Pancavimsatisahasrika Prajnaparamita

·期刊原文

A comparative study of the 'Pancavimsatisahasrika

Prajnaparamita'

Shogo Watanabe

The Journal of the American Oriental Society

Vol.114 No.3 (July-Sep 1994)

pp.386-396

COPYRIGHT American Oriental Society 1994

---

It is a well-known fact that manuscripts of the

Pancavimsatisahasrika Prajnaparamita (PV) exist in both a revised

and unrevised form. The former are those of the revised text that

has been divided on the basis of Maitreya's Abhisamayalamkara (AA)

and Haribhadra's Abhisamayalamkaraloka (AAA),(1) while the latter

correspond to the original text of the PV, which is no longer extant

in its complete form. Our manuscript from Central Asia is of the

unrevised text of the PV.

The purpose of the following comparative study of the revised and

unrevised PV is to elucidate the manner in which the original PV

underwent changes - that is to say, the policies adopted and process

undergone in revising the text. Furthermore, by examining this

manuscript and clarifying its relation to other texts, we shall be

able to ascertain an aspect of the style of the original PV that has

been regarded as lost.

It is common knowledge that the PV is quoted in Indian

Buddhist texts, as well as in their Tibetan translations, simply as

the [Maha-]prajnaparamita.(2) In addition, we do not find quotations

deriving from the Astadasasahasrika (AD) and PV in the same text,

and it is not in fact possible to distinguish clearly manuscripts of

the unrevised PV from those of the AD. We may conclude from these

facts that the PV did not develop from the AD, and that both the AD

and PV are variant texts deriving from the same source. We could,

therefore, assume the existence of an Urtext of both the AD and the

PV.

First, in order to facilitate our comparative study of the Sanskrit,

Tibetan and Chinese versions, we shall list all the texts of the

"AD-PV line,"(3) which represent variants deriving from the same

source, and indicate the page or folio numbers and lines relevant to

our Central Asian fragment.

1) AD: Astadasasahasrika Prajnaparamita, chapters 70 to 82,

corresponding to the 6th, 7th and 8th Abhisamayas. Ed. E. Conze,

Serie Orientale Roma 46 (Roma: IsMEO, 1974), p. 5, 1. 33-p. 8, 1.

22.

2) PV: Aryapancavimsatisahasrikayam Bhagavatyam

Prajnaparamitayam

Abhisamayalamkara-nusarena samsodhitayam,

Anupurvabhisamaya-dhikara.h. Ed. Sh. Watanabe, Toyo Daigaku

Diagakuin Kiyo, no. 25 (Tokyo, 1989), p. 167, 1. 1-p. 172, 1. 21.

3) Ta: 'Phags pa shes rab kyi pha rol tu phyin pa khri brgyad stong

pa. Tibetan Tripitaka (Peking Edition), vol. 20, no. 732, ch. 70;

Phi ff. 50a8-53b1.

4) Tk: Shes rab kyi pha rol tu phyin pa stong nyi su lnga pa. Ibid.,

vol. 19, no. 731, ch. 59; Di ff. 75b3-80a8.

5) Tt: Shes rab kyi pha rol tu phyin pa stong nyi su lnga pa. Ibid.,

vol. 89, no. 5188, ch. 63; Ca ff. 224a5-229a3.

6) T220(2): Ta pan-jo po-lo-mi-to ching. Taisho Shinshu Daizokyo,

vol. 7, no. 220-2, ch. 73; pp. 35665-358b15.

7) T220(3): Ta pan-jo po-lo-mi-to ching. Ibid., vol. 7, no. 220-3,

ch. 28; pp. 710624-712a13.

8) T221: Fang-kuang pan-jo ching. Ibid., vol. 8, no. 221, ch. 76; p.

121a11-c2.

9) T222: Kuang-tsan ching. Ibid., vol. 8, no. 222; omits this

chapter.

10) T223: Mo-ho pan-jo po-lo-mi ching. Ibid., vol. 8, no. 223, chs.

75-76; pp. 385c5-386b6.

Eight of the texts in the above list have affiliations with the

unrevised PV, namely, the Sanskrit text of the AD (no. 1), the two

Tibetan versions Ta and Tk (nos. 3 and 4), and the five Chinese

versions (nos. 6-10). The remaining two texts - namely, the Sanskrit

text of the PV (no. 2) and the Tibetan Tt found in the Tanjur (no.

5) - are of the same type as the revised PV.

Our fragment corresponds to part of chapter six of the revised PV.

We shall, accordingly, first survey the structure of this chapter on

the basis of the classification added in the revised PV itself and

deriving from the AA and AAA.

Chapter Six: Anupurvabhisamaya (Gradual Intuition)

O. samanyenanupurvabhisamaya (Gradual Intuition in General)

The Six Perfections

1. danaparamita (The Perfection of Giving)

2. silaparamita (The Perfection of Morality)

3. ksantiparamita (The Recollection of Patience)

4. viryaparamita (The Perfection of Energy)

5. dhyanaparamita (The Perfection of Meditation)

6. prajnaparamita (The Perfection of Wisdom)

The Six Recollections

7. buddhanusmrti (The Recollection of the Buddha)

8. dharmanusmrti (The Recollection of the Dharma)

9. sanghanusmrti (The Recollection of the Samgha)

10. silanusmrti (The Recollection of Morality)

11. tyaganusmrti (The Recollection of Renunciation)

12. devatanusmrti (The Recollection of Deities)

13. sarvadharmabhavasvabhava (The Nonexistence of Own-being in

Dharmas)

This chapter may be said to first explain the nonexistence of all

beings, after which it goes on to describe the six perfections, the

recollections (anusmrti), and progressive understanding by means of

the knowledge that all beings have nonexistence as their own-being.

