Tibetan Texts of Garbhāvakrāntisūtra: Diﬀerences and Borrowings

Robert KRITZER Tibetan Texts of Garbhāvakrāntisūtra: Diﬀerences and Borrowings

The International Research Institute for Advanced Buddhology Soka University Tokyo・2012・Hachioji JAPAN

Garbhāvakrāntisūtra (“Scripture on the Entry into the Womb”) is a Buddhist text that describes the rebirth process. It begins by explaining the conditions necessary for conception and why a woman might fail to conceive. It then describes the oedipal fantasies of the transmigrating being and its state of mind as it enters its mother’s womb.

The heart of the sūtra is a very detailed, week-by-week account of the 38 weeks of gestation, which is followed by descriptions of both miscarriage and successful childbirth. The remainder of the text is devoted to the suﬀering that befalls the newborn, including infestation by worms, attack by demons, and aﬄiction by many illnesses.

The longer versions of the sūtra also include a variety of additional topics, most notably an exposition of the four diﬀerent ways of entering the womb. The ﬁrst three types of garbhāvakrānti are those of virtuous people, some of whom are aware while entering, dwelling in, and emerging from the womb, some while entering and dwelling, and some only while entering. The fourth is that of the unvirtuous person, who loses his memory and awareness at the moment of death in the previous lifetime and is unaware at all three stages.

Garbhāvakrāntisūtra is signiﬁcant for a number of reasons. Its description of gestation is the most detailed in Indian medical and religious literature. Moreover, it appears to be the ﬁrst account divided into thirty-eight weeks; the nine-month account in the medical literature and in non-Buddhist religious texts is the norm. The sūtra seems to have both inﬂuenced and been inﬂuenced by the Indian medical tradition, with which it perhaps shared common sources (see Kritzer “Life”). The text also illustrates certain Buddhist attitudes toward the body: its detailed and pejorative descriptions of the mother’s vagina are representative of the more misogynistic strains of Buddhist asceticism (see Kritzer “Childbirth”).

The sūtra is also important as a source for descriptions of rebirth in Buddhist texts, including: meditation texts such as the Yogācārabhūmi of Saṃgharakṣa (Hsiu-hsing tao-ti ching 修行道地經, translated by Dharmarakṣa, and Tao-ti ching 道地經, translated by An Shih-kao); abhidharma texts such as the Vibhāṣā (A-p’i -ta-mo ta p’i-p’o-sha lun 阿毘達磨大毘婆沙論), Pañcavastukavibhāṣāśāstra (Wu-shih p’i-p’o-sha lun 五事毘婆沙論), and Abhidharmakośabhāṣya (Api-ta-mo chü-she lun 阿毘達磨倶舍論); Yogācārabhūmi (Yü-ch’ieh shih-ti lun 瑜伽師地論) (see Yamabe; Kritzer “Garbhāvