sautrantika in the abhidharmakosabhasya

8/10/2019 Sautrantika in the Abhidharmakosabhasya

http://slidepdf.com/reader/full/sautrantika-in-the-abhidharmakosabhasya 1/59



ARTIKEL 201



SAUTR A NTIKA IN THE ABHIDHARMAKO S ABH A∑YA

ROBERT KRITZER

I. Introduction

The term “Sautr antika” appears in the

Abhidharmako s abh a Ò yaapproxi-mately twenty times. In almost every case, the opinion attributed to the

Sautr antikas contradicts the orthodox Sarvastivadin/VaibhaÒika position,and, as Kato shows (75-78), represents Vasubandhu’s own opinion. Sau-tr antika is closely associated with Dar Ò†antika, which is often consideredto be either the same as Sautr antika or its immediate predecessor, andmany of Vasubandhu’s “Sautr antika” opinions strongly resemble onesattributed to Dar Ò†antika by the *Vibh a Òa .1 Recently, however, scholarshave begun to notice that some of the same opinions can also be foundin the Yoga ca rabh umi. In this paper, I examine the occurrences in the

Abhidharmako s abh a Ò ya of the term “Sautr antika,” refer to relevant pas-sages in the *Vibh a Òa and Harivarman’s *Tattvasiddhi , and show that themajority of the positions labelled Sautr antika have correspondences inthe Yoga ca rabh umi, most frequently in theVini s cayasa µ graha ∞i on thePañcavijñ a nak a yamanobh umi. I also discuss the possible implications of Vasubandhu’s evident reliance on theYoga ca rabh umi.

I limit myself here to those passages in which the word “Sautr antika”

actually appears in the Sanskrit text. There are many other positions thathave been identified by commentators such as P’u-kuang and Yas omitraas Sautr antika, not to mention the more than 200 references by Saµ-ghabhadra to theching-chu or “sutra-master,”2 a term used to sig-nal Vasubandhu’s departure from Sarvastivadin orthodoxy (see Kritzer

Journal of the International Association of Buddhist StudiesVolume 26 • Number 2 • 2003

1 On the other hand, Vasubandhu does not agree with all the Dar Ò†antika positions.It seems, in fact, as though he himself uses the term Dar Ò†antika pejoratively. When he agreeswith a Dar Ò†antika/Sautr antika opinion, he labels it Sautr antika (see Cox 37-39, which isbased largely on Kato; see also Harada).2 I borrow Cox’s translation (56).



Comparison , “Preliminary Report”).3 I also ignore references to yog a ca ra s (practitioners of yoga) and purva ca rya s (earlier teachers),except when they occur within a larger argument attributed to Sautr an-tika.

II. Summary of Results

Of the nineteen positions that I have isolated,4 eleven involve Vasuban-dhu’s rejection ofdharma s that the Sarvastivadins classify as real andindependent entities, including a number of thecittaviprayuktasa µ sk a ra s(forces not associated with mind), theasa µ sk ® tadharma s (unconditioneddharma s), theanu s aya s (latent defilements), andvijñaptir u pa (manifestedmatter) andavijñaptir u pa (unmanifested matter). Vasubandhu often assertsthat the functioning ofbi ja s (seeds), themselves merely prajñapti (pro-visional entities), is sufficient to explain the phenomena in question. Other positions concern the process of perception, the nature of the Buddha’sknowledge, the reality of the past, and the possibility of a fall from arhat-ship. In a number of these cases, too, Vasubandhu appeals to eitherbi jaor the closely related idea ofsaµ tatipari ∞amavi s eÒa (transformation of thelife-stream)5 in his unorthodox statements.

One of the nineteen positions seems to have nothing at all corre-sponding to it in theYoga ca rabh umi. In two other cases, theYoga ca ra-bh umi contains no argument similar to Vasubandhu’s, but its generalposition on the subject is in agreement with his. In the remainingsixteen cases, a correspondence between the two texts is more or less

clear.Six of the positions identified as Sautr antika by Vasubandhu are attrib-uted to Dar Ò†antika by the *Vibh a Òa ; in three other cases, statementsrelated to the Sautr antika positions are attributed to Dar Ò†antika. How-

332 ROBERT KRITZER

3 In an ongoing project, I am comparing every passage that Saµ ghabhadra identifiesas the position of theching-chu with theYoga ca rabh umi and compiling a list of correspon-dences. So far, I have completed the comparison for the first four chapters of the Abhi-dharmako s abh a Ò ya ; the results of the first three chapters have been published (Kritzer Comparison ).

4 Kato identifies 17 (74-78).5 For this translation, see Cox (95).



consciousness simply arises in dependence on the organ and the object.There is no action here, merelydharma s, merely causes and results. How-ever, in worldly discourse one can say that the eye sees.9

*Vibh a Òa 10 – The *Vibh a Òa attributes to the Dar Ò†antikas the positionthat the coming together of certain factors is equivalent to “seeingru pa .”11

*Tattvasiddhi 12 – Consciousness sees, not the organ.

Yoga ca rabh umi13 – TheYoga ca rabh umi contains a number of statementsto the effect that cognition is really the result of the laws of cause andeffect, not of something seeing and something else being seen. In particu-lar, the Vini s cayasa µ graha ∞i on the Pañcavijñ a nak a yamanobh umi saysthat, at the highest level, neither the organ nor the consciousness per-

334 ROBERT KRITZER

9 Saµ ghabhadra identifies this as the opinion of the Dar Ò†antikas. He does not use theappellation sutra-master (T. 1562: 367b25; see Kato 75). Fukuhara comments that this isvery close to a Mahayana way of thinking (159).

10 (T. 1545: 61c10-11).11 However, Kato shows that the Dar Ò†antika position in the *Vibh a Òa , which is fully

explained in thePañcavastukavibha

Òa s a

stra , is actually completely different from theSautr antika position in the Abhidharmako s abh a Ò ya , since the factors that come together in the Dar s †antika position are consciousness anddharma s associated with mind, not organ,object, and consciousness (23-24).

12

(T. 1646: 267a7-9; the argument continues until268a10).

13 mig gis gzugs rnams mtho n ba nas yid kyis chos rnams rnam par s es so zes bya ba’ibar du ji skad gsu ns pa de la / ci mig la sogs pas mtho n ba nas rnam par s es pa’i bar du

yin nam / ’on te de dag gi rnam par s es pa dag gis mtho n ba nas rnam par s es pa’i bardu yin ze na / smras pa / don dam par ni mig la sogs pas kya n ma yin la / de dag gi rnam

par s es pa dag gis kya n ma yin no / de ci’i phyir ze na / d nos po rnams ni rten ci n ’brelbar ’byu n ba’i phyir da n skad cig pa’i phyir da n / g.yo ba med pa’i phyir ro / brda’i tshuldu ni gtso bo yin pa’i mig la sogs pa la mtho n ba po la sogs pa ñe bar gdags pa ches rigsso / de ci’i phyir ze na / mig la sogs pa dba n po rnams yod na ni rnam par s es pa ’byu nba ma tsha n pa med par nes kyi / rnam par s es pa’i rgyun ni yod du zun kya n dmig lasogs pa dba n po rnams tsha n ba ’am ma tsha n bar dmigs pa’i phyir ro / lta ba la sogs patsam la mtho n ba la sogs pa ñe bar gdags pa ga n yin pa de ni don dam pa yin no (Yoga ca ra-bhumi t : zi 83a6b3).

(T. 1579: 610a19-27; cited in Saeki 1: 88). See also theParam a rthag a thas and their commentary (Wayman 168, 174, 178;Yoga ca rabh umi t : dzi 236b3-4, 238a4-6; T. 1579:363a27-b1, 364a27-b1).



ceives. It gives three reasons: because thesvabh a va (real nature) of alldharma s arises due to a multiplicity of causes; because nothing enduresfor more than a moment; because there is no real action. At the worldlylevel, however, one can say that the eye sees, because whenever there isan organ, consciousness will definitely not be lacking. On the other hand,it is possible for the organs to be lacking even when the stream of con-sciousness exists.

Comment – Here the Sautr antika argument closely follows theYoga ca ra-bhumi in its ultimate rejection of anything that perceives and its accep-tance on the worldly level of the notion that it is the organ that perceives.The conclusion of the*Tattvasiddhi is completely different.

2. P ® thagjanatvam (the state of being an ordinary person) is not a realdharma . It is simply thesa µ tati (life-stream) in which thea ryadhar-mas (the attributes of a noble or spiritually accomplished person) havenot yet arisen.

Abhidharmakos abh

a Ò ya

14

– Vasubandhu approves of the Sautr antikadefinition of p ® thagjanatvam , according to which p ® thagjanatvam is the

sa µ tati in which thea ryadharma s have not yet arisen.15

*Vibh a Òa 16 – The *Vibh a Òa attributes a denial of the real existence of p ® thagjanatvam to the Dar Ò†antikas.

SAUTR A NTIKA IN THE ABHIDHARMAKO S ABH A∑YA 335

14 eva µ tu s a dhu yath a sautr a ntik a na m / katha µ ca sautr a ntik a na m / anutpann a rya-dharmasantati Ì p ® thagjanatvam iti ( Abhidharmako s abh a Ò ya : 66.16-18; T. 1558: 23c2-3;La Vallée Poussin 1: 193; Saµ ghabhadra identifies this as the opinion of the sutra-master and criticizes Vasubandhu for denying the real existence of p ® thagjanatvam [T. 1562:399b10-c7; Cox 203-206; Kato 75].)

15 Cox points out that Vasubandhu does not specifically state here that p ® thagjanatvamis unreal, but she says that its unreality is implied in its definition as asa µ tati , “which,as a composite entity, cannot be real” (224 n. 109).

16 (T. 1545: 231b26-27; see Cox 224 n. 109).17 (T. 1646: 289c3-4; the

argument continues until 289c13).18 Katsura (86) cites T. 1646: 289a-c, in which variouscittaviprayukta s are said to lack

separate existence.



*Tattvasiddhi 17 – There is no p ® thagjanatvam different from the p ® thag- jana , the ordinary person himself. (The *Tattvasiddhi says that all of thecittaviprayuktasa µ sk a ra s are prajñapti .)18

Yoga ca rabh umi19 – TheVini s cayasa µ graha ∞i on thePañcavijñ a nak a ya-manobh umi defines p ® thagjanatvam as a designation for the state in whichthe lokottara (supermundane) aryadharma s have not yet arisen.Comment – The definitions of p ® thagjanatvam in the Abhidharmako-s abh aÒ ya and theYogaca rabh umi are essentially identical. While it is pos-sible that the Dar Ò†antika position is the source for the common defini-tion in these two texts, the *Vibh a Òa does not give us any details.

3. Thesa µ sk ® talak Òa∞as (marks of the conditioned) are not realdharma s.

Abhidharmako s abh aÒ ya20 – Thesaµ sk ® talak Òa∞as are not real entities since,unlike realdharma s such asru pa , they cannot be known by perception,inference, or scripture.*Vibh a Òa 21 – The *Vibh a Òa attributes a denial of the real nature of the

sa µ sk ® talak Òa∞as to the Dar Ò†

antikas.

336 ROBERT KRITZER

19 so so’i skye bo gnas skabs ga n la gdags / rnam pa du yod ce na / smras pa / ’jig rtenlas ’das pa ’phags pa’i chos ma bskyed pa’i gnas skabs la’o (Yoga ca rabh umit : zi 77a8).

(T. 1579: 607c8-9). The other passage in theVini s cayasa µ graha ∞i that deals with thecittaviprayuktasa µ sk a ra s says that p ® thagjanatvam refers to the seeds ofdar s anaheyadharma s in the three worlds that have not yet been destroyed (Yoga ca rabh umit : zi 26b1-2;T. 1579: 587b25-26). I have argued that there is no contradiction between the two defi-nitions in theVini s cayasa µ graha ∞i, or between this and the one favored by Vasubandhuin the Abhidharmako s abh a Ò ya (Kritzer Rebirth 246-248).

20 tad etad a k a s aµ pa † yata iti sautr a ntik a Ì / na hy ete j a tya dayo dharm a dravyata Ìsaµ vidyante yath a ’bhivyajyante / ki µ k a ra ∞am / pram a ∞a bha va t / na hy e Òa µ dravyato ’stitvekiµ cid api pram a ∞am asti pratyak Òam anum a nam a pt a gamo v a yath a ru pa d i na µ dharm a ∞a miti ( Abhidharmako s abh a Ò ya : 76.20-23; T. 1558: 27b24-26; La Vallée Poussin 1: 226; Sa µ-

ghabhadra identifies this as the opinion of the s u tra-master [T. 1562: 406b16-20; Cox 311-312] and criticizes it on the grounds that Vasubandhu must accept the provisional reality of the saµ sk ® talak Òa∞as; according to Sa µ ghabhadra, their provisional reality cannot be provenby perception or scripture, while proof by inference of their provisional reality would implyproof by inference of their ultimate reality [T. 1562: 406b20-29; Cox 312]; Kat o 75.)

21

(T. 1545: 198a14-17; see also Cox 358 n. 32).22



*Tattvasiddhi 22 – J a ti (birth),vya ya (destruction),sthiti (continued exis-tence), andanyath a tva (change of state, i.e., aging) simply refer to the fiveskandha s at various points. They are not separatedharma s.

Yoga ca rabh umi23 – TheVini s cayasa µ graha ∞i on thePañcavijñ a nak a ya-manobh umi says that the foursaµ sk ® talak Òa∞as, including j a ti, are not realentities, separate fromru pa and the otherskandha s.Comment – Later in his presentation of the Sautr antika argument, Vasu-bandhu gives his own explanation of the fourlak Òa∞as: the first arising of

the series ofsaµ sk a ra s is j a ti; the series in the state of cessation is calledvyaya (=anityat a [impermanence]); the procession of the series is called


(T. 1646: 289b18-29).This position closely resembles Vasubandhu’s, except that Vasubandhu refers tosa µ sk ® ta-dharma s, notskandha s: tatra prav ahasy adir utp ado niv ® ttir vyaya Ì / sa eva prav aho ’nuvart-tam a na Ì sthiti Ì / tasya p urva paravi s eÒa Ì sthityanyath a tvam ( Abhidharmako s abh a Ò ya :77.6-8; T. 1558: 27c11-12; La Vallée Poussin 1: 227; Saµ ghabhadra identifies this as the

opinion of the sutra-master, who, he says, is following the school of the Sthavira [T. 1562:407c9-11; Cox 320] and criticizes it at length [T. 1562: 407c17-408b28; Cox 321-326]).

Vasubandhu’s position, in turn, resembles that of theVini s cayasa µ graha ∞i on thePañca-vijñ a nak a yamanobh umi: de lta bas na skye ba la sogs pa ya n ’du byed rnams la [bta ]gs

pa’i yod pa yin par rig par bya’o / de la rgyu yod na ra n gi mtshan ñid s non ma byu n ba’grub pa ni ’du byed rnams kyi skye ba zes bya’o / s na ma las phyi ma g zan ñid du g zandu ’gyur ba ñid ni ’du byed rnams kyi rga ba zes bya’o / skye ba’i dus tsam la gnas pa ni’du byed rnams kyi gnas pa zes bya ste / de lta bas na skye ba’i skad cig gi ’og tu ’jig pa’iskad cig ni ’du byed rnams kyi ’jig pa zes bya’o (Yoga ca rabh umit : zi 22a2-4).

(T. 1579: 585c24-28; seeKritzer Rebirth 234-235).

23 ci’i phyir gzugs la sogs pa ’du byed rnams las skye ba da n / rga ba da n / gnas pada n / mi rtag pa ñid dag rdzas g zan du yod pa ma yin par kho n du chud par bya ze na(Yogaca rabh umit : zi 21b1-2). (T. 1579: 585c9-10 [this question is answered in the passage that follows: T. 1579: 585c10-28;Yoga ca ra-bhumi t : zi 21b2-22a4]).

