 8, Ratnagotravibhaga chapter 2 (see
;Davidson, pp. 132-144 for recent speculations on the chronology of the "Mai-
Itreya" corpus), and the Buddhabhumisutra (the four verses near the end of the
!satra on dharmadhatuviSuddhi. The relative dating of the MSA and Buddhabhumi-
:Wtra is presently somewhat controversial, but that does not affect the argument
'here). Like MSA 9, all of these texts call the. first kaya "svabhavikaMya." Of all
,'the early Yogacarafastras to teach three kayas, only the Dharmadharmatavibhaga
'refers to the first as "dharmakaya." But this text mentions the three kayas only
{in passing, and obviously drew the theory from other sources. It is in the
commentaries and subcommentaries to these texts, ascribed to Vasubandhu,
. .I\.svabhava, Sthiramati, etc., that the term "dharmakaya" begins regularly replac-
ing the term "svabhavikakaya" in the list of three kayas. And this becomes the
.)lorm in later texts such as the Kayatrayavatarafastra, Kayatrayasutra, Kayatraya-
stotra, and Madhyamakavatara.
36. By "Large PP sutra" I mean the suira which Conze identified as
existing in three versions: 18,000, 25,000, and 100,000 flokas, all of which are
hi.rgely the same in content but differ in the extent to which they repeat the
:same PP formulas regarding the emptiness of all dharmas. Conze, PP Literature,
'pp. 10-11. There are three reasons for identifying the 25,000 floka version
as the basis of the AA. First, upon analysis, the 8000-floka sutra does not provide
an adequate textual basis for the last three and a half chapters of the AA,
ihile the Large PP sutra does (this will be detailed in my forthcoming disser-
tation). Thus, only the Large PP could have been the textual basis fm the AA
in its entirety. Secondly, as far as we know, the 25,000 PP was the first PP
sUlra identified by classical Indian scholars as the AA's textual basis. And it
'took more than two hundred years before any other PP sutra was so identified
(the 8000 PP by Haribhadra). Thirdly, more commentaries associate the AA
.with the 25,000-floka version of the PP than any other, including the three
: earliest AA commentaries.
37. Nancy R. Lethcoe, "Some Notes on the Relationship between the
,Abhisamayalamkara, the revised Paiicavimiatisahasrika, and the Chinese transla-
'tions of the unrevised Paitcavimfatisahasrika," JAOS, 96.4 (1976), 499. Conze
(PP Literature, p. 36): calls the revised version of the 25,000 sloka PP sutra the
"recast version of the Paiicavimiati. PP" (Pk #5188 in the Tibetan canon). I
,shall refer to it as the "revised 25,000 PP." It is a redaction of the 25,000 PP,
composed, I believe, after Arya Vimuktisena, which shows the correspondence
between the passages of the Large PP sutra and the topics of the AA. In it,
each portion of toe sutra is labelled with the name of the AA section for which
that portion of the sutra was thought to be the textual basis. The reason I
think it appeared after Arya Vimuktisena will become clear in what follows.
38. I am referring here to Ratnakarasanti's Suddhamati, Pk 5199,281-5-2
ff. and Saratama, Pk 5200, 92-4-4 ff., Uaini's Sanskrit edition, p. 172); Abha-
,yakaragupta's Marmakaumudi, Pk 5202, 198-5-6 ff. and Munimatala'T{Lkara, Pk
5299,232-1-3 ff.; Bu ston's Lung gi snye ma, vol. 2, p. 204 ff.; gYag ston's Rin
po che'i phreng ba blo gsal mgul rgyan, vol. 4, p. 382 ff.; Tsong kha pa's Legs
70 JIABS VOL. 12 1,ro. 2
bshad gser phreng, vol. 2,465-4 ff.; rGyal tshab's rNam bshad snying po'i rgya
p. 549 ff., Serarje btsun chos kyi rgyalmtshan's Chos sku spyi don, 14b3 to 15b7 n,
39. This passage is numbered VIlLI, VIII.2, and VIII.3 in Conze, La; e
Sutra, pp. 653-4 and Mahii-Prajiiiipiiramitii Sutra, fols. P523a8-P523b5. Whe;e
Conze translates "aniisrava" as "without outflows," I have substituted "unde_
filed" in order to keep the terminology of this paper consistent.
40. The titles: "sviibhiivikal; kiiyal;, " "siimbhogikal; kiiyal;," and "nairmii'IJ-ikah
kiiyal;" appear in the revised 25,000 PP as the titles of their respective p a s s a g ~ S
(Collze, Mahii-Prajiiiipiiramitii Sutra, fols. P523a8-523b5). Ratnakarasanti and
Abhayakaragupta quoted these passages as the PP textual basis for AA 8, and
as evidence that the AA teaches three buddha kiiyas (Siirattamii, Jaini, ed.,
p. 172; Marmakaumudi, Pk 5202, 198-5-&--199-1-1). The title ''jiiiiniitmako
dharmakiiyal;" labels the portion of the passage which Tibetan scholars believed
Haribhadra took as the textual basis for his description of the second bUddha
kiiya (the kiiya which consists of the collection of buddha's gnoses, ''jiiiiniitmako
dharmakiiyal;"). Haribhadra in his Aloka (W ogihara, pp. 914-916) and Spu{iirtha
(Amano, pp. 262-271) delineated the jiiiiniitmako dharmakiiyal;, and Tibetan
scholars identified Haribhadra's PP basis for it as it is labelled in the passage
above (Bu ston's Lung gi snye ma, Vol. 2, p. 206; rGyal tshab's rNam bshad
snying po'i rgyan, p. 551; Sera rjetsun pa's Chos sku phyi don, fol. 15a5); I have
put the title ''jiiiiniitmako dharmakiiyal;" in brackets because it does not actually
appear in the revised 25,000 PP, while the names of the other three kiiyas do.
41. E. Obermiller's groun?breaking study of the AA analyzed the AA by
referring to Haribhadra's AA - Aloka and AA - Sputiirthii and by relying heavily
on several majorTibetan AA commentaries (Analysis of the Abhisamayiila7[lkiira,
1933, pp. vii-viii). His report that AA 8 taught four buddhakiiyas was based on
these sources. But the Tibetan commentators upon whom he relied (Bu ston,
Tsong kha pa, rGyal tshab, '] am dbyang bshad pa) all identified revised 25,000
PP passages VIII.I-VIIL3 (quoted above) as the sutra basis for AA 8's buddha-
kiiya teaching (see note 38). A number of influential scholars since Obermiller
have followed his lead, based on similar sources, reporting simply that AA 8
teaches four kiiyas (see note 3). In order to arrive at a proper interpretation
of AA 8, it is important first to identify its actual textual basis in the PP sidra,
and then to see if this can shed light on its teaching of the buddhakiiyas. This
is what I will attempt to do in what follows. Because no modern scholar has
yet done this, there has been a tendency to repeat what scholars such as
Obermiller have said without realizing that the Tibetan sources upon which.
he relied had misidentified the PP sutra basis of AA 8, and that this has a
bearing on the interpretation of AA 8.
42. Lethcoe, op. cit., pp. 499-504.
43. For a description of extant PP sutras, see Conze, PP Literature, 31-74.
The revised 25,000 PP is extant in 18th and 19th century Nepalese Sanskrit
manuscripts and in the Tibetan canon (Pk 5188). Although included among
siistras in the bsTan 'gyur of the Tibetan canon, I am treating it here not as
a siistra but as a sutra. There are several reasons for this. Nancy Lethcoe, using
Chinese translations of the 25,000 PP sutra, has charted the development of
this sutra over a period of several centuries and has clearly shown that its
CONTROVERSY OVER DHARMA.KAYA 71
"sed version, extant only in Sanskritand Tibetan, lies within that continuum
It is a late version of the 25,000 PP sutra, revised by the
'. ertion of AA topic names, and less obviously, by occasional transpositions,
and deletions which bring the sutra more closely into line with the
.. 37). It gives app.earance of being just the 25,0?0
pp sutra dIstmgUIshed only Its the names of the AA topICS
'serted into It after the correspondmg sutra passages. The passages are the
dialogues between the Bhagavan, Subhuti, Sariputra, etc., without any
Y tervening exegesis or commentary whatsoever. At some point Indian com-
:entators on the AA began quoting this revised version of the 25,000 PP
when giving the sutra basis for the AA. Ratnakarasanti and Abhayakaragupta
quoted revised 25,000 PP passages VIII.I-VIII.3 as a sutra basis for AA 8 (see
Jlote 37). It may be that by the time of Ratnakarasanti (c. 1000 C.E.), and
perhaps som:what ea:lier, Indian_scholars found the revised. 25,000 PP the
most convement verSIOn of the sutra to use when commentmg on the AA,
since only this version of the PP had its passages marked with the AA topic
names' for ready reference. The Tibetan commentators then followed them
iri this. When Indian and Tibetan scholars quoted the revised 25,000 PP
(passages VIII.I-VIII.3), they referred to it as "mahiitz bhagavatz," or "sutre"
(Tib., "mdo las"), which means they were treating it as a sutra, not as a fiistra
j(see note 38).
. Since later Indian and Tibetan scholars quoted the revised 25,000 PP
ias sutra, why was it put into the bsTan 'gyUT section of the Tibetan Tripitaka
(the collection of fiistras, commentaries) rather than the bKa' 'gyur section (the
/wllection of sutras)? We can only surmise, but it would appear that because
the topic names of the AA had been inserted into the sutra, it could not be
considered simply the Buddha's word. After all, words of fiistra (the AA), even
if only topic titles, are not the word of the Buddha. Furthermore, in the
,Tibetan translation of the revised 25,000 PP, Haribhadra is identified as the
. compiler. Sutras are not supposed to have a compiler apart from the Buddha
'and those in dialogue with him. Some such considerations probably required
'that, for classification purposes, the revised 25,000 PP not be put into the
bKa' 'gyur.
44. AA - vrtti, Pk 5185, Vol. 88, pp. 92-100. At 92-4-6 ff. Arya Vimukti-
sena identifies the three kiiyas as the first three topics of AA 8. At 98-4-7 and
,98-5-1 to 98-5-3, he explicitly identifies the fourth topic of AA as "sprul pa'i
sku'i phrin las," "the activity of the nairmii1Jikakiiya."
45. Only the first chapter of Arya Vimuktisena's commentary is available
to me in Sanskrit (Pensa's critical edition of the AA - vrtti), but it can be used
to find the correlative Tibetan terms in the Tibetan translation of the rest of
the commentary (Pk 5185): "zhes gang gsungs pa yin no," "zhes bya ba la sogs pa
gang gsungs pa yin no," ''ji skad du," "mdo las," and "zhes bya ba."
For 51 of the 66 topics comprising the AA's first seven chapters, Arya
Vimuktisena makes clear that he is quoting sutm directly, rather than para-
phrasing, because he uses the vocative forms of the names of one or more
characters from the siltm ("Bhagavan," "Subhute," "Sariputra," etc.). But even
in the other 15 cases, he explicitly indicates he is either quoting or paraphrasing
72 JIABS VOL. 12 NO.2
by always using one or more of the markers mentioned above. Itis
only when we come to the first two of AA 8 that .no such Sanskrit
markers appear. He suddenly stops quotIng or paraphrasmg S1.ltra and JUSt
presents his own explanations. ,
46. Pk 5185, p. 98-4-6 to 98-4-7; sDe dge ka, fo1. 205b2-3. The Tibetan
reads: 'di gnyis kyi bshad pa ni sprul pa'i sku'i phrin las ston pa'i mdo las Jig rten
las 'das pa'i chos kyi sbyin pa'i bsdu ba'i dngos po nyid kyis ston par 'gyur te des na"
dang po ma gsungs so.
47. Arya Vimuktisena identifies the PP textual basis for all of AA 8
be the portions of the LargePP sYira which Conze numbers "VIII 4" and
"VIII 5" (Conze, Large Sidra, pp. 573-643; Maha-Prajiiaparamita Sutra, fol
s
.'
P523b6 to P594al). These portions are indeed found in all versions of the
Large PP sutra extant in Chinese, Sanskrit, and Tibetan.
48. Relevant portions 6f Arya Vimuktisena's own introductory remarks
on each of the three kiiyas in his AA - vrtti are as follows:
svabhavikakaya (commenting on AA 8 vss. 1-6): chos kyi dbyings dang
, ldan par gyur ba zag pa med pa'i chos thams cad kyi rnam pa thams cad du rnam par
dag pa'i rang bzhin te ngo bo nyid gang yin pa de ni bcos ma ma yin pa'i don
na bcom ldan '005 kyi ngo bo nyid kyi sku yin par shes par bya stel ... gang dag gis
rnam pa thams cad du shintu rnam par dag pa'i ngo bo nyid chos kyi skur 'gyur ba
zag pa med pa'i chos de rnams kyang gang zhig yin zhe nal byang chub phyogs mthun
tshad med dangl rnam par thar dang mthar gyis nil . .. etc. (quoting AA 8 vss. 2-6
listing the buddhadharmas), Pk 5185, fols. 92-4-8 ff.
sambhogikakaya (commenting on AA 8 vs 12): sku des sangs rgyas bcam
ldan 'das byang chub sems dpa' chen po sa chen po La zhugs pa rnams dang thabs cig
tu ,ka na ma tho ba med pa theg pa chen po'i chos kyi longs spyod kyi dga' ba dang
bde ba so SOT myong bar mdzad pa yin nol, Pk 5185, fols. 96-2-6 ff.
nairmii1;Iikakaya (commenting on AA 8 vss. 33-34a): 'bras bu'i gnas skabs
rnam pa thams cad legs par yongs su rdzogs pa'i chos kyi sku thob pa ni 'khor ba ji
srid par phyogs bcu'i Jig rten gyi khams rnams su sprul ba rnams kyis lhun gyis grub
ching rgyun mi 'chad, par sems can gyi don sna tshogs pa 'jug par byed pa'i sgo nas
gnas yongs su gyur ba'i phrin las kyi dbang du mdzad do I , Pk 5185, fols. 98-5-1 ff.
The revised 25,000 PP passages VIII.I-VIII.3 read as follows (I quote
the Tibetan for comparison to Arya Vimuktisena's passages above): VIIU-
sviibha,vikakiiya: rab 'byor gzhan yang zag pa med pa'i chos rmi lam lta bu dngos
po dang mi ldan pa dngos po med pa'i rang bzhin can rang gis mtshan nyid kyis stong
pa rnam pa thams cad yongs su dag par 'gyur ba de dag thams cad kyi rang bzhin
gang yin pa mtshan nyid gcig po 'di lta stel mtshan nyid med pa de ni de bzhin gshegs
pa dg;ra bcom pa yang dag par rdzogs pa'i sangs rgyas yin par rig par bya stel rab
'byor byang chub sems dpa' sems dpa' chen po de ltar shes rab kyi pha rol tu phin pa
La bslab par bya'ol Irab 'byor gyis gsol bal bcom ldan 'das zag pa med pa'i chos thams
cad kyang gang dag lagsl bcom ldan 'das kyis bka' stsal bal byang chub kyi phogs kyi
chos sum cu rtsa bdun dangl tshad med pa bzhi dangl ... etc. (listing all the buddha-
dharmas)l rab 'byor 'di ni zag pa med pa'i chos thams cad ces bya'ol lrab 'byor de ltar
byang chub sems dpa' sems dpa' chen po shes rab kyi :pha rol tu ph yin pa 'di La bslab
pal' bya'o zhes bya ba ni ngo bo nyid kyi sku yin noll, Pk 5188, fols. 3-4-1 ff. This
passage of the sutra appears to have been written based on Arya Vimuktisena's
CONTROVERSY OVER DHARM.A.l01YA 73
arks on AA 8 vss. 1-6 above, while also making use of the terminology of
section VIII.4 which immediately follows interpolated passages VIII.1-
.3 (PP VIII.4: ... chos thams cad rmi lam lta bu dngos po ma mchis pa dngo
mchis pa'i ngo bo nyid rang gi mtshannyid kyis stong pa roams la 'di dag ni
te., fols.4-1-1 ff.). Note that Arya Vimuktisena raises a hypothetical
. n in his- comments on svabhavikakaya: "gang dag gis roam pa thams ead
<kin tu roam par dag pa'i ngo bo nyid chos kyi skur 'gyur ba zag pa med pa'i chos
rims kyang gang zhig yin zhe nal." "What are those undefiled dharmas whose
pletely purified nature is the As the answer to this he quotes
vs. 2-6. The, author of PP VIII.1 puts Arya Vimuktisena's hypothetical
into the mouth of Subhuti. If Arya Vimuktisena had been quoting
Wi'VIIIJ, rather than the other way around, he would have indicated so with
quotation and by names "Bhagavan"
In the vocatIve, as he_had done m SImIlar cases throughout hIS
To my knowledge, Arya Vimuktisena never raises a question as
when it was actually by a in the In such
tases he always quotes the character m the sutra askmg the questIOn.
VIII.2-sambhogikakaya: rabk 'byor gzhan yang shes rab kyi pha rol tu phyin
(#.'di la bslabs shing chos de dag thams cad thob nas bla na med pa yang dag par
pa'i byang chub tu mngon par rdzogs par sangs rgyas tel thams cad tu thams
I,"'t',:cfd, mam pa cad nas thams du de bzhin gshegs fa dgra bcom pa yang daFf
pa z sangs rgyas roams kyz sku skyes bu chen po'z mtshan sum cu rtsha gnyzs
brgyan pal dpe byad bzang bo rgyad cus pal. byang chub sems dpa' sems
W4prz'chen po roams La theg pa chen po mchog gz chos kyz longs spyod bla na med pa
'lifdga'ba dangl bde ba dangl tshim pa dangl rab tu dga' ba ston par mdzad par 'gyur
rab 'byor de Itar byang chub sems (dpa'sems dpa' chen po shes rab kyi pha rol
pa La bslab par bya'o zhes bya ba ni longs spyod rdzogs pa'i sku yin noll)
5188,[0Is. 3-5-2 ff. There is no passage like this anywhere else in the PP
[miras. It is clearly modelled on Arya Vimuktisena's remarks.