Our manuscript corresponds to the latter half of the chapter, from

section 7, buddhanusmrti to section 13, sarvadharmabhavasvabhava.(4)

An examination of this part of the revised Sanskrit text and the

manuscript of the original PV, etc., enables us to point out the

following techniques employed in revising the text.

1) The insertion of a set interrogative phrase calling for a reason

- i.e., tat kasya hetoh (and why?) - and the addition of explanatory

sentences before and/or after this phrase.

2) The expansion of the list of dharmas.

3) Additions and omissions. We find the addition of divisional

indicators, based on the AA and AAA, only in the text of the revised

PV; revisions of the PV may have occurred when it was recast in

order that it might accord with the divisions of the AA.(5)

4) Formal adjustments to set phrases relating to the categorized

dharmas.

We may, accordingly, cite the following passages as exemplifying the

above techniques used in revising the text.

1.1 The set phrase on nonexistence (abhava)

tha hi tesa dharmanam svabhavo nasti yasya svabhavo nasti so abhavo

*asmrti ama- (PV MS, recto 1. 8)

[ta]tha hi team dharmanam svabhavo nasti yasya sva-bhavo nasti. sah

abhavah.* a[bhava]smrty ama[- nasikaro hi dharmanusmrti.] (AD, p. 6,

11. 18-20)

[ta]tha hi tesam dharmanam svabhavo nasti yesam sva-bhavo nasti te

'bhavah / tat kasya hetor asmrtir ama[- nasikara dharmanusmrtih/]

(PV, p. 12, 11. 16-17)

'di ltar chos de dag la ngo bo nyid med de / gang la ngo bo nyid med

pa de ni dngos po reed pa'i phyir ro //* [de bas na chos rjes su

dran pa ni] dran par bya ba ma yin shing yid la bya ba ma yin no /

(Ta, Phi 51a3)

de ni 'di ltar chos de dag la ngo bo nyid med do / gang la ngo bo

nyid med pa de ni dngos po med pa'o / dngos po med pas ni dngos po

med pa dran par mi nus te [ de ci'i phyir zhe ha/de ni 'di ltar dran

pa med cing yid la bya ba reed pa [ni chos rjes su dran pa'o] (Tk,

Di 76b1-2)

de ni 'di ltar chos de nag la ngo bo nyid med do / gang la ngo bo

nyid med pa de ni dngos po med pa'o / dngos po med pas ni dngos po

med pa dran par mi nus te / de ci'i phyir zhe na / de ni 'di ltar

dran pa reed cing rjes su dran pa reed pa [ni chos rjes su dran

pa'o] (Tt, Ca 225a4-5)

In reply to Subhuti's question on how the bodhisattva-mahasattva

should develop the "recollection of dharmas (dharmanusmrti),"

Sakyamuni expounds the concept of the non-own-beingness of harmas

from the standpoint of the recollection of dharmas. This is a

passage typical of the PV line of texts.(6)

In both the AD and the PV MS the exposition consists of three

stages: [1] [Because] these dharmas have no own-being. [2] What has

no own-being is nonexistent. [3) Because the recollection of dharmas

is a nonrecollection [and nonattention]. In neither case is there,

however, the phrase tat kasya hetoh, asking the reason for the

nonexistence (abhava) of that which is without own-being. The

Tibetan Ta is identical to the AD and PV MS except for the addition

of the word "therefore (de bas na)" (= tasmat) between [2] and [3].

By way of contrast, the revised PV contains the set phrase tat kasya

hetoh, while the two Tibetan versions Tk and Tt insert the following

explanatory comment before this phrase: "It is impossible to

recollect the nonexistent by means of the nonexistent. And why? In

this manner.... (dngos po med pas ni dngos po med pa dran par mi nus

te/ de ci' i phyir zhe ha/de ni 'di ltar)." (It will also be noted

that in [3] the Tt has rjes su dran pa reed pa [ananusmrti] instead

of yid la bya ba med pa [amanasikara].)

Ordinarily speaking, to recollect dharmas means to be attentive to

and recall dharmas. But the later compilers of the

Prajnaparamita-sutra feared that the paradoxical expression "the

recollection of dharmas is a nonrecollection and nonattention

(asmrty-amanasikaro hi dharmanusrmrti[h])" would be misunderstood,

and they accordingly added the above explanatory comment in

accordance with their own understanding of this passage. This

passage may be assumed to have evolved in this manner.

The corresponding passages in the Chinese translations are as

follows:

[UNKNOWN TEXT OMITTED]

A glance at the above translations reveals that the more recent the

date of translation, the longer the translation becomes. Firstly,

apart from some differences in terminology, T221, the oldest

translation, and T223, the second-oldest, basically consist of three

stages and tally with the PV MS and AD, which omit the interrogative

phrase tat kasya hetoh. A variant reading of T223, however, contains

the corresponding phrase, namely, ho wei ku, thus evidencing a

closer resemblance to the PV. The T220(3) differs from the T220(2)

in that it adds the words marked with a double underline in stage

[2], and its exposition consists in fact of four stages. In

addition, the T220(2) and T220(3) also contain the words

"nonrecollection and nonattention (pu-k'o-nien pu-k'o-ssu-wei)"

(underscored with a wavy line), corresponding to the explanatory

comment added to the Tibetan versions Tk and Tt. In conclusion it

may be said that the T220(3) represents the most developed version

of the AD-PV line of texts.