24 tatra prav a hasy a dir utp a do niv ® ttir vyaya Ì / sa eva prav a ho ’nuvarttam ana Ì sthiti Ì / tasya p urva paravi s eÒa Ì sthityanyath a tvam ( Abhidharmako s abh a Ò ya : 77.6-8; T. 1558:27c11-12; La Vallée Poussin 1: 227; Saµ ghabhadra identifies this as the opinion of thesutra-master, who, he says, is following the school of the Sthavira [T. 1562: 407c9-11; Cox320] and criticizes it at length [T. 1562: 407c17-408b28; Cox 321-326].) See KritzerCom-

parison 39.25 de lta bas na skye ba la sogs pa ya n ’du byed rnams la [bta ]gs pa’i yod pa yin par

rig par bya’o / de la rgyu yod na ra n gi mtshan ñid s non ma byu n ba ’grub pa ni ’du byed



sthiti ; the difference between earlier and later moments of the stream iscalledanyath a tva .24 This explanation is very similar to one found later inthe passage from theVini s cayasa µ graha ∞i on thePañcavijñ a nak a yama-nobh umi mentioned above:25 when, due to causes, formerly non-existentsa µ sk ® tadharmas (conditioneddharma s) arise, this is called j a ti.26 Whenthe sa µ sk ® tadharmas that arise later are different from the earlier ones,this is called jar a (old age). When thesesa µ sk ® tadharma s, having arisen,persist for a limited time, this is calledsthiti . And when, after the momentof arising, the characteristics of thesesa µ sk ® tadharma s are destroyed,

this is called cessation oranityat a .Again, the basic positions of the Sautr antika in the Abhidharmako s-abh a Ò ya and of theYoga ca rabh umi agree with those of Dar Ò†antika andHarivarman. However, we have no record of the Dar Ò†antika explanationof the individuallak Òa∞as, and Harivarman does not use expressions likeabh utva bha vati (“not having existed, it exists”), which are found through-out the discussions in the Abhidharmako s abh a Ò ya and theYoga ca rabh umi.

4. The Buddha did not say that an asa µ sk ® tadharma can be a cause. Abhidharmako s abh a Ò ya 27 – The Sautr antikas deny that the Buddha saidthat anasa µ sk ® ta could be a cause. On the contrary, he said that all causesare impermanent and hencesa µ sk ® ta .

338 ROBERT KRITZER

rnams kyi skye ba zes bya’o / s na ma las phyi ma g zan ñid du g zan du ’gyur ba ñid ni ’dubyed rnams kyi rga ba zes bya’o / skye ba’i dus tsam la gnas pa ni ’du byed rnams kyignas pa zes bya ste / de lta bas na skye ba’i skad cig gi ’og tu ’jig pa’i skad cig ni ’dubyed rnams kyi ’jig pa zes bya’o (Yoga ca rabh umi t : zi 22a2-4).

(T. 1579: 585c24-28; not in Tibetan;see Kritzer Rebirth 234-235). See KritzerComparison 39.

26 Miyashita finds in theYoga ca rabh umi the origin of the pen wu chin you ( )theory in the Abhidharmako s abh a Ò ya .

27 naiva hi kvacid asa µ sk ® ta µ bhagavat a hetur ity uktam/ ukta µ tu pary a ye∞a heturiti sautr a ntik a Ì / katham uktam / ye hetavo ye pratyay a ru pasyotp a d a ya te ’py anity a Ì / anity a n khalu hetupratyay a n prat i tyotpanna µ ru pa µ kuto nitya µ bhavi Ò yati / eva µ ya vad vijñ a nam iti ( Abhidharmako s abh a Ò ya : 91.13-17; T. 1558: 33c22-26; La Vallée Poussin 1:277; not mentioned in Kato).

28 (T. 1545: 103c21-23).



*Vibh a Òa 28 – The *Vibh a Òa mentions an opinion thatasa µ skrta s cannotbe k a ra ∞ahetu , but it does not attribute it to any specific group.*Tattvasiddhi – (nothing relevant)

Yoga ca rabh umi – (nothing relevant)

Comment – This sentence marks the beginning of a very long passagein which Vasubandhu criticizes the VaibhaÒika definitions ofasa µ sk ® ta-dharma s as real entities. Although theYoga ca rabh umi does not seemto include any statement similar to this one, see the following item for correspondences between the Abhidharmako s abh a Ò ya and theYoga ca ra-bhumi regarding the unreality of theasa µ sk ® ta s.

5. The asa µ sk ® tas are not real and separatedharma s.

Abhidharmako s abh a Ò ya 29 – According to the Sautr antika, theasa µ sk ® tasare not real and separatedharma s like ru pa, vedan a (feeling), etc.

*Vibh a Òa (1)30 – The *Vibh a Òa quotes the Bhadanta as saying thata k a s a(space) is prajñapti 31 and refutes him.


29 sarvam ev a sa µ sk ® tam adravyam iti sautr a ntik a Ì / na hi tad r u pavedan a divat bh a-va ntaram asti ( Abhidharmako s abh a Ò ya : 92.3-4; T. 1558: 34a12-14; La Vallée Poussin 1:278; Saµ ghabhadra identifies this as the statement of the sutra-master [T. 1562: 429a21-23] and criticizes it [T. 1562: 429a28]; Kato 75-76).

30

(T. 1545: 388c24-29). The other *Vibh aÒa texts (T. 1546and T. 1547) attribute this position to Buddhadeva and Dharmatr ata, respectively (Kato 22,128 n. 47).

31 The Samyukt a bhidharmah ® daya attributes a similar opinion to Dar Ò†antika:(T. 1552: 944a7-9).

32

(T. 1545: 161a10-12).33 (T. 1646: 343c12-

14 [the argument is similar to that of theVini s cayasa µ graha ∞i on thePañcavijñ a nak a ya-manobh umi; see below]);

(T. 1646: 343c17-18).(T. 1646: 369a23-25 [a similar denial of the real existence ofnirodhasam a patti together with an admission that it is not totally nonexistent is found in theVastusa µ graha ∞i ; seebelow]).



*Vibh a Òa (2)32 – The *Vibh a Òa attributes to the Dar Ò†antikas the opinionthat three types ofnirodha (cessation), including pratisa µ khya nirodha(cessation resulting from knowledge) andapratisa µ khya nirodha (cessa-tion not resulting from knowledge), are not real and refutes them.

*Tattvasiddhi 33 – Ak a s a and nirv a ∞a are not realdharma s.

Yoga ca rabh umi (1)34 – TheVini s cayasa µ graha ∞i on thePañcavijñ a na-k a yamanobh umi says thata k a s a is simply an appellation expressing theabsence ofru pa . If some place does not contain anything, the notion

arises that the place containsa k a s a . Therefore,a k a s a is only a prajñaptiand is not real.Yoga ca rabh umi (2)35 – The Vastusa µ graha ∞i , in a definition of * pha-laprajñapti ( , provisionally real by way of being a result), mentions

pratisa µ khya nirodha , which, it says, is not non-existent, since it is a resultof the path, but is not really existent, since it is simply a designation for

340 ROBERT KRITZER

34 de la nam mkha’ ga n ze na / gzugs med pa tsam gyis rab ti phye ba ni nam mkha’

yin te / ’di ltar ga n la gzugs yi rnam pa mi dmigs pa de la nam mkha’i ’dus es ’byu n bar’gyur pas de’i phyir de ya n btags pa’i yod pa yin par rig par bya’i rdzas su ni ma yin no

(Yoga ca rabh umi t : zi 39b1-2).(T. 1579: 593a15-18).

See alsoVastusa µ graha ∞i (T. 1579: 879a14-18; not in Tibetan). Yamabe has noted thesimilarity between this passage and the Sautr antika opinion in the Abhidharmako s abh a Ò ya(personal communication).

35

(T. 1579: 879a5-8; not in Tibetan). The Hsien-yang sheng-chiao lungives a definition of pratisa µ khya nirodha that is similar but mentions prajñ a :

(T. 1602: 484c3-4).36 so sor btags pa ma yin pa’i ’gog pa ga n ze na / de las g zan pa skye ba’i rkyen m non

du gyur pa na de las g zan pa skye bas / de las g zan pa mi skye zin ñe bar zi ba’i ’gog patsam ni so sor btags pa ma yin pa’i ’gog pa zes bya ste / ga n de’i tshe na ma skyes s inskye bas’i dus las thal ba de ni de’i tshe na ma ya n skye bar mi ’gyur bas / de’i phyir de

ya n btags pa’i yod pa yin gyi rdzas su yod pa ni ma yin te / de’i ra n gi mtshan ñid ni g zancun zad kya n mi dmigs so / de ya n chos kyi rnam pa da n ma bral ba’i phyir dus g zan gyitshe rkyen da n phrad na ’byu n bar ’gyur bas de’i phyir so sor btags pa ma yin pa’i ’gog

pa de ni gtan du ba ma yin no (Yoga ca rabh umi t : zi 39b2-5).

(T. 1579: 593a19-25). See alsoVastu-sa µ graha ∞i (T. 1579: 879a18-20; not in Tibetan). Yamabe has noted the similarity betweenthis passage and the Sautr antika opinion in the Abhidharmako s abh a Ò ya (personal commu-nication).



the total non-arising in the future ofkles as (defilements) that have alreadybeen destroyed.Yoga ca rabh umi (3)36 – TheVini s cayasa µ graha ∞i on thePañcavijñ a na-k a yamanobh umi says that apratisa µ khya nirodha is simply an appella-tion expressing destruction or pacification when a certaindharma , itsconditions for arising having been actualized, does not arise becauseanother dharma arises. Whenever the time for adharma to arise isexceeded, thatdharma is destroyed and will not arise. (However, if theconditions for arising are encountered, it may arise in the future, so thisis not a perma-nent destruction.) Because it has no separatesvalak Òa∞a (characteristicmark),apratisa µ khya nirodha is a prajñapti , not a real entity.

Comment – I have not found any passage in the Yoga ca rabh u mi that explic-itly states that the category ofasa µ sk ® ta is not really existent. However,in the passages referred to, the Yoga ca rabh u mi questions or denies the realexistence ofa k a s a , pratisa µ khya nirodha , andapratisa µ khya nirodha . Itsdefini tion ofa k a s a is very similar to that of the Sautr antika, according towhom a k a s a is nothing more than the absence of that which is tangible. 37 Thedefinitions of pratisa µ khya nirodha are also similar, although the Sautr a ntikadef inition stresses the role of knowledge,38 which is not mentioned in theYoga ca rabh umi. Finally, the Sautr antika definesapratisa µ khya nirodha asthe non-arising ofdharma s due not to knowledge but to a lack of causesfor their arising.39 Like the definition in theYogaca rabh umi, this insists that


37 spra Ò† avya bh a vam a tram a k a s am / tadyath a hy andhak a re pratigh a tam avindantaa k a s am ity a hu Ì ( Abhidharmako s abh a Ò ya : 92.4-5; 34a14-16; La Vallée Poussin 1: 279;Saµ ghabhadra identifies this as the statement of the sutra-master [T. 1562: 429a23-25] andcriticizes it at great length [T. 1562: 429a28-430a7]). See KritzerComparison 53.

38 utpann a nu s ayajanmanirodha Ì pratisa µ khya balen a nyasy a nutp a da Ì pratisa µ khya -nirodha Ì ( Abhidharmako s abh a Ò ya : 92.5-6; T. 1558: 34a17; La Vallée Poussin 1: 279;Saµ ghabhadra identifies this as the statement of the sutra-master [T. 1562: 429a25-26] andcriticizes it at very great length [T. 1562: 430a18-434b7]). See KritzerComparison 54.

39 vinaiva pratisa µ khyay a pratyayavaikaly a d anutp a do ya Ì so ’pratisa µ khya nirodha Ì / tad yath a nik a yasabh a ga s eÒasy a ntar a mara ∞e ( Abhidharmako s abh a Ò ya : 92.7-8; T. 1558:34a18; La Vallée Poussin 1: 279; Saµ ghabhadra identifies this as the statement of the sutra-master [T. 1562: 429a26-27] and criticizes it at great length [T. 1562: 434b8-435b2]). SeeKritzerComparison 55.

40 naimittiko hi n a ma bhagav a n sy a d eva µ sati na puna Ì sa k Òa tk a r i / tasm a t sarvamicch a ma tre ∞a bhagav a n j a na t i ti sautr a ntik a Ì / acintyo hi buddh a na µ buddhavi Òaya ity



apratisa µ khya nirodha , as the non-existence of something, can only be a prajñapti , not a realdharma . However, the phrasing is somewhat differ-ent, and I am not sure that the two definitions are completely in agreement.

6. The Buddha knows the future directly.

Abhidharmako s abh a Ò ya 40 – Concerning the question of the Buddha’sknowledge of the future, the Sautr antikas say that the Buddha knows itdirectly. Vasubandhu adds that the Buddha knows by merely wishing andexplains that the Lord has said, “the Buddha-range of the Buddhas isacintya (unimaginable).”

*Vibh a Òa 41 – The *Vibh a Òa refutes two other theories, which it doesnot attribute to a specific group, of how the Buddha knows the future (byinference or by means of a mark in beings’sa µ tati s that indicates thefuture results of their actions) and accepts the idea that he knows itdirectly.

342 ROBERT KRITZER

ukta µ bhagavat a ( Abhidharmako s abh a Ò ya : 99.9-11; T. 1558: 37a2-4; La Vallée Poussin1: 304-305; Kato 76).41

(T. 1545: 51b15-c6).42

(T. 1646: 240a26-b10). See Kritzer “Unthinkable”69-71.

43 gzhan ya n dgra bcom pa ni sñoms par zugs na mñam par b zag pa yin la la ns na mñam par ma b zag pa yin gyi / de b zin g s egs pa la ni gnas skabs thams cad du sems mñam parma b zag pa med pa da n (Yoga ca rabh umi t : ’i 114a5-6);

(T. 1579: 738c7-9).



is an action that is free from desire pertaining toru pa , the stream of con-sciousness will evolve without requiringru pa as its support.*Vibh a Òa 46 – The *Vibh a Òa mentions an opinion that the stream of con-sciousness ina ru pyadh a tu does not have a support. It does not attributethis opinion to a specific school but simply refutes it with theA bhidhar-mika position mentioned above.*Tattvasiddhi 47 – Rebirth inru padh a tu (the subtle material realm) after the ru pasa µ tati (material continuity) has been interrupted by birth ina ru pyadh a tu

is mentioned as an example of something arising without a

344 ROBERT KRITZER

(T. 1646: 262b10-15).