PP rab 'byor gzhan yang shes rab kyi pha rol tu
'" . pa La sob pa na chos de dag thams cad rtogs par byas nas bla na pa yang
, par rdzogs pa'i byang chub tu mngon par rdzogs par sangs rgyas nas phyogs bcu'i
tftg rten gyi khams dpag tu med mtha' med par dus thams cad du de bzhin gshegs pa
[gra bcom pa yang dagpar rdzogs pa'i sangs rgyas kyi sku (sDe dge: "sku'i") sprul
!Pfi sna tshogs kyi sphrin gyi (sDe dge: "gyis"j sems can thams cad gyi don mdzad pa
llle ltar ni.b 'byor byang chub serns dpa' sems dpa' chen po shes rab kyi pha rol tu ph yin
lpa la bslab par bya'o zhes bya ba ni sprul pa'i sku yin noll Pk 5811, fols. 3-5-6 ff.
i!\gain, this is clearly based on Arya Vimuktisena's remarks. Arya Vimuktisena
the P P textual basis for nairmaTJikakaya after making his own comments
;011 the subject; and that textual basis is PP VIII.4. He had never heard of PP
Arya Vimuktisena's identification of PP VIII.4 and VIII.5 as the
bases for AA chapter 8 is reasonable. These passages are found in all
recensions and translations of the Large PP Sutra in Sanskrit, Chinese
'f-nd'TIbetan. They are found in the Gilgit manuscript of the 18,000 PP which
dated to 5th or 6th century C.E., and in Molq;ala's Chinese translation of
!he 25,000 PP, dated 291 C.E. So there is no reason to doubt that they were
74
JIABS VOL. 12 NO.2
part of the Large PP Sutra at the time that the AA was composed, c. 4th-5th
century C.E. Within the Large Pp, passages VIII.4-VIII.5 comprise the last
. part of the sutra, and immediately follow the passages identified by Arya
Vlmuktisena and later commentators as the bases for AA chapter 7.
The likelihood is that the author of the AA did indeed base his chapter 8 all.
them.,;,
50. Conze, Large Sutra, Motilal edition, pp. 576-643, especially pp. 578-.
587. Conze, Mahii-prafniipiiramitii sUtra, VIII 4-VIII 5. Pk 731, pp. 137-2-4'
to 187-3-3, especially pp. 139-1-1 to 145-5-5.
51. Conze, Large Sutra, Motilal edition, p. 578.
52. Dutt, Conze and Lethcoe have all noted that the revised 25,000 pp
sutra is a recast version ofthe 25,000 PP sutra, the section headings of the AX
having been inserted into the corresponding sections of the sutra. In addition'
Conze and Lethcoe noted that the sutra in its revised edition was altered id
certain places (by additions and transpositions) to bring it more closely into
line with the AA (Nalinaksha. Dutt, ed., The Paiicavi1[!Solisiihasrika Praj-
iiapiiramitii, Edited with Critical Notes and Introduction, [London: Luzac & Co"
1934] pp. v-xiii. Conze, The Prajiiiipiiramitii Literature, pp. 37-39. Lethcol
"Some Notes," pp. 500 ff.): With specific reference to revised 25,000 PP pas:
sages VIII.1-VIII.3, Conze, noting that these passages are missing in the Gilgit'
Manuscript of the 18,000 PP, believed they were later to the pi
sutra (Conze, The Gilgit Manuscript of the NtiidaSasiihasrikiiprajiiiipiiramitii-sUtra,
p. xvii). Elsewhere, however, based on the report ofTaranatha, Conze surmised
that the revised 25,000 PP belonged to the 5th century, and that Arya Vimuk";
tisena consulted the revised PP before writing his own commentary on the'
AA (P P Literature, p. 37). Lethcoe found that revised P P passages VIII.l-VIII$
were missing in all Chinese translations of the 25,000 PP (Lethcoe, p.
No scholar, up to the present time, has noticed the evidence of Arya
tisena's commentary, which proves thatPPpassages VIILl-VIII.3 were added
after Arya Vimuktisena (and were composed taking his remarks as basis). Nor
has anyone noticed the implications of this for the interpretation of AA
The revised 25,000 PP Sanskrit manuscripts do not identify its compiler: .
The Tohoku index of the bsTan 'gyur identifies Haribhadra a,s the com pilei]
and in the final lines of the Tibetan translation, the compiler does
identify himself as "Seng ge bzang po," Tibetan for "Haribhadra" (Pk 518S;;
fols. 61-3-1 to 61-3-2). Conze's surmise that the revised 25,000 PP predateCii
Arya Vimuktisena (early 6th century) cast its attribution to Haribhadra (Iate::1
8th century) into doubt. The proof presented here that passages VIII.-
post-date Arya Vimuktisena may indicate that the entire text post-dates him,';
thereby lending some further support for its attribution to
Dutt, noting that the indexes to the Tibetan bsTan 'gyur give Haribhadra,%
as the compiler, tentatively identified him as the author (referring to him as;:
."Sirphabhadra," a mistaken restoration of the Tibetan "Seng ge bzangpo"):;'
However, he was very tentative about it, because, he claimed,
nowhere identified himself in the revised PP as its compiler (Dutt, ed. Thi,]
Paiicavi1[!Satisiihasrikii Prajiiiipiiramitii, p. viii). However, as noted
Haribhadra does identify himself as the compiler at the very end of the
CONTROVERSY OVER DHARMAKAYA 75
has.come down to us in its Tibetan translation.
53. At this point I would refer the reader back to the second section of
paper where AA 8 vss. 1-6 were and translated.
54. Pk 5185, fols. 92-4-6-100-3-7. Arya Vimuktisena is dated to the
6th centu.ry C.E. (Ruegg, Literature of Madhyamaka, p. 87).
55. As far .as we know, Haribhadra (late 8th century C.E.) was the first
this interpretation by newly proposing that the AA taught not three
four kayas.
56. In a future paper, I will examine AAchapter 8's place within the
of the AA as a whole. Special attention will be paid to the AA's table
COIlteJrltS and concluding verses. Certain idiomatic Sanskrit word construc-
will be analyzed and compared to similar constructions found in other
iN texts. This sort of philological and comparative textual analysis will
the theory formulated in this paper, by providing further evidence
the AA teaches three (and not four) buddhakayas.
BIBLIOGRAPHY
. PK = Peking edition of the Tibetan Tripitika. Tibetan Tripitaka
Research Institute, Tokyokyoto, 1956
AA = Abhisamayala'T[1ltara
DhDh V = Dharmadharmatavibhaga
MA V = Madhyantavibhaga
MSA = Mahayanasutralarrtkara
Msg = Mahayanasarrtgraha
P P = Prajnaparamita
RGV = Ratnagotravibhaga
2iWIC MATERIALS
If:'
Shastri, Swami Dwarikadas, ed. Abhidharmakosa & Bhiya
rJ. of Acarya Vasubandhu with Sphutartha Commentary of Acarya Yaomitra .
. .. Bauddha Bharati Series, No.5. Varal).asi: Bauddha Bharati, 1971.
f#hisamayalarrtkara-Aloka by Haribhadra. Wogihara, U. ed. Abhisamayalarrtkar'a-
laka Prajnaparamitavyakhya. Tokyo: Toyo Bunko, 1934.
M-Durbodha-Aloka by Dharmakirtisri. Pk 5192, Vol. 91.
ifA-Prasphutapada by Dharmamitra. Pk 5194, Vol. 91.
rM-Sputartha by Haribhadra. Sanskrit reconstruction based on Aloka and TIbet-
I:; an translation by Hirofusa Amano: A Study on the Abhisamaya-alarrtkara-
jro-rika-sastra-vrtti. Tokyo: Japan Science Press, 1975. Pk 5191, Vol. 90.
by Ratnakarasanti. Pk 5199, Vol. 91.
M-varttika by Bhadanta Vimuktisena. Pk 5186, Vol. 88
M-vrtti by Arya Vimuktisena. Pensa, C. ed. Serie Orientale Roma XXXVII.
'76
]IABSVOL.12NO.2
Rome: ISMEO, 1967 (just chapter 1 in Sanskrit is edited.) Pk 5185, Vol. 88.
. Wogihara, ed. with AA-Aloka. Conze, Ed.
ward, trans. The Peifection of Wisdom in Eight Thousand Lines
and tra?s. The Gilgit.
Manuscrzpt of the Correspondmg to the 6th'':'
7th and 8th Abhisamayas. Rome: ISMEO, 1974. '.
Kiiyatrayastotra. Pk 2015, Vol. 46. Roerich, George N. ed.and trans. The Blue>
Annals, pp. 1-2.
Kiiyatrayasutra. Pk 949, Vol. 37.
Kiiyatrayavrtti, by Jiianacandra. Pk 5291, Vol. 101.
KiiyatrayiiramukhaSiistra by Nagamitra. Pk 5290, Vol. 101.
Madhyiintavibhiiga. Nagao, Gadjin M., ed. Tokyo: Suzuki Research Foundation
1964. .,
Nagao, Gadjin M., ed. with Kiirikii.
Mahiiyiinasa7[Lgraha Lamotte, Etienne, ed. and.trans. La Somme du Grand
d'Asariga. Publications de l'Institute Orientaliste de Louvain, No.8. Lou>
vain: Institut Orientaliste, 1973.
Pk 5551, Vol. 112
Mahiiyiinasa7[Lgrahopanibandhana by Asvabhava. Pk 5552, Vol. 113.
Mahiiyiinasutriila7[Lkiira. Levi, Sylvain, ed. Mahiiyiina-Sutriila7[Lkiira-Expose de
Doctrine du Grand VChicule, Tome L Bibiotheque de l'Ecole des HauteS';
Etudes, fasc. 159. Paris: Librairie Honore Champion, 1907. ..
Levi, Sylvain, ed. with MSA.
Mahiiyiinasutriila7[Lkiiratikii by Asvabhava. Pk 5530, Vol. 108.
Markaumudi, by Abhayakaragupta. Pk 5202, Vol. 92.
Munimatiila7[Lkiira by Abhayakaragupta. Pk 5299, Vol. 101.
Paiicavi1[!Satisiihasrika-prajiiiipiiramitii-sutra, revised. Conze, Edward. Mahii-Pra':
jiiiipiiramitii Sutra. Unpublished (Conze's typescript romanization of thl
Sanskrit, University Microfilms, Ann Arbor). Dutt, Nalinaksha ed., The'
Paiicavi1[!Satisiihasrikii Prajiiiipiiramitii, Edited with Critical Notes and Introduc:
tion. London: Luzac & Co., 1934. Pk 5188, Vols. 88-90..
Ratnagotravibhiiga and Ratnagotravibhiiga-vyiikhyiina. Johnston, E.H., ed. Patna:
.. The Bihar Research Society, 1950. D
Samiidhiriijasutra. Regamey, K., ed. and trans. chapters 8, 19,22. Warsaw: The:
Warsaw Society of Sciences and Letters, 1938..:
Sa7[Ldhinirmocanasutra. Lamotte, Etienne, ed. and trans. Sa7[Ldhinirmocanri
Sutra-l'Explication des Mysteres. Louvain: Bibliotheque de l'UniversitY;
1935.
Siirattamii by Ratnakarasanti. Jaini, P. ed. TIbetan Sanskrit Works No. XVIII:
Patna: K.P. Jayaswal Research Inst., 1979. Pk 5200, Vol. 92.
by Sthiramati. Pk 5531, Vol. 108.
Suvar7J.aprabhiisottama-sutra. Nobel, Johannes, ed. Leiden: E.J. Brill, 1958.
Vajracchedikii Praiiiipiiramitii. Conze, Edward, ed. Rome: Is.M.E.O., 1957.
CONTROVERSY OVER DHARMAK4YA
TIBETAN WRITINGs
'rf:;
77
rin-chen-grub. Sher 'grel rgyal cher bshad pa lung gi snyema. Varal}.asl:
i;i'i Pleasure of Elegant Sayings, 1979 ..
'bo-ram-pa bsod-nams seng-ge. sBas don zab mo'i gter gyi kha 'byed. n.p., n.d.
. . . Yum don rab gsal. n.p., n.d.
dar-ma-rinchen. rNam bshad snying po'i rgyan. Varal}.asl: Pleasure
of Elegant Sayings, 1980. ,
:Sera-Ije-otsun-pa. Chos sku phyi don. n.p., n.d.
'{TI;ong-kha-pa. Legs bshad gser phreng. Dharamsala n.p., n.d.
gYag-ston sangs-rgyas-dpal. Sher phyin mngon rtogs rgyan gyi 'grel pa rin po che'i
phreng ba blo gsal mgul rgyan. Rajpur (n.p., n.d.).
l1WORKS IN EUROPEAN LANGUAGES
ti:Conze, Edward. The PrajfiaparamitaLiterature. 2nd ed. Tokyo: Reiyukai; 1978.
. The Large Sutra on Perfect Wisdom. (translation of the revised
SCI; Paficavi'T[l5ati. PP sutra with portions of the Sata. and 14tadaSa. PP sutras).
Jt Delhi: in Eight Thousand Lines (translation
t',. of the 14tasahasrikaprajfiaparamita.) Bolinas, CA: Four Seasons Founda-
liX. tion, 1975. _ /
Ronald Mark. : "Afraya-parivrttil-paravrtti Among the Yogacara." Un-
published Ph.D. dissertation, University of California, Berkeley, 1985.
P. "B usshin," in H obogirin: Dictionnaire Encyclopedique du Bouddhisme
d'apres les Sources Chinoise et Japonaises. Tokyo: Maison Franco-Japonais,
1931.
Nalinaksha. Mahayana Buddhism. Delhi: Motilal Banarsidass, 1978.
tJLa Vallee Poussin, Louis de. Vijfiaptimatratasiddhi: La Siddhi de Hiuan-tsang.
;;,:" Paris: Paul Geuthner, 1928-29.
if . L'Abhidharmakosa de Vasubandhu-Traduction et Annotations.
y;.
;{, Paris: Paul Geuthner, 1924.
;Lancaster, Lewis R. "The Oldest Mahayana Siitra: Its Significance for the
.... Study of Buddhist Development." Eastern Buddhist, VIII.l (197 5), 30-41.
Nancy R. "Some Notes on the Relationship between the
, AbhisamayalarrWira, the revised PaficavimSatisahasrika, and the Chinese
translations of the unrevised PaficavimSatisahasrika." JAOS, 96.4 (1976),
499-511.
!,Nakamura, Hajime. Indian Buddhism: A Survey with Bibliographical Notes.
, Hirakata City, Osaka: KUFS Publication, 1980. .
;Obermiller, E. Analysis of the AbhisamayalaT{!kara. London: Luzac, 1933.
Andrew. "The Position of the 14tasahasrika Prajfiaparamita in the
. Development of Early Mahayana," in Prajfiaparamitii and Related Systems:
Studies in Honor of Edward Conze. Edited by Lewis Lancaster. Berkeley,
CA: Berkeley Buddhist Studies Series, 1977. ' .
78 JIABS VOL. 12 NO.2
Regamay, K. Three Chapters from the Samiidhiriijasutra. Warsaw: The Warsaw
Society of Sciences and Letters, 1938.
Ruegg, David Seyfort. La Thiorie du Tathiigatagarbha et du Gotra: Etude Sur la
Soteriologie et la Gnoseologie du Bouddhisme. Pu?lications de I'Ecoie Fran<;aise
d'Extreme-Orient, vol. 70. Paris: KF.E.O., 1969
_______ . The Literature of the Madhyamaka School of Philosophy in India.
Wiesbaden:. Otto Harrassowitz, 1981.
Schopen, Gregory. "The Phrase 'sa prthiv'ipradesas caityabhuto bhavet' in the
Vajracchedikii: Notes on the Cult of the Book in Mahayana." Indo-Iranian
Journal 17 (1975), 147-181. .
_______ . "Sukhavatl as a Generalized Religious Goal in Sanskrit
Mahayana Sutra Literature." Indo-Iranian Journal 19 (1977), 177-210.
Takasaki, Jikido. A Study on the Ratnagotravibhiiga. Serie Orientale Roma, Vol.
XXXIII. Rome: ISMEO, 1966.
Jhiina and Buddhist Scholasticism
:;by Martin Stuart-Fox
Buddhism teaches as its highest truth a path of meditative
.practice for the attainment of a series of altered states of con-
sciousness culminating in enlightenment and liberation. The-
. central place accorded this course of meditative techniques in
.iearly Buddhism is reason enough to examine carefully and crit-
)cally the various descriptions of it given in the Buddhist canon.
. An examination of the texts,. however, reveals both in-
:;adequacies and discrepancies. The more advanced techniques
rare too sketchily described to serve as guides to practicing
descriptions of stages are repeatedly presented in
stereotyped terms, discussed or elaborated upon only in much
later commentaries; the meanings of words are often unclear.
listings of stages on the path to enlightenment are fre-
quent. 1 And, in certain cases, textual descriptions contain what
appear to be outright contradictions.
The tendency has been for believers and scholars alike to
attempt to explain away such discrepancies, rather thari to ex-
plain how they came to be present in the canon. In part this
has been due to the concern of Buddhist scholars to extract
\from the texts some definitive statement of Buddhist theory
-and practice in order to reveal the "true nature" of Buddhism.
-Unfortunately, this often entails an exaggerated and uncritical
respect both for the texts and .for those who compiled them,
. together with a reluctance to question their accuracy, especially
where they pertain tohigher meditative practices. Thus, it has
been claimed that if paradoxes occur, these must have been
<deliberately designed by the ancient compilers to shake us out
of established patterns of thought, thus preparing our minds
for the revelation of Dhamma.
2
Other modern scholars either
79
--- -- -----------------
80 JIABS VOL. 12 NO.2
have been content to accept the attempts of earlier commen_
. tators to reconcile evident contradictions, or have simply ign()red
them.
Instead of explaining away textual discrepancies, however
a more productive line of inquiry would be to examine
cI."itically for any light they may shed on the evolution of Buddhist
thought and institutions. Accordingly, the purpose of this paper
is to examine two related discrepancies: the first between two
descriptions of the composition of first jhiina; the second, the
insertion of an additiomiljhiina, designated in this paper asjhiina
la, in certain later texts.
A comparison is first made between i:he descriptions of first
jhiina in the fourfold series given in the Sutta-pitaka, and the
descriptions of the first two jhiinas in the fivefold series in the
Abhidhamma-pitaka. It will be maintained that neither the con-
tradiction evident between the two descriptions of first jhiina,
nor the insertion of jhana la have been satisfactorily accounted
for either in the commentarialliterature or by modern scholars.
3
It will be argued on various grounds that the description of first
jhiina in the Abhidhamma account is phenomenologically ques-
tionable. A discussion then follows of how the conflicting descrip-
tions are likely to have come into existence, given what we know
of the historical conditions under which early Buddhism
It is suggested that the Abhidhamma listing is probably a product
of Buddhist scholasticism, having no basis in meditative experi-
ence. The paper concludes by drawing out certain implications
this study has for our understanding of the development of
early Buddhism, and for the methodology of Buddhist studies.
1. Jhana in the Suttapitaka
The importance of the jhiinas as stages in Buddhist medita-
tion is made abundantly clear time and a,gain in the Sutta-pitaka.
Together the jhiinas comprise the last stage, right concentration
(samma-samadhi) , of the Noble Eightfold Path.