In the above we have been able to shed light on the process whereby

the original text was gradually augmented with the addition of an

interrogative phrase of reason and explanatory comments. Another

example of the absence of the phrase tat kasya hetoh is found in the

PV MS, recto 1.2.

1.2 Repeated revisions

samgho anusmartavya yava sarvvakaramjnata anupra-punisyati anusmrti

amanasikarena (PV MS, recto 1. 11)

samghanusmrtih. (xxx)srapnoty[?] asmrty-amanasika-rena. (AD, p. 6,

11. 29-30)

samgho 'bhavasvabhavato 'nusmartavyo yo so bhaga-vatah

sravakasamghas catvarah purusayugah / astau mahapurusapudgalas tat

kasya hetos tatha hi tesam svabhavo nasti yesam svabhavo nasti te

'bhava abhavasvabhavapratibhavitas caivocyante catvarah purusayuga

astau mahapurusapudgalah / tat kasya hetoh / asmrtir amanasikara

samghanusmrtih sa iha samghanusmrtyam siksitvabhavasvabhavayogena

ya-vat sarvakarajnatam anuprapsyati (PV, p. 12, 1. 24-p. 13, 1.2)

dge 'dun rjes su dran par bya ste / bcom ldan 'das nyan thos kyi dge

'dun skyes bu zung bshi dang / skyes bu gang zag brgyad pa gang yin

pa de dag gi ngo bo nyid reed de / skyes bu zung bshi dang / skyes

bu gang zag brgyad po de dag ni ngo bo nyid med cing dngos po med

pa'i ngo bo nyid kyis phye ba yin te / de ni dngos po med pa nyid du

yid la byas te / rnam pa thams cad mkhyen pa nyid thob pa nas (Ta,

Phi 51a7-8)

dge 'dun rjes su dran pa sgom ste / bcom ldan 'das kyi nyan thos kyi

dge 'dun / tshul khrims dang /ting nge 'dzin dang / shes rab dang /

rnam par grol ba dang / rnam par grol ba'i ye shes mthong ba phun

sum tshogs pa'i skyes zung bshi dang / skyes bu gang zag brgyad po

de dag dngos po med pa'i ngo bo nyid du yid la byed do // de ci'i

phyir zhe na / ... / de ci'i phyir zhe na / de ni 'di ltar dran pa

med cing yid la bya ba med pa ni dge 'dun rjes su dran pa'o // rab

'byor de ltar byang chub sems dpa' sems dpa' chen po dge 'dun rjes

su dran med pa de la bslabs sing / dngos po med pa'i ngo bo nyid kyi

tshul gyis rnam pa thams cad mkhyen pa nyid kyi bar du yongs su

rdzogs par byed do / (Tk, Di 76b6-77a2) (Tt, Ca 225b 1-4)

This passage constitutes part of the section on the "recollection of

the sangha (samghdnusmrti)," among the six recollections. In the

preceding line of the PV MS (1. 10) Sakyamuni asks, "Subhuti, how

should the bodhisattva[-mahasattva develop the recollection of the

sangha]?" and this is followed by an explanation of the method for

practicing the recollection of the sangha. According to the PV MS,

"the sangha is to be recollected [as being nonexistent], and by

neither recollecting nor being attentive to the sangha one will

attain to 'knowledge of all forms (sarvakarajnata)'." Among the

AD-PV line of texts, this corresponds most closely to the AD.

The PV, on the other hand, is marked by considerable accretions.

Following the statement that "the sangha has nonexistence as its

own-being," it goes on to explain that the sangha of the Blessed

One's disciples consists of the so-called "four pairs and eight

categories of men," all of whom are without own-being, and that "the

recollection of the sangha is to be learnt by the method of having

nonexistence as own-being" (italicized passages). This exposition

contains two interrogative phrases of reason, and in the case of the

first, in particular, it is preceded by an explanatory comment on

the sangha and then followed by a stereo-typed explication of

non-own-beingness typical of the pV.(7)

Among the Tibetan versions, the Ta is in this instance closer in

content to the PV than to the AD, stating as it does that the sangha

of the disciples of the Blessed One consisting of the four pairs and

eight categories of men is without own-being and has nonexistence as

its own-being. The Tk and Tt, on the other hand, both affiliated to

the PV, are virtually identical and have, in addition to the

elaboration found in the Ta and PV, the additional comment that "the

disciples of the Blessed One are endowed with [the five aggregates

of the Dharma consisting of] morality, meditation, wisdom,

liberation, and the knowledge-and-vision of liberation" (marked with

underlined italics), thereby raising the status of the disciples on

the basis of Abhidharma doctrine and suggesting repeated revisions

of the text in later times.

The corresponding Chinese translations are as follows:

[UNKNOWN TEXT OMITTED]

Apart from the phrase "from the initial generation of the intent

(ts'ung ch'u fa i)" not found in any of the other translations, T221

is the most succinct in its treatment of the recollection of the

sangha and contains no accretions. T223, on the other hand, adds a

short phrase corresponding to the reference to the four pairs and

eight categories of men in the PV (fen-pieh yu fo ti-tzu chung), but

there is as yet no evidence of the exposition of non-own-beingness

corresponding to the underlined section in the PV.