(T. 1646: 266b7-12).48 gal te rnam par s es pa gzugs kyi sa bon da n ldan pa ma yin du zin na / so so’i skye

bo gzugs med pa rnams su skyes pa tshe zad ci n las zad nas de nas ’ci ’pho zin ya n ’ogtu skye pa’i gzugs kyi sa bon med pas ’byu n bar mi ’gyur ba zig na ’byu n ste / de lta basna gzugs kyi sa bon da n ldan pa’i rnam par s es pa de la brten nas / de’i gzugs ’byu n barrig par bya’o

(Yoga ca rabh umi

t : zi 16b4-6; see Schmithausen 21, 288 n. 172 b).(T. 1579: 583c7-10). An

explanation of the mutual dependence ofvijñ a na and na mar u pa in the Savitark a dibh umi prat i tyasamutp a da exposition contains a somewhat similar explanation of howru pa canresume: a ru pye Òu punar n a ma s rita µ ru pa µ bi j a s rita µ (corrected from j i va s rita µ on thebasis of the manuscript by Schmithausen [ 469 n. 1135])ca vijñ a na µ vijñ a na s rita µna mar u pab i ja µ ca pravartate / yata Ì punar b i j a t samucchinnasy a pi r u pasy a yaty a µ

pr adurbh a vo bhavaty ayam atr a pi vi s eÒa Ì (Yoga ca rabh umi: 200.1-3).gzugs med pa rnamsna ni rnam par s es pa mi n la brten ci n / gzugs kyi sa bon la ya n brten la / gzugs kyi sabon da n min ya n rnam par s es pa la brten ci n ’jug ste / ’di ltar gzugs yo ns su chad pa las

phyis sa bon las ’byu n bar ’gyur te / ’di la ya n bye brag de yod do (Yoga ca rabh umit : dzi

116b1-2).(T. 1579: 321b14-17; see YamabeYugashichiron ). TheVastu-

sa µ graha ∞i also says that thebi ja s of ru pa exist in a ru pyadh a tu consciousness:gzugsmed pa dag ni rnam par s es pa mi n la ya n brten la gzugs kyi sa bon la ya n brten to / mi nda n gzugs kyi sa bon ya n rnam par s es pa la brten ci n ’dug ste / gzugs kyi rgyun chad zin

pa las kya n gzugs kyi sa bon de las phyi ma la ’byu n bar ’gyur te / ’di la ya n bye brag ni[Derge addsde] yod do (Yoga ca rabh umit : ’i 285b7-286a1).

(T. 1579: 827c29-828b3).49 tatra v a san a hetvadhi Ò† ha nam adhi Ò† ha ya k Òepahetu Ì prajñ a pyate / tat kasya heto Ì /

tath a hi / s ubh a s ubhakarmaparibh a vit a Ì sa µ sk a ra s traidh a tuke Ò† a niÒ† agati Òv iÒ† a niÒ† a tma-bha va n a k Òipanti. (Yoga ca rabh umi: 107.20-108.2).de la rgyu’i gnas bag chags la brtennas / ’phen pa’i rgyu ’dogs par byed de / de ci’i phyir ze na / ’di ltar dge ba da n / mi dge



cause. Consciousness ina ru pyadh a tu is said to be without support:dhar-ma s are able to exist without support.Yoga ca rabh umi (1)48 – TheVini s cayasa µ graha ∞i on thePañcavijñ a na-k a yamanobh umi says that, if consciousness did not contain the seeds of ru pa , rebirth after falling froma ru pyadh a tu would be impossible.

Yoga ca rabh umi (2)49 – The Savitark a dibh umi specifies that thesa µ sk a ra s perfumed bykarma , i.e.,a k Òepahetu , which consists ofbi ja s,project a new lifetime in all three realms.

Yogaca rabh umi (3)50 – TheVinis cayasa µ graha ∞i on theCint amayi Bhumiincludesnik a yasabh a ga and j i vitendriya , along with the othercittavipra-

yuktasa µ sk a ra s, in a long list ofsa µ sk ® tadharma s that are prajñapti andthus not real.

Comment – The issue here is the support of consciousness ina ru pya-dh a tu , where its usual support, the body, cannot exist. The VaibhaÒikasolution is that twocittaviprayuktasa µ sk a ra s, nik a yasabh a gat a and j i vi-tendriya , ensure that the disembodied consciousness doesn’t simply diein this realm. In theYoga ca rabh umi, on the other hand, consciousnessitself is the support. However, as Schmithausen points out, the explanation


ba’i las kyis yo ns su bsgos pa’i ’du byed rnams kyis khams gsum du sdug pa da n / mi sdug pa’i ’gro ba rnams su / sdug pa da n mi sdug pa’i lus rnams ’phen par byed pa da n / deñid kyi dba n gis phyi rol gyi d nos po rnams kya n / phan sum tshogs pa da n / rgud par ’gyurba’i phyir te / de bas na ’du byed rnams kyi dge ba da n mi dge ba’i las kyi bag chags labrten nas ’phen pa’i rgyu ’dogs so (Yoga ca rabh umit : dzi 64b5-8).

(T. 1579: 301b28-c3). For other relevant definitions ofa k Òepahetu in the Yoga ca rabh umi, see Kritzer Rebirth 155-165.

50 ’dus byas kyi mi n can gyi d nos po la skye ba da n / rga ba da n / gnas pa da n / mirtag pa da n / sa bon da n rnam par rig byed da n / rnam par rig byed ma yin pa da n / thob

pa da n / ’thob pa ma yin pa da n / srog gi dba n po da n / ris mthun pa da n / mi n gi tshogsda n / tshig gi tshogs da n / yi ge’i tshogs rnams da n so so’i skye bo ñid da n / tshogs pada n ma tshogs pa da n / ’jug pa so sor nes pa da n / sbyor ba da n / mgyogs pa da n / gorims da n / dus da n yul da n gra ns ñe bar ’dogs pa da n (Yoga ca rabh umi t : zi 208a4-6).

(T. 1579: 659a12-16).51 LamotteSaµ dhinirmocana 55 (5.2); Yoga ca rabh umi t : ’i 58a2-5; T. 1579: 718a17-

23; see Schmithausen 47, 320 ns. 329, 330.52 Yoga ca rabh umi t : zi 4a8-4b5; T. 1579: 580a2-12; Schmithausen 51.



of how consciousness acts as a support develops over the course of thecompilation of the text. InYoga ca rabh umi (1), quoted above, conscious-ness is taken to mean the six ordinary types of consciousness, which aresaid to contain the seeds ofru pa . Schmithausen thinks that the conceptof bi ja here, according to which consciousness and the material sensefaculties contain each other’s seeds, predates the theory ofa layavijñ a na(21, 285-288, ns. 170-172).

Schmithausen, on the other hand, also refers to two other passages, onein the Saµ dhinirmocanas utra, 51 the other in what he calls the “Prav ® tti

Portion ” of the Vini s cayasa µ graha ∞i on the Pañcavijñ a nak a yamanob-humi.52 He infers that in both passages thea layavijñ a na (store-conscious-ness) acts as thea s raya (support) of the new being ina ru pyadh a tu. If Iunderstand him correctly, he thinks these passages are significant becausethey are on different sides of a dividing line in the development of the con-cept ofa layavijñ ana . The relevant chapter of theSaµ dhinirmocana, accord-ing to Schmithausen, still conceives of thea layavijñ a na (or a d a navijñ ana[the appropriating consciousness]) as “sticking in the body” (50). However,it also states that, ina ru pyadh a tu, the a d a navijñ ana does not appropriatethe body, which does not exist there. Thus, thead anavijñ ana mentioned hererepresents an intermediate stage between the six ordinary consciousnessescontaining the seeds ofru pa and the more fully developeda layavijñ a na of the “Prav ® tti Portion ,” in which the association of thea layavijñ a na withthe physical body no longer applies (51). Thisa layavijñ a na supplants thephysical body as thea s raya of all beings in the realms in which a physi-cal body exists, as well as providing ana s raya for beings without bodies.

Vasubandhu’s claim that thecittasa µ tati (mental continuity) is a suffi-cient support for beings ina ru pyadh a tu shows that he thinks that con-sciousness can function asa s raya . However, it is difficult to show a clear connection between Vasubandhu’s statement and the passages I haveidentified in theYoga ca rabh umi. His statement about the projecting causeis perhaps related toYoga ca rabh umi (2), but the context of that passage,an explanation of the ten types of causes, is quite different. As for Yoga ca rabh umi (3), Vasubandhu’s implicit denial ofnik a yasabh a ga and

j i vitendriya is in line with theYoga ca rabh umi’s denial of the reality of

cittaviprayuktasa µ sk a

ra s. However, his focus here is not onnik a yasa-bha ga and j i vitendriya .

346 ROBERT KRITZER



The termcittasa µ tati suggests the six ordinary consciousnesses, whichpoints toYoga ca rabh umi (1), but Vasubandhu does not explain the mech-anism by which ordinary consciousness could again produce a physicalbody when a being is reborn in a lower realm. Elsewhere, in a well-knownpassage concerning the question of how consciousness can resume after unconscious states likenirodhasam a patti (trance of cessation), Vasuban-dhu quotes the opinion of the purva ca rya s, who make an analogy to thequestion of how the sense faculties and body can resume when one isreborn in a lower realm after a period ina ru pyadh a tu. According to these

purva ca rya s, the fact that consciousness and the sense faculties containeach other’s seeds answers both questions.53 This is the same theory of bi ja that underliesYoga ca rabh umi (1). However, in our current passage,Vasubandhu does not mention mutual seeding.

Nor does Vasubandhu refer to mutual seeding in yet another passagein which he explains the resumption ofru pa after rebirth froma ru pya-dh a tu into a lower realm. Here he states that the arising ofru pa is duesolely to consciousness, the consciousness that was impregnated by thevipa kahetu (cause of fruition) of thatru pa .54 Kato , who points out thatPur ∞avardhana identifies this as a Sautr antika opinion, thinks that thispassage is another expression of a theory of mutual seeding (261), andYamabe (Yugashichiron ) seems to agree with Kato . However, the wordbi ja does not appear. Nor does Vasubandhu refer to the other aspect of mutual seeding, the arising of consciousness fromru pa . I think that thispassage is more similar to the Sautr antika statement under discussion here(that there is no support for consciousness ina ru pyadh a tu besides con-sciousness itself) than to the opinion of the purva ca rya s.

Therefore, it is possible that Vasubandhu distinguishes between theidea of the purva ca rya s and that of the Sautr antikas. Since Vasubandhu


53 apare punar a hu Ì / katha µ t a vad a ru pyopapann a na µ ciraniruddhe ’pi r u pe punarapi r u pa µ j a yate / citt a d eva hi taj j a yate na r u pa t / eva µ cittam apy asm a d eva sendriy a t k a ya j j a yate na citt a t / anyonyab i jaka µ hy etad ubhaya µ yaduta citta µ ca sendriya s cak a ya iti p urva ca rya Ì ( Abhidharmako s abh a Ò ya : 72.18-21; T. 1558: 25c22-26; La ValléePoussin 1: 212; Saµ ghabhadra identifies this as the opinion of the sutra-master and criti-cizes this statement along with the seed theory that underlies it [T. 1562: 404a2-20; Cox273-274]).This passage and its relationship to theYoga ca rabh umi have been discussed byHakamaya, Schmithausen (285 n. 170), and Yamabe (Yugashichiron ).

54 ru pasy a citt a d evotpattis tadvip a kahetuparibh a vit a l labdhav ® ttitas ( Abhidharmako-s abh a Ò ya : 435.20: T. 1558: 146b2-3; La Vallée Poussin 5: 142).



does not use the termsa layavijñana or a d a navijñ ana , there is no obviousconnection between the Sautr antika position here and the passages men-tioned by Schmithausen. Nevertheless, Vasubandhu says that the beingsin a ru pyadh a tu have “no support other than the stream of conscious-ness”; he does not say that their support is the seeds ofru pa containedin the stream of consciousness. Thus it seems as though the Sautr antikastatement is based on a notion of a consciousness that has already “tran-scended its original feature of essentially being bound, and somehowsubordinate, to corporeal matter, and has rather in its turn become a fun-

damental constituent of personality” (Schmithausen 51). Furthermore, inthe Karmasiddhiprakara ∞a , Vasubandhu adduces the inability ofnik a-

yasabh a gat a and j i vitendriya to act as a support for consciousness ina ru pyadh a tu as proof that there exists a consciousness that can containseeds and that is different from the six ordinary consciousnesses, namelythe vipa kavijñ a na (maturation consciousness) ora layavijñ a na (LamotteTraité 198.34-199.13, 248-249.)55 If we admit the possibility that the termcittasa µ tati can stand fora layavijñ a na ,56 the connection between the pas-sage in the Abhidharmako s abh aÒ ya and theYoga ca rabh umi becomes more

348 ROBERT KRITZER

55 The same argument also is found in the Mah a ya nasa µ graha (Lamotte La Somme1: 39.1-4; 2: 61-62).

56 Schmithausen takes Odani to task for equatingcittasa µ tati witha layavijñ a na in thecontext of the Maul i bhumi of the Yoga ca rabh umi on the grounds that this is “inadmissi-ble if we are to understand the materials of theYoga ca rabh umi in their original sense, andnot from the point of view of later systematization” (342 n. 442). However, in the caseof the Abhidharmako s abh a Ò ya , Vasubandhu, unlike the author of the Maul i bhumi, waspresumably familiar with a concept ofa layavijñ a na that was systematized to at least some

degree. Therefore, it is not impossible that he intentionally substituted the termcittasa µ tati ,which was current inabhidharma texts, fora layavijñ a na , which, of course, was not.57 These are gods living in the seconddhy a na heaven ofru padh a tu . See La Vallée

Poussin 2: 18-20.58 sautr a ntik a vya cak Òate / s u tra ukta µ yath a te n a na tvasa µ jñina Ì / tatra ye sattv a a bha s-

vare devanik a ye ’ciropapann a bhavanti naiva sa µ varttan i kus ala na vivarttan i kus ala asya lokasyate t a m arci Òaµ d ®Ò† va bhi t a Ì santa udvijante sa µ vegam a padyante / sahaivai Òa ’rci Ì s u nyaµbr a hmaµ vima naµ dagdhv a ’rva g a gami Ò yat i ti / tatra ye sattv a a bha svare devanik a ye ciropa-

pann a Ì saµ varttan i kus ala vivarttan i kus ala s ca sya lokasya te t a n sattv a n bh i t a n a s va sayanti / m a

bhai Ò† a m a rÒa Ì ma bhai Ò† a m a rÒa Ì / pu rvam apy e Òa ’rci Ì s u nyaµ br a hmaµ vima naµ dagdhv a

’traiv a ntarhite ti / ato ’rci Ì a gamavyapagama sa µ jñitva t bh i t a bhi tasa µ jñitva c ca te n a na t-vasa µ jñino na sukh a duÌ kha sukhasa µ jñitv a d iti

( Abhidharmako s abh a Ò ya

: 116.16-23; T. 1558:43a10-19; La Vallée Poussin 2: 20; Sa µ ghabhadra identifies this as the position of the s u tra-master [T. 1562: 463b5] and criticizes it [T. 1562: 463b6-11]; Kat o 76).



likely. However, I shall have more to say later about the absence of theterm a layavijñ a na in the Abhidharmako s abh a Ò ya .

8. The A bhasvara gods57 have different ideas because some of them havethe idea of fear, while others do not.

Abhidharmako s abh a Ò ya 58 – The A bhasvara gods are said to have dif-ferent ideas because, at the time of the destruction of the universe, someof them have the idea of fear, while others do not. (According to theVaibhaÒikas, it is because their feelings alternate between pleasant and nei-ther-pleasant-nor-unpleasant.)*Vibh a Òa (1)59 – The *Vibh a Òa says that theA bhasvara gods have diffe-rent ideas, because their feelings alternate between pleasant and neither-pleasant-nor-unpleasant.*Vibh a Òa (2)60 – The *Vibh a Òa also mentions thesu tra 61 that states thatsome A bhasvara gods are not afraid of the conflagration.*Vibh a Òa (3)62 – It quotes the samesu tra describingA bhasvara gods asbeing afraid.*Tattvasiddhi – (nothing relevant)


59

(T. 1545: 707b2-6).60

(T. 1545:386b5-9).

61

The Saptas uryav a kara ∞a of the D i rgh agama (T. 1: 429a22-29; see La Vallée Poussin2: 20).62

(T. 1545: 690b21-25).63 ’od gsal gyi lha gnas na ni s na phyir skyes pa rnams / tsha ns pa’i ’jig rten tshig pa’i

me lce mtho n ba las ’jigs pa da n / mi ’jigs pa’i ’du s es su ’gyur bas / de dag ni ’du s esmi ’dra bar rig par bya’o (Yoga ca rabh umi t : dzi 211a5-6).