4
This indicates
that the jhanas are stages in the practice of sa math a , or meditation
for calm, where the mind is prepared for vipassana, the practice
of insight.
5
This interpretation of the position of the jhiinas in
the Buddhist path is supported by canonical accounts of
JRANA AND BUDDHIST SCHOLASTICISM 81
iGotama's own enlightenment, according to which attainment of
>ihe fourth jhiina prepared his mind for developing the three
forms of supernormal knowledge, (tisso vijja), the last of which
'appears to constitute enlightenment.
6
.
,',. Descriptions of the fourjhiinas occur frequently throughout
'the Sutta-pitaka, always in the same stereotyped form. The stand-
,ard description for the first two jhiinas, literally translated, and
with certain key terms retained in their Pali forms, reads as
followS.
(1) Detached indeed from desires, detached from unwholesome
states, attaining the with-vitakka, with-vicara, detachment-
born, piti-sukha firstjhana, he abides [therein].
(2) From the suppression of vitakka-vicara, attaining inner tran- .
quillity, one-ness of mind, the non-vitakka, non-vicara, con-
centration-bornpiti-sukha secondjhana, he abides [therein].7 .
;:the terms left untranslated-vitakka, vicara, pili and sukha-are
'those identified in the Abhidhamma as four of the five jhana
factors (jhiinariga), to be discussed below.
~ . , Any analysis of the descriptions of- the jhiinas is hampered
by difficulty in determining the meanings of key terms. Neverthe-
less, it is possible from these brief descriptions to gain some idea
of (a) what constitutes firstjhiina; and (b) how the progression
:{rQm jhiina 1 to jhiina 2 is achieved. To begin with, first jhiina is
;tharacterized by separation (vivicca) from desires and unwhole-
:some states. These are traditionally summed up in the five "hin-
drances" (nivara1J,a): sensory desire; malice, sloth and torpor,
distraction and remorse, and doubt.
8
In addition first jhiina is
described as "detachment-born" or "separation-born" (vive-
luLja'T[t), reinforcing the notion of separation from unwholesome
mental states. On the positive side, first jhiina is characterized
by the presence of vitakka, vicara, pili and s'Jl,kha. Pili (usually
translated as 'joy") is subsequently transcended in the transition
from second to third jhiina, and sukha ("pleasure") in the tran-
sition from third to fourth jhiina. As neither pili nor sukha are
involved in the transition from first to second jhiina, they will
hot be considered further in this discussion.
Vitakka and vicara together constitute that characteristic
which is present in first jhiina but not in second. The meaning
82 JIABS VOL. 12 NO.2
of these terms is therefore crucial to an understanding of What
is entailed in that transition. Let us, therefore, look first at what
light the textual description of second jhiina may shed on the
meaning of vitakka and viciira. The impOFtance of the elimination
of vitakka and viciira for the attainment of the second jhiina is '
made clear by, the repetition involved in the statement that the
attainment of second jhiina is achieved through the suppression
of vitakka-viciira, and that the resulting state is non-vitakka, and
non-viatra. Now when the description of second jhiina is com-
pared with the structurally similar description of first jhiina, it,
is clear that JUSt as first jhiina is born of the detachment or
separation (viveka) necessary to counter desires and unwhole-
some states, so second jhiina is born of the concentration
(samiidhi) necessary to suppress vitakka-viciira. The quality of
concentration is indicated by the statement that second jhiina is
characterized by inner tranquillity (ajjhatta1[L sampasiidanarrt) and
one-ness of mind (cetaso ekodibhiiva1[L).
We are nqw in a position to investigate further the meaning
of the two terms vitakka and viciira. In the Sutta-pitaka
J
vitakka
often stands alone to mean "reflection, thought, thinking,,,g,
whereas viciira is only rarely found alone, and then in texts
which reveal evidence of early Abhidhamma analysis, such as, the
description of three types of samiidhi, to be discussed below.'
Vitakka is thinking about something: for example, kiimavitakka
translates as "thoughts about 10ve."IO Viciira, according to the
definition given by Rhys Davids and Stede in their Pali-English
Dictionary is "investigation, examination, consideration, deliber-
ation,"U implying a deeper, more focused form of thinking.
However Rhys Davids and Stede note that vitakka and viciir,a,
when used together in the combined form vitakka-viciira found,
in the description of second jhiina, denote "one and the same;
thing: just thought, thinking, only in an emphatic way (as they'
are semantically synonyrrious) ... one has to take them as one:
expression." 12 The suggestion here seems to be that when vitakka .
and viciira were used in combination, the effect of adding viciira
was to reinforce or emphasize the denotation of vitakka, perhaps
extending it to cover all varieties of thinking, including sustained
and focused thought. It is thinking in this inclusive sense that
the meditator suppresses through concentration when he attains
one-ness of mind and thus moves from first to second jhiina.
JHANA AND BUDDHIST SCHOLASTICISM 83
So much can be gleaned from the stereotype description of
and sec?nd in the
rj;specific detaIls, thIS descnptIOn does provide certam essential
for the practicing meditator: to attain first jhiina,
ipractice detachm.ent to and men-
states; to attam second Jhana, practice concentratIOn to sup-
fpress addition, th.e description specifies the positive
rgualiues that success m endeavours, most n.otably
presence of mward tranqUillity and one-ness of mmd as
attainment of second jhiina.
l .
rtf. Jhana in the Abhidhamma
(c Elsewhere in the Tipitaka are found two other descriptions
both differing in important respects from the Sutta
ticcount. They are formally set out only in the Abhidhamma-pitaka,
they either have the same general form as the Sutta ac-
['tount, or take the form of lists of ''jhiina factors.,,13 These lists
are clearly not meant to be a comprehensive statement
i?hfthe characteristics of the mental states constituting the various
'jhanas, as they omit some of the qualities included in the Sutta
taccount.
14
Instead these lists of jhiina factors name only those
:characteristics that are involved in the transition from each jhiina
)0 the next. The device of listing jhiina factors as a means of
the sequence of jhanas was a relatively late de-
;velopmep.t, a typically Abhidhammic mode of analysis and pre-
sentation which effectively reduced the jhiina description to its
'barest essentials.
Sometimes four jhiinas are listed in the Abhidhamma; some-
times the number is extended to five by interpolating an addi-
tional jhiina (here called for convenience la) between the first
and second jhiina of the Sutta account. Both the fourfold and
fivefold Abhidhamma lists include mental one pointedness (ekag-
gata) as a characteristic (or factor) in all the jhiinas, thus conflict-
ing with the Sutta account, which makes no mention of ekaggata
.in first jhiina. In the fivefold list,jhiina lais characterized as with
vicara but without vitakka. The two Abhidhamma descriptions,
together with their counterpart from the Sutta-pitaka are there-
fore as depicted in the following three tables.
84 JIABS VOL. 12 NO.2
Table 1
The jhiina factors in the Sutta-Pitaka
jhiina 1 vitakka-viciira p'iti sukha
sukha
sukha
jhiina 2 p'iti
jhiina3
jhiina4
ekaggatii. (= ekodibhava*)
(ekaggatii)
(ekaggatii)
*Ekodibhiiva is specifically mentioned only in jhiina 2. Though not
mentioned in jhiinas 3 or 4, it is clearly to be taken as characterizing
these as well.
Table 2
The jhiina factors in the Abhidhamma fourfold jhiina
jhiina 1 vitakka-viciira p'iti sukha ekaggatii
jhiina2 p'iti sukha ekaggatii
jhiina3 sukha ekaggatii
jhiina4 ekaggatii
Table 3
The jhiina factors in the Abhidamma fivefold jhiina
jhiinal vitakka-viciira p'iti sukha ekaggatii
jhiina la viciira p'iti sukha
e ~ a g g a t i i
jhiina2 p'iti sukha ekaggatii
jhiina3 sukha ekaggatii
jhiina4 ekaggatii
JHANA AND BUDDHIST SCHOLASTICISM 85
There are in these tables two distinct, though related, dis-
ccrepancies which require explanation: the addition in both
1!bhidhamma versions of ekaggata in first Jhana; and the interpo-
in the. fivefold Abhidhamma series of an additional stage
[ijMna la), brought about through splitting vitakka-vicara into
ttWo separate factors.
A comparison of tables 1 and 2 reveals that the only essential
between the Sutta version and the fourfold
version lies in the addition of ekaggata as a factor
fu.jhiina 1. But this is a most curious addition. Ekaggata (mental
:linepointedness) is synonymous with "one-ness of mind" (cetaso
and, as noted above, is that characteristic of second
Jhfi,na arises with. the suppression of
:i:oncentratlOn, and whIch thereafter charactenzes the remammg
hfi,nas. It is synonymous with cittass' ekaggata.
15
One would ex-
:pect, therefore, that ekaggata, if it is to be recognized as a Jhana
[actor, would appear only in second, third and fourth Jhanas.
fact, however, in both the fourfold and the fivefold
*-bhidhamma lists of Jhana factors ekaggata is included in first
Jhana as well, along with the very factor, vitakka-vicara, it is said
;iIl the Suttas to suppress. This obvious anomaly clearly requires
i, One possible explanation might be that the ekaggata that
the Abhidhamma ascribes to Jhana 1 may be somehow qualitatively
:different from that of the other Jhanas. It seems reasonable to
the mental one-pointedness of the lowerJhanas to be less
lWell-developed than that of the higher Jhanas, less "stable,,,16 so
;hiore likely to break down through the intrusion of "hindering
'thoughts." Credence is lent to this view by the existence, accord-
:ing to the Abhidhamma of a "weak" form of ekaggata defined as
of thought,,17 or "stability of mind,,,18 which is said
,to characterize other mundane states of consciousness.
19
About
form of ekaggata Buddhaghosa comments that none of the
bther characteristics of ekaggata apply to it.
2o
Buddhaghosa in
fact recognizes three degrees or kinds of ekaggata. The weakest
kind is that present in "original consciousness." A degree
stronger than this is the kind of ekaggata present in the transi-
tional state of consciousness known as access-ihana, which
characterizes the moment of entry into first Jhana. The third
and strongest kind of ekaggata is that characterizing firstJhana.
21
86
JIABS VOL. 12 NO.2
In firstjhiina, ekaggata has already developed to the point where
it is "touching the object well, as the lid above tOl.lches the surface
of the box below.,,22 In other words, Buddhaghosa believed
both that the ekaggata which characterizes firstjhana was qualita_
tively different from the weak form present in mundane states
of consciousness, and that it was qualitatively identical with that
characterizing the higher jhanas. Thus Buddhaghosa's accOunt
with its three different grades of ekaggata, provides no
of the anomaly of the presence of ekaggata, as a factor in the!
first jhana.
This brings us to the second discrepancy noted above
namely that in the fivefold Abhidhamma series only vitakka
su ppressed in movirig from first jhana to jhana lao Vicara is sepa-
rately suppressed only in the transition to the next stage again
(secondjhana). This description makes sense onlyifit is in prac-
tice possible separately to suppress first vitakka then vicara. In
the Sutta-pitaka, the term vicara was used only to reinforce the
meaning of vitakka. However, according to Rhys Davids and
Stede: "With the advance in the Sangha of intensive study of
terminology these terms become distinguished mutually. Vitakkl{
became the inception of the mind, or attending, and was no
longer applied, as in the Suttas, to thinking in general.,,23 The
Vibhariga distinguishes vitakka as "meditation, thinking, thought,
fixation, focussing, application of the mind, right thought" froni
,vicara, which is "searching, examining, constant examining,'
scrutinizing, constant connection of (and) constant inspection
by consciousness.,,24 In other words, by the time of the
Abhidhamma, vicara ,had already taken on the sense of steady,
focused thinking. By the time of Buddhaghosa (fifth century
CE), the distinction had become well established. According to
Buddhaghosa, Vitakka "is literally 'one thinks about,' or a 'think-
ing about' .... Its [main] characteristic is the lifting of conscious-
ness on to the object .... It has the junction of impinging, of
tircumimpinging .... Its manifestation is bringing the mind near
to the object." By contrast, vicara is "discursive work upon, or
traversing of the object. It has threshing out (or contemplation) ",
of object as characteristic, the linking of co-existent states to the
object as junction, and continuous binding as manifestation.,,25.
The question is, of course, whether the differentiation between
vitakka and vicara in the Abhidhamma reflected a more refined;
JHANA AND BUDDHIST SCHOLASTICISM 87
riAtrospective phenomenological description of mind, or was
fhierelY a scholastic distinction made in the process of intellectual
%analysis. Here the commentarialliterature is unhelpful. As Rhys
and Stede warn: "The explanations of Commentators
mostly.of an edifying nature and based more on popular
on
ThIs termmologlcal dlstmctlOn m the fivefold senes between
[tVitakka and viciira actually makes inclusion of ekaggatii in first
For thOl:gh there seems t? some
m the claIm that sustamed thought (vzcara) can
fcoexist with onepointednessof mind (in jhiina la), it is clearly
for the mental process of casting around and alight-
on an object of thought (vitakka) to be able to exist with
i"bnepointedness (in firstjhiina). In this connection, it is perhaps
t'ilOt surprising to note that in the Abhidhamma listing there is
Ydisagreement over the means of transition from first jhiina to
Yhiina lao The Vibhanga states that jhiina la is vivekajaT{l (born of
fdetachment), as is firstjhiina in the fourfold series;27 while the
WDhammasangar;,i states it is samiidhijaT{l, (born of concentration),
is jhiina 2 in the fourfold series.
28
This suggests, at the very
neast, that the monastic compilers were in disagreement not only
how the interpolated jhiina ought to be characterized, but
:also over how it should be attained. 29
tIll. Attempts at Reconciling the Discrepancies
In view of these anomalies in both the four- and fivefold
;Abhidhamma lists, one might have expected Buddhists generally
:10 have given preference to the Sutta description of first jhiina
,as being the "correct" version. Surprisingly, however, early com-
and modern scholars alike have consistently opted
Jor the Abhidhamma account. For. example, Buddhaghosa, while
explicitly recognizing that ekaggatii is not present in first jhiina
in the Sutta accounts, prefers the Abhidhamma version as superior
even to that of the Buddha himself.
Among the factors, although collectedness of mind [ekaggata] is
not shown in this [Sutta] reading, as "wherein is thinking applied
and sustained," yet it is a factor, as is stated in the Vibhanga:
''jhiina is applied thinking, sustained thinking, rapture, bliss, collected-
88 JIABS VOL. 12 NO.2
ness of mind." Whatever may have been the intention of the Blessed
One in making the outline, it is revealed in the Vibha;liga.
30
In fact, so eager is Buddhaghosa to paper over the difference
between the Sutta and Abhidhamma accounts that in the chapter
of his Visuddhimagga where he quotes the Sutta description, he
goes on to refer in the next line to "the First Jhana, which has
put away five factors, is endowed with five factors ... ".31
Modern scholars have tended to follow Buddhaghos
a
.
Paravahera Vajiraiia.Q.a Mahathera, in his Buddhist Meditation in
Theory and Practice, agrees that whatever the suttas say, ekaggatii
was meant to be included in firstjhana,32 So too does Henepola
Gunaratana, whose doctoral thesis on the jhanas is the most
detailed modern study devoted to this most important aspect
of Buddhist teaching. Gunaratana lists the four factors in first
jhana as described in the Sutta accounts, but then comments,
"the fifth, one-pointedness, is added elsewhere.,,33 Instead of
discussing this discrepancy, he merely states that it is "more
than obvious" that ekaggata ought to be included in firstjhiina.
34
.
To account for the omission of ekaggata from the Sutta a(
count of jhana 1, Gunaratana suggests that "the prominence of
ekaggata in the attainment of jhiina [by which he means specifi-
cally first jhiina] was so evident that it was felt unnecessary to
mention it separately.,,35 This suggestion finds little textual sup-
port. Ekaggata is certainly prominent as a characteristic ofjhiinas
2,3 and 4, but its prominence in them derives from the complete
suppression of discursive thought; in jhiina 1 discursive thought
is still present. Elsewhere, Gunaratana suggests that ekaggata is
not mentioned in the Sutta account of first jhana because it is
not until second jhana that "concentration first acquires emi-
nence." He supports this with the observation that: "The concen-
tration of the firstjhana, being subjectto the disturbing influence
of applied thought [vitakka] and sustained thought [vicara], is
still imperfect.,,36 But these two sugges!ions are based on con-
tradictory premises. He cannot have it both ways: ekaggata can-
not both be so prominent in firstjhana as not to warrant mention
and not acquire eminence until second jhana.
There do exist, in the Sutta-pitaka, three references to the
occurrence of ekaggata in firstjhana. It is conceivable, therefore,
. that the Abhidhamma description is merely the formalization of
JHA.NA AND BUDDHIST SCHOLASTICISM 89
supported description. How-
lfiver, cntIcal exammatIon of these three references reveals that
':11 are textually suspect, late interpolations or additions to the
"ali corpus. Only one of the three references attributes a state-
on the occurrence of ekaggatii in firstjhana to the Buddha
;Himself. It is found in the Saliiyatana-vagga of the Sarrtyutta and
in the context of a miraculous appearance by the Buddha
tb the disciple Moggallana,37 a context which already suggests
liliat the passage constitutes a later textual interpolation. In the
t()urse of this appearance, the Buddha urges the meditating
}disciple to practice mental onepointedness, repeating an identi-
(tal exhortation for each jhiina. Thus, for the first jhiina the
lfofIllula becomes: "Make steadfast thy mind in the first trance
':UhiinaJ. In the first trance, make the mind one-pointed [cittam
IIi the first trance compose the mind.,,38 That
same set formula is repeated without distinction for each
r'jhiina could well be a consequence of faulty memorizing: refer-
to onepointedness in subsequent jhanas may have been
inadvertently to first jhana as well. But in view of the
fll1agiographic reference to Moggallana it seems more likely that
text is late, and was composed under Abhidhammic influ-
tJ ::
ig: Support for this conclusion comes from another source,
[:one whose importance was appreciated by A.K. Warder,39 but
has not been used as often as it might have been by Pali
That source is the Chinese counterpart of the four
tpikiiyas (the Chinese iigamas). In the Chinese texts, this reference
fo the practice of onepointedness, together with "virtually the
Moggalliina-sarrtyutta" is missing,40 thus indicating that the
'entire section represents a late addition to the Theravadin
canonY
A second example of a reference in the Sutta-pitaka to the
"occurrence of ekaggatii in first jhiina occurs in the Mahiivedalla-
of the Majjhima.