Once we come to T220(3), translated by Hsuan-tsang, however, there

are signs of excessive elaboration. In particular, the latter half

of the passage in question (enclosed in square brackets) may be

considered to have been added at the time of the compilation of

T220(3): that is, as follows. The bodhisattva-mahasattva, who

courses in perfect wisdom, trains the recollection of the sangha

just as the nonexistence. The bodhisattva-mahasattva, who just

trains the recollection of the sangha, will understand gradual

activity, training and progress. The bodhisattva-mahasattva, who

just thus understands gradual activity, training and profess,

fulfills the progressive stages from beginning with the four fields

of mindfulness (catvari smrtyupasthanani) to the knowledge of all

forms (sarvakarajnata). And according to the fulfillment, the

bodhisattva-mahasattva is fully awake to the 'knowledge of the

all-knowing one (sarvajnajnana)'.

The term 'knowledge of the all-knowing one (i-ch'ieh-chih-chih)' is

based on the doctrine of the three kinds of knowledge that developed

in the T221 (the sra-vaka's 'knowledge of everything

[i-ch'ieh-chih],' the bodhisattva's 'knowledge of the aspect of the

path [tao-hsiang-chih],' and the Buddha's 'knowledge of all forms

[i-ch'ieh-hsiang-chih]').(8) Although the T223 contains no reference

to this knowledge, the T221 mentions 'sa-yun-jo' (sarvajnata =

i-ch'ieh-chih). Hsuan-tsang's translation, on the other hand, has

i-ch'ieh-hsiang-chih, corresponding to sarvakarajnata / rnam pa

thams cad mkhyen pa nyid in the PV MS and the three Tibetan

translations, and it further adds the still more developed term

i-ch'ieh-chih-chih (sarvajnajnana). Generally speaking, the T220(3)

resembles the PV in content, but in this instance it is even more

detailed than the PV. But its elaboration is of the same stereotyped

content as that found in its sections on the other recollections,

and it is possible that in this case the accretion is merely the

result of a "formal adjustment."

The text with the most marked additions is T220(2), again

translated by Hsuan-tsang. For example, where T220(3) has "the

assembly of the Buddha's disciples is endowed with various merits,"

T220(2) adds that the Buddha's disciples are endowed with morality

and the other five aggregates of the Dharma. This tallies with the

aforementioned Tibetan translations of the PV (Tk and Tt), and

represents an example of composite textual revision. In this manner,

with the passage of time, considerable use was made of the

interrogative phrase of reason tat kasya hetoh, preceded or followed

by various explanatory comments.

2. Et cetera (pyalam)

smrtyupasthanesu siksitavya pyalam yava mahakarunayam siksitavya so

bodhimarga ... (PV MS, verso, 1. 7) x x x mahakaru)nayam

siksitavyam. sa ca bodhimarge

[siksamana] (AD, p. 7, II. 30-31)

smytyupasthanasamyakprahanarddhipadendriyabalabo-dhyangmargesu

siksitavyam / aryasatyapramana-dhyanarapyavimoksasamadhisamapattidharanimu-khesu siksitavyam / s

u siksitavyam / abhavasvabhavayogena / dasasu paramaitasu dasasu

bodhisattvabhumisu siksitavyam / mahakarunayam siksitavyam / sa evam

bodhimarge [siksamano] (PV, p. 14, II. 14-19)

dran pa nye bar gzhag pa bshi la bslab par bya'o // de bzhin du

sbyar te / snying rje chen po'i bar la bslab par bya'o // de de ltar

byang chub kyi lain la [slob na] (Ta, Phi 52b1-2)

dran pa nye bar gzhag pa rnams la bslab par bya'o / /yang dag par

spong ba dang//rdsu 'phrul gyi rkang pa dang / dbang po dang / stobs

dang / byang chub kyi yan lag dang / lam la bslab par bya'o //

'phags pa'i bden pa dang / bsam gtan dang / tshad med pa dang /

gzugs med pa'i snyoms par 'jugs pa dang / rnam par thar pa brgyad

dang / mthar gyis gnas pa'i snyoms par 'jug pa dgu dang / stong pa

nyid dang / mishan mamed pa dang / smon pa reed pa'i ting nge 'dzin

dang / mngon par shes pa dang / ling nge 'dzin dang / gzungs kyi sgo

dang / de bzhin gshegs pa'i stobs bcu dang / mi 'jigs pa dang / so

so yang dag par rig pa dang / byams pa chen po dang / shying rje

chen po dang (Tk, Di 78b1-4)

dran pa nye bar gzhag pa rnams la bslab par bya'o // yang dag par

spong ba dang / ... / ting nge 'dzin la snyoms par 'jug pa dang /

gzungs kyi sgo rnams la bslab par bya'o // stong pa nyid dang /

mtshan ma med pa dang / smon pa reed pa'i ting nge 'dzin dang /

dngos po med pa'i ngo bo nyid kyi tshul gyis de bzhin gshegs pa'i

stobs dang / rai 'jigs pa dang / so so yang dag par rig pa dang /

sangs rgyas kyi chos ma 'dres pa bcu brgyad la bslab par bya' o //

pha rol tu phyin pa bcu po rnams la bslab par bya'o // shying rje

chen po la bslab par bya'o (Tt, Ca 227a5-227b1)