(T. 1579: 354c20-22).64 Prat i tyasamutp a da is the principle of conditioned origination, often expressed in a

twelve-membered formula; according to thea vasthika interpretation, each member of theformula represents a different state (a vasth a ) of the fiveskandha s.

65 atra tu sautr a ntik a vijña payanti / ki µ khalv et a iÒ† aya ucyante y a yasye Ò† ir a hosvit sutr a rtha Ì / s u tr a rtha ity a ha / yadi s utr a rtho nai Òa s u tr a rtha Ì / katha µ k ® tva / yat t a vad



Yoga ca rabh umi63 – According to theSrutamay i Bhumi, the A bhasvarashave different ideas because when they see Brahma’s conflagration, someare afraid and some are not.Comment – Vasubandhu and theYoga ca rabh umi clearly agree here. The*Vibh a Òa does not seem to recognize a controversy about what it meansfor these gods to have different ideas.

9. The a vasthika (static) interpretation of prat i tyasamutp a da 64 cannot be justified bysu tra .

Abhidharmako s abh a Ò ya 65 – Vasubandhu criticizes thea vasthika inter-pretation of prat i tyasamutp a da. He says that it cannot be justified bysu tra because thePrat i tyasamutp a das u tra is ni t a rtha (of explicit mean-ing).

*Vibh a Òa – (nothing relevant)*Tattvasiddhi – (nothing relevant)

Yoga ca rabh umi – (In the exposition of prat i tyasamutp a da in the Savi-t a r -k a dibh umi, the explanation of the individual members essentially followsand comments upon thePrat i tyasamutp adas utra [T. 1579: 322b2-324a15;Yoga ca rabh umi: 204.1-212.3; see Kritzer Rebirth 33-52 for a summaryof this section]. Due to the length of the passage, I do not provide the text.)

350 ROBERT KRITZER

uktam a vasthika e Òa prat i tyasamutp a do dv a da s apañcaskandhik a avasth a dva da s a nga ni tyetad uts utram / s utre ’nyath a nirde s a d / avidy a katam a / yat tat p urva nte ’jñ a nam iti vista-re∞a / yac ca n i t a rtha µ na tat punar neya µ bhavat i ti nai Òa s utr a rtha Ì ( Abhidharmako s a-bhaÒ ya : 136.14-18; T. 1558: 50a7-13; La Vallée Poussin 2: 75; Saµ ghabhadra identifiesthis as the opinion of the sutra-master, relates it to the last of the Sthavira’s six argumentsagainst thea vasthika interpretation, and refutes it [T. 1562: 495c22-496a10; Kato 76]).

66 See Kritzer Rebirth 183-189.67 sa ngi tipary a ye cokta µ mah a samudr a d aud a rik a Ì pr a ∞ino jal a t sthalam abhiruhya

sikat a sthale ’ ∞∂ a ni sth a payitv a sikat a bhir ava Ò† abhya punar api mah a samudre ’vataranti / tatra y a sa µ ma t r ˜ ∞a m a ∞∂ a ny a rabhya sm ® tir na mu Ò yate t a ny a ∞∂ a ni na p ut i bhavanti ya sa µ tu mu Ò yate t a ni p u tibhavanti / tad etan na var ∞ayanti sautr a ntik a Ì / m a bhut para-k i ye∞a ha re∞a ha ra iti / eva µ tu var ∞ayanti /ye Òa m a ∞∂ a na µ ma taram a rabhya sm ® tir namuÒ yate t a ni na p ut i bhavanti / ye Òaµ tu mu Ò yate t a ni p ut i bhavanti / tasy a Ì spar s avasth a ya Ìsmarant i ti ( Abhidharmako s abh aÒ ya : 154.4-7; T. 1558: 55b28-c1; La Vallée Poussin 2: 125;Kato 76-77).



Comment – Although theYoga ca rabh umi does not contain an explicitcriticism of the VaibhaÒika a vasthika interpretation, Vasubandhu seemsto follow theSavitark a dibh umi in relying on thePrat i tyasamutp adas utra. 66

10. The Sa µ gi tipary a ya is criticized for its statement that large sea-beings, after they lay their eggs on the shore, provide nourishment inthe form ofmana Ì sa µ cetan a (mental action or volition) by thinkingof their eggs.

Abhidharmako s abh a Ò ya67

– Vasubandhu criticizes theSaµ gi tipary a yafor its statement that large sea-beings, after they lay their eggs on the shore,provide nurture in the form ofmana Ì sa µ cetan a by thinking of their eggs.The Sautr antikas say that one being’s thought cannot be nourishment for another being. Instead, the eggs, by thinking of their mother, provide suchnourishment for themselves.68


68 La Vallée Poussin says that some commentators specify “Sautr antika” here, but so far

the only texts that I’ve found that use the word are the Sanskrit text of the Abhidharmako s-

abh a Ò ya and Paramartha’s translation (T. 1559: 212c8). Neither the Tibetan (gu 162b1-4)nor Hsüan-tsang’s translation mentions whose opinion this is. Yas omitra, Saµ ghabhadra,and Pu-kuang likewise fail to attribute the position to any teacher or school. Fa-pao men-tions only that it is Vasubandhu’s preferred opinion (T. 1822: 612b14-15). According toYamaguchi and Funahashi, none of the commentators attributes it to Sautr antika (343 n. 8).

Kato notes that Hsüan-tsang omits “Sautr antika” for reasons unclear. He also mentionsthat Saµ ghabhadra does not identify whose opinion this is (76-77).

As Saeki notes (2: 450), the opinion that Vasubandhu quotes from theSaµ gi tipary a yais an alternate opinion in the *Vibh a Òa (T. 1545: 676b20-21). But Saeki does not note thatthe opinion that Vasubandhu prefers, which is attributed to the Sautr antikas in the Sanskritand in Paramartha’s translation, is actually the preferred opinion of the *Vibh a Òa (T. 1545:676b16-20) and is what is said in theSaµ gi tipary a ya (T. 1536: 400c7-11). The *Vibh a Òa(T. 1545: 676b21-23) refutes the alternate opinion with the same argument as Vasubandhu’s.Yamaguchi and Funahashi do not mention this discrepancy.

The opinion favored by Vasubandhu is also given in the *Saµ yukt a bhidharmah ® daya(T. 1552: 952c8-9).

Van den Broeck suggests that Hsüan-tsang altered the text of the Saµ gi tipary a ya on the basisof Vasubandhu’s opinion, but he does not give any real basis for his suggestion (100 n. 7).

69

(T. 1545: 676b16-23).70 na sti sa µ sth a na µ dravyata iti sautr a ntik a Ì / ekadi nmukhe hi bh u yasi var ∞a utpanne

d i rgha µ ru pam iti prajñapyate / tam ev a pek Ò ya lp i yasi hrasvam iti / caturdi s aµ bhu yasi



*Vibh a Òa 69 – The *Vibh a Òa quotes the same passage from theSaµ gi tipa-rya ya and refutes it using the same argument as Vasubandhu.*Tattvasiddhi – (nothing relevant)

Yoga ca rabh umi – (nothing relevant)

Comment – Most texts do not consider this a Sautr antika opinion; fur-thermore, the actual position of theSaµ gi tipary a ya is not clear. There-fore, it is hard to know what to make of this passage.

11. Saµ stha nar u pa is a prajñapti.

Abhidharmako s abh a Ò ya 70 – Saµ stha nar u pa (matter as shape) is only a prajñ a pti , because if it were real, then a singlesa µ stha nar u pa would beperceived by tworu p i ndriya s (material sense organs), i.e., the eye, whichwould see the shape, and the faculty of touch, which would feel it. In fact,it is simply a designation for quantities ofvar ∞a (matter as color) arrangedin various ways.

*Vibh a Òa – (see the following item)

352 ROBERT KRITZER

caturasram iti / sarvatra same v ® ttim iti / eva µ sarvam / tad yath a ’la tam ekasy a µ dis ides a ntare Òv anantare Òu nirantaram a s u d ® s yam a na µ d i rgham iti prat i yate sarvato d ® s ya-ma na µ ma ∞∂ alam iti na tu khalu j a tyantaram asti sa µ sth a nam / yadi hi sy a t dvigr a hya µsya t cak ÒuÒa hi d ®Ò† va d i rgham ity avas i yate k a yendriye ∞a pi sp ®Ò† veti dv a bhy a m asyagraha ∞aµ pr a pnuy a t / na ca r u pa yatanasya dv abhya µ graha ∞am asti / yath a va spra Ò† avyed i rgh adigraha ∞aµ tath a var ∞e sa µ bhavyat am ( Abhidharmako s abh aÒ ya: 194.14-21; T. 1558:68b1-11; La Vallée Poussin 3: 8-9; Saµ ghabhadra identifies this as the opinion of the sutra-master [T. 1562: 535c23-536a4] and criticizes it at length [T. 1562: 536a4-b5]; not men-tioned in Kato).

71 athedan i µ k a yasya gati µ nir ak ® tya sa µ sthana µ ca tatra bhavanta Ì sautr a ntik a Ì k a µk a yavijñapti µ prajñapayanti / sa µ stha nam eva hi te k a yavijñapti µ prajñapayanti / na tu

punar dravyata Ì ( Abhidharmako s abh a Ò ya : 195.15-17; T. 1558: 68c8-9; La Vallée Poussin3: 12; Saµ ghabhadra says that the sutra-master is stating the opinion of his own school[T. 1562: 537a25-26] and criticizes it at length [T. 1562: 537a27-b13]; Kato 77). In thecourse of this argument, Saµ ghabhadra refers to Sautr antika three times (T.1562: 537b3, b7,b8). It seems as though he is all but identifying Sautr antika as the school of the sutra-master.

72 As Kato points out (77), the implication of the whole Sautr antika argument aboutvijñapti ( Abhidharmako s abh a Ò ya : 195.15-196.2) is that allkarma is cetan a . The *Vibh a Òaattributes such a position to the Dar Ò†antikas: (T. 1545:587a7-8; see Kato 71).

73(T. 1545: 634c26-28). We know that refers to the Dar Ò†antikas because



*Tattvasiddhi – (see the following item)

Yoga ca rabh umi – (see the following item)

Comment – (see the following item)

12. K a yavijñ a pti (physically manifested matter) issa µ stha na , which isa prajñ a pti .

Abhidharmako s abh aÒ ya71 – The Sautr antika answer to the Sarvastivadins’question regarding the nature ofk a yavijñ a pti is that it issaµ stha na , whichis, however, prajñ a pti , not dravya (substantially real).72

*Vibh a Òa 73 – The Dar Ò†antika objects: “Ifvijñapti and avijñapti are ru pa ,then what are blue, yellow, red, and white?” The *Vibh a Òa answers thatit is not the case that there is noru pa besidesvar ∞a . K a yavijñapti is sa µ-sth a na , not var ∞a.

*Tattvasiddhi 74 – Saµ sth a na is nothing other thanru pa (i.e., var ∞a ).If there is no color, there can be no perception of shape, while if shape


the stated purpose of the whole section is to refute the Dar Ò†antika opinion thatvijñaptiand avijñapti are unreal (T. 1545: 634c9-10).

74

(mistakenly for ?)

(T. 1646: 273a23-28).75 ri n po da n thu n nu ñid la sogs pa ga n dag dbyibs zes bya ba de dag kya n ci rdzas

su yod pa ’am / btags pa’i yod pa yin par brjod bar bya ze na / smras pa / btags pa’i yod pa yin par brjod par bya’o / de ci’i phyir ze na / bsags pa las gnas pa ni dbyibs s es byaba’i nes pa’i tshig yin pa da n bsags pa tsam dmigs pa da n / mtshan ñid las g zan pa’i donmi dmigs pa da n / bltos s in bltos na no bo ñid ’dres par ’gyur ba da n (Yoga ca rabh umi t :zi 56a1-3).

(T. 1579: 599b7-11). Yamabe has noted the similarity between this passage andthe Sautr antika opinion in the Abhidharmako s abh a Ò ya (personal communication).

76 sa µ sth a na µ katamat / yo r u papracayo d i rgh a di-paricched a k a ra Ì (Yoga ca rabh umi:5.2). dbyibs ga n ze na / ga n gzugs rgyas par ri n po la sogs par yo ns su bcad [correctedfrom gcad on the basis of the Derge] pa’i rnam pa’o (Yoga ca rabh umi t : zi 3a6).

(T. 1579: 279b8-9).77 de blos bye bas s in rta la sogs pa da n ’dra ba’i phyir ro (Yoga ca rabh umi t : zi 56a3).

(T. 1579: 599b11).



no further details about their argument.79 As for the nature ofkarma ,the Yoga ca rabh umi does not say, in so many words, thatkarma is essen-tially volition. However, a definition in theSavitark a dibh umi of thereal nature ofkarma seems to suggest the primacy of volition: “What isthe real nature ofkarma ? When a dharma arises, that which is charac-terized as (mental) determination also arises, and due to its arising, phys-ical and vocal determination proceed later. This is the real nature of karma .”80

13. Avijñapti does not really exist, for three reasons.

Abhidharmako s abh a Ò ya 81 – Vasubandhu attributes to the Sautr antikasthe statement thatavijñapti does not really exist, for three reasons: it issimply the non-performance of an action that one has undertaken not todo; it is a prajñapti based on pastmah a bhu tas (great elements, i.e., thefour types of matter [earth, etc.] in its most basic form), which them-selves do not exist; it lacks the characteristics ofru pa .


82 (T. 1545: 634c6-7; see above).83 (T. 1646:

290b9-10). (T. 1646: 304a17-18). See Katsura88.

84 dharm a yatanapary a panna µ pun a ru pa µ dvividha µ dravyasat prajñaptisac ca / yat prabh a vata Ì samadhigocaraµ nirmitavat tatphalaµ tadviÒayaµ tatpratisa µ yuktavij-ña navi Òaya µ ca tad dravyasat / sa µ var a sa µ varasa µ g ® hi ta µ tu prajñaptisat (accordingto Matsuda [personal communication], this passage appears in a Sanskrit manuscript frag-ment of theVini s cayasa µ graha ∞i preserved in St. Petersburg, and Matsuda has recon-structed it as above [non-italicized portions represent Matsuda’s reconstruction]).chos kyiskye mched du gtogs pa’i gzugs ni rnam pa gñis te / rdzas su yod pa da n btags pa’i yod

pa’o / mthu las byu n ba’i ti n ne ’dzin gyi spyod yul sprul pa lta bu de’i ’bras bu da n / de’i yul da n de da n mtshu ns par ldan pa’i rnam par s es pa’i yul ga n yin pa de ni rdzas su yod pa yin no (Yoga ca rabh umit : zi 51a8-b1).

(T. 1579:597b6-9).

85 These are special types ofavijñapti (Hirakawa 191-193).86 Yoga ca rabh umi t : zi 208a4-6; T. 1579: 659a12-16. See note 50 above.87 Yoga ca rabh umi t : zi 19a1-21b1; T. 1579: T. 1579: 584c18-585c8. Due to the length

of the passage, I have not included the text.



*Vibh a Òa 82 – The *Vibh a Òa attributes to the Da rÒ† a ntikas the argumentthat avijñapti is not real because, likevijñapti , it is notru pa in the waythat var ∞a is.

*Tattvasiddhi 83 – Avijñapti is a cittaviprayuktasa µ sk a ra , and thus it hasno separate existence.

Yogaca rabh umi (1)84 – TheVini s cayasa µ graha ∞i on thePañcavijñ anak a-

yamanobh umi says thatsaµ varar u pa (a form of matter produced by takinga virtuous vow) andasa µ varar u pa (a form of matter produced by taking

an evil vow)85 are merely prajñapti .Yoga ca rabh umi (2)86 – The Vini s cayasa µ graha ∞i on the Cint a may i

Bh umi includes avijñapti , along with thecittaviprayuktasa µ sk a ra s,vijñapti, and bi ja, as prajñapti s, saying that they are nominal designa-tions forsa µ sk ® tadharma s.