42
There, the disciple Sariputta states that
,firstjhana is "five-factored," (counting vitakka-viciira as two fac-
:!ors instead of one and including ekaggatii). But Sariputta's de-
scription contains an inconsistency. When asked what charac-
.terizes firstjhiina, Sariputta answers by listing four factors: vit-
{j,kka, viciira, pZti and sukha, but then, on being asked how many
factors are to be found in first jhana, he replies that there are
90 JIABS VOL. 12 NO.2
five: vitr:kka, viciira,pUi,sukha andekaggatii! Now, as Pande
out, thIS Sutta bears all the hallmarks of a late text.
43
What i?
more, although its counterpart in the Chinese canon is
all but identical, it lacks precisely this section on the composiiioff\)
of the jhiinas.
44
This section therefore constitutes an even
interpolation in a late text, almost certainly to be attributed toX
the influence of early Abhidhammic analysis...,;;:
The third reference to ekaggata in first jhiina is found in
Anupada-sutta. There ekaggatii is included in a list of sixteen;
characteristics of first jhiina. 45 The list itself is full of
being both repetitive and inconsistent. It first follows
stereotype description of first jhiina with only vitakka-viciira, Pit!
andsukha, but then goes on to list these same factors again, with
ekaggatii. Other qualities listed include equanimity (upekkha) and
. "desire" (chanda). Equanimity is out of place because it is not'
supposed to be attained until third jhiina. Desire is out of
both because it conflicts with equanimity, and because it should.
be overcome with the attainment of firstjhiina.
46
Not surpris)
ingly, the entire Anupada-:sutta does not exist in the Chinese
canon, thus confirming Pande's identification of it as a demon
j
'
strably late text.
47
Hi
We can only conclude that none of these three Sutta-Pitaka'
references constitutes evidence that the Buddha himself ever
taught that ekaggatii was present in first jhana. By including
ekaggatii, Buddhaghosa, and a number of modern scholars as
well, have without valid reason preferred the Abhidhamma
scription to that of the Buddha-a choice which itself is perhaps
in need of explanation.'
When we turn to attempts to reconcile the second
ancy, concerning the interpolation of jhiina la in the fivefold
Abhidhamma listing, we encounter another set of similarly un':
satisfactory explanations. To the question "Why are four and
five meditations taught?" the Vimuttimagga replies: "because the
result depends on two sorts of men."
Q. How does a yogin induce the second meditation from
the first?
A. He considers the coarseness of initial and sustained ap:
plication of thought, knows the disadvantages of initial
JHANA AND BUDDHIST SCHOLASTICISM 91
and sustained application of thought, and induces the
second meditation, which is free from initial and sus-
tained application of thQught. This is the way of prog-
ress in the four meditations.
And again, there is another man. He is able to induce
freely the second meditation out of the first meditation.
He considers the coarseness of initial application of
thought and knows the disadvantages of initial applica-
tion of thought. He discerns the state of being free from
initial application of thought. Possessing restricted sus-
tained application of thought, he induces the second
meditation. This is the way of progress in the five medi-
tations. Therefore, the five meditations are taught.
48
this, Buddhaghosa, in the Atthasiilinz, adds a further reason:
the teaching." This he explains as follows:
Those conditions of the Law by which, because they have
been thoroughly penetrated, the teaching is adorned-
those conditions were thoroughly penetrated by the
Tathagata. Hence, because of the vastness of his knowl-
.edge, the teacher, who is skillfulin arranging his teaching;
and who has attained the [art of] embellishing it, fixes that
teaching by whatever factor that has come to hand, and in
any way he chooses. Thus here he has classified a First
Jhana of five factors, a fourfold Second Jhana 'without
initial and with only sustained application of mind', a
threefold Third Jhana, a twofold Fourth Jhana and a
twofold fifth Jhana. This we have called embellishing the
teaching.
49
Neither ofthese two commentarial explanations can be said
'!to be convincing. Modern scholars offer a variety of suggestions
t.as to how the two lists arose. According to Rhys Davids and
:;!Stede, the jhiinas form "one series of mental states, and the
,stages might have been fixed at other points in the series."so
;This is to make Gotama's division into four stages all but arbi-
:trary, which is hardly convincing. Narasabho says: "It should
be noted that the fivefold system is given with a view to varying
92 JIABS VOL. 12 NO.2
mental endowments as well as simplicity for cultivation of the
aspirants [sic]. To some only vitakka appears gross whereas the
remaining factors appear calm ... ,,51 Gunaratana agrees.52
Pande simply remarks that: "In the Abllidhamma-stage the four
Jhanas were turned, for the sake of greater system, into a five-fold
[ sic]. ,,53 Other scholars ignore either one account or the other. 54
Again, it might be suggested that the interpolation ofjhiina
la in the Abhidhamma fivefold listing merely formalizes earlier
distinctions between vitakka and viciira drawn in the Sutta-pitaka ..
On five occasions in the Suttas, the following threefold classifi-
cation of samadhi is given: (i) with vitakka and vicara; (ii) without
vitakka but with viciira; and (iii) without vitakka or vicara.
55
Since
the jhiinas constitute stages in the attainment of samma-samiidhi,
this classification could possibly have led to the insertion ofjhiina
la into the Abhidhamma.
56
(At the same time, if ekaggatii is taken
as the defining characteristic of samadhi, this classification could
also suggest the possible presence of ekaggata in first jhiina.)
Reference to the Chinese texts throws some interesting light
on these five references. The references in the Sarrtyutta and the
Ariguttara are both late, as neither Pali sutta has any Chinese.
counterpart. Chinese counterparts do exist for those suttas in
the Dzgha and MaJjhima in which the remaining three references
occur. In the case of the Sarigiti-sutta, itself a demonstrably late
text comprising a series of Ariguttara-like numerical groups of
short doctrinal statements,57 the Chinese text closely follows the
Pali sequence, except at just the point where the reference to
the three kinds of samadhi occurs.
58
Of this there is no sign. In
the Dasuttara-sutta, a slightly different situation pertains. The
reference to three kinds of samadhi based on the presence or
absence of vitakka and vicara is replaced by a reference to three
kinds of samadhi characterized by emptiness, desirelessness, and
signlessness.
59
In this case, it would appear that an earlier, rather
cryptic reference to three kinds of samadhi preserved in the
Chinese rescension was replaced in the Theravadin canon by a
simpler, but later classification.
It is the single reference to the threefold classification of
samadhi that occurs in the MaBhima that permits us to narroW
down the probable date of this curious doctrinal development.
The reference occurs in the Upakkilesa-sutta, and is also found
in the Chinese canon-with a single significant difference. The
JHANA AND BUDDHIST SCHOLASTICISM 93
;:secorid samadhi is described as one in which vitakka is absent and
is reduced.
60
In the Pali version, vicara is simply stated to
tte present. The Chinese description is repeated several times
sutta, so it would appear that this may constitute a tran-
veT-sion dating from the period when the distinction
I'between vitakka and vicara was being drawn on the basis that
"initial thought" was eliminated, it took time to' eliminate
thought." Now since it has been shown that the
rchinese counterparts of the Dzgha and Ariguttara were probably
from the Dharmaguptaka canon, whereas the Maj-
were fr?y? the Sa:vastivadin
tiano
n
,61 It IS possIble to date thIs transItIOnal versIOn" of the
rthreefold classification of sam(idhi to the period between the
lbreakaway of the Dharmaguptakas (no sign of the doctrine in
Dzgha or Ariguttara) and the division between the Therava-
I'llins and the Sarvastivadins (occurrence in the Majjhima, but not
its final Theravadin form) .
. j-:", !
flV: Resolving the Discrepancy
1Jt; For modern scholars trained in the logic of textual analysis,
!fue discrepancies evident between the Sutta and Abhidhamma
of the jhanas are too obvious to be disregarded.
!Either ekaggatii can coexist with vitakka-vicara, or it cannot. Either
fthere exists an intermediate stage (jhiina la) which is without
but with vicara, or there does not.
62
Logic alone suggests
ifliat vitakka, understood as discursive thought, cannot exist in .
state of consciousness entailing one-pointed mental concen-
,'ITation: if the mind is casting around for an object upon which
16focus, or is following one train of thought after another "like
,gwild monkey,,,63 it cannot be said to be one-pointed.
;if.: Whether or not vitakka-vicara can coexist with ekaggata in
:f!rst jhiina clearly has a lot to do with how the terms themselves
::are understood. It is admittedly difficult to be sure exactly what
and processes the terms used in early Buddhist psychology
;actually referred to, but as already indicted, change in the mean- .
;ipg of terms is insufficient to resolve the problem. Even as "initial
vitakka retains a discursive component. The change
r!ll meaning of vicara noted above, while it may explain the
tf","
[,.
94 JIABS VOL. 12 NO; 2
interpolation of jhiina la in the Abhidhamma series, does nothin
J
to eluddate the problem of first jhiina. Nor can change in
meaning of ekaggatii account for the presence of this factof"
together with vitakka in first jhiina. Even Buddhaghosa did not'
accept that ekaggatii as a factor of first jhiina was some weak
form of attention such as was said to characterize less developed
states of consciousness, including access-jhiina.
64
,
In the Sutta and Abhidhamma accounts we have two different'
descriptions of what is purported to be the same mental state.
But because the descriptions are different, different interpreta_
tions are possible. First jhiina, is usually interpreted as a state
of deep concentration, achievement of which is beyond the cap,}
acity of all but the most advanced meditators. This interpretation
is based on the Abhidhamma account. From the Sutta account,;
however; a rather different interpretation is required. Firstjhiind
in the Sutta account is the stage before mental one-pointedness
is established. Rather than being a state of deep concentration:'
therefore, it seems to be a preliminary, stage preceding a series
of such states (the higher jhiina and arupa jhiina). In the Suttd
account, vitakka-viciira and ekaggatii do not coexist precisely be2:
cause it is through the elimination of vitakka-viciira in the
ition to second jhiina that one-ness of mind is attained. The firsl
jhiina of the Suttas is evidently a state that can be readily attained
by anyone who has practiced right mindfulness, a state
many wandering samar]as would have been conversant with. '.'
In principle, it ought to be possible to test the Sutta descrip:
tion of the transition from first to second jhiina through intro':
spective analysis.
65
Gotama learned the lower jhiinas from his
earliest meditation masters, practiced them even as a child, and
enjoined his disciples to do the same.
66
It should be possible;
therefore, for present-day practitioners of meditation to apply
similar techniques to attain similar elementary concentrated
states, and thereby test the accuracy of the textual description(
Though introspective analytical reports "of the kind developed,
in modern cognitive psychology67 could not be taken as in them;,
selves providing conclusive empirical evidence in support of
Buddhist claims for the effectiveness of meditation techniques,
they would lend strong suppor( to textual accounts so con"
firmed.
68
"
It is perhaps debatable whether this method of empiricar
JHANA AND BUDDHIST SCHOLASTICISM 95
would be applicable !or stages the Buddhist
fliieditatIve path, as the Jhiinas or the Knowl-
however Its IS .much for the
IntrospectIve analytIcal of
states could be checked agamst the findmgs of
tiikt-based approaches such as those which make it possible to
tllistinguish an earlier "primitive" Buddhism which .might be
l:scribed to Gotama himself from later accretions through the
rd1.ting of texts on the basis of language or content;69 or through
Wotm-criticism of the kind pioneered by Biblical scholars.
70
One way to obtain empirical verification of whether or not
can possibly coexist with vitakka-viciira in. first jhiina
be to conduct a survey of Buddhist meditation masters
"$D"" ..
!from which presumably a clear consensus would emerge which
!.!$ould resolve the contradiction between the Sutta and
accounts offirstjhiina. There are, however, practical
in the way of conducting such a survey. Those under-
W:loing training in Buddhist meditation are usually under strict
not to discuss their experiences in the presence of
but their meditation master. Masters themselves are likely
It; be reluctant to advance any claim to have achieved higher
states, if only for fear of the negative effect such a
is believed to have on spiritual progress toward nibbiina,
of the skepticism it might well provoke. However, meditation
might be less reluctant to report on their introspective
of lower meditative stages..:.-especially if this took
form of commenting upon published accounts by non-Bud-
applying Buddhist techniques.
71
Descriptions of concentration practice by non-Buddhists
'provide a possible alternative means of verification, though only
:p1jma facie evidence could be so adduced in support of one
description or another. Such prima facie evidence is avail-
in fact, to anyone willing to embark on a course of elemen-
;fary concentration practice. The immediate goal of such practice
achieve mental one-pointedness through concentrating at-
upon some object of perception, such as the tactile sen-
of the breath at the right nostril. 72 The most notable
of this concentrated state, when one reflects upon
experience, is that the chatter of thought is temporarily
Most people have probably experienced this phenome-
96
JIABS VOL. 12 NO.2
i
non on occaSIOns when they have become totally engrossed fo
sensory exam pIe when listening intently
mUSIC, or when III the beauty of.a sunset: N?rmalIy,'
flow of thought qUIckly resumes, but wIth It is
sIble to extend the concentrated state to endure mmutes at it
time. Even this elementary experiment in concentration indi!
cates that mental one-pointedness cannot coexist with
thought. Phenomenological analysis thus confirms what lOgi2!
would lead us to expect, namely that the Sutta description ih4
which first vitakka-viciira and ekaggatii do not coexist in firstjhiind!
is the correct one.
, 'Now it is just conceivable that confusion over the
ition of firstjhiina could have arisen from exegesis of the
of the SaT{tyutta-nikiiya quoted above in which the meditator
instructed to make the mind one-pointed in firstjhiina. Accord>:
ing to the Sutta-pitaka account, the meditator must suppress air
discursive thought in order to attain second jhiina. This would
require that preliminary attempts to establish one-pointedness'
be made in firstjhiina.
73
In this sense, ekaggatii could
be said to occur here. Even so, this one-pointedness of mind:
would never coexist with discursive thought. During those short'l
periods when was achieved, discursive thought!
would necessarily stop. One-pointedness of mind of significant:
duration could only be said to be present when discursive
1
thought no longer disrupted the concentrated state. If firstjhiind';
is characterized by the presence of discursive thought, it can hardly;
also be characterized by mental one-pointedness, even if in thfi
course of elementary concentration practice discursive thoughtl
were to be momentarily restrained. Only when
thought is fully suppressed through concentration could
ness of mind" be termed a factor, that is, a permanent
teristic of the state attained-and that is said to occur only iii)
second jhiina. Thus, on the basis of the logic of definition,
would have to conclude that the Abhidha,mmic first jhiina wa(i:
inaccurately characterized, and that the Sutta description shoulq,
be preferred.
If logic and introspective analysis of concentration practic(!,i
both confirm the Sutta description of first jhiina, and
exegesis and change in the meaning of terms cannot
the presence there of ekaggatii as a characterizing factor,
"'\i;:i
,",.'i'
JHANA AND BUDDHIST SCHOLASTICISM 97
us
t
ask why all schools of Buddhism have accepted the later
idhamma account in preference to what was in all probability
2
ta
ma's own earlier description. This is not the same as asking,
+e fundamentally, how ekaggata came to be included in first
c.:.' a in the first place. Once ekaggata had become included in
canon, Buddhists very naturally accepted the new descrip-
11 without question. The Abhidhamma-pitaka, as one of the
ee "baskets/, not only is scripturally as authoritative as the
< -pitaka; it even purports to be more analytically exact. If
" Abhidhamma says ekaggata is present in first jhana, no school
contradict it. To elaborate the doctrine is one thing; to
IDe issue with the most authoritative texts on Buddhist analyt-
psychology would be quite another.
Ii.' We should. not be that once ekaggata had come
11 be m jhiina, was by all schools of
What IS noteworthy IS that thIS development neces-
led to a reinterpretation of first jhiina. Once ekaggata had
included, firstjhana could hardly be taken to be an elemen-
1& stage in practice. it came to be con-
as far more exalted whICh monks could
to attam-a VIew that would have been remforced by the
that gradual decline of the Dhamma was inevitable.
The interesting question, however, is not why believing -
1'B'llddhists accepted the Abhidhamma account once ekaggata had
a factor of first jhiina, but rather how it came to be
fffiduded as a factor in the first place. As we have seen, prima
evidence that mental one-pointedness and discursive
cannot coexist makes the possibility that the change in
was based on more refined introspective analysis
I!fhlikely. It is possible that more refined introspective analysis
W<iS responsible for drawing the distinction found in the Abhi-
J!-amma fivefold series between vitakka as the "initial application"
and vicara as "sustained" thinking about it. This
pistinction between two modes of thought is one which most
people would be familiar with, and could hardly have been
by those responsible for the kind of psychological
we find in the Abhidhamma. Anyone who has thought
geeply about anything knows that focusing attention on content
prevent the arising of random mental images.
74
The description of jhiina la in the fivefold series as charac-
t:
98 ]IABSVOL.12NO.2
terized by both sustained thought and mental one-pointedness
could, therefore, conceivably be defended as phenomenologi_
cally accurate on the grounds that sustained thought constitUtes
a form of concentration, that concentrated focus on the COntent
of thought constitutes mental one-pointedness. However, this
description of jhiina la could still be questioned on the grounds
that ekaggatii as one-pointed concentration actually eliminates
all thought. In any case, while it may well be that more refined
Abhidhammic introspective analysis led to differentiation be-
tween vitakka and viciira, and even to the inclusion of ekaggata
as a factor of the additional inserted jhiina la, this cannot with
any plausibility explain how ekaggatii came to be considered to
coexist with vitakka in first jhiina.
N either changes in the meanings of words, nor refinements
in psychological analysis, can provide, with any plausibility, an .
explanation for the discrepancies associated with the Sutta and
Abhidhamma descriptions of the J"hiinas. Nor, as indicated above,
did references in the Sutta-pitaka provide precedent for the in-
clusion of ekaggatii in first jhiina or the insertion of jhiina la, since
the relevant sections did not form part of the early corpus of
memorized texts upon which early Abhidhamma formalization
would have been based.
75
On the contrary, it is much more
likely that both references to ekaggatii in first jhiina and the
threefold classification of samiidhi were products of Abhidham-
mic scholasticism only later interpolated into the canon.
This conclusion would be further strengthened if it could
be shown how the earlier Sutta description came to be altered
to produce the Abhidhamma version. Unfortunately, conclusive
historical evidence of this kind simply does not exist. What the
historian can do, however, is attempt to construct a hypothetical
account of how the alteration might have occurred, given what
we know of the historical development of early Buddhism, and
offer some assessment of the likelihood and coherence of such
an account.
The following explanation for why the Abhidhamma listsfive
factors in first jhiina takes particular account of Buddhist scho-
lastic mentality. Let us begin with the jhiina factors. These are
known collectively as the jhiinarigas, a term which does not o c c ~ r
in the Sutta-pitaka. Together with the concept it connotes, thIS
term is a product of Abhidhammic scholasticism.