This line of the PV MS may be interpreted as follows: "The fields of

mindfulness should be learnt, and as far as great compassion should

be learnt. He who [learns] the path to enlightenment [in this

manner] ..." Unfortunately, the corresponding expanded portion of

doctrinal categories in the AD is damaged (x x x), but it is to be

surmised that it did not differ greatly from the PV MA. The Tibetan

translation of the AD (Ta) tallies so closely with the PV MS that it

could be almost described as a translation of it. But it is

difficult to find correspondences in the remaining texts affiliated

to the PV, and, judging from the context, it is to be assumed that

the lists of doctrinal categories found in the PV and its Tibetan

translations (italicized passages) correspond to the abbreviated

expression pyalam yava of the PV MS (corresponding to de bzhin du

sbyar te in the Tibetan Ta).

As for the four Chinese translations, if we focus on the abbreviated

expression pyalam yava of the PV MS, we find that T221, although

employing a different expression, summarizes the various doctrinal

categories with the statement that "one should learn the

thirty-seven factors [of enlightenment] and great compassion and

great pity" [UNKNOWN TEXT OMITTED] (121b16). There is, however, no

corresponding passage in T223 (386a20) [UNKNOWN TEXT OMITTED]. The

corresponding passage in T220(3) [UNKNOWN TEXT OMITTED]. (711c19-21)

is considerably shorter than that of T220(2) [UNKNOWN TEXT OMITTED]

(357c25-358a4) and is close to that of the PV MS, while T220(2)

gives the most detailed list of doctrinal categories, and these find

correspondences only in the PV.

It may be noted here that there are numerous instances in which

the abbreviated expression yava sarvvakaramjnata of the PV MS is

expressly expanded.(9) Hsuan-tsang's translations, in particular,

show evidence of having added for stylistic reasons categories that

were originally missing and of having listed all the dharmas of a

particular category when they had been abbreviated, although it is

not clear whether these features were characteristics of the

Sanskrit text employed by Hstian-tsang or whether they are the

results of Hsuan-tsang's own embellishments. Be that as it may, a

distinctive feature of Hsuan-tsang's translations is that among the

various versions of the Prajnaparamita-sutra, with its many

repetitions, they preserve a reasonably consistent style.

3.1 The recollection of dharmas (dharmanusmrti)

tavya yatra anopi smrti nasti ka pu[na]rvvadam. smrtikarmo? *katham

ca Subhuti bodhisa ... (PV MS, recto, 1. 10) *kathan ca Subhu)te

bodhisattvena (AD, p. 6, 1. 24)

(bhavayi)tavya yatranv api smrtikarma nasti kah punar-vadah

smrtivigamah // iti dharmanusmrtih // *katham ca Subhute

bodhisa(ttvena) (PV, p. 12, 11. 21-23)

[bsgom par ]bya ste / de la ni dran pa'i chos phra rab tsam yang med

na / dran pa dang bral ba lta ci smos / *rab 'byor byang chub [sems

dpa'... ] ji ltar [bsgom zhe na] (Ta, 51a5-6)

dran pa'i las rdul tsam yang yod pa'am / dran pa dang bral bar 'gyur

ba lta ci zhig smos / *rab 'byor ji ltar na byang chub (Tk, 7664-5)

dran pa'i ias rdul tsam yang yod pa'am / dran pa dang bral bar 'gyur

ba lta ci zhig smos zhes bya ba ni chos rjes su dran pa yin no / */

rab 'byor ji ltar na byang chub (Tt, 225a7-8)

This passage coincides with a division in the text, with the first

half belonging to the section on the recollection of dharmas and the

second half belonging to the section on the recollection of the

sangha. The second half, from katham (marked with an asterisk)

onwards, represents the opening words of the section on the

recollection of the sangha, and it poses no particular problems

since it is found in all versions of the text. Here we shall

consider the pveceding first half of this passage, representing the

conclusion to the section on the recollection of dharmas, for it

merits our attention, in that it clearly demonstrates the process of

elaboration undergone in the PV line of texts. We shall first

translate the different versions of the passage linking the two

sections on the recollection of dharmas and the recollection of the

sangha.

AD "[... is to be cultivated.] *Subhuti, how [should] a

bodhisattva[-mahasattva cultivate the recollection of the samgha?]"

PV MS "[... is to be....] There is no mindfulness there whatsoever,

let alone any act of mindfulness. *Subhuti, how [should] a

bodhisattva[-mahasattva cultivate the recollection of the samgha?]"

PV "[... is to be cultivated]. There is no act of mindfulness

[there] whatsoever, let alone any cessation of mindfulness. The

above is the recollection of dharmas. *Subhuti, how [should] a

bodhisattva[-mahasattva cultivate the recollection of the samgha?]"

Ta "[... is to be cultivated]. There is no dharma of recollection

there whatsoever, let alone any cessation of mindfulness. *Subhuti,

how [should] a bodhisattva[-mahasattva cultivate the recollection of

the samgha?]"

Tk "[There is no thought of existence there, nor even any thought of

nonexistence.] There is no act of mindfulness whatsoever, let alone

any cessation of mindfulness. *Subhuti, how [should] a

bodhisattva[-mahasattva cultivate the recollection of the samgha?]"

Tt "[There is no thought of existence there, nor even any thought of

nonexistence.] There is no act of mindfulness whatsoever, let alone

any cessation of mindfulness. The above is the recollection of

dharmas. *Subhuti, how [should] a bodhisattva[-mahasattva cultivate

the recollection of the samgha?]"