Yoga ca rabh umi (3)87 – TheVini s cayasa µ graha ∞i on thePañcavijñ a na-k a yamanobh umi denies the reality of pastdharma s.

Comment – As in the case ofvijñapti , Vasubandhu and theYoga ca ra-bhumi agree thatavijñapti is a prajñapti . Although theYoga ca rabh umidoes not give any reasons, Vasubandhu bases his second argument on adenial, which he shares with theYoga ca rabh umi, of the reality of the past.

14a. When the su tra mentionsr u pa that is invisible and not subject tocollision (apratigha ), it is referring not toavijñapti but to r u pa thatis produced by meditation.

Abhidharmako s abh a Ò ya 88 – The VaibhaÒikas give many different typesof arguments in support of the real existence ofavijñapti , but they arewrong. One argument in support is that thesu tra says that there are three

356 ROBERT KRITZER

88 atra sautr a ntik a a hu Ì bahv apy etac citram apy etat / naiva µ tv etat / yat t a vad ukta µ trividhar u pokter iti / tatra yog a ca ra upadi s anti / dhy a yina µ sam a dhivi Òayo r u pa µsam a dhiprabh a va d utpadyate / cak Òurindriy a viÒayatv a t anidar s anam / de s a na vara ∞atv a d apratigham iti / atha matam / katham id a ni µ tat r u pam iti / etad avijñaptau sam a nam( Abhidharmako s abh a Ò ya : 197.3-7; T. 1558: 69a29-b4; La Vallée Poussin 3: 18;S a µ g h a -bhadra identifies this as the opinion of the sutra-master [T. 1562: 540c22-24] and criti-cizes it [T. 1562: 540c24-541a8]).



types of ru pa , one of which is invisible and not subject to collision(apratigha ). According to the VaibhaÒikas, this must beavijñapti ( Abhi-dharmako s abh aÒ ya : 196.9-11; T. 1558: 69a2-4; La Vallée Poussin 3: 14).Vasubandhu, in making what he describes as the Sautr antika argumentagainst VaibhaÒika, quotes those who practice yoga ( yoga ca ra Ì ) as sayingthat, due to the power of meditation,ru pa that is the object of meditationis produced in meditators. Thisru pa is invisible because it is not theobject ofcak Òurindriya , and it is not subject to collision because it doesnot cover any place. Vasubandhu defends this statement against a possi-

ble Sarvastivadin objection.*Vibh a Òa – (nothing relevant)

*Tattvasiddhi – (nothing relevant)

Yoga ca rabh umi89 – TheVini s cayasa µ graha ∞i on thePañcavijñ a nak a ya-manobh umi includes, in the category of really existent,ru pa that, due tosupernatural power, is the object ofsam a dhi (meditation), like a magicalcreation,ru pa that is the result of thatsam a dhi, r u pa that is the object of

that sam a dhi , and ru pa that is the object of the consciousness associatedwith thatsam a dhi . This is contrasted withsa µ varar u pa and asa µ vara-ru pa , which are merely prajñapti .


89 Yoga ca rabh umi t : zi 51a8-b1; T. 1579: 597b6-9. See note 84 above.90 yad apy uktam an a sravar u pokter iti tad eva sam a dhiprabh a vasa µ bh uta µ ru pam

an a srave sam a dha v an a srava µ var ∞ayanti yog a ca ra Ì ( Abhidharmako s abh a Ò ya : 197.7-8;T. 1558: 69b4-6; La Vallée Poussin 3: 18-19; Saµ ghabhadra identifies this as the opinionof the sutra-master [T. 1562: 541a11-13] and criticizes it [T. 1562: 541a13-19]). Saµ gha-bhadra questions the identity of these yog a ca ra s and expresses surprise that Vasubandhuquotes from them in interpretingsu tra (T. 1562: 541a14-15).

91

(T. 1646: 343b17-19).92 tat puna Ì sam a dhigocara µ ru pa µ yatpratisa µ yuta Ì sam a dhi Ì tatpratisa µ yukt a ny

eva tanmah a bhut a ny up a d a ya laukika µ sa srav a na srava µ sam a dhim up a d a yotpadyate natu lokottara µ / saprapa µ ca k a rasam a dhihetukatv a t tasya (according to Matsuda [personalcommunication], this passage appears in a Sanskrit manuscript fragment of theVini s ca-

yasa µ graha ∞i preserved in St. Petersburg).tin ne ’dzin gyi spyod yul gyi gzugs de ya n tinne ’dzin ’byu n ba chen po dag rgyur byas pa’i gzugs ga n da n mtshu ns par ldan pa de dagñid da n de ya n mtshu ns par ldan pa yin no / ’jig rten pa’i ti n ne ’dzin zag pa da n



Comment – There is a clear correspondence here between the statementof those who practice yoga , quoted by Vasubandhu, and theYoga ca ra-bhumi.

14b. Ana sravar u pa (undefiled matter) is notavijñapti . Ther u pa producedby the power ofsam a dhi is an a srava if the sam a dhi in which it isproduced isan a srava.

Abhidharmako s abh a Ò ya 90 – Another Sarvastivadin argument in supportof the reality ofavijñaptir u pa is that thesu tra says that there is anan a s-ravar u pa . Vasubandhu again quotes those who practice yoga , who say thatthe ru pa produced by the power ofsam a dhi is an a srava if the sam a dhiin which it is produced isan a srava .

*Vibh a Òa – (nothing relevant)*Tattvasiddhi 91 – The *Tattvasiddhi does not make this argument con-cerningavijñapti , but in another context it refers to the fact that ana ru pyasam a dhi can produce anan a srava r u pa .

Yoga ca rabh umi92 – TheVini s cayasa µ graha ∞i on thePañcavijñ a nak a ya-manobh umi says that theru pa that is the object ofsam a dhi arises on thebasis of themah a bh u ta s associated with thatsam a dhi , and it arises onthe basis oflaukikasam a dhi (mundane meditation), whethersa srava or an a srava . However, it does not arise on the basis oflokottarasam a dhi(supermundane meditation) because it is caused by asam a dhi in which

prapañca (conceptual proliferation)93 is present.

358 ROBERT KRITZER

bcas pa da n zag pa med pa la brten nas skye ba yin gyi / ’jig rten las ’das pa las ni ma yin te / de ni spros pa’i rnam pa da n bcas pa’i ti n ne ’dzin gyi rgyu las byu n ba’i phyirro (Yoga ca rabh umi t : zi 51b1-3).

(T. 1579:597b9-12).

93 For this translation, I follow Ña∞ananda as quoted in Schmithausen’s long note onthe term (509 n. 1405).

94 yad apy ukta µ pu ∞ ya bhiv ® ddhivacan a d iti tatr a pi p urva ca rya nirdi s anti dharmat ahy eÒa yath a yath a d a t r ˜ ∞a µ d a ya Ì paribhujyante tath a tath a bhokt r ˜ ∞a µ gu∞avi s eÒa d anu-grahavi s eÒa c c a nyamanas a m api dat r ˜ ∞a µ tad a lambanad a nacetan a bha vit a Ì sa µ tataya Ìsuk Òmaµ pari ∞amavi s eÒaµ pr a pnuvanti yen a yaty aµ bahutaraphal abhini Ò pattaye samarth abhavanti ( Abhidharmako s abh a Ò ya : 197.15-19; T. 1558: 69b14-20; La Vallée Poussin 3:20; Saµ ghabhadra identifies this as the opinion of the sutra-master [T. 1562: 541c8-14]and explains and criticizes it at very great length [T. 1562: 541c14-542b6]).



Comment – The passage in theYogaca rabh umi implies what Vasubandhustates more clearly, thatan a sravar u pa can arise due tosam a dhi .

14c. Merit increases, not due to avijñapti , but due to a gradual transfor-mation of the sa µ tati of the giver.

Abhidharmako s abh a Ò ya 94 – A third Sarvastivadin argument in support of the reality ofavijñaptir u pa is that thesu tra says that merit increases.


95 yeÒu sa µ sk a reÒu yac chubh a s ubha µ karmotpannaniruddha µ bhavati tena hetun a tena pratyayena vi s iÒt a saµ sk a rasantati Ì pravartate s a va sanety ucyate / yasy a Ì prabandhapatit a ya

iÒ† a niÒ† aphala µ nirvartate (Yoga ca rabh umi: 128.2-4).’du byed ga n dag la dge ba da n midge ba’i las skyes nas ’gags pa yod la / rgyu de da n rkyen des ’du byed bye brag can gyirgyud ’jug pa de la ni bag chags zes bya ste / de rgyun du gnas pa las sdug pa da n mi sdg

pa’i ’bras bu grub par ’gyur ba’i phyir (Yoga ca rabh umi t : dzi 75b4-5).

(T. 1579: 305b3-6).96 ji ltar na sbyin pa las lo ns spyod can du ’gyur ba yin ze na / ’di ltar ’di na la la

snon gyi tshe rabs g zan dag tu sbyin pa las byu n bai bsod nams bya ba’i d nos po byas s inbsags par gyur te / de da ltar phyug pa’i khyim da n / nor che ba nas mdzod da n / ba n

mdzod kyi tshogs ma n ba’i bar gyi khyim du skye bar ’gyur ba ’blta bu’o (Yoga

ca

rabhu-

mit : dzi 269a2-4).(T. 1579: 375b13-16).

97 dharmat a -yukti Ì katam a / kena k a ra ∞ena tath a bh u t a ete skandh a (s) tath a bh u tolokasa µ nive s a Ì kena k a ra ∞ena khara-lak Òa∞a p ® thiv i dravalak Òa∞a a pa Ì uÒ∞alak Òa∞aµteja (Ì ) samud i ra ∞alak Òa∞o vayu Ì / anity a Ì skandh a (Ì ) / kena k a ra ∞ena s a nta µ nirv a ∞amiti / tath a ru pa [∞a]lak Òa∞aµ ru pam anubhavana-lak Òa∞a vedan a sa µ j a nan a lak Òa∞a sa µ jñaabhisa µ skara ∞a-lak Òa ∞a Ì sa µ sk a r a vij a nan a lak Òa ∞a µ vijñ a nam iti / prak ® tir e Òa µdharm a ∞a µ ida µ svabh a va e Òa i d ® s a Ì dharmatai Òa (µ ) caiva c a sau dharmat a / saiv a tra

yuktir yoga up a ya Ì eva µ va etasm a t / anyath a va naiva v a sma t sarvatraiva ca dhar-mataiva pratiprasara ∞aµ dharmataiva yukti Ì / cittanidhyapan a ya cittasa µ jñapan a ya iyamucyate dharmat a -yukti Ì (Sra vakabh umi: 143.4-16; Sra vakabh umi

t : Wayman 79).chos ñid

kyi rigs pa ga n ze na / ci’i phyir phu n po rnams de lta bur gyur pa yin / ’jig rten gnas pade lta bur gyur pa yin / ci’i phyir sa’i mtshan ñid sra ba yin / chu’i mtshan ñid g s er ba

yin / me’i mtshan ñid tsha ba yin / rlu n gi mtshan ñid g.yo ba yin / ci’i phyir phu n po rnamsmi rtag pa yin / ci’i phyir mya nan las ’das pa zi ba yin / de b zhin du ci’i phyir gzugs kyimtshan ñid gzugs su ru n ba yin / tshor ba’i mtshan ñid myo n ba yin / ’du s es kyi mtshanñid kun s es par byed pa yin / ’du byed rnams kyi mtshan ñid m non par ’du byed pa yin / rnam par s es pa’i mtshan ñid rnam par s es par byed pa yin ze na / de ni chos ñid yinte / chos de dag gi ra n bzin de yin zin / de dag gi no bo ñid de lta bu yin pas chos ñid dega n kho na yin pa de ñid ’dir rigs pa da n / sbyor ba da n / thabs yin no / de b zin du de ltabu ’am / g zan nam / g zan du ma ’gyur pa ni sems la b zag par bya ba da n / sems la gobar bya ba’i phyir thams cad du ya n chos ñid kho na la brten pa da n / chos ñid kho na’irigs pa yin te / de ni chos ñid kyi rigs pa zes bya’o (Yoga ca rabh umi t : wi 68b6-69a4).



may be that of a Sautr antika group that preceded Vasubandhu and thatcannot be identified with Yogacara. However, Hakamaya does not pro-vide any evidence from, for example, the *Vibh a Òa , the *Tattvasiddhi s a s-tra , or the * Nya ya nus a ra , and he leaves the question open.

15. Asa µ vara does not really exist separately (from volition).

Abhidharmako s abh a Ò ya 98 – The Sautr antikas say thatasa µ vara does notreally exist separately (from volition). It is the intention to do something

bad, an intention that continues until it is destroyed.*Vibh a Òa 99 – The *Vibh a Òa does not deal with this issue directly. How-ever, it mentions that those who assert the unreality ofvijñapti andavijñapti would be unable to establish the differences between those whoare established insa µ vara, asa µ vara, or naivasa µ var a sa µ vara (neither sa µ vara nor asa µ vara ).

*Tattvasiddhi – Avijñapti is a cittaviprayuktasa µ sk a ra , and thus it has noseparate existence (see above).


(T. 1545: 634c24-26).100 sdom pa ma yin pa’i rigs su skyes pa ji lta ba b zin du ga n su ya n ru n ba ga n da n

ga n nas ’o ns kya n run ste / sems skyed par byed pa ya n de b zin du rgyas par rig par bya’o / de ni ji srid du sdom pa ma yin pa’i sems pa spo n bar mi byed pa de srid du ma bsdams par brjod par bya ste / de ni ñin gcig b zin du sems pa de ma n du sogs pa da n / las de kuntu sbyor bas bsod nams ma yin pa m non par ’phel bar rig par bya’o (Yoga ca rabh umi t : zi31b1-3).

(T. 1579: 589c3-7). See note 104 below.101 yath a bhyupagama µ vikalo ’pi sy a t pr a de s iko ’py asa µ vara Ì sa µ vara s ca nya-

tr a Ò† avidh a d iti sautr a ntik a Ì tanm a tra s i ladau s i lyapratibandh a t ( Abhidharmako s abh a Ò ya :222.5-6; T. 1558: 79a3-6; La Vallée Poussin 3: 93); Saµ ghabhadra quotes this statement[T. 1562: 563b17-19] and criticizes it [T. 1562: 563b20-24]; Kato 77).

102

(T. 1545: 608b20-27).103

(T. 1646: 303a20-25).104 de la sdom pa ma yin pa’i rigs su skyes zin pa ga n la la ’tsho ba ’dis ’tsho ba[r]

bya’o zes ra n gi sems m non par ’du byed ci n / de ’tsho ba de la ’dod pa bzod par byed



*Vibh a Òa 102 – According to the *Vibh a Òa , the Gandhara teachers saythat asa µ vara can be incomplete, while the Kasm i ra teachers say that itcannot.

*Tattvasiddhi 103 – The *Tattvasiddhi says thatsa µ vara cannot be partial.

Yoga ca rabh umi104 – The explanation of the unrestrained person in theVini s cayasa µ graha ∞i on thePañcavijñ a nak a yamanobh umi suggests thatbecoming unrestrained is a gradual process resulting from the accumula-tion of bad actions based on bad thoughts or intentions.

Comment – Although theYoga ca rabh umi does not contain a similar argument or an explicit statement thatsa µ vara or asa µ vara can be par-tial or incomplete, its description of the gradual process of becomingasa µ vara may imply that one can be unrestrained toward certain beingsand not others or with regard to certain rules and not others. In this case,Harivarman seems to disagree with Vasubandhu and perhaps theYoga-

ca rabh umi.


iti ( Abhidharmako s abh a Ò ya : 278.17-22; T. 1558: 99a1-9; La Vallée Poussin 4: 6-7; Saµ-ghabhadra identifies this as the opinion of the sutra-master [T. 1562: 596c24-597a2] andcriticizes it [T. 1562: 597a2-15]; Kato 78).