74
It seems
JRANA AND BUDDHIST SCHOLASTICISM 99
faikely- that for the jhiinas, the characterizing factors first listed
rwere those which necessarily had to be overcome in moving
Wsuccessfully to higher jhanas (see table 1). Ekaggata would have
included as a factor gained, not lost, because of its promi-
Jnence in characterizing the higher jhiinas and the emphasis
rJplaced i: by practicing With the of
;('Vitakka-vzcara mto two factors, first Jhana, the stage attamed
overcoming the five hindrances, was characterized by
factors. But for the scholastic mind, there existed an uncom-
asymmetry where five hindrances were juxtaposed with
jhiina factors. Five hindrances needed to be paired with a
of five factors, a compelling reason for discovering an addi-
factor in first jhiina-and the most readily available
comparison of tables 1 and 2 shows) was ekaggata.
f. Scholastic concern over the relationship between the hin-
[drances and thejhiina factors provides the key to understanding
thow the discrepancy between the Sutta and Abhidhamma descrip-
of the jhiinas is likely to have arisen. A direct relationship
first stated in theMahavedalla-sutta, already referred to. There,
replies as follows to the question how many "factors"
abandoned and how many possessed in first jhiina :
Your reverence, in regard to the first meditation, five factors are
abandoned, five are possessed: if a monk has entered on the first
meditation, desire for sense-pleasure is abandoned, malevolence
is abandoned, sloth and torpor are abandoned, restlessness and
worry are abandoned, doubt is abandoned, but there is initial
thought and discursive thought, rapture and joy and one-poin-
tedness of mind. Thus, your reverence, in regard to the first
meditation, five factors are abandoned, five factors are pos- .
sessed.
75
tJt is never explicitly stated in the Tipitaka that each of the five
Yhiina factors is instrumental in overcoming a specific hindrance.
states that a direct one-to-one correspondence
t;between the five jhiina factors and the hindrances is given in
nhe Petakopadesa.76 But in this he is mistaken; all we in fact find
:in the Petakopadesa is a statement that the "five-factored medita-
)tion [jhana]" is the "opposite" of the five hindrances." N everthe-
tiess, by the time Buddhaghosa was writing, these equivalences
;'were well established: one-pointedness (ekaggata) was said to be
100 JIABS VOL. 12 NO.2
opposed to sensory desire (kamacchanda),joy (Jiiti) to malice (vya_
. pada) , initial application of mind (vitakka) to sloth and torpor
(tMnamiddha), bliss (sukha) to distraction and remorse (uddhacca_
kukkucca) , and sustained application of'mind (vicara) to dOUbt
(vicikiccha).78 .
. Now, some of these e.quivalences seem quite inapprop_
nate.
79
For example, one mIght have expected ekaggata to neut.
ralize uddhaccakukkucca (distraction and remorse) rather than
sensory desire. As for vitakka, it is hard to see how it could be
thought of as neutralizing anything. Apologists explain that the
vitakka which counters sloth and torpor is of a special
Nor does it seem likely that vicara would neutralize doubt. Or!'
the contrary, doubt could actually be encouraged by sustained
thought. Here apologists claim that viciira counteracts doubt
only when it is "directed to jhiina.,,81 ....
The Vimuttimagga provides an even more bizarre
of the lists of one-to-one correspondences so dear to the schoJ
lastic mind. There the relevant passage states: "The hindrances
are overcome by the perfection of the five jhana factors. The
overcoming of the first hindrance is the first meditation, jhiina:
Thus the overcoming of the five hindrances results in five medi-'
tations,jhiinas.,,82 The five hindrances are not overcome by five'
jhana factors in firstjhana. Rather, the hindrances are
as the jhana factors are lost in moving through the series of five
jhanas. This account is obviously inconsistent with the
tion of the first jhiina as characterized by separation from
wholesome states (all five hindrances), and makes no sense in'
terms of the jhiinas as a sequence of ever more concentrated.
mental states. .""
. The Vimuttimagga provides an excellent exam pIe of two rein-:
forcing scholastic tendencies-to draw up neat and regular lists.
wherever possible, and to equate lists so that individual items.
in each are paired in symbolic relationship.83 Both tendencies
are already evident throughout the later sections of the Tipitaka.
It was this penchant in Indian scholasticism (for it is not foun<t
only in Buddhist writings) for composing lists and drawing symj;
bolic parallels that best accounts for both the inclusion of ekag;{
gatii in first jhana, and for the insertion of jhiina lao . '0.
The description of the jhanas in the Sutta-pitaka
the presence of ekaggatii only in second jhiina, but it is
JHANA AND BUDDHIST SCHOLASTICISM lOl
understood as continuing to characterize third and fourth
f'hii,nas. If ekaggata is included in jhanas 3 and 4, the asymmetrical
kab1e 1 results. It would be natural for the scholastic mind to
:"complete" the table by including ekaggata in first jhana. Sub-
.vitakka-vicara was divided into two separate factors
'almost certainly in order to "match" the previously existing set
five hindrances with the necessary number of jhana factors,84
tather than as a result of more refined introspective analysis.
irogether, these scholastic exercises would have given rise first
;to the Abhidhamma fourfold jhana setout in Table 2, and then
t() the even neater and still more formally satisfying fivefold
'irrangement of Table 3.
',,' That the inclusion of a jhana stage in which vitakka is missing
but vicara retained probably resulted from scholastic formalizing
1tather than introspective analysis is further indicated by the
:ionflicting descriptions of this jhana in the Dhammasanga1'}i and
.. fhe Vibhanga remarked on above, and by the treatment of the
:Jhanas in the Katha-vatthu
85
where the "Theravadins". are said
Ito argue, against adherents of other schools, that no intervening
stage exists between first and second jhanas in the Sutta account. .
yet in the Theravadin Abhidhamma the fivefold listing clearly
'does include jhana la as just such an "intermediate stage." It
appears that by this time the jhanas had for some monks hecome
,110 more than another "point of controversy."
" The suggestion that the Abhidhammic description of first
Jhiina resulted from scholastic elaboration rather than constitut-
ing a phenomenologically accurate reporting of an attained
:meditative state is unthinkable only for those who approach the
Abhidhamma as sacred scripture or with exaggerated deference
for the wisdom of the arhats. In fact, we have strong historical
evidence for the development of Buddhist scholasticism. Soon'
after Gotama's death, the sangha changed from being a band of
wandering mendicants to become a settled monastic order. 86 At
the same time there developed an immense body of oral litera-
lure, all of which had to be memorized until the canon was
written down, some time after the reign of Asoka. To memorize
these lengthy records groups of bhikkhus were responsible for
different sections. Dutt describes the process:
Each group would then memorize and also specialize in its own
102
JIABS VOL. 12 NO.2
section, not as mere reciters (bhiinakas), but as professors, ex-
positors, commentators-in short as custodians of both the texts
and their true meaning.
87
Thus, we find reference to Dhammakathikas (expounders of
Dhamma), Vinayadharas (experts in the Vinaya), and Suttantikas
(specialists on the Suttas). Another group were designated the
Jhiiyins, literally those who practice the jhiinas. At Anguttara iii.
355 the tension is revealed that existed between the Jhayins
(which Hare translates as "musers") and Dhammayogas (Hare's
"Dhamma-zealots", followers of the Dhamma as texts to be
studied).88 Each group apparently had been criticizing the other: .
each considered its particular way of practice to be the only true
way to nibbiina. The lesson of the text is that such disputes should
end: each group should respect the methods of the other, for
both lead to the same goal, though few enough of either group
will attain it.
Two things should be noted about this text: first, that such
mutual criticisms were being voiced; second, that there had
already evolved an influential group of monks seeking to ap-
prehend the supreme reality by means of the intellect,&9 rather
than by the meditative techniques pioneered by the Buddha. A
contest was taking place for the soul of the sangha between on
the one hand, the Dhammayogas, those "puffed up, proud, excit-
able fellows, mouthy speechifiers, forgetful of mindfulness, lack-
ing self-possession and composure, with their thoughts a-wander
and their sense-governance rude," and on the other hand the
Jhiiyins, those who had "touched with the body the deathless
state.,,90 In this contest, the Jhiiyins lost. .
Further evidence for a steady decline in the practice ofjhiina
in the sangha comes from the Vinaya-pitaka. As C.A.F. Rhys
Davids points out, the Vinaya contains few references to the
jhiinas as a system of meditation a monk should pursue, and
only four references to Jhiiyins and their special needs. She con-
cludes that the practice of jhiina had already seriously declined
for:
there is no doubt that had the Sangha, during the centuries when
the Vinaya was growing by accretions, held Jhana in its original
worth, it would have produced a disciplinary chronicle glowing
with Jhana atmosphere throughout.
91
JRmA AND BUDDHIST SCHOLASTICISM 103
From the admittedly fragmentary evidence that is available,
!t is dear that the Jhayins within the early sangha soon became
J
it Il1inority, a trend undoubtedly accelerated by the rapid expan-
sion of the sangha under Asoka.
92
As meditators, they were
'probably as unconcerned with the organization and administra-
tion of the sangha as they must have been with speculative de-
bates on aspects of Dhamma, or the compiling and memorizing
of texts. This was left to the scholastically inclined.
Long before the time of Buddhaghosa, the Buddhist sangha
had become predominantly a worldly organization, concerned
above all with its own preservation, with maintaining its popular
'a.ppeal and princely patronage. By t h a ~ time the meditative tra-
dition may well have been reduced to httle more than an eccen-
tric group of recluses. 93 Since most textual commentators stood
squarely in the Dhammayoga tradition, it seems likely that most
were not Jhayins but scholars and exegetes who elaborated scholas-
tic discussions of the path while lacking acquaintance with the
:higher stages of meditative practice. It is not hard to see how, in
the hands of such monks, the scholastic equating of five jhana
Jactors with the five hindrances in firstjhana might have occurred.
V. Conclusion
This paper has argued that the two descriptions of first
jhana, the four-factor Sutta listing and the five-factor Abhidhamma
listing, are contradictory and cannot be reconciled. Attempts to
achieve such a reconciliation, both in the commentarialliterature
and by modern scholars, are unconvincing. Textual analysis
alone suggests that the inclusion of ekaggata in first jhana is
logically incompatible with the presence of vitakka, even given
. later modifications of meaning of both terms. Elementary con-
centration practice confirms that coexistence of ekaggata with
.vitakka is at least phenomenologically questionable. There is a
strong prima facie case, therefore, for supposing the later
Abhidhamma description to be invalid, and the Sutta description
to be the correct one.
This conclusion suggests that by the time the Abhidhamma
texts came to be written down, a high degree of scholasticism
characterized Buddhist thinking. Evidence of the incorrect de-
scription of firstjhana thus supports C.A.F. Rhys Davids' conclu-
104 JIABS VOL. 12 NO.2
sion that various "psychic" states are described in the
in such a way as to indicate that "the compilers had not themselves:;;
at first hand of what they .were recording. ,,94 By
mg the Abhzdhamma texts as canonIcal, all later
were faced with the problem of explaining away evident dis:";
crepancies. Rather than do this, one would be better advised u;i;
treat Abhidhamma texts and the commentarial literature With";;
more critical suspicion than has usually been the case, even;
where the subject matter is the descriptive psychology of those;
altered states of consciousness that the texts purport to reveal. '$;
Two wider implications should therefore be drawn front
the above analysis, touching upon both the historical
ment of early Buddhism and the methodology of Buddhist
studies. It would appear that the gap between those who spent
their time in the sangha practicing meditation (the Jhayins) and.
those who discussed and commented upon the Dhamma (the'
Dhammayogas) was already wide and deep by the time th/'
Abhidhamma-pitaka had taken shape. In part, no doubt, this wat
due to different abilities and interests. But it was probably also'
exacerbated by the form of esoteric transmission by which the.:
meditative tradition was communicated to adepts. Divorced asci
they most probably were from experience of those states of
consciousness attained through application of advanced medita=!
tive practices, Buddhist scholastics pursued their own course of
elaborating increasingly complex lists of categories such as we
find in the Abhidhamma. What they have to say about altered,
states of consciousness should therefore be treated with caution .. '!
The second implication is that textual contradictions must
be recognized as such. They must not be dismissed on the
grounds that accounts of experiential states of mind "elude mere
intellectual treatment.,,95 Contradictions arise as historical de
J
velopments and require historical explanation. We cannot as!
sume meditative practices to have remained constant during the .
millennium from the time of Gotama to that of Buddhaghosa/
any more'than we can assume textual compilation over this
period to have been unaffected by the divisions and debates,;
that were occurring both within the sangha with the rise of the
Mahayana schools, and between Buddhism and resurgent
duism. As scholars, we must be even more critical than we have'
been in studying the texts. By so doing, we will be in a position
JHANA AND BUDDHIST SCHOLASTICISM 105
to further light on shaping
IlJid to to a. better understandmg of medita-
itti\re In thIS way, s.cholarly study may e::cplicate stages
iiAt
he
BuddhIst path to enhghtenment of practIcal benefit to
iftthodern day meditators.
iii""
IROTES
1. This question has been dealt with by Rod BuckneUin "The Buddhist
i5ifath to Liberation: an analysis of the listing of stages" ,Journal of the International
(t1.hociation of Buddhist Studies 7 (1984), pp. 7-40. I gratefully acknowledge
valuable criticisms of successive drafts of this paper. .
2. Cf. Mircea Eliade, Yoga: Immortality and Freedom, 2nd ed. trans. by
R. Trask (London: Routledge and Kegan Paul, 1969), p. 250.
il;t
i
3. Recen.t studies of include: Cousins, jhana: its
fpature and attamment accordmg to the Pah sources," Relzgwn 3 (1973): 115-
rq31; Winston L. King, Theravada Meditation: The Buddhist Transformation of
(University Park and London: Th,; Pennsylvania State University Press,
1}980), pp. 47-48; Donald K. Swearer, Control and Freedom: The structure
'6fBuddhist Meditation in the Pali Suttas," Philosophy East and West 23 (1973)
and Paul Griffiths, "Buddhistjhana: a form-critical study," Religion
"P(1983): 55-68.
Ili" 4. Dii.313.
5. The relationship between the two paths, of con.centration and
fJsstill a matter for debate among scholars. See Paul Griffiths, "Concentration
Insight: The Problematic ofTheravada Buddhist Meditation-Theory,"Jour-
fjia,l of the American Academy of Religion 49 (1981): 605-624.
6. Cf M i. 22.
7. (1) vivicc'eva kiimehi vivicca akusalehi dhammehi savitakkaT{t savicaraT{t
yaT{t pUisukhaT{t patht;LmaT{t jhanaT{t upasampajja viharati.
(2) vitakka-vicaranam vupasama ajjhattaT{t sampasadanaT{t cetaso ekodibha-
. avitakkaT{t avicaraT{t samadhijaT{t pUisukhaT{t dutiyaT{t jhanaT{t upasampajja viha-
li. This description occurs frequently, e.g., at D i.182ff.
8. Listed at, e.g., D i. 71-73, D iii.49, etc. The DhammasangaTji (hereafter
lists six hindrances, including ignorance (# 1152).
9. T.W. Rhys Davids and William Stede, Pali-English Dictionary (Lon-
Luzac, 1959), p. 620.
10. M i.114f. cf. Rune E.A. Johansson,. The Dynamic Psychology of Early
(London: Curzon Press, 1979), pp. 185-190 for an analysis of the
," ...... of vitakka.
II. Pali-English Dictionary, p. 615
12. Ibid., p. 620.
13. Vibhanga 263-266, (hereafter Vibh), translated as The Book of Analysis
Ashin Thittila (London: Luzac, 1969), pp. 344-348.
106 JIABS VOL. 12 NO.2
14. Cf. the list .of of in Petakopadesa
translated as The Pztaka-Dzsclosure (London: Luzad
1964), p. 186. In the Anupada-sutta, Sanputta Identifies numerous other facto"!;
of the jhiinas (M iii. This list extended still further in
Abhzdhamma and later commentanes. Cf. Narada Maha Thera, A Manual ifi
Abhidhamma (Kandy: Buddhist Publication Society, 1968), pp.
15. At D ii. '217, also M 1. 301, samiidhi is defined as cittass' ekaggatii.
. also Dhs 11 and 24 where cittass' ekaggatii and sammiisamiidhi are defined r '2
identical terms. (Cfalso Dhs # 287 and 291.)n
j
16. Vibh #575.
17. Dhs #424 (translated by Caroline A.F. Rhys Davids as A
Manual of Psychological Ethics, 3rded. (London: Pali Text Society, 1974),
. 18. The Expositor, vol II, p. 345 (translation of the Atthasiilini by Pe Maung':
Tm (London: Luzac, 1920). ...iJ
19. Dhs #556. Cf. Atthasiilini p. 293. Bhikkhu J. Kashyap in The;
Abhidhamma Philosophy (Delhi: Bharatiya Vidya Prakashan, 1982), p. 46, states!!
that ekaggatii varies in strength from "very feeble" to "fully steady" in
jhiinas. 'J
20. Atthasiilini #259. Cf. A Buddhist Manual of Psychological Ethics, p. 105:1
21. # 147 (hereafter Vism) Translated as The Path of
by Pe Maung Tm (London: Luzac, 1923).
22. Ibid.,rj
23. Pali-English Dictionary,p. 620.
24. Vibh #257 The Book of Analysis, p. 335. In A Buddhist Manual ij'
Psychological Ethics Caroline A.F. Rhys Davids states that, in her opinion, vitakkli;
is a "distinctively mental procedure at the inceptiqn of a train of thought, thJ)
,deliberate movement of voluntary attention" (p. 8, note 1). Vicara is "the'
movement and maintenance of the voluntary thought continuum," .something'
which includes the senses of investigation, analysis, and discursive thought
(p. 9, note 4). It
25. Atthasiilinz#1l4-1l5; The Expositor, pp. 151-152. Cf. also Vism
pp. 164-165.
26. Pali-English Dictionary, p. 620 A later text, the Abhidhammattha-Sanga::
ha, adds little or nothing to our understanding of the meaning of the terrm
l
under discussion. Vitakka in first jhiina is said to be directing of the mind to
the "after-image," a meaning apparently compatible with the presence of
ekaggatii, while the presence of ekaggatii in mundane states of consciousness
is explained by denying that in such cases it connotes "concentration." C[
Compendium of Philosophy (London: Luzac, 1972), p. P8, note 5. Such modifi,.
cations of meaning constitute attempts to explain away, rather than
the presence of both vitakka and ekaggatii together in first jhiina in the
Abhidhamma.