In the passage in question, our Central Asian fragment begins

with the statement that "[the recollection of dharmas is to be

cultivated.] There is no mindfulness (smrti) there whatsoever, let

alone any act of mindfulness (smrti-karma)." In the AD this portion

is completely missing, and this would suggest that the PV MS added a

concluding statement to the text of the AD. But in the Ta, which

ought to represent a translation of the AD, not only do we find

words corresponding to the first half of the passage in question in

the PV MS, but the form they take - namely, "there is no dharma of

mindfulness (dran pa'i chos) there whatsoever, let alone any

cessation of mindfulness" - bears a resemblance rather to the PV or

PV MS.

The PV itself, with its statement that "there is no act of

mindfulness there whatsoever, let alone any cessalion of mindfulness

(smrti-vigama)," adopts an emphatic negative to negate

'mindfulness,' thus differing somewhat from the PV MS with its

negation of 'mindfulness' and the 'act of mindfulness,' but this

does not alter the fact that it is still 'mindfulness' that is being

negated. But in the context of the extant PV there does not appear

to be any impelling need to refer to the 'cessation of mindfulness,'

and it may be regarded as a purely formal adjustment. In addition,

the PV inserts the divisional indicator "the above is the

recollection of dharmas," thus providing further evidence of later

revision of the text.

Among the Tibetan translations, the Tt tallies completely with the

PV except for the initial sentence enclosed in square brackets,

while the Tk lacks the divisional indicator of the Tt and preserves

a slightly older form than that of the PV and Tt.

The corresponding passages in the Chinese translations are as

follows:

[UNKNOWN TEXT OMITTED]

Among the Chinese translations, T221 lacks the first half of this

passage, while T223 tallies with the PV MS with the difference that

the former has "recollection of dharmas (nien-fa [dharmanusmrti])"

where the latter has "act of mindfulness (smrti-karma)," As for

T220(2) and T220(3), the former introduces the concept of

non-own-beingness as constituting the basis of nonrecognition, and

whereas the concept mediating between non-own-beingness and the

negation of the recollection of dharmas is smrti in T223, in T220(2)

and T220(3) it is dharma. It is maintained, in other words, that

"(if everything is without own-being) dharmas (/mindfulness) cannot

be recognized, let alone the recollection of dharmas," and hence the

import of the argument of T220(2) and T220(3) is identical to that

of T223. A clear distinction should perhaps, however, be made

between these Chinese translations and the Tibetan Ta and Tk, with

their reference to the "cessation of mindfulness." Furthermore, in

view of the fact that a divisional indicator has been inserted only

in the PV and Tt, these two texts may be regarded as representing

the final stage of development. Hence, on the basis of a textual

comparison of this section alone, it is possible to postulate the

following sequence of textual revision: AD, T221 [right arrow] PV

MS, T223, T220(3), T220(2) [right arrow] Ta, Tk [right arrow] PV,

Tt.

3.2 The recollection of the samgha (samghdnusmrti) karmo nasti

katham ca Subhuti bodhisatvena maha-satvena silanusmrti

manasikarttavya iha [[center dot]]u... (PV MS, recto, 1. 12)

kathan ca Subhute bodhisattvena mahasattvena sila)-nusmrtir

bhavayitavya? iha Subhute (AD, p. 6, 11. 32-33)

[yatranv api smrti]karma nasti kah punarvadah smr-tivigamah // iti

samghanusmrtih // katham ca Subhute bodhisattvena mahasattvena

silanusmrtir bhavayitavya / iha Subhute (PV, p. 13, 11. 4-7)

[de la ni dran pa'i chos phra rub tsam yang] med do // rub 'byor

byang chub sems dpa' sems dpa' chen pos tshul khrims rjes su dran pa

ji ltar bsgom zhe na / rub ['byor] (Ta, 51b1-2)

[dran pa'i] las rdul tsam yang yod pa 'am / dran pa dang bral bar

'gyur ba lta ci zhig smos / rub 'byor ji ltar na byang chub sems

dpa' sems dpa' chen po tshul khrims rjes su dran pa sgom pa yin zhe

na / rub ['byor] (Tk, 77a3-4)

[dran pa'i] las rdul tsam yang yod pa'am / dran pa dang bral bar

'gyur ba lta ci zhig smos / rub 'byor ji ltar na byang chub sems

dpa' sems dpa' chen po tshul khrims rjes su dran pa sgom pa yin zhe

na / rub ['byor] (Tk, 77a3-4)

[dran pa'i] las rdul tsam yang yod pa'am / dran pa dane bral bar

'gyur ba lta ci zhig smos zhes bya ba ni dge 'dun rjes su dran pa

yin no // rab 'byor ji ltar na byang chub sems dpa' sems dpa' then

po tshul khrims rjes su dran pa sgom pa yin zhe na / rub ['byor]

(Tt, 22565-6)

This passage corresponds to the concluding portion of the section on

the "recollection of the sangha (samghanusrmrti)" quoted in 1.2

above and to the opening portion of the following section on the

"recollection of morality (silanusmrti)." The first half of the PV -

namely, "there is no act [of mindfulness (smrti-karma) there

whatsoever], let alone any cessation of mindfulness (smrti-vigama)"

- is, however, missing from the AD, while if we compare the PV and

PV MS, we find that the divisional indicator and the immediately

preceding words of the PV - namely, "... let alone any cessation of

mindfulness. The above is the recollection of the sangha" (marked

with underlined italics) - is not found in the PV MS. (The Tk and

Tt, the two Tibetan translations corresponding to the PV, tally with

the PV except for the omission of the divisional indicator in the

Tk.)