106

(T. 1545: 313a1-3).107 (T. 1646: 258c7-8).108 Katsura points out that, according to the *Tattvasiddhi , caitta s are not realdharma s (44).109 ñon mo ns pa’i kun nas ñon mo ns pa’i rab tu dbye ba rnam par g zag pa ga n ze na /

mdor bsdu na ñon mo ns pa da n ñe ba’i ñon mo ns pa ji skad bstan pa rnams kyis ni rgyugñis kyis sems can rnams kun nas ñon mo ns par byed de / ’di lta ste / kun nas dkris pada n bag la ñal gyis so / de la ñon mo ns pa kun tu ’byu n ba m non du gyur pa ni kun nasdkris pa zes bya’o / de ñid kyi sa bon ma spa ns s in ya n dag par ma bcom pa ni bag lañal zes bya ste / gnas ngan len kya n de yin no / ma sad pa’i phyir ni bag la ñal yin la sad

pa’i gnas skabs kyi phyir ni kun nas dkris pa yin no (Yoga ca rabh umi t : zi 118a8-b3).

(T. 1579: 623a20-24). Yamabe has noted that the passage from the Abhi-dharmako s abh a Ò ya is directly based on this passage (personal communication).

110 sarvalaukikotkar Òab i j a nugamy a tva d anu s aya Ì (Yoga ca rabh umi: 167.6). ’ jig rten pa’i yar ’phel ba thams cad kyi sa bon da n ldan pas na bag la ñal rnams so (Yoga ca ra-bhumi t : dzi 97b8-98a1). (T. 1579: 314b25-26).

111 de la da n ba’i gzugs da n / sems da n sems las byu n ba’i chos ji skad bstan pa thamscad la ñon mo ns pa’i sa bon ya n dag par ma bcom pa da n / ma spa ns pa ga n yin pa deni bag la ñal zes bya ste / gnas nan len kya n de yin no (Yoga ca rabh umi t : zi 215a5-6).



17. Anus ayas are kles as in the state of seeds, not separate entities ( dravya s).

Abhidharmako s abh a Ò ya 105 – The Sautr antikas defineanu s aya s as kles asin the state of seeds and say that they are not separatedravya s. Anu s aya sare dormant, i.e., not actualized, while pary a vasth a nas (active defilements)are awakened.

*Vibh a Òa 106 – According to the Vibhajyavadins, anu s ayas are the seedsof paryavasth a na s, and they are dissociated from mind (cittaviprayukta ).

*Tattvasiddhi 107 – Anu s aya s are cittasa µ prayukta .108

Yoga ca rabh umi (1)109 – The Vini s cayasa µ graha ∞i on the Savitark a di-bh umi contains an explanation ofanu s aya and paryavasth a na almostidentical to that in the Abhidharmako s abh a Ò ya .

Yoga ca rabh umi (2)110 – The Savitark a dibh umi identifiesanu s aya s asbeing the seeds ofkles as.

364 ROBERT KRITZER

(T. 1579: 661b26-29).112 Yoga ca rabh umi t : zi 208a4-6; T. 1579: 659a12-16. See note 50 above.113 naiva hi sautr a ntik a at i t a t karma ∞a Ì phalotpatti µ var ∞ayanti / ki µ tarhi / tatp urva-

k a t sa µ t a navi s eÒa d ity a tmav a daprati Òedhe sa µ pravedayi Ò ya ma Ì ( Abhidharmako s abh aÒ ya :300.19-21; T. 1558: 106a11-13; La Vallée Poussin 4: 63; Saµ ghabhadra identifies thisas the opinion of the sutra-master [T. 1562: 629b3-5], refers to Vasubandhu’s longer expla-nation at the end of Chapter 9 [ Abhidharmako s abh a Ò ya : 477.7-18] ofsa µ t a ∞apari ∞a ma[T. 1562: 629b5-17] and criticizes it at very great length [T. 1562: 629b18-630a11]; Kato78).

114

(T. 1545: 393a9-12).115

(T. 1646: 255c24-26). See Katsura, who points out that Harivarman does notmentionsa µ t a napari ∞a mavi s eÒa in this respect (41).

116 yad apy uktam asty at i ta µ karma yata Ì sattv a h savy a baddh a vya ba dha µ vedayan-t i ti / tatr a pi tadv a san a ya µ tadastitvopac a ram abhipretyokta µ / ye Òu sa µ sk a reÒu yac chu-bha s ubha µ karmotpannaniruddha µ bhavati tena hetun a tena pratyayena vi s iÒt a sa µ sk a ra-santati Ì pravartate s a va sanety ucyate / yasy a Ì prabandhapatit a ya iÒ† a niÒ† aphala µnirvartate iti na yujyate / tato ’pi n a sti do Òa Ì (Yoga ca rabh umi: 127.19-128.4).’das pa’ilas yod do zes gsu ns pa ga n yin pa de la ya n / bag chags de la / de yod pa’i ’dogs pa ladgo ns nas gsu ns pa yin te / ’du byed ga n dag la dge ba da n mi dge ba’i las skyes nas ’gags

pa yod la / rgyu de da n rkyen des ’du byed bye brag can gyi rgyud ’jug pa de la ni bag

chags zes bya ste / de rgyun du gnas pa las sdug pa da n mi sdug pa’i ’bras bu grub par’gyur ba’i phyir mi ru n ste (Yoga ca rabh umi t : dzi 75b3-5).



Yoga ca rabh umi (3)111 – In theCint a ma yi prajñ a bhumi of theVini s caya-sa µ graha ∞i , the undestroyed seeds ofkles as are calledanu s aya s.

Yoga ca rabh umi (4)112 – According to theCint a ma yi prajñ a bhumi of theVini s cayasa µ graha ∞i , bi ja s are prajñ a pti .

Comment – Vasubandhu and theYoga ca rabh umi explainanusaya s inthe same way. Harivarman’s explanation is very different.

18. A result arises due to a sa µ t a navi s eÒa (a special state of thesa µ tati )

based on a past action, not directly due to a past action. Abhidharmako s abh a Ò ya 113 – A result does not directly arise from apast action; instead, it arises due to asa µ t a navi s eÒa based on a pastaction.


(T. 1579: 305b1-6).117 bcom ldan ’das kyis las ’das pa ni yod de gal te las ’das pa med du zin na ’di na

la las gnod pa da n bcas pa da n / gnod pa med pa’i tshor ba myo n ba mi ’gyur zes ga ngsu ns pa de la dgo ns pa ga n ze na / ’das pa’i tshor bas rnams su las dge ba da n mi dgeba bskyed ci n ’gags pas phyi ma la ’bras bu ’dod pa da n mi ’dod pa m non par ’grub parde’i sa bon gyis ’du byed kyi rgyun phyi ma phyi ma yo ns su bsgom pa las dgo ns nas(Yoga ca rabh umi t : zi 20b4-6).

(T. 1579: 585b7-13).118 arhattv a d api n a sti parih a ∞ir iti sautr a ntik a Ì / e Òa eva ca ny a ya Ì / katham ida µ

gamyate / a gam a d yuktita s ca ( Abhidharmako s abh a Ò ya : 375.10-11 [but the whole argu-ment continues until 377.5]; T. 1558: 130a16-130c16; La Vallée Poussin 4: 258 [-265];Saµ ghabhadra identifies this as the opinion of the sutra-master [T. 1562: 711c2-3] and criti-cizes it at exceedingly great length [T. 1562: 711c7-716a13; I have not distinguished herebetween his brief quotations of Vasubandhu’s opinions and his lengthy criticisms]; Kato78).

119 yadi t a vad arhatas tadr u pa Ì pratipak Òa utpanno yena kle s a atyantam anutpatti-dharmat a m a ppan a Ì / katha µ puna Ì parih i yate / atha notpanna Ì / katha µ k Òi ∞a sravobhavati / atyantam anayoddh ® t a ya µ tadb i jadharmat a ya m ak Òi ∞a sravo v a puna Ì kathamarhan bhavat i ty eva µ yukti Ì ( Abhidharmako s abh a Ò ya : 376.17-20; T. 1558: 130c2-4; LaVallée Poussin 4: 263-264; Saµ ghabhadra identifies this as the opinion of the sutra-master [T. 1562: 716a1-4] and criticizes it [T. 1562: 716a4-13]).120



*Vibh a Òa 114 – At the beginning of its long defense of the reality of thethree times, the *Vibh a Òa identifies Dar Ò†antika and Vibhajyavada as theopponents. However, it does not refer to the theory ofsa µ t a navi s eÒa .

*Tattvasiddhi 115 – Although pastkarma gives rise to real results, it doesnot exist.

Yoga ca rabh umi (1)116 – The Savitark a dibh umi explains that when theBuddha said that a past action exists, he was really talking about impres-

366 ROBERT KRITZER

(T. 1545:312b8-14; La Vallée Poussin 4: 264 n. 2).

121

(T. 1545: 313a14-25).

122 For example:

(T. 1646: 258a24-b1).123 bcom ldan ’das kyis ji skad du dge slo n dag dgra bcom pa ya n tshe ’di la lhag pa’i

sems las byu n ba bde bar gnas pa b zi po de dag las ga n ya n ru n ba las yo ns su ñams parna smra zes ga n gsu ns pa de la / gal te de’i ñon mo ns pa can gyi chos thams cad kyi sabon dag ya n dag par bcom na ni / ji ltar de la ’og ma pa’i ñon mo ns pa ’byu n bar ’gyur / gal te mi ’byu n na ni ji ltar de yo ns su ñams par ’gyur ze na / yongs su ñams pa ni gñis

po ’di dag yin te / spo n ba’i yo ns su ñams pa da n / gnas pa’i yo ns su ñams pa’o / de laspo n ba’i yo ns su ñams pas ni so so’i skye bo kho na yo ns su ñams par ’gyur ro / gnas

pa’i yo ns su ñams pas ni ’phags pa da n / so so skye bo ya n yons su ñams pa ’gyur ro / dela ’jig rten pa’i lam gyis ñon mo ns pa spa ns pa ya n mnon du byed pa ni spo n ba’i yo nssu ñams pas yo ns su ñams par ’gyur te / gnas pa’i yo ns su ñams pas yo ns su ñams par’gyur ba ya n de yin no / ’jig rten las ’das pa’i lam gyis ñon mo ns pa spa ns nas / de lasgzan pa’i phral gyi bya ba dag la rab tu chags pa’i blo can yid la mi byed pa’i rgyus de’imjug thogs su tshe ’di la bde bar gnas pa la s non ji lta bar phyis kya n de b zin du m nondu byed mi nus la / sa ’og ma pa’i ñon mo ns pa ni / m non du mi byed pa ga n yin pa deni de lta na gnas pa’i yo ns su ñams par ’gyur ba yin gyi spo n ba’i yo ns su ñams pa ni ma

yin no / gal te dgra bcom pa ñon mo ns pa thams cad spa ns pa’i ñon mo ns pa can gyi chosde dag thams cad kyi sa bon ya n dag par ma bcom na ni / ji ltar na dgra bcom pa semss in tu rnam par grol ba da n / zag pa zad par ’gyur / gal te ya n dag par bcom na ni de’isems kyi rgyud ñon mo ns pa can gyi chos thams cad kyi [corrected fromkyis on the basisof the Derge]sa bon med pa la tshul b zin ma yin pa yid la byed pa tsam ya n ’byu n barmi ’gyur na / ñon mo ns pa lta smos kya n ci dgos te / de lta bas na ’jig rten las ’das pa’i

lam gyis ñon mo ns pa spa ns pa la ni yo ns su ñams ba med par kho n du chud par bya’o(Yoga ca rabh umi t : zi 17b7-18b1).



sions of the action, not the action itself. These impressions endow thesa µ tati with the potential to yield results.Yoga ca rabh umi (2)117 – The Vini s cayasa µ graha ∞i on the Pañcavijñ a-

nak a yamanobh umi gives a similar explanation, but with the addition of the termkarmab i ja (seed ofkarma ).

Comment – See item 14c.

19. One cannot fall from arhatship.

Abhidharmako s abh a Ò ya 118 – The Sautr antikas say that one cannot fallfrom arhatship. They argue that the definition of anarhat is one whosekles as are completely destroyed. This implies that the seeds of hiskles asare likewise destroyed, in which case it is impossible for thekles as to ariseagain.119

*Vibh a Òa (1)120 – The *Vibh a Òa mentions the view of the Vibhajyava-

dins, who say that thekles as cannot arise again after having been des-troyed.*Vibh a Òa (2)121 – The *Vibh a Òa attributes to the Dar Ò†antika the viewthat parih a ∞i (fall from arhatship) is a prajñapti and not a realdharma .

*Tattvasiddhi 122 – The *Tattvasiddhi gives many arguments to the effectthat once thearhat has destroyed thekles as, they cannot arise again. Butit does not mention the destruction of thebi ja s of kles as.

Yoga ca rabh umi (1)123 – The Vini s cayasa µ graha ∞i on the Pañcavi- jña nak a yamanobh umi says that thearhat , who has destroyed thekles asand theirbi ja s, cannot fall from arhatship.


(T. 1579:

584b3-19).124 rnam par bya n ba’i phyogs da n mthun pa’i chos rnams kyis de yo ns su bstan to /



On the other hand, an examination of the passages in which Vasu-bandhu attributes a doctrinal position to Sautr antika shows that, in almostevery case, a closely related, if not identical, position can be found some-where in theYoga ca rabh umi. Corresponding passages appear most fre-quently in theVinis cayasa µ graha ∞i on thePañcavijñ a nak a yamanobh umi,followed by theSavitark a dibh umi of the Maul i bhumi and other sectionsof theVini s cayasa µ graha ∞i . If, however, Vasubandhu is actually follow-ing theYoga ca rabh umi, one must explain why he uses the term Sautr an-tika and why he never refers to theYoga ca rabh umi or its characteristic

doctrine,a layavijñ a na , in the Abhidharmako s abh a Ò ya .

B. The Traditional Explanation: Vasubandhu’s Position in the Abhidhar-mako s abh a Ò ya is Sautr antika

The commonly held view concerning Vasubandhu’s philosophicaldevelopment is the one sarcastically described by Lamotte in his intro-duction to theKarmasiddhiprakara ∞a: “Who can believe that Vasubandhuwithout mentioning his acquaintance with the Sam khya, was a VaibhaÒikain his youth, a Sautr antika in his mature years, a Vijñanavadin in his oldage, and a Pure Land follower of Amitabha at his death?” (Lamotte His-tory 39 [English translation of Histoire 179]). Having asked this acuterhetorical question, Lamotte indicates that he accepts at least the traditionaldescription of the mature Sautr antika, whom he identifies as the author of both the Abhidharmako s abh a Ò ya and theKarmasiddhiprakara ∞a. 126

Since Lamotte’s exposition of Vasubandhu’s Sautr antika standpoint inthe Karmasiddhiprakara ∞a is the most explicit that I know, and since

most of his arguments apply equally to the Abhidharmako s abh a Ò ya , I dis-cuss it as representative of the traditional explanation of Vasubandhu’sposition. Lamotte begins by saying that the purpose of both texts is “tocombat, within the framework of the H i nayana and relying on the best of the Sautr antika, the exaggerated realism of the VaibhaÒikas and the spiri-tualism of the Vats i putr i yas” (Lamotte History 40 [English translation of


Samayabhedoparacanacakra use the term Sautr antika, neither text uses it to refer tothe same group as Vasubandhu.126 For a discussion of Vasubandhu’s career, see Kritzer Rebirth 198-199.