27. Vibh #264; The Book of Analysis, p. 345.
28. Dhs #168.
29. Henepola Gunaratana, A Critical Analysis of the Jhiinas in the Theraviida
Buddhist Meditation, Ph.D dissertation, American University, Washington 1980
JHANA AND BUDDHIST SCHOLASTICISM 107
'[\e.214-215, suggests an editorial error may have occurred in one or the
description of jhiina la but is unable to deci_de which might be correct.
.. ......... 30. Vism #147. The Path of Purity, p. 170. Na1).amoli translates the last
somewhat differently: "for the intention with which the Blessed One
the summary [i.e., the Sutta version] is the same as that with which he
the exposition that follows it [iein the Abhidhamma]." The Path of Purifi-
fi'tion vol. 1, p. 153 .
... ca 31. Vism #139, ThePath of Purity, p. 161.
";2 32. Paravahera Vajiraiia1).a Mahathera, Buddhist Meditation in Theory and
i.Practice (Colombo: M.D. Gunasena, 1962), p. 38.
;" 33. Gunaratana, A Critical Analysis of the Jhiinas, p. 105.
, '34. Ibid., p. 196.
(f' 35. Ibid., p. 143.
36. Ibid., p. 174. Cf Vism #126.
37. S iv. 263, translated as The Book of Kindred Sayings by F.L. Woodward
(London: The Pali Text Society, 1927), p. 180.
'r' 38. Ibid. .
39. A.K. Warder, Indian Buddhism (Delhi: Motilal Banarsidass, 1970),
Bucknell, "The Importance of Pali/Chinese Comparisons in
,studies of Early Buddhist Doctrine," paper presented at the IXth annual
.;conference of the Australian Association for the Study of Religions, Canberra,
A,ugust 1984, p. 4.
;;".41. Pande agrees that the "Sayings about Moggallana" are almost cer-
tainly late. G.C. Pande, Studies in the Origins of Buddhism 2nd ed. (Delhi: Motilal
;Banarsidass, 1974), p. 229.
42. Mi. 294.
,... 43. Pande concludes that the whole sutta is a late composition. See Pande,
.prigins of Buddhism, p. 134.
44. Taisho 26 (211). See Bucknell, "The Importance of Pali/Chinese
'Comparisons," p. 4.
45. M iii. 25-29.
46. The translation of chanda as "desire" is LB. Horner's. A better trans-
lation might be "impulse (towards something)." But in any case chanda should
be overcome in first jhiina.
47. Pande, Origins of Buddhism, p. 318, calls this an early Abhidhammic
text.
48. Vimuttimagga, translated as The Path of Freedom by N.R.M. Ehara,
Soma Thera and Kheminda Thera (Colombo: D. Roland D. Weerasuria, 1961),
p. 46. The "second meditation" referred to is our ''jhiina la."
49. Atthasiilini 179, The Expositor, vol. 1, p. 240.
50. Pali-English Dictionary, p. 286.
51. Phra Maha Singathon Narasabho, Buddhism: A Guide to a Happy Life
(Bangkok: Mahachulalongkornraja-vidyalaya, 1971), p. 66.
52. Gunaratana, A Critical Analysis of the Jhiinas, p. 211 Gunaratana illus-
trates the relationship of the two systems through the metaphor of two moun-
tain climbers who climb the same mountain using different ascent stages (pp.
211-212).
108 JIABS VOL. 12 NO.2
53. Pande, Origins of Buddhism, p. 534.
54. For .example, Edward J. Thomas does not mention fivefold jhii:na in
either The Life of Buddha (London: Routledge and Kegan Paul, 1949), or The
History of Buddhist Thought (London: Routledgt; and Kegan Paul, 1951).
does Winston King, in Theravada Meditation. Neither does Nalinaksha Dutt
Early Monastic Buddhism (Calcutta: Oriental Book Agency, 1960); but whereas"
King quotes the Sutta account of first jhana, Dutt describes the fourfold
Abhidhamma account with ekaggata included (p. 188). .
55. At D iii. 219; D iii. 274; M iii. 162; S iv. 360, 363; A iv. 300-301.
56. Buddhaghosa specifically refers to these three forms of samadhi.
Atthasalini: #179.
57. Pande, Origins of Buddhism, p. 115.
58. Taish6 i. 50. See Bucknell, "The Importance of PalilChinese Com-
parisons." p. 5. .
59. Ibid. Cf. TaishO i. 53.
60. Ibid. Cf. TaishO i. 538.
61. See Warder, Indian Buddhism, pp. 7-9.
62. For believing Buddhists, the problem may not seem so clear cut.'
The Indian tetralemma permits the two further alternatives of "both ... and
... " and "neither ... nor .... "
63. S ii. 94.
64. Vism #147.
65. Cf. Roderick S. Bucknell and Martin Stuart-Fox, The Twilight Lan-
guage: Explorations in Buddhist Meditation and Symbolism (London: Curzon Press
and New York: St Martins Press, 1986). . ,
66. CfM i. 246; D i. 74-76; Mi. 347.
67. As developed in, for example, Alan Richardson, The Experiential
Dimension of Psychology (St Lucia: University of Queensland Press, 1984). .
68. Cf. Frank Hoffman, "The B.uddhist Empiricism Thesis," ReligioUs
Studies 18 (1982), pp. 151-158.
69. Cf. Pande, Origins of Buddhism. Also Kogen Mizuno, Primitive Bud-
dhism, trans. by Kosho Yamamoto (Ube: The Karin Bunko, 1969).
70. Griffiths, "Buddhist Jhfma." Griffiths analyzed the 86 occurrences'
in the Sutta-Pitaka of the stereotype description of the four jhanas in order to,
define the position of thejhanas in alternative soteriological paths. His
sions, though of interest; do not, however, go far enough, for the form-critical
method fails to take account of historical context. We obtain no hint from
Griffiths' study as to the origins of these different paths and the goals to which
they lead. Nor is any light shed on the historical circumstances that led to the
incorporation of variant paths in the Pali Canon. " '.
71. For examples of descriptions of meditation, see R.N. Walsh, "Initial
meditative experiences," parts I & II, Journal of Transpersonal Psychology 9
(1977), pp. 157-192 and 10 (1978), pp. 1-28; Rod Bucknell, "Experiments
in Insight Meditation," Australian Journal of Transpersonal Psychology 3 (1983),
pp. 96-117; and chapters 3 and 4 of Bucknell and Stuart-Fox, The Twilight
Language.
72. Cf. the Anapanasati-sutta M iii. 82.
JRANA AND BUDDHIST SCHOLASTICISM 109
'73. Gunaratana suggests that the term samiidhija'l'{t (literally bDrn Df con-
l:fentratiDn) in the descriptiDn Df secDnd jhana cDuld be understDDd as meaning
'that secDnd jhana is bDrn Df the cDncentratiDn applied in first jhana
A of the jhanas, p. 175). he equates
hamjuj)u,ja'l'{t WIth ekodzbhava as synDnymDus terms remfDrcmg the ImpDrtance
one-pointedness in secDnd jhana. Ibid., p. 178. FurthermDre, the analDgy
:'with first jhana, as vivekaja1'{1. makes it clear that the reference Df
twords in "-ja'l'{t" is to. the factDr that effects the transitiDn, and.not to. a factDr
the preceding jhiina.
i1. 74. Such meditatiDn is practiced in all the majDr wDrld religiDns as, fDr
:e;cample, in Christian meditatiDn Dn the cross.
75. See notes 37 to 47 above.
t 76. Vism # 190.
f 77. Mi. 294-295, translated as The Middle Length Sayings by LB. Horner
Luzac, 1954), p. 354 .
. ', 78. Vism #141. The Path of Purification, p. 147.
79. Petakopadesa, 161. The Pitaka-Disclosure, p. 220
", 80. E.g., at Vism #141. The Path of Purification, p. 147. See also. The
if,xpositor, p. 221 Gunaratana argues that the hindrances were limited to. five
of many such pDssible "factDrs Df abandDning" because there were Dnly
:)ive jhana factors with which to. cDrrelate them (Gunaratana, A Critical Analysis
for the jhiinas, p. 58). I argue just the DppDsite, that five jhiina factDrs were
i,tequired because there already existed a well knDwn far mDre ancient list Df
1five hindrances which were said to. be DverCDme in first jhana. Cf. fDDtnDte
below.
81. The jhiina factDrs are nDt alDne in counteracting the hindrances. At
"S v. 105-106 there Dccurs a set of five "wise cDnsideratiDns" which are alSo.
said to. eliminate the five hindrances.
82. Gunaratana, fDr example, says the vitakka which CDunters SIDth and
torpDr is "Df a high quality and specialized functiDn"! Gunaratana, Critical
'Analysis ofthejhanas, p. 151.
83. Ibid., p. 151. The idea that viciira cDuld be "directed to. jhana" seems
part Df an attempt to. provide viciira with a new meaning, as a fDrm Df CDncen-
tratiDn, in Drder fDr it to. be cDmpatible with ekaggatii.
84. The Path of Freedom, p. 93.
85. FDr a detailed discussiDn Df this develDpment see Bucknell and
Stuart-FDx, The Twilight Language, chapter 6.
. 86. The five hindrances are fDund in the earliest sectiDns Df the Sut-
tap#aka, e.g., at D i. 71-73, in the Siimaiiiiaphala-sutta'. See Pande,. Origins of
1Juddhism, p. 114.
87. Katha-vatthu, translated as Points of Controversy by S.Z. 1\ung and
C.A.F. Rhys Davids (LDndDn: Luzac, 1960), pp. 327-330.
88. In the summary which fDIIDWS, I have drawn heavily Dn the wDrk
ofSukumar Dutt. See bDth The Buddha and Five After-Centuries (LDndDn: Luzac,
1957) and Buddhist Monks and Monasteries of India (LDndDn: GeDrge Allen and
Unwin, 1962).
89. Dutt, The Buddha and Five After-Centuries, p. 92.
110 JIABS VOL. 12 NO.2
90. A iii. 355. The significance of this text was remarked upon by Louis
.de la Vallee Poussin in "Musila et Narada: Ie chemin du Nirvana," Melanges
chinois et bouddhiques 5 (1937): 189-222. .
trend culminating in the method of,Nagarjuna .. It. seems likely
that thIS mtellectual trend was encouraged by the BrahmamstIc concept of
jnana-yoga.
92. A iii. 355, translated as The Book of Gradual Sayings, vol 3 by E.M.
Hare (London: The Pali Text Society, 1934), p. 252.
93. C.A.F. Rhys Davids, "Dhyana in Early Buddhism," Indian Historical
Quarterly 3 (1927), pp. 695-696. I cannot, however, accept the conclusion of
this paper that jhana had value over and above that of a state of mental
preparation. Rhys Davids criticizes Ananda (A ii. 195) for takingjhana to be
"pure and simple mind practice," and one of four "factors for utter
tion." But these samadhi, panna, vimutti-together comprise a
statement of the Path. Samadhi (= four jhanas) is here, as elsewhere, a prepara_
tory stage.
94. :Further indirect evidence for the decline of the jhayin tradition comes
from the decline both in status and numbers of the arhat in early Buddhism.
Pande calls this decline "the most hotly debated point in the whole range of
early sectarian controversy." Pande, Origins of Buddhism, p. 564.
95. Rod Bucknell and I have argued that knowledge of how to practice
the higher meditative techniques became confined to an esoteric transmission
in early Buddhism. See Rod Bucknell and Martin Stuart-Fox, "Did the Buddha
impart an Esoteric Teaching?", journal of Indian History 61 (1983), pp. 1-17;
also Bucknell and Stuart-Fox, The Twilight Language, chapter 2.
96 .. C.A.F. Rhys Davids, The Birth of Indian Psychology and its Development
in Buddhism (London: Luzac, 1936), p. 333. .'
97. Gunaratana, Critical Analysis of the jhanas, p. 216, makes this statement
after devoting a whole thesis to precisely such an intellectual treatment ofthe
jhanas.
II. BOOK REVIEWS
- Rationality and Mind in Early Buddhism, by Frank J. Hoffman. Delhi:
Motilal Barnarsidass, 1987. 12 + 126 pp.
It often has been remarked that when European scholars "dis-
covered" Buddhism in the late 19th and early 20th centuries, they
tended to gravitate to that form of Buddhism that most closely
reflected their own philosophical preoccupations. Thus, the
rationalist French initiated the study of Indo-Tibetan scholasticism
and metaphysics, while the British, with their long tradition of
empiricism, found in the Theravada of occupied Sri Lanka and
Burma what seemed like a perfect mirror of their own concern
with ethics, equality and evidence. It is, of course, likely that what
the Europeans encountered was not a mirror but something more
akin toa projected film image. Nevertheless their interpretations
of Buddhism have been influential, not only on subsequent gener-
ations of vVestern scholars, but also on some of the very Buddhists
they sought to describe. This has been particularly important in
the case of Sri Lanka, where native scholars educated in both their
own and British traditions, most notably K.N. Jayatilleke (Early
Buddhist Theory of Knowledge), David J. Kalupahana (Causality: the
Central Philosophy of Buddhism) and Walpola Rahula (What the Buddha
Taught), provided interpretations of Buddhism that emphasized its
compatibility with modern scientific and empirical approaches to
the world. The fact that these interpretations have been proffered
by "real Asian Buddhists" gave them an added authoritativeness,
and it is probably safe to say that the majority of Westerners have
had their understanding of Theravada (hence what the Buddha
allegedly "really" taught) shaped by what is sometimes called "the
Buddhist empiricism thesis."
This thesis, which often entails the correlate assertion that
Buddhism is really more a "philosophy of life" than a "religion,"
has begun.to come under attack in recent years, from both Asian
and Western scholars. For example, A.D.P. Kalansuriya, a Sri Lan-
kan, has argued that advanced Buddhist meditative experiences
cannot possibly be understood empirically, because they trans-
cend the ordinary senses that are the only meaningful basis for
"empiricism"; and David Snellgrove, who is English, devotes con-
siderable effort in the early chapters of his recent indo-Tibetan
Buddhism to criticizing the notion that the Buddha is "an agnostic
teacher of ethics of entirely human proportions who was later
divinized by the enthusiasm of his followers," which is, he argues,
simply "a nineteenth-century European creation, corresponding
to the similar efforts that were made to find a purely human
111
----------------------
112 JIABS VOL. 12 NO.2
ethical teacher behind the Jesus Christ of the Gospel accounts.
In both cases the 'mythological' interests are primary, and since
they dictate ... the form in which the story is told, not only does
the story become trite when deprived by critical scholarship of
its religious significance, but also a gap begins to yawn between
the 'founder' himself and his believing followers" (vol.I,p.8).
There is, in short, a growing belief among scholars not only that
Buddhist philosophy per se cannot easily be subsumed under
Western categories, but that the religious context out of which
that philosophy emerges makes its identification as "scientific"
or "empirical" particularly problematic.
This belief is at the heart of Frank J. Hoffman's Rationality
and Mind in Early Buddhism, which is the most sophisticated and
many-pronged attack on the Buddhist empiricism thesis that has
yet appeared. In this review, I will briefly layout, and comment
upon, the argument of each of the book's six chapters, then
remark on the book's style and structure, and conclude with some
general observations.
In his first chapter, "Understanding Early Buddhism,"
Hoffman carefully qualifies the term "early Buddhism," indicat-
ing that it is simply a "shorthand" for "the Buddhism of the five
nikiiyas," which are themselves assumed to be neither of the same
chronological stratum, nor necessarily the words of Gautama
Buddha. He goes on to describe his general purpose, which is
to analyze certain philosophical issues and conceptual problems
in early Buddhism. To do so, one must avoid both pure textual
exposition-for this approach tends toward apologetics, and is
insufficiently critical-and the temptation to "straightjacket" the
texts into "an alien and perhaps pre-conceived philosophical
framework" (p.2)-approach that fails sufficiently to respect the
texts' own words and meanings. Hoffman's own approach, then,
is to combine both emic and etic approaches, showing "on the
one hand, sympathetic understanding of what is internally coher-
ent and linguistically precise in the language of the Asian texts
studied, and, on the other hand, attention to Asian thought from
a critical philosophical point of view" (p.7). The "texts" Hoffman
proposes to use are the nikiiyas, without recourse to their commen-
taries; "the critical philosophical point of view" is provided mostly
by contemporary Anglo-American philosophers of religion, with
a backward glance at Wittgenstein.
I am essentially in sympathy with Hoffman's stated goal and
methodology: it seems a prudent and balanced basis for cross-cul-
tural philosophical analysis. The only point with which I might
quarrel is his insistence on dispensing entirely with the commen-
BOOK REVIEWS 113
'tarial tradition. Granted, it is' naive to suppose that commentaries
give us the "true purport" of the texts on which they comment;
Buddhaghosa, for example, wrote his commentaries on the
nikiiyas several hundred years after the texts were composed, and
brougJ:t to his analysis concerns peculiar to his own time and
situation. Nevertheless, if commentary is not "pure" exposition
of earlier texts, it is usually a part of the same intellectual con-
tinuum, and often can provide illumination of the possible mean-
ings of the texts. Certainly, it is a "closer continuer" of the original
texts than our own efforts are, and while attention to commen-
taries sometimes can lead us away from the meaning of the orig-
inal, just as often it will provide contextual insights that enable
us to check our tendency to impute our own radically different
concerns into texts from another time and culture.
In chapter 2, "Rationality and Logic," Hoffman focuses on
the "fourfold pattern" -often called the tetralemma-employed
by the Buddha in analyzing the famous "unanswered questions."
He is particularly concerned to rescue Buddhism from the accusa-
tion that it is philosophically incoherent because it invokes a
logical "principle of contradiction" and then seeI;Ils to violate that
principle by setting as the third and fourth of the four possible
positions, "both X and non-X" and "neither X nor non-X."
Hoffman argues (contra Jayatilleke) that early Buddhism has no
term for, hence no real concept of, "propositions." Therefore,
the tetralemma and the "principle of contradiction" are not ele- .
ments of a formal logic, applied to propositions, but, rather, heuris-
tic rules to be applied to utterances, which "can be understood
properly as existential statements" that cannot be "formally sym-
bolized" (p.21). Thus, when we consider, e.g., whether or not the
tathagata exists after death, the third position is not that "he both
exists and does not exist" -which is self-evidently contradictory-
but, rather, that "part of the tathiigata survives death and part
does not," and if the surviving part is taken as a "permanent
iitman surrogate," it must be rejected (p.21).