[UNKNOWN TEXT OMITTED]

The Tibetan Ta, on the other hand, omits the phrase "let alone any

cessation of mindfulness" found in the previous example and adopts

an intermediate form, namely, "there is no dharma of mindfulness

(smrti-dharma) there whatsoever." It is to be surmised that the

"dharma of mindfulness" of the Ta changed to the "act of

mindfulness" in the PV MS and further separated into the "act of

mindfulness" and "cessation of mindfulness" in the PV. This would

suggest a four-stage process of textual revision and elaboration,

namely, from the AD to the Ta, from the Ta to the PV MS, and from

the PV MS to the PV. Let us now translate the four different

conclusions to the section on the recollection of the sangha.

AD "Subhuti, in this manner the bodhisattva-mahasattva should

cultivate the recollection of the samgha."

Ta "Subhuti, in this manner the bodhisattva-mahasattva should

cultivate the recollection of the samgha. There is no dharma of

mindfulness there whatsoever."

PV MS "Subhuti, in this manner the bodhisattva-mahasattva should

attend to the recollection of the samgha. There is no act of

mindfulness there whatsoever."

PV "Subhuti, in this manner the bodhisattva-mahasattva should

cultivate the recollection of the samgha. There is no act of

mindfulness there whatsoever, let alone any cessation of

mindfulness."

This format will lead to more or less the same result as in our

first example. The corresponding Chinese translations adopt the same

form as the conclusion to the foregoing recollection of dharmas.

T221 thus tallies with the AD, while T223, T220(2) and T220(3)

resemble one another. Although there is a difference as to whether

it is 'mindfulness' (T223) or 'sangha' (T220[2], T220[3]) that is

first negated, in both cases this is followed by "let alone any

recollection of the sangha," and the final result is the same

negation of the recollection of the sangha. Although their context

differs somewhat, they may be said to be close to the PV MS.

Next, as regards the mode of expression, there are minor differences

between these three Chinese translations and the PV MS and Ta, but

since this emphatic form of negation also appears in the PV, Tk and

Tt, these texts may be said to represent a similar stage of textual

revision. But as regards the object of this negation, one may

perceive a similarity between Ta (smrti-dharma) and T223 (nien).

This accords approximately with our previous conclusion.

It is also possible to draw a similar conclusion from the passage

bridging the concluding portion of the section on the "recollection

of abandonment (tyaga-nusmrti)" and the opening portion of the

section on the "recollection of the gods (devatanusmrti)."(10)

1 See Sh. Watanabe, "Pancavimsatisahasrika Prajnaparamita, VII:

Ekaksanabhisamayadhikarah (1)," Toyo Daigaku Daigakuin Kiyo 27

(1991), esp. pp. 131-36.

2 We consider that both the AD and the PV were originally subsumed

under the title Prajnaparamita. As the number of Prajnaparamita

texts multiplied in India, it became necessary to give them separate

designations for the sake of identification. This was accomplished

at a later date by naming them according to the number of lines that

they contained, and so we have rather mundane titles such as

Satasahasrika (100,000), Pancavimsatisahasrika (25,000) and

Astadasasahasrika (18,000). But this method of designation was not

employed in China. For example, one Chinese version of the PV,

translated by Moksala, was entitled the Fang-kuang (T221) on the

basis of the title of its first chapter, while another version was

called the Kuang-tsan (T222), on the same principle, as was also the

Tao-hsing (T224), a translation of the AD. The appellations Ta-p'in

(T223) and Hsiao-p'in (T227) for Kumarajiva's translations are

designations merely for the sake of distinction, and the proper

title of both is Maha-prajna-paramita (-sutra).

3 Japanese scholars have generally regarded the AD and PV as

belonging to the same line of texts, which they have referred to as

the "Fang-kuang line" or "Ta-p'in line" after the Chinese titles.

Since the Sanskrit texts and Tibetan translations of the AD and PV

belong to the same category, we consider this classification to be

very appropriate. But being based on the Chinese versions

corresponding to the PV, the terms "Fang-kuang line" and "Ta-p'in

line" may give the misleading impression that this grouping excludes

the AD and its Tibetan and Chinese translations. In the present

case, we therefore advocate the use of the designation "AD-PV line"

for this group.

4 See Sh. Watanabe, "Pancavimsatisahasrika Prajnaparamita,

VI:Anupurvabhisamayadhikarah," Toyo Daigaku Daigakuin Kiyo 25

(1989): 180-81.

5 See Nancy R. Lethcoe, "Some Notes on the Relationship between the

Abhisamayalamkara, the Revised Pancavimsatisahasrika, and the

Chinese Translations of the Unrevised Pancavimsatisahasrika," JAOS

96 (1976): 511.

6 We have discussed the subject matter of this passage typical of

the theory of non-own-beingness in our paper "'Hannya-kyo' ni okeru

abhava no yoho abhava (Abhava in the Prajnaparamita Sutras),"

Indogaku Bukkyogaku Kenkyu 37.2 (1989): 121-25.