Lamotte Histoire 180]). Furthermore, he points out that the two textsignore Mahayana, including Yogacara idealism, and goes on to refute theChinese and Tibetan tradition that theKarmasiddhiprakara ∞a is a Maha-

yana work, disagreeing in particular with Bu-ston’s characterization of thetext as idealistic.

Then Lamotte presents what he considers to be internal evidence sup-porting his claim that the text is Sautr antika. First, he calls attention tothe fact that the teachers and schools mentioned and the majority of thescriptures quoted in the text, with the exception of two quotations from

the Saµ dhinirmocanas u tra , belong to H i nayana. However, this in itself does not prove anything. After all, many portions of theYoga ca rabh umialso fail to quote Mahayana sutra s. TheVinis cayasa µ graha ∞i , which doesquote Mahayana su tra , to the best of my knowledge quotes primarily theSaµ dhinirmocanas u tra .

370 ROBERT KRITZER

127 See item 11 above.128 de la gzugs kyi phu n po thams cad ni skad cig pa yin par brjod par bya’o / de’i ci

phyirze na / skyes nas ’jig [corrected fromna jig on the basis of the Derge] pa dmigs pa’i phyir ro / skye ba’i rgyu ni ’jig pa’i rgyu yin par mi ru n ste / mtshan ñid mi ’dra ba’i phyir

ro / skyes pa gnas pa’i rgyu de las g zan pa ya n mi dmigs pas de’i phyir ’du byed thamscad ni ra n gi nan gis ’jig pa yin par rig par bya ste / de’i phyir skad cig pa ñid rab tugrub po (Yoga ca rabh umi t : zi 58a4-6).

(T. 1579: 600a18-22). See also the Abhidharma-kos abh a Ò ya : sa µ sk ® tasy a vas ya µ vyaya t akasmiko hi bh a va na µ vina s a Ì / ki µ k a ra ∞am / k a ryasya hi k a ra ∞aµ bhavati / vin a s as ca bha va Ì / ya s ca bha vas tasya ki µ kartavyam / so’sa v a kasmiko vin a s o yadi bh a vasyotpannam a trasya na sy a t pa s cad api na sy a d bh a vasyatulyatv a t / ath anyath i bhuta Ì na yukta µ tasyaiv anyath a tvam / na his sa eva tasm ad vilak Òano

yujyate / ko ’ya µ k Òa∞o n a ma / a tmal a bho ’nantaravin a s i / so ’sy a st i ti k Òa∞ikam / da ∞∂ i-kavat ( Abhidharmako s abh a Ò ya : 193.5-10; T. 1558: 67c17-20; La Vallée Poussin 3: 5;Saµ ghabhadra identifies this as the opinion of the sutra-master [T. 1562: 533b21-22] andcriticizes it, saying that, because of the validity of thesa µ sk ® talak Òa∞as, destruction musthave a cause [T. 1562: 533c10-21]; see Rospatt 180-181).

The passage from theYoga ca rabh umi is translated by Rospatt, who also provides theTibetan text and the text from the Sanskrit manuscript (181-182 n. 399). Rospatt thinks thatthe argument here is somewhat different from that of the Abhidharmako s abh aÒ ya (181-182).

129 de la lus yul nas yul g zan du ’byu n ba tsam da n / de ñid na ’gyur ba ’byu n ba tsamni lus kyi rnam par rig byed do / nag tsam ni nag gi rnam par rig byed do / de b zin dusems m non par ’du byed pa skyes pa’i sems pa tsam ni yid kyi rnam par rig byed do / deci’i phyir ze na / ’du byed thams cad ni skad cig pa yin pa’i phyir yul nas yul g zan du ’phobar mi rigs pas (Yoga ca rabh umit : zi 31a4-5).



Next, Lamotte identifies and characterizes as Sautr antika a number of positions asserted by Vasubandhu (Traité 177-179), many of which are alsofound in the Abhidharmako s abh a Ò ya . Below I summarize these positionsand provide corresponding passages from theYoga ca rabh umi:

1) According to Lamotte, Vasubandhu adopts Sautr antika positions on major issues regardingkarma .

a) Positions concerningvijñapti and avijñapti :

Karmasiddhiprakara ∞a – Sa µ sthana does not exist separately from color.Yoga ca rabh umi – The same position is found in theVinis cayasa µ graha ∞i

on the Pañcavijñ a nak a yamanobh umi.127

Karmasiddhiprakara ∞a – Destruction is spontaneous, without a cause.Yoga ca rabh umi – The same position is found in theVinis cayasa µ graha ∞i

on the Pañcavijñ a nak a yamanobh umi.128

Karmasiddhiprakara ∞a – There is no duration or movement.Yoga ca rabh umi – A similar statement is found in theVini s cayasa µ gra-ha ∞i on the Pañcavijñ a nak a yamanobh umi.129

Karmasiddhiprakara ∞a – The essence of body and speech karma is volition.Yoga ca rabh umi – A similar position may be implied by theSavitark a di-bhumi.130

Karmasiddhiprakara ∞a – Avijñapti proceeds from volition, not from matter.Yoga ca rabh umi – The definition ofasa µ vara in theVini s cayasa µ graha ∞i

on the Pañcavijñ a nak a yamanobh umi implies thatasa µ vara is based onvolition.131

b) Positions concerning action and retribution:


(T. 1579: 589b18-22).130 See note 66, comment on item 12.131 See item 15.132 See item 18.133 See item 14c.134 ’jug pa’i rnam par s es pa tsam ñe bar zi bar zad kyi / kun g zi rnam par s es pa ñe



Karmasiddhiprakara ∞a – The a layavijñ a na appropriates a body that con-sists of ru pa , the reality of which is not questioned.Yoga ca rabh umi – One of the proofs ofa layavijñ ana in theVinis cayasa µ-graha ∞i on the Pañcavijñ a nak a yamanobh umi is that, without ana laya ,there could be no appropriation of the body.136 Furthermore, theYoga ca-

rabh umi generally does not question the reality ofru pa .137

Karmasiddhiprakara ∞a – The explanation of why the Buddha did notteach a laya to his disciples is different from Asanga’s in the Mah a ya na-

sa µ graha , which is predicated on the unreality of the external object.In the Karmasiddhi , Vasubandhu quotes theSaµ dhinirmocana , sayingthat ignorant people would mistake thea laya for a soul.Yoga ca rabh umi – The Vini s cayasa µ graha ∞i on the Pañcavijñ a nak a ya-manobh umi quotes the same passage at the beginning of its exposition of a layavijñ a na to explain why it has not been taught before.138

As we can see, all of these positions supposedly characteristic of Sautr an-tika can be traced more or less clearly to theYoga ca rabh umi, particularly

to theVini s cayasa µ graha ∞i on thePañcavijñ anak a yamanobh umi. And notone of these positions can be traced to a text earlier than the Abhi-dharmako s abh a Ò ya in which it is identified as Sautr antika. Thus, theSautr antika positions in the Abhidharmako s abh a Ò ya and theKarmasiddhi-

prakara ∞a , a text that is considered to be later than the Abhidharma-kos abh a Ò ya and more developed, i.e., closer to classical Yogacara, arecomparable if not identical. The only striking difference is that theKarma-siddhiprakara ∞a mentionsa layavijñ a na .139

Therefore, Lamotte’s “internal evidence” that theKarmasiddhipraka-ra ∞a is a H i nayana Sautr antika text can equally well be viewed as tes-timony to Vasubandhu’s reliance on theVini s cayasa µ graha ∞i of theYoga ca rabh umi, a reliance similar to that which we have seen in the

Abhidharmako s abh a Ò ya . Lamotte calls the positions that he cites “Sautr an-tika” because Vasubandhu has identified them as such in the Abhidhar-mako s abh aÒ ya or because the commentator on theKarmasiddhiprakara ∞a


139 Schmithausen disagrees with Lamotte, who thinks that thea layavijñ a na that appears



have seen, written records of these sources, if they ever existed, are nolonger extant.

According to the traditional view of his career, Vasubandhu wrote the Abhidharmako s abh a Ò ya after he had come to accept Sautr antika ideas.Some time later, having learned the doctrine ofa layavijñ a na , perhapsdirectly from Asanga, perhaps from a text like the Mah a ya nasa µ graha ,he converted to Mahayana and became a Yogacara/Vijñanavadin. In thiscase, it is difficult to explain his mentioninga layavijñ a na in the Kar-masiddhiprakara ∞a . According to the traditional explanation, Vasubandhu

wrote this text before his conversion. Did he invent a non-Vijñanavadinversion ofa layavijñ a na independently, as a sort of logical developmentof his Sautr antika seed theory, in the same process that Schmithausendescribes with respect to theYoga ca rabh umi? Did he borrow the termfrom one of the no longer extant Sautr antika sources that I postulatedabove? If Schmithausen is correct that the theory ofa layavijñ a na devel-oped within theYogaca rabh umi, both of these hypotheses seem far-fetched.

It is far more likely that Vasubandhu was, in fact, familiar with theYoga ca rabh umi. If so, he would have known the positions that he callsSautr antika from that text and perhaps from the lost Sautr antika sourcesas well. In either case, according to the traditional explanation, at the timeof writing the Abhidharmako s abh a Ò ya , Vasubandhu must have adoptedthese ideas, while not yet accepting the theory ofa layavijñ a na , which hewould have also known from theYoga ca rabh umi. By the time he wroteKarmasiddhiprakara ∞a , he tentatively believed in a not fully developed,Sautr antika-likea layavijñ a na , which he knew from theYoga ca rabh umiand perhaps from a lost Sautr antika source. Finally, he wrote texts likethe Trim s ik a after his conversion to Mahayana. In other words, the devel-opment of Vasubandhu’s belief ina layavijñ a na paralleled the develop-


140 Schmithausen 34-65. I have undoubtedly grossly oversimplified his complex argument.141 Schmithausen 10, 259 n. 92. He identifies the section asYoga ca rabh umi t : zi 1b2-

10b6; T. 1579: 579a7-582a12.142 Schmithausen identifies the section asYoga ca rabh umit : zi 2b2-4a4; T. 1579: 579a14-

c22 (300 n. 226). He also shows that the various proofs are not completely consistent inthe ideas ofa layavijñ a na on which they are based (194-196).



ment of the theory ofa layavijñ a na in theYoga ca rabh umi but some timeafter theYoga ca rabh umi was already completed.

However, it seems strange that Vasubandhu would repeat the entireprocess of the discovery ofa layavijñ a na . A fairly complete version of thetheory must have been available to him in what Schmithausen calls the“alayavijñana treatise” at the beginning of theVini s cayasa µ graha ∞i .141

It is hard to believe that Vasubandhu would have been satisfied with atheory like mutual seeding once he had been exposed to the eight-fold proof of a layavijñ a na , which includes among its arguments a refutation of that

very theory.142 Nevertheless, Vasubandhu does, in fact, present the theoryof mutual seeding on at least one occasion, evidently with approval.143

If Vasubandhu already believed ina layavijñ a na when he wrote the Abhi-dharmako s abh aÒ ya , we must explain why he introduces such positions thatare superseded by the more developed theory ofa layavijñ a na .

C. An Alternative Explanation: Vasubandhu’s Sautr antika Position inthe Abhidharmako s abh a Ò ya Reflects his Yogacara Beliefs

I have previously speculated that Vasubandhu was a Yogacara whenhe wrote the Abhidharmako s abh a Ò ya (Kritzer “Vasubandhu,” Rebirth199-204), and I have not changed my mind. In this article, I hope to pres-ent more persuasive arguments based on further evidence.

In my earlier works, I reasoned primarily on the basis of two argu-ments in the Abhidharmako s abh a Ò ya : Vasubandhu’s “Sautr antika” defi-nition ofvijña na as a member of the prat i tyasamutp a da formula and hiscriticism of certaincittaviprayuktasa µ sk a ra s. Since then, I have begun a

more systematic examination of Vasubandhu’s unorthodox opinions inthe Abhidharmako s abh a Ò ya . In addition to my search for the term Sautr an-tika, the results of which I have presented in Section III of this paper,I have also searched for opinions identified by Saµ ghabhadra as those of the sutra-master. Saµ ghabhadra uses this appellation with reference notonly to most of the positions that Vasubandhu himself labels Sautr antika,

376 ROBERT KRITZER

143 See section III, item 7.144 I have published the results regarding the first three chapters of the Abhidharma-



but to many others as well. Many of these passages also correspond moreor less clearly to passages in theYoga ca rabh umi.144

Although some of the correspondences that I have identified are lesscertain than others, their sheer number strongly suggests that Vasubandhurelies heavily on theYoga ca rabh umi for his criticism of Sarvastivada.However, the great majority of the correspondences between the Sautr an-tika positions in the Abhidharmako s abh a Ò ya (and the Karmasiddhipra-kara ∞a ) and theYoga ca rabh umi involve passages that do not appear tobe based on a theory ofa lay a vijña na .

In trying to account for this, I run the danger of reading too much intoVasubandhu’s statements. In essence, I argue that Vasubandhu favors posi-tions in theYoga ca rabh umi that do not mentiona layavijñ a na because heinfers a theory ofa layavijñ a na underlying them. This is clearly a riskyproposition, especially since Sautr antika is traditionally seen as precedingYogacara, both historically and in the development of Vasubandhu’sthought. With reference to theYoga ca rabh umi, Schmithausen warns us notto “lightly interpret our text on the lines of later sources and developments”(205). This very principle is what enables Schmithausen to challenge thetraditional view that theYoga ca rabh umi is a coherent composition of oneman, Asanga, and I believe that Schmithausen is correct in his approach.


kos abh a Ò ya (KritzerComparison ).145 Kritzer Rebirth 200. As far as I know, Aramaki has not published this observation,

which he conveyed to me personally. However, in the meantime, he has publicly statedhis ordering of the strata of theYoga ca rabh umi, which differs significantly from Schmit-hausen’s. According to Aramaki, the Maul i bhumi (excluding theSravakabh umi and Bodhi-sattvabh umi), which contains much of the traditionalabhidharma material found in theYoga ca rabh umi, is later than theVini s cayasa µ graha ∞i , in whicha layavijñ ana is taught andthe Saµ dhinirmocanas u tra is quoted.

146 de la kun rdzob kyi tshul rnam par b zag pas ’jug pas ni ’di lta ste / yid kyi sar s narbstan pa zin du rig par bya’o / de la don dam pa’i tshul rnam par b zag pa b s ad par byaste / don dam pa’i tshul rnam par b zag pas ’jug pa ga n ze na / mdor bsdu na rnam pars es pa ni rnam pa gñis te / kun g zi rnam par s es pa da n / ’jug pa’i rnam par s es pa’o(Yoga ca rabh umi t : zi 189a8-b2).

(T. 1579: 651b11-15; see Schmithausen 689-690).147 sa bon rnam par b zag pa’i tshul ’di ni kun g zi rnam par s es pa rnam par ma g zag

pa la rig par bya’o / rnam par b zag pa la ni mdor bsdu na de la chos thams cad kyi sabon yod par rig bar bya ste / sa bon de dag ni ma spa ns pa da n spa n bar bya ba ma yin



Nevertheless, there are two major differences between theYoga ca ra-bhumi and the Abhidharmako s abh a Ò ya that I feel justify some departurefrom Schmithausen’s principle. First, several of the most important sourcesfor Yogacara doctrine, including theSaµ dhinirmocanas u tra , the com-pleted Yoga ca rabh umi, the Abhidharmasamuccaya , and the Mah a ya na-saµ graha are generally considered to predate Vasubandhu’s work. In other words, Vasubandhu must have known the doctrine ofa layavijñ ana , whether he agreed with it or not. Second, to the best of my knowledge, no one,not even Schmithausen, has suggested that the Abhidharmako s abh a Ò ya is

not a coherent composition of one man, Vasubandhu.I have proposed that Vasubandhu’s reason for not mentioninga layavij-ña na in the Abhidharmako s abh a Ò ya has to do with the nature of the text,which is an exposition and criticism of traditionalabhidharma , not a pre-sentation of Yog a ca ra ideas (Kritzer Rebirth 203-204). I referred to AramakiNoritoshi’s idea that portions of theYoga ca rabh umi present the Yogacaraexposition of ultimate truth, that is, the doctrine ofa layavijñ a na , whileother portions, which do not mentiona layavijñ a na , represent provisionaltruth.145 My conclusion was that the Abhidharmako s abh a Ò ya , like theabhidharma portions of theYoga ca rabh umi, is an exposition of provi-sional truth. Since the Yogacaras seem to have arisen from a Sarvastivadinmilieu (Yamabe “An Shigao”), it is not surprising that much Yogacara

378 ROBERT KRITZER

pa’i chos de dag da n ci rigs su ldan par rig par bya’o (Yoga ca rabh umi t : zi 17b6-7).