I think Hoffman performs a service here by pointing out
clearly that the tetralemma must be located within the context
of actual utterances and discussions, and that it should not there-
fore too quickly be homologized to Western formal logic, nor, by
the same token, should Buddhist utterances be trans-
lated into "propositions" that can quickly be "proved" or "re-
futed." I confess,' however, that I do nOt entirely understand the
problem he seeks to resolve or agree with his particular solution
to it. In the first place, Hoffman does not clearly identify those
critics who feel that the third and' fourth members of the tet-
114 JIABS VOL. 12 NO.2
ralemma threaten to undermine Buddhism's philosophical coher-
ence. Granted; the third and fourth members are not easily
explained, but I do not think that they need to be rephrased in
the manner-not clearly explicated in the original texts-that
Hoffman suggests. Why should they not invoke a principle of
contradictiori that simply says, e.g., the tathiigata cannot both exist
and .not exist after death because survival and non-survival are
mutually exclusive? This may, as some have suggested, involve
setting up a philosophical straw-man, but it does not, I think risk
the sort of self-contradiction Hoffman is concerned about-espe-
cially when we consider that the Buddha consistently rejects the
"bothland"-as well as the "neither/nor"-alternative.
Further, though Hoffman is probably correct that there is
no term for "proposition" in the nikiiyas, and though there no
doubt is a legitimate distinction to be made between propositions
and utterances, I do not think that this should obscure the fact
that there is considerable interest in early Buddhism in distin-
guishing between "true" (sacca) and "false" (micchii) , with the
former in some sense corresponding to, and the latter straying
from, "things as they are" (yathiibhUtam). Now it may be more
proper to regard what is "true" as an utterance than as a propo-
sition, but I suspect that to do so is to underestimate the implicit
propositionality of utterances, and to insist on a distinction that
is in some ways more seman tical than real. Buddhist utterances
may exist within a non-propositional, e.g., religious, context, but
that does not mean that Buddhists believed their utterances to
be relative to their particular context; after all, whether or not
tathiigatas arise to explain it, the way things are is the way they'
are. Hence, Buddhists descriptions of the way things are do con-
form to some meaningful sense of the term "proposition," if not
necessarily to the most restrictive one.
Chapter 3, "Rationality and Pessimism," seeks to refute the
common perception that Buddhism-whose first noble truth,
after all, is that of suffering-is "wholly pessimistic and in that
way 'irrational'" (p.27). Hoffman does riot specify why a "wholly
pessimistic" outlook would be "irrational"; perpaps it is because.
it would eclipse all meaningful discussion of "better" and "worse."
In any case; he carefully examines the meaning of dukkha in a
number of nikiiya passages, concluding that there is a narrower
sense in which dukkha is "anguish"-i.e., overt suffering-and a
broader, more fundamental sense in which it is descriptive of all
experience insa1'{1Siira, where conditions are impermanent, hence
unreliable. This, Hoffman observes, "is in contrast to the 'no
BOOK REVIEWS
115
. arising' and 'no falling' characteristics of ... nibbiina [which] is
not characterized by impermanence" (p. 29). After examining a
number of different versions of pessimism, Hoffman concludes
that "pessimism a.dmits no consolation," whereas Buddhism, with
its b e l ~ e f the possibility of a virtuous life here, a better rebirth in
the hereafter, and the ultimate attainment of nibbiina, "sees ...
many sources of consolation," and so "cannot accurately be called
pessimistic" (p.37). He ends the chapter by arguing (contra
Radhakrishnan), that the Buddhist assertion that "all is dukkha"
or "all is impermanent" must be understood as evalu.ative rather
than descriptive and scientific-since "to see the world yatka
bhutam is ... not to see what a video-camera would record but is
in part to see in a hopeful manner the possibility of liberation"
(pp. 42-43). There is, in short, in Buddhist "descriptions" of
reality the deliberate inclusion of an element of value, and a lack
of concern with issues of "verification" or "falsification," which
clearly take those descriptions outside what we could comfortably
characterize as "scientific"-and because Buddhism makes no
pretense to being "scientific," it cannot be accused of confusing
fact and value.
I am in complete agreement with Hoffman's refutation of
Buddhist pessimism: it is a careful, lucid and compelling analysis.
His analysis of factual and evaluative elements in Buddhism,
which serves as a sort of coda to the discussion of pessimism, is
important, but less convincing. He argues successfully, I think,
that early Buddhism is not "scientific" or "factual" in precisely
our sense of those terms: the Buddha is not setting forth "hypoth-
eses" that must stand or fall on empirical evidence, and he clearly
informs Buddhist descriptions of "the way things are" with pre-
suppositions that are in turn rooted in "values," such as the hope
of nibbiina. Once again, however, I think that the fact that early
Buddhist assertions do not pretend to be "factual" in a sense that
a modern scientist would recognize does not mean that Buddhists
regard their assertions as, say, context-dependent conceptual
schemes or language games; they clearly regard their assertions
as universalizable, hence as independent of their particular as-
sumptions. This, of course, entails precisely the sort of interweav-
ing of facts and values of which Hoffman seeks to exculpate early
Buddhism. I have no simple solution to this problem: perhaps
there really is a confusion; or perhaps one could point to the
incorporation of "values" and presuppositions into "factual" de-
scriptions as a way of narrowing the apparent gap between the
two types of assertions, thereby beginning to undercut the critique
116 JIABS VOL. 12 NO.2
of any set of assertions that overtly contains evaluative elements.
1 will saya bitmore about this in my comments on chapter 5.
In chapter 4, "Mind and Rebirth," Hoffman addresses two
major questions: (1) How does early Buddhism describe the re-
birth process? and (2) Is there a systematic attempt in early Bud-
dhism to explain. how there might occur "reidentification of per-
sons across lives"? (1) In his analysis of nikiiya accounts of the
rebirth process, he is careful to note that "there are several can-
didates for the rebirth link (sankiirii, Citta, gandabba, viiiiiii'TJ,a), and
no consistent, technical view about this matter in early Buddhism"
(p.5l). He also observes that "[i]n early Buddhism there is no
'doctrine of moments' (!?Ja'TJ,aviida)," no division of perceptions
into "distinct existences" as in the later Buddhist Sautrantika
school or, in the West, Hume (p.56). If there is no doctrine of
momentariness, then the problem of radical discontinuity-with
its potential for vitiating any account of causality-is not posed
for early Buddhism. Nor, however, does Buddhism "offer a sub-
stitute for the concept of iitman" -no permanent substance is
asserted (p.54). How then is continuity explained? Hoffman finds
the most helpful image in Wittgenstein's Philosophical Investiga-
tions, where "the strength of a cord does not always depend on
there being a single strand which runs from end to end, but
sometimes depends on the interrelationship between overlapping
and criss-crossing fibers, none of which runs the entire length of
the cord. The early Buddhist doctrine of rebirth may be viewed
like this: there is no permanent, unchanging iitman linking up
successive lives with its continuous psychic fiber, but there is
nevertheless continuity which is assured by overlapping and criss-
crossing fibers" (p.5l). (2) Searching the nikiiyas for criteria for
reidentification of persons across lives, Hoffman finds them lack-
ing. He considers in detail a variety of possible criteria, including
those of consciousness, memory and body, and sees all of them
as problematic: "viewed externally there is no good reason to
accept 'there is rebirth,' for it is not clear that the obstacle of
providing conditions for the meaning of 'the same person' can
be overcome" (p.76). This is not a problem, however, as long as
one recognizes that early Buddhism is not interested in providing
such criteria, but, rather, regards the doctrine of rebirth as part
of its own "conceptual background," a presupposition "against
which other beliefs may fit or fail to fit, since the tests are devised
in terms of the background" (p.75).
Hoffman's discussion in this chapter is impressively detailed
and generally convincing. I find his description of the processes
BOOK REVIEWS
117
of change and rebirth in terms of the strands of a rope quite
striking, and I suspect that it probably does capture the spirit of
the early Buddhist attitude. I would be more comfortable with
itif there were an exactly equivalent image in the nikayas, but I
do not think Hoffman provides one: the closest parallel he can
find is the image of the flame, which does express the concept of
continuity within change, but does not account for the problematic
interaction of various impermanent elements in the way that the
rope image does. I am in general agreement with Hoffman's
analysis of the problem of establishing criteria for identifying
. persons across lives, though I am not entirely in accord with his
rejection of the memory criterion. He argues that the memory
of past lives is problematic in part because it is available to so
few, in part because the lives one can in principle remember are
infinite, hence the memory incapable of supplying "all the data," .
leaving Buddhism with, at best, a "weakly quantified law" (p.70).
The fact that the memory of past lives is available to so few,
however, does not mean that it is not accessible in principle to
everyone;. while the absence of "all the data" on rebirth, karma,
etc., should not (problems of subjectivity aside) prevent supernor-
mal perceptions of them from acquiring the same kind of eviden-
tIal force as other inductive generalizations. This may still leave
unresolved problems about the precise mode of reidentification, .
but it would seem to leave the possibility open in principle. Finally,
I agree with Hoffman that "proving" rebirth is not a major con-
cern of early Buddhism. It should be noted, however, that not
all Buddhists have considered it an unassailable presupposition:
under attack from materialist critics, later Indian Buddhists, such
as Dharmaklrti, provided elaborate rational.defenses of the doc-
trine of rebirth-and thereby implicitly recognized that the "fac-
tuality" of what might otherwise be relegated to the "evaluative"
realm had become a concern of Buddhists-even if those concerns
were, as Hoffman insists, latent or nonexistent in Buddhism's
early period.
It is in chapter 5, "Mind and Verification," that Hoffman
mOst directly attacks the Buddhist empiricism thesis. His discus-
sion is threefold, dealing in turn with the role of saddhii ("faith"),
the function of abhiiiiiii, and the degree to which early Buddhism
fulfills the criteria for "empiricism." He argues first of all against
those-most notably Jayatilleke-who claim thatsaddhii is invari-
ably a "faith" in the Buddha subsequent to checking the truth of
what he teaches, a "rational" faith that is to be contrasted with
the "blind" faith ofbrahmins. Hoffman marshals textual evidence
118 JIABS VOL. 12 NO.2
for the view that there are at least some instances where saddhii
. precedes and, indeed, facilitates one's understanding of what the
Buddha teaches. "Thus," concluded Hoffman, "there is an affec-
tive element in saddhii which is ignored if one treats believing in
the Buddha as equivalent to believing that what he says is true"
(p.83). If saddhii is in many cases affective and prior to understand_
ing the teaching, it loses the purely rational,a post(!riori quality
with which Jayatilleke sought to invest it. Hoffman continues by
arguing that neither the specific, paradigmatic event of the
Buddha's enlightenment J.mder the bodhi tree, nor abhiiiiiii in
general, should be seem simply as experiences that serve to con-
firm beliefs, but must be understood as informed by doctrinal
assumptions and issuing in inspired activity in the world. Thus,
abhiiiiia does not verify propositions, for "it is a mistake to think
that there is a body of propositions which can be rightly labeled
'religious knowledge', in a sense even remotely analogous to sci-
entific knowledge. Unlike 'religious knowledge', there may indeed
be 'religious wisdom', but if there is, it is to be found embodied
in the lives of religious people, and as with 'philosophical wisdom',
cannot be embodied in a set of propositions but is embodied in
practices" (pp. 95-96). Finally, Hoffman directly attacks the Bud-
dhist empiricism thesis, arguing that "empiricism" requires
criteria of falsification, and that early Buddhism simply does not
provide these: even the Buddha's famous invitation to "come and
see" (ehi passako) "will not be falsified by the assiduous meditating
monk who meditates and yet does not 'see' rebirth .... At no
point will the meditation teacher agree that the student has falsi-
fied the doctrine in case the student came and did not see" (p. 98).
I think that Hoffman's arguments succeed in refuting the
Buddhist empiricism thesis-at least in the strong form that
Jayatilleke, Kalupahana and others have presented it. There
clearly is in early (as in later) Buddhism a place for. a priori ;'faith":
not only is "confidence" not always "subsequent to checking,"
but one's checking may not "verify" doctrines if one does not
possess that a priori faith. By the same token, "religious e x p e r i ~
ences" (as Steven Katz and others have argued) cannot be isolated
either from their informing assumptions nor from the "life" of
which they are a part, and so cannot be regarded as having quite
the same evidential status as scientific, empirical studies do. Nor
do we see many (if any) instances where religious doctrines are
"falsified" on the basis of experience. For all these reasons, it
seems safe to say that early Buddhism is not "empiricism" as that
term is used in Western philosophy. However, the fact that early
BOOK REVIEWS
119
'Buddhism involves a priori assumptions and attitudes, is not over-
tly propositional, and may not fulfill the definition of empiricism,
does not-as I have indicated before-mean that early Buddhists
were not making truth-claims of any sort. They were not and did
not ci,!-im to be philosophical rationalists, but they did believe
that their doctrines accurately described things as they are, that
happiness depended on seeing things as they are, and that under-
standing of the way things are-while no doubt embedded in
religious aspiration and practice-was in some sense verifiable
by examination, whether through the senses or the abhiiiiiii, of
actual conditions in the world. This may not be "propositional"
or "empirical" in the strong sense, butl think that it is an outlook
that entails certain propositions and a certain confidence in "check-
ing" those propositions. _
Hoffman fears, I suspect, that if we concede that early Bud-
dhism is attempting to be propositional it too easily may be shown
- to be irrational. No doubt the admission of philosophical motives
does raise problems about presuppositions, coherence, etc., but
they may not be insurmountable. Hoffman seems to me to have
an exaggerated notion of the purity of scientific and empirical
inquiry-as if they were not themselves informed by presupposi-
tions and embedded within ideological (and even cultural) con-
texts. I am not trying to argue for a strong. Kuhnian position, or .
maintain that "religion" and "science" are simply two alternative
conceptual schemes, each valid in its own domain; I do think,
however, that the degree to which we now appreciate the contex-
tual nature of all understanding, including the scientific, has
altered our notion of the conditions under which truth-claims
may be made: because we are more modest about the "purity"
of any and all evidence, we may, paradoxically, entertain truth-
claims (even-shudder!-propositions) of more various types than
we could, say, during the heyday of logical positivism. There are,
of course, many issues left undecided about the commensurability
between and relative merits of various types of truth-claims, but
I do not feel that religions-early Buddhism included-require
the "protection" afforded by the assertion of their non-cognitive,
non-propositional nature. The relationship between "belief' and
"understanding" is a complex and undoubtedly problematic one,
but it is thus in all realms of "knowledge," and I think we may
do religions as much of a disservice by understating the serious-
ness of their religious truth-claims as we do by overstating them-
for if the latter results too often in oversimplification, the former
risks trivialization.
120 JIABS VOL. 12 NO.2
The sixth and final chapter is devoted to "The Deathless
(Amata)," i.e., to nibbiina, of which "the deathless" may be both a
synonym and a quality. Hoffman argues first (contra Kalansuriya)
that parinibbiina is not a transcendental state of immortality, on
the grounds that such a view is not entailed by the Buddha's
denial of the materialist uccedavada. He goes on to defend (contra
Peter Masefield) the meaningfulness of the distinction between
nibbiina and parinibbiina, arguing that the occasional application
of the latter to a living arhant does not vitiate the distinction, but
simply indicates a variant usage that is an exception to the general
rule, i.e., that parinibbiina is applied after the death of an arhant.
Finally, Hoffman analyzes the meaning of amata as it applies to
both nibbiina and parinibbiina, arguing that when applied to the
former, it simply means "the destruction of what defiles," i.e.,
greed, hatred and delusion (p.l13), and that when applied to the
latter it indicates "the limit of a Buddhist stream of life, not an
experience in that stream" (p.l14). The Buddhist view thus be-
comes that one lives not an "endless life" but an "eternal life" in
which "it is possible to live in such a way that one is not limited
by, but independent of, death" (p.l15). Since death, for Buddhists,
is inextricably bound with rebirth, once one has achieved a state
where rebirth no longer will occur, one is, by definition, free
from death, for one's passing will not be like the death of unen-
lightened beings. Returning to the conundrum of the tathiigata's
survival after "death," Hoffman concludes that from an emic
perspective, there simply is no thesis that can be found, while
from an etic perspective "the Early Buddhist position does suggest
that there is no question of anything surviving in parinibbiina
once the conditions for rebirth are gone" (p.l16). He ends by
suggesting that Buddhists denied that parinibbiina is extinction
as vociferously as they did (a) to avoid speculation on what did
not yield to speculation, (b) to avoid confusion with materialism's
uccedavada and (c) because one's own extinction cannot be experi-
enced, hence cannot be asserted on philosophical grounds.
Hoffman's arguments and textual analyses in this chapter
are generally solid. He gives a sound defense of the nibbiina-pari-
nibbiina distinction. His interpretation of amatais persuasive with
regard to nibbiina, and I am almost prepared to go along with
his rather ingenious reading with regard to parinibbiina, too. Cer-
tainly, when we consider that parinibbiina entails the elimination
of the aggregates, and the aggregates are the only com prehensible
basis for experience, there seems little room left for the continu-
ation of "anything" after the passing of an arhant. This does not,
BOOK REVIEWS
121
. however, it seems to me, completely rule out the possibility that
parinibbiina is an endless transcendental condition-as long as
we recognize that "transcendental" means "incomprehensible in
terms of either existence, non-existence, both and neither," and
that the specification "endless" could have very little meaning
within such a transcendental state (any more than "temporal" is
a meaningful characteristic of the Judeo-Christian God prior to
creation). Some light might be shed on the issue by further exami-
nation of the meaning of amata (oramrta) in early Indian culture.
Rationality and Mind in Early Buddhism is a detailed and
highly nuanced book, and I cannot claim to have explicated or
understood it in its entirety. Part of the fault lies with my own
limited powers of comprehension, but part, I think, must be
attributed to Hoffman. The book covers a great many topics of
great importance, yet Hoffman's style is so compressed that he
is sometimes difficult to follow. He refers continually to people
and positions (including some of his own, in articles) that the
reader is somehow expected to know. For example, he talks in
the last chapter of defending Kalupahana's view of parinibbiina
against Jayatilleke's, but never clearly states what those two views
are. By the same token, he refers in his discussion of falsifiability
in early Buddhism to his own "parable of the bhikku," but never
sets it forth explicitly. In a work of serious philosophical analysis
like Hoffman's, this sort of omission is a disservice to thereader,
who needs as clear and complete an exposition of the positions
under consideration as the author can afford. The book also
would have gained a bit more coherence had Hoffman (who
clearly understands Pili) himself translated the nikiiya passages
he cites, rather than relying on a variety of extant translations,
which inflict upon the non-Piliglot reader an avoidable bewilder-
ment as to the meanings of terms. My final stylistic criticism is
that the book-ranging both as wide and deep as it does-needs
a unifying conclusion. It's not that Hoffman's points don't come
through, or that a careful reader will not be able to construct his
or her own sense of Hoffman's overall position; it is simply a
matter of authorial helpfulness to provide for the reader who
has followed him through dense philosophical thickets a conclud-
ing view from on high that provides a perspective on the terrain
that has been covered.