7. See n. 6.

8 See Kajiyoshi Koun, Daijo Bukkyo no seiritsu-shiteki kenkyu (A

study of the history of the development of Mahayana Buddhism)

(Tokyo: Sankibo Busshorin, 1980), 635-51.

9 Cf. PV MS, recto 1. 11; PV, p. 12, 1. 25-p. 13, 1. 1; T221, p.

121a27 (simpler than the AD and PV MS); T223, p. 386a1 (also impler

than the AD and PV MS); T220(2), p. 357a9-20 (close to the PV).

10 Cf. PV MS, verso 1. 1; AD, p. 7, 11. 12-13; PV, p. 13, 11. 26-27;

Tk, Di f. 7768; Tt, Ca f. 226b3; Ta, Phi f. 52a1-2; T221, p. 121b7;

T223, p. 386a12-13; T220(2), p. 357c1-2; T220(3), p. 711b23-25.

CONCLUSIONS

Manuscripts of the PV line

The first mention of the unrevised PV in Chinese records is the

reference to Chu Shih-hsing [UNKNOWN TEXT OMITTED] having obtained

the Sanskrit manuscript of T221 in Khotan in A.D. 260.(11)

Approximately four hundred years then elapse until Hsuan-tsang's

translation of the T220 in 660-63, and there is evidence of notable

developments and modifications in texts affiliated to the PV during

this period. Our manuscript (PV MS) is a manuscript of an old type

belonging to the line of the unrevised PV, and it preserves a

slightly more developed form than the T221, also originating from

Khotan. As has been pointed out on more than one occasion in the

above, it bears a closer resemblance to the Tibetan translation of

the AD (Ta) than to the Tibetan translation of the unrevised PV

(Tk), while, among the Chinese translations, it is closest to the

T223 and closer to the T220(3) than to the T220(2). Hence, although

our manuscript is closely related to the AD line, it clearly differs

from the AD originating from Gilgit, and in view of the existence of

passages tallying with the PV, we have identified it as an old

manuscript of the PV. It is likely that several versions of the

Prajnaparamita-sutra similar to this type of old manuscript were

circulating at the time in Khotan.

Modes of Textual Revision

Following a comparison of our manuscript with the various texts of

the AD-PV line, it has been possible to shed light on the process

whereby the texts affiliated to the PV underwent revision. The modes

of textual revision adopted were as follows:

1) The adjustment of format on the basis of traditional doctrinal

categories. In other words, the phrase "all dharmas" is replaced by

detailed lists of various categories of dharmas, giving the

impression that the Prajnaparamita-sutra wished to present itself as

an exposition of Mahayana Abhidharma.

2) The explanatory elaboration of earlier texts. The phrase "and

why?" is added to a statement in the original text, and by

explaining the reason in reply to this interpolated question the

surrounding context is consequently coordinated and expanded.

3) Standardization and formal adjustments towards this end within

each text. The various texts of the AD-PV line represent different

stages of textual development. As the Prajnaparamita-sutra with its

many repetitions evolved, there arose various irregularities, and

the compiler(s) or translator(s) of each text made internal

adjustments so as to remove these irregularities, resulting in

repetitions of similar patterns.

4) The insertion of divisional indicators in the revised PV. This is

a modification found only in the revised PV, in which divisional

indicators are added to the unrevised PV and the surrounding context

is coordinated.

In our foregoing discussions we have been able to gain a glimpse of

these four modes of textual revision, and as a result we have come

to hold the following views in regard to the Urtext of the PV.

Conclusion Based on the Differences between the Various AD-PV Texts

To date, comparative studies of texts affiliated to the PV have

presupposed the existence of a stereotypical PV. But the

discrepancies between these texts as discussed in our present study

are sufficiently diverse to cause one to abandon this assumption,

and these discrepancies are so great that the texts in question

cannot be regarded as different translations of a single base text.

Furthermore, it is impossible to differentiate clearly between texts

affiliated to the PV and those affiliated to the AD. We have

accordingly proposed that all these texts be referred to

collectively as the AD-PV line, and we have examined them on the

basis of this line of thinking.

The conclusion that we have reached is that there exists no standard

text of the AD or PV and that these texts ought to be regarded as "a

group of affiliated sutras that developed independently of one

another." They do, of course, have points in common in regard to the

content of the thought to which they give expression, and

recognizing common points in this respect, we have suggested the

need for constructing an Urtext.

Therefore, this Urtext is not a text existent in the normal meaning;

it represents nothing more than a provisional concept to assist,

when considering the evolution of the many texts of the

Prajnaparamita-sutra presently extant, in clarifying the current of

development and elucidating the process of historical inevitability

that produced the extant texts. By postulating a single source, it

becomes possible to explain systematically the variations in the

philosophical development of Prajnaparamita literature as a whole.

Our Urtext is a factitious and hypothetical text born of such

requirements.

This means, in particular, that, in this sense, the extant AD and

the unrevised PV have a common Urtext. This Urtext evolved into what

came to be historically known as the Prajnaparamita, which was then

quoted in various treatises. The extant texts of the AD-PV line

represent variations born of this Urtext, and the old manuscript

examined in the above and regarded as being affiliated to the PV is

one such early example.

11 "Fang-kuang ching hou-chi (Postscript to the Fang-kuang ching),"

Ch'u san-tsang chi chi (Collection of notes concerning the

translation of the Tripitaka), Taisho Shinshu Daizokyo, vol. 55, p.

47a.