(T. 1579: 584a27-b1;see Schmithausen 271 n. 131).

148 de la ’gog pa’i sñoms par ’jug pa ga n ze na / ci ya n med pa’i skye mched kyi ’dod chags da n bral go n ma’i ’dod chags da n ma bral ya n ru n / ’dod chags da n bral ya n ru nba’i gnas pa’i ’du s es s non du bta n ba’i yid la byed pas sems da n sems las byu n ba’i chosrnams ’gog pa tsam da n / ñe bar zi zin mi ’byu n ba tsam ni ’gog pa’i sñoms par ’jug pazes bya ste / ’jug pa’i rnam par s es pa tsam ñe bar zi bar zad kyi / kun g zi rnam par s es

pa ñe bar zi ba ni ma yin no (Yoga ca rabh umi t : zi 39a3-6).

(T. 1579: 593a1-4; see Schmithausen 272 n. 131).149 ’gog pa’i sñoms par ’jug pa ni ci ya n med pa’i skye mched kyi ’dod chags da n bral

ba’i gnas pa’i ’du s es s non [corrected frommnon on the basis of the Derge and the Chi-nese] du bta n ba’i yid la byed pas mi n ’gog pa’i gnas skabs la’o / de ya n rnam pa gsumste / no bo ñid las ni dge ba ñid yin no / ga n zag las ni ’phags pa’i rgyud du gtogs te / slob pa’i rgyud dam mi slob pa’i rgyud du gtogs pa yin no / skye ba las ni kun g zi rnam

par s es pa rnam par ma g zag ni da n por ’dir skyes ci n / de’i og tu gzugs kyi khams sumnon du byed do / m non du byed pa ni gzugs kyi lus la rag las pa yin pa’i phyir ro / kun



and Sarvastivadin teachings on this level essentially agree. Thus, Vasuban-dhu is able to use the general framework of the Sarvastivada abhidharma ,while he “corrects” those details that seriously conflict with the Yogacaraabhidharma .

I have noticed that Schmithausen refers to a passage in theYoga ca ra-bhumi that supports Aramaki’s idea. In theVini s cayasa µ graha ∞i on theSacittakabh umi, it is said that the traditional way of establishing con-sciousness is taught in the Manobh umi (of the Maul i bh umi), while theultimate teaching is that there are two types of consciousness, thea laya-

vijña na and the prav ® ttivijñ a na s.146 Although Schmithausen mentions thisonly to prove that the mention ofa layavijñ ana in the Manobh umi is a later addition to the text, the passage shows that the author(s)/compiler indeeddistinguishes between levels of teaching within the text.

Three passages in the Vini s cayasa µ graha ∞i on the Pañcavi- jña nak a yamanobh umi, identified by Schmithausen, may also be relevant.The first comes at the end of a long explanation of mutual seeding (seesection III, item 7), in which mutual seeding is said to be taught onlywhen a la-

yavijñ a na is not yet established.147 The next states that the definitionof nirodhasam a patti as a state in which allcitta and caitasika s (mentaldharma s) are suppressed refers only to the prav ® ttivijñ a na s, not toa laya-vijña na .148 In the third passage,nirodhasam a patti is said to be obtainablein ru padh a tu after it has been obtained ink a madh a tu (the realm of desire).


gzi rnam par s es pa rnam par g zag na ni m non du byed pa thams cad du ’gro ba yin parblta bar bya’o (Yoga ca rabh umi t : zi 76b2-5).

(T. 1579: 607b4-10; see Schmithausen 271 n. 131).150 The text actually uses the near-synonym,a d a navijñ a na .151 len pa’i rnam par s es pa zab ci n phra / sa bon thams cad chu bo’i klu n ltar ’bab /

bdag tu rtog par gyur na mi ru n zes / byis pa rnams la ıas ni de ma bstan (LamotteSaµ dhinirmocana 58 [5.7]; Lamotte supplies the Sanskrit, presumably from theTriµ s ika-bha Ò ya : a d a navijñ a na gabh i ras uk Òmo / ogho yath a vartati sarvab i jo / b a la na e Òo mayi na

prak a s i / m a haiva a tma parikalpayeyu Ì ). See also the version of thesu tra contained inthe Vini s cayasa µ graha ∞i on the Bodhisattvabh umi (Yoga ca rabh umi t : ’i 60a1; T. 1579:718c2-3). See note 138.

152 prajñ a ’mal a sa nucar a ’bhidharma Ì tatra prajñ a dharmapravicaya Ì / amaleti an as-rav a / s a nucareti sapariv a ra / evam an a srava Ì pañcaskandhako ’bhidharma ity ukta µ



However, according to the text, whena layavijñ a na has been established,nirodhasam a patti must be obtainable ina ru pyadh a tu as well.149

Schmithausen mentions all of these references toa layavijñ a na as exam-ples of later additions made by the compiler (271-272 n. 131). In the caseof the first passage, at least, he also states that he does “not hesitateto take this systematical statement of the compiler historically, viz. in thesense that thisbi ja theory was devised whenalayavijñana had not yetbeen introduced” (288 n. 173). Schmithausen thus does not connect thesestatements with that of theVini s cayasa µ graha ∞i on the Sacittakabh umi

concerning two levels of teaching. Given his confidence that theYoga-ca rabh umi is not the work of a single author, this is understandable.However, Schmithausen does assume that there was a compiler who puttogether the various strata into the text we now have, and he allows for the possibility of the compiler’s having inserted his own comments intothe text.

Therefore, in light of the fact that either the author of theVini s caya-sa µ graha ∞i on the Sacittakabh umi or the compiler explicitly refers toa layavijñ a na as an ultimate teaching, it does not seem unreasonable thatthe author/compiler’s comments on the three passages in theVini s caya-sa µ graha ∞i on the Pañcavijñ a nak a yamanobh umi reflect his judgementregarding the level at which their statements apply rather than, or as wellas, his understanding of the historical development of the relevant doc-trines. Furthermore, theSaµ dhinirmocanas utra contains the famous state-ment that the Buddha did not teacha layavijñ a na 150 to fools who mightmistake it for a soul (a tman ).151 The Saµ dhinirmocanas utra , which Lopez(6) describes as providing for Yogacara the criteria for determining “what

380 ROBERT KRITZER

bhavati / e Òa t a vat p a ram a rthiko ’bhidharma Ì / s a µ ketikas tu tatpr a ptaye y a pi ca yac cas a stram ( Abhidharmako s abh a Ò ya 2.3-6; La Vallée Poussin 1: 3-4).

153 atas tadhetos tasya dharmapravicayasy a rthe s a str a kila buddhe ∞a bhidharma ukta Ì( Abhidharmako s abh a Ò ya 3.1; La Vallée Poussin 1: 5-6).

154 pr a ye∞a hi k a s mi ravaibh a Òik a ∞a µ ni tya disiddha e Òo ’sm a bhir abhidharma a khy-a ta Ì ( Abhidharmako s abh a Ò ya : 450.1-2; La Vallée Poussin 5: 223).

155

(T. 1562:775b20-23; La Vallée Poussin 5: 223 n. 1b).

156 See Jaini’s introduction to the Abhidharmad i pa (second ed. 111); Rhys Davids andStede (488).157 See, for example, Vasubandhu’s definition ofbi ja in the context of his denial of the



constitutes the definitive (ni t a rtha ) and the interpretable (ney a rtha ),”implies that the doctrine ofa layavijñ a na , besides being dangerously diffi-cult to understand, is a more definitive teaching than that of the tradi-tional six ordinary types of consciousness. Although the main force of thestatement in theSaµ dhinirmocanas utra is to explain why the crucial terma layavijñ a na cannot be found in thea gama s (scriptures), it also suggestsa reason for the limited use of the term in Yogacara abhidharma .

It is true that, in the context of Sarvastivada, abhidharma is the defini-tive teaching. But early in the Abhidharmako s abh a Ò ya , Vasubandhu dis-

tinguishes betweenabhidharma in its ultimate meaning, i.e., pure wisdom,and abhidharma in a conventional sense, namely impure wisdom as wellas the s a stra s (doctrinal treatises) that result in the attainment of purewisdom.152 Furthermore, as is well known, Vasubandhu denies that theabhidharma s a stra s are the words of the Buddha.153 Therefore, from Vasu-bandhu’s point of view, although most of the Sarvastivadin abhidharmathat he describes without criticism in the Abhidharmako s abh a Ò ya isconducive to pure wisdom, it is not necessarily a statement of all that isknown by pure wisdom.

In other words, the purview of the Abhidharmako s abh a Ò ya does notinclude Mahayana teachings such asa layavijñ a na . Vasubandhu accu-rately describes his own work asabhidharma based in general on theteaching of the Kas m i ra VaibhaÒikas.154 Saµ ghabhadra elaborates on this,quoting Vasubandhu as saying that, in addition to VaibhaÒika teachings,he has also taught a bit of another path.155 Saµ ghabhadra gives as exam-ples Vasubandhu’s statements aboutsa µ sth a nar u pa and the past and




reality of pr a pti : kiµ punar ida µ bi ja µ na ma / yan n a mar u pa µ phalotpattau samartha µ

sa k Òa t p a ra µ parye ∞a v a / santatipari ∞a mavi s eÒa t ( Abhidharmako s abh a Ò ya : 64.4-5).158 ka Ì Sautr a ntik a rtha Ì . ye s utra-pr a ma ∞ik a Ì na s a stra-pr a ma ∞ik a Ì . te Sautr a ntik a Ì( Abhidharmako s avya khya : 11.29-30).

382 ROBERT KRITZER



whether or not they can be characterized as Yogacara, contradict ortho-dox VaibhaÒika, for which the texts of the Sarvastivadinabhidharmapi† akaare authoritative. If Vasubandhu admitted that they are indeed authorita-tive, he would not be able to criticize them as fundamentally as he does.By siding with “those who takesutra as authority,” he is free to reject thatwith which he disagrees, implying that it has not been taught in thesutras.On the other hand, he is still free to accept those Sarvastivadin opinionswith which he agrees. As Yasomitra makes clear, much ofabhidharmacan be found in thesutras, particularly ones like the Arthavinis cayasutra

that illuminate the characteristics ofdharmas.159 Furthermore, thesutrascan often be interpreted in more than one way. For example, the wordscitta, manas, andvijña na are mentioned together in the Di ghanik a ya,160

apparently as synonyms, which is in fact the way that they are understoodby the VaibhaÒikas.161 But theYoga ca rabhumi famously differentiatesthem:vijña na refers to the six traditional forms of consciousness,manasis kliÒ† amanas, andcitta is a layavijñana. If Vasubandhu’s intention in the Abhidharmakos abha Ò ya is secretly to reinterpretabhidharma,it is perhapsno wonder that he refers to his opinions as “Sautr antika.”162

Refe rences

Primary Sources

Abhidharmad i pa with Vibha Òa prabhav ® tti. Ed.P.S. Jaini. Tibetan Sanskrit WorksSeries 4. Second edition. Patna: Kashi Prasad Jayaswal Research Institute,1973.

Abhidharmakos abha Ò ya. Vasubandhu. Ed. P. Pradhan. Tibetan Sanskrit WorksSeries 8. Patna: K.P. Jayaswal Research Institute, 1967.

Abhidharmakos abha Ò ya. Vasubandhu. Peking Bstan ‘gyur 5591 (mdo’grelgu).

SAUTR A NTIKA IN THE ABHIDHARMAKOS ABH A∑YA 381

159 sutre ’pi hy abhidharma-pi† akaÌ pa† hyate. traipi† ako bhik Òur iti naiÒa doÒaÌ . sutra-vis eÒa eva Ì y Arthavinis cayadayo ’bhidharma-saµ jña Ì yeÒu dharma-lak Òa∞aµ var∞ yate( Abhidharmakos avyakhya : 11.32-12.1).

160 Yañ ca kho idaµ vuccati cittan ti va mano ti va viñña∞an ti va ayaµ att a nicco dhuvasassato avipari∞ama-dhammo sassati-samaµ tath’eva† hassat i ti ( Di ghanik a ya: 21.20-23).

161 cittaµ mano ’tha vijñanam ek a rthaµ ( Abhidharmakos a II 34ab [ Abhidharmakos a-bha Ò ya: 61.20]).

162 SeeYogacarabhumi: 11.3-8 (Yogacarabhumit : dzi 6a8-b3; T. 1579: 280b5-8; Schmit-hausen 117).



— . Traité de la Démonstration de L’Acte(edition and French translation ofKar-masiddhiprakara∞a). Mélanges chinois et bouddhiques4 (1935-36).Lopez, Donald S., ed. Buddhist Hermeneutics. Delhi: Motilal Banarsidass, 1993.Reprint (First edition: University of Hawaii Press, 1988).

Miyashita Seiki . “Kusharonni okeru hon mu kon u ron no haikei”. Bukkyogaku Semina 44 (1986): 7-37.

Mizuno Kogen . “Hiyushi to J o jitsuron ” . Komaza-wadaigaku Bukky o Gakkai nenp o 1 (1930): 134-156.

Nishi Giyu . Abidatsuma bukkyo no kenkyu . Tokyo:Kokusho kanko kai, 1975.

Rhys Davids,T.W. and William Stede.The Pali Text Society’s Pali-English Dic-tionary.London: The Pali Text Society, 1972. Reprint (First edition 1921-1925).

Rospatt, Alexander von.The Buddhist Doctrine of Momentariness.Institut für Kul-tur und Geschichte Indiens und Tibets an der Universität Hamburg, Alt- undNeu-Indische Studien 47. Stuttgart: Franz Steiner Verlag, 1995.

Saeki Kyokuga .Kand o Abidatsumakusharon .3 vols. Kyoto: Hozokan, 1978.

Schmithausen, Lambert. Alayavijña na. 2 vols. Studia Philologica Buddhica:Monograph Series IV. Tokyo: The International Institute for Buddhist Stud-ies, 1987.

Van den Broeck, José. La Saveur de l’Immortel( A-p’i-t’an Kan Lu Wei Lun).Publications de l'Institut Orientaliste de Louvain 15. Louvain-la-Neuve:Université Catholique de Louvain, 1977.

Wayman, Alex. Analysis of theSra vakabhumi Manuscript . University of Cali-fornia Publications in Classical Philology 17. Berkeley: University of Cali-fornia Press, 1961.

Yamabe Nobuyoshi . “An Shigao as a Precursor of the Yogacara Tradi-tion: A Preliminary Study.”Watanabe Takao Kyo ju kanreki kinen ronshu: Bukkyoshiso bunkashi ronso

Kyoto: Nagata Bunshodo , 1997. 785-826.

— . “Yugashichironni okeru zenaku inka setsu no issokumen”. Nihon Bukkyo Gakkai Nenpo 65 (2000): 127-146.Yamaguchi Susumu and Funahashi Issai .Kusharon no Genten

Kaimei, Seken-bon . Kyoto: Hozokan, 1955.

384 ROBERT KRITZER

sautrantika in the abhidharmakosabhasya

Documents