My stylistic and philosophical criticisms notwithstanding, I
want to make it clear that I regard Rationality and Mind in Early
Buddhism as an important and insightful advance in our under-
standing of the ideology of the nikiiyas. Hoffman makes a vital
122 JIABS VOL. 12 NO.2
contribution with his analysis of the importance, in Buddhist
. doctrine and experience, of conceptual and affective presuppos_
itions; he is quite correct in locating Buddhist doctrines and
experiences within a particular lebenswelt; and I think he effec-
tively demolishes the strong form of the Buddhist empiricism
thesis, demonstrating that early Buddhism is not simply reducible
to propositions that can be verified or falsified in the way that
scientific hypotheses traditionally can. Although, as should by
now be clear, I feel that Hoffman underestimates the degree to
which (a) Buddhists are making universalizable truth-claims and
(b) those claims can-very cautiously-be treated propositionally,
I am very much indebted to him for helping to clarify my own
views on the complex problems he tackles, and I cannot conceive
that any further discussions of the philosophical standpoint of
early Buddhism (or even Buddhism as a whole) could proceed
without reference to his book.
Roger Jackson
J .W. de J ong's Review of Jeffrey Hopkins'
Meditation on Emptiness: An Exchange
Jeffrey Hopkins Replies to J. W. de Jong
Concerning Professor ].W. de Jong's review (JIABS, vol. 9,
no. 1, pp. 127-128) of my Meditation on Emptiness, I was surprised
to find a large number of his points to be unfounded. The ten
critical points of his review fall into four groups; he (1) is wrong
about my translation of siidhyadharma, )"igs lta, avyabhiciirin, biidhii,
siddha, and adhyiiSaya; (2) has m,issed the point of my note on rdo
rje gzegs ma, (3) has understandably been confused by my inventive
translation of the passage from the Ghanavyuhasutra, and (4) is
right about the passage from the Dhiira1Jzsvarariijapariprcchiisutra
and the translation of samyaktvaniyiima.
I cannot consider all of these points in the space allotted to
a response to a review, but a few will illustrate my countercriticism.
For instance, Professor de Jong makes an unsubstantiated re-
translation: (p. 126)
"Wrong is Hopkins' rendering of siidhyadharma by 'predicate of
the probandum' (p. 508). The siidhyadharma is the 'property to be
proved', i.e., the probandum."
This is all he says....:....he gives no sources; and he could not be
more wrong. De Jong's rendition is simply outside the pale of a
ba&ic understanding of Buddhist logical terminology. As Masaaki
Hattori says in the introduction to his Digniiga, On Perception: 1
"His great contribution to the cause ofIndian logic is the invention
of the hetucakra, that is, the table which shows nine possible rela-
tions between the Reason (hetu) and thesiidhyadharma or predicate
of the Thesis siidhya) to be proved."
De Jong has confused the siidhyadharma (the predicate of the
thesis, or predicate of the probandum) with the siidhya (the thesis
to be proved, the probandum). The compound siidhyadharma is
not to be interpreted as a karmadhiirya, meaning the "property
to be proved", but as a genitive meaning the "predicate
of that which is to be proved, i.e., of the probandum". This
reading of the terms is confirmed also by the context, but it would
123
124 JIABS VOL. 12 NO.2
take too much space here to make the point.:'!
In an attempt at correction, de Jong shows carelessness and
lack of comprehension: (p. 127)
"Hopkins has misread the text in Ngawang Gelek Demo's edition,
p. 906.5: )"ig lta sangs rgyas- kyi gdung chad pa lta bu de dag gis ...
Hopkins' text has )"igs lta, etc., and he translates this as follows:
'through fear [of the suffering of cyclic existence Foe Destroyers
have forsaken helping others, and thus] their Buddha lineage has
been severed' (p. 604). In this passage the Arhats are compared -.
to those whose Buddha-lineage has been severed on account of it
false nohon of personality ()"ig-lta,
First, de Jong's citation of Ngawang Gelek Demo's edition is
flawed; the text reads )ig ita sangs rgyas kyi gdung bshad pa lta bu
de dag gis, not )ig lta sangs rgyas kyi gdung chad pa ita bu de dag
gis as cited by de Jong. Thus, that edition, if taken at face value,
should be translated, in de J ong's vocabulary, as, "The false notion
of personality is described as the Buddha lineage," not that the
Buddha lineage is severed by such a false notion, as de Jong has
it. Though such makes sense in the context of the Vimalak'irti-
nirdesa where this notion does indeed appear, it makes no sense
here in the context of 'Jam dbyangs bzhad pa's explaining the
meaning behind passages in the Sa'f[Ldhinirmocanasutra that indi-
cate that certain sravakas never attain the highest enlightenment.
For, it would indicate just the opposite-that even such a false
view can contribute to their eventually becoming fully en-
lightened, thereby suggesting that these sravakas do indeed attain
the enlightenment of Buddhahood. The passage has to be
emended to make sense, and de Jong has indeed emended it,
but he should have cited the original accurately and indicated
his emendation.
Second, four of the five editions consulted read )igs lta
(though such is to be expected, given that evidence shows that
they stem from the same edition) whereas N gawang Gelek Demo's
edition reads )ig lta, as de Jong prefers. (Indeed, if I had simply
read )igs lta without it occurring to me that it might be )iglta
[satkayadr,5#] I would deserve the reviewer's scorn, nevermind criti-
cism.) My decision that the former reading is preferable, either
as 'Jam dbyangs bzhad pa's intended meaning or as a more sen-
sible interpretation, was based on an annotation by the Mongolian
scholar Ngag dbang dpalldan, cited in the emendations (p. 968)
and in note 555 (p. 871), neither of which de Jong apparently
noticed. That note reads, "The bracketed material is from Ann,
BOOK REVIEWS
125
(lbu 62a.8." Since the bracketed material immediately follows the
word "fear" which de Jong claims is a misreading of 'jigs for 'jig,
I would have expected him to check out the reference before
launching a criticism, but he obviously did not. The reference is
to the Annotations
3
of N gag dbang dpal Idan, used throughout
my translation, who says (emphasis added):
"When one forsakes activities for the benefit of others due to
viewing the suffering of cyclic existence fearfully, the Buddha
line or lineage is severed. ('khor ba'i sdug bsngalla 'jigs par 1ta ba'i
dbang gis gzhan don gyis bya ba dor na sangs rgyas kyi gdung ngam rigs
chad b[sic]ar 'gyur ro)."
Ngag dbang dpal ldan, in a brilliant display of erudition, has
ferreted out the meaning of 'Jam dbyangs bzhad pa's unusual
terminology; he corrects not 'jigs to 'jig but (as is indicated in my
emendations to the text, p. 968) gdung bshad to bdung chad. I
speak of Ngag dbang. dpal ldan's "erudition" because 'Jam
dbyangs bzhad pa is addressing teachings in the Sarrtdhinirmoca-
nasilatra that indicate that certainsnivakas never attain the highest
enlightenment, and although Ngag dbang dpal ldan does not
cite the passage, it must be one found in the seventh chapter
where it is explained that a sravaka who proceeds solely to peace-
fulness (samathaikayanika, zhi ba'i bgrod pa gcigpu pa) cannot attain
full enlightenment because of being deficient in compassion and
because of being very afraid of sufferIng (dul;khiitibhayatas, sdug
bsngal gyis shin tu 'jigs pa). That such sravakas are incapable of the
highest enlightenment is depicted in 'Jam dbyangs bzhad pa's
text (as corrected by Ngag dbang dpal ldan) by "their Buddha
lineage has been severed" (sangs rgyas kyi gdung chad pa). Also,
that the reasons for this include these sravakas' being very afraid
of suffering is indicated by "fear" or, more literally, "viewing with
fear" ('jigs lta, i.e., 'jigs par Ita ba). Ngag dbang dpalldan's contex-
tual reading, based on philological analysis (i.e., associating the
word 'jigs in the two texts), is most sound. Professor de Jong's
criticism, however, turns out to be careless for not pursuing refer-
ences in a note and an emendation and for mistakenly citing the
reading that he prefers. Had he taken greater care, he would
have perceived the appropriateness of Ngag dbang dpal ldan's
explanation and, thereby, would not have wanted to emend 'jigs
to 'jig. .
One of Professor de Jong's criticisms speaks directly to the
important issue of style of translation, the reviewer disagreeing
with my preference for rendering terms more literally. Referring
126 lIABS VOL. 12 NO.2
to my translation of the controversy between Bhavaviveka and
Candraklrti from the first chapter of the Prasannapadii, he says;
(p. 126)
"In part five of his book, Hopkins translates and explains the
controversies between on the one hand, Buddhapalita, and on the
other, Candrakirti .... For instance, Hopkins translates badhii (Tib.
gnod-pa) by 'damage, harm' (cf pp. 502, 526 arid note 395),
whereas the technical meaning of the Sanskrit term 'refutation
annulment' is well-known from Sanskrit philosophical texts, both
Buddhist and non-Buddhist."
First, the controversy is not "between on the one hand, Buddha-
palita, and on the other, Candraklrti", but between Bhavaviveka
and Candraklrti, who is defending Buddhapalita.
Since I chose this particular translation-equivalent, despite
its obvious awkwardness, after much reflection, the term provides
a good instance of what, at least on the surface, appears to be a
clash of translation-paradigms. Simply put, I often find that the
re-rendering of Sanskrit and Tibetan philosophical terminology
into what some contemporary translators have identified as its
philosophical meaning loses much of the psychological punch.
De J ong does not consider the fact that the eleventh century
Indian and Tibetan translators-Mahasumati and Pa tshab nyi
ma grags-and revisors-Kanakavarman and, again, Pa tshab
nyi ma grags-who were well aware of Sanskrit technical ter-
minology, undoubtedly consciously chose to translate the term
biidhii as gnod, "damage" or "harm". The interpretation of Bud-
dhist technical terminology by such Indian and Tibetan scholars
strikes me as important and valuable because it provides a fascinat-
ing source for the understanding of Sanskrit terminology nine
hundred years ago. Specifically, the psychological dimension of
the Tibetan gnod, "damage" or " i ~ u r e " , as can be gained from
contact with the oral tradition, is that the adherence that a person
has to a wrong view needs to be counteracted, to be harmed, to
be damaged. The martial imagery is not by chance; the aim of
the battle is to be so affected by a good argument that one's own
position is damaged. This does indeed mean to be "refuted", but
such a translation does not convey the implications of the term.
As is obvious, oral traditions are often wrong and thus cannot
simply be accepted at face value. However, in this case, we have
the evidence that nine hundred years ago Indian and Tibetan
scholars (not just those who translated this text but many other
translators, too) avoided the many possible Tibetan equivalents
BOOK REVIEWS
127
for "refute" and chose to stick with "damage, injure", which we
know to be one of many meanings of the Sanskrit. Thus, given
the obvious connection with "refute" but in an earthier way,
"damage" or "injure" is a better translation, for it at least has a
chance of conveying (or contributing to conveying) the cultural
background of the term. I am not putting forward a general
theory that we should return to older, non-technical meanings
of technical terms; I am asserting that it is helpful to check these
supposedly non-technical meanings in order to overcome preju-
diced adherence to translation-equivalents that, no matter how
much we have become used to them, are actually sanitized ver-
sions that fail to communicate cultural dimensions. Thorough-
going philology needs to take account of cultural context.
In conclusion, though I appreciate the corrections that are
valid, I have been disappointed by the many errors and the fre-
quent lack of substantiation, these being surprising in work by
such an eminent scholar.
NOTES
1. (Cambridge: Harvard, 1968), p. 4.
2. In another criticism without substantiation, de Jong says:
(p. 126)
' ~ l s o , in other instances Hopkins' renderings of technical terms are not
very satisfactory, for example, 'un mistaken' (p. 485) foravyabhicarin (21.5)."
This is hard to answer since this is all the reviewer says! Suffice it to cite
Vaman Shivaram Apte's The Practical Sanskrit-English Dictionary on vya-
bhicarin, the term without the negative prefix a; Apte gives "straying or
deviating from, going astray, erring, trespassing; irregular, anomalous;
untrue, false; faithless, unchaste, adulterous; profligate, wanton; depart-
ing from its usual meaning, having several secondary meanings; change-
able, inconstant."
In another undocumented criticism that is a mere quibble about
choice of translation terms: (p. 127)
"In the translation of this text, Hopkins is careless too in his renderings
of technical terms. Probably nobody will recognize in 'unusual attitude'
Tibetan lhag-bsam, Sanskrit adhyiisaya" (p. 604).
Suffice it to cite (1) Franklin Edgerton's Buddhist Hybrid Sanskrit Grammar
and Dictionary which for adhyasaya gives "mental disposition; (strong)
purpose, intent determination (esp. religious)" and, in connection with
128 JIABS VOL. 12 NO.2
lhag pa'i bsam pa, gives "superior (adhi) thought, will, inclination" and
(2) Etienne Lamotte's L'Enseignement de Vimalakfrti (Louvairl: 1962),
p. 406, which, in a classic note, gives "haute resolution", "pensee profonde",
and "haut sentiment". These are all very close, to "unusual attitude"!
3. Annotations for Uam-jang-shay-ba's) "Great Exposition of Tenets",
Freeing the Knots of the Difficult Points, Precious Jewel of Clear Thought (grub
mtha' chen mo'i rrichan 'grel dka' gnad mdud grol blo gsal gees nor), (Sarnath:
Pleasure of Elegant Sayings Press, 1964).
J. W. de J ong Replies to Jeffrey Hopkins
I am sorry that I have not been able to convince Professor
Jeffrey Hopkins on a number of points. Let me begin by correct-
ing an error of mine. The controversy in the Prasannapadii is
between Bhavaviveka and Candrakirti, and not between Budda-
palita and Candrakirti.
Hopkins goes into much detail in order to defend his trans-
lation of the following passage on p. 604 of his book 'jigs lta sangs
rgyas kyi gdung chad pa lta bu de dag gis. Hopkins rightly remarks
that N gawang Geleg Demo's edition has gdung bshad pa, notgdung
chad pa. Having read the Tibetan text at the end of the book
(p. 26) and the correction on p. 969, I corrected the text of
N gawang Geleg Demo's edition without pointing this out because
we both agreed on this reading. As to the confusion between 'jigs
lta and jig lta this is so common that it is almost unnecessary to
draw attention to it. Hopkins uncritically follows Ngag dbang
dpal ldan's fanciful explanation based on his failure to correct
the wrong reading 'jigs lta. Hopkins refers to the seventh chapter
of the SaT{ldhinirmocana but this does not say anything about "view-
ing with fear".
According to Hopkins I am wrong about his translation of
siidhyadharma, avyabhiciirin, biidhii, siddha and adhyasaya. I remain
unrepentant and continue to find it inadmissible to use such
Tibetan Hybrid English renderings as "ren'owned" for siddha,
"being renowned to the other [party]" for parata/:t prasiddhi, and
"harm, damage" for biidhii. As to siidhyadharma, Hopkins has not
taken into account the ambiguity of the term siidhya.
1
The context
has to be taken into account. Candrakirti declares that if, as does
Bhavaviveka, one rejects a subject or substratum (dharmin), one
cannot prove the existence of a quality (dharma) such as anutpiida.
2
BOOK REVIEWS 129
The translation "unmistaken" for avyabhiciirin is not precise
enough because it suggests a Sanskrit term abhriinta. As to the
rendering of adhyiiSaya by "unusual attitude" this is a most awk-
ward translation which is not found anywhere else.
Let me repeat once more that I have a high regard for the
work done by Hopkins. It is exactly for this reason that I feel
justified in critically examining his work in the hope that: "Du
choc des opinions jaillit la verite."
NOTES
1. See the remarks by H.N. Randle in his Indian Logic in the Early
Schools (Oxford, 1930), p. 185, nJ, and his translation of siidhyadharma
on p. 170. .'
2. See the translation of chapter one of the Prasannapada by
Teruyoshi Tanji (Kansas University Press, 1988), p. 23. His translation
i ~ by far the best-of all existing translations of this often difficult chapter.
130 JIABS VOL. 12 NO.2
Errata to Vol. 10.2
Brian Galloway's "Notes on Nagarjuna and Zeno on Motion"
(journal of the Inten:ational Ass.ociation of 10(2)
(1987) 80-87) reqmres correctIOn on the followmg pomts. .'
Figs. 1 and 2, first referred to on p. 81, were omitted
are now given here.
On p. 81 l. 10, for 'by' read 'but'.
On p. 83 l. 30, for the comma read a semicolon.;
On p. 86, in the second displayed equation, the
h and v should be understood as being of normal type size (not:
small as printed); the superscript numbers belong with the vari-.
abIes and not with the subscripts. Thus for examfle the second;
term should be understood as v;'/v! orv,h/(v) . To the rig4t;
of the summation sign read a fraction bar between the tWb
terms. Thus we should have
h
+ + +
+ ...
The fraction bars in the first equation should also have been
horizontal. <
On p. 87 l. 1, for 'notion' read 'noting'; second line froni
bottom, for 'de la Poussin' read 'de La Vallee Poussin'.
The title of the article, given correctly on p. 80, is incorrect;
on the contents page of JIABS 10(2); for 'in' read 'on'. .'
a x c
Fig. 1.
a c x
b
.,.
Fig. 2.
ERRATA 131
Erratum to Vol. 12.1
':Jh
e
name of the reviewer of Lambert Schmithauserr's Alayavi-
'!jna
na
(vol. 12, no. 1, pp. 170-177) was inadvertently omitted.,
is Paul J. Griffiths.
132 jIABSVOL.12NO.2
CONTRIBUTORS
Prof. Robert E. Buswell, Jr.
Dept. of East Asian Languages
and Cultures
OJ01
290 Royce Hall
UC.L.A.
Los Angeles, CA 90024
Prof. J.w. de Jong
Faculty of Asian Studies
Australian National University
Box 4, Post Office
Canberra, A.CT
Australia
Mr. Brian Galloway
2108 Shattuck Ave. #4
Berkeley, CA 74704
Prof. Peter N. Gregory
Program for the Study of
Religion
College of Liberal Arts and
Sciences
3014 Foreign Languages
Building .
Univ. of Illinois at
Urbana-Champaign
Urbana, IL 61801
Prof. Jeffrey Hopkins
Dept. of Religious Studies
Cocke Hall .
University of Virginia
Charlottesville, VA 22903
Prof. Roger R. Jackson
Dept. of Religion
Carleton College
Northfield, MN 55057
Mr. John J. Makransky
Dept. of South Asian Studies
1242 Van Hise Hall
Univ. of Wisconsin-Madison
Madison, WI 53706
Prof. Martin Stuart-Fox
Dept. of History
University of Queensland
St. Lucia, Queensland
Australia, 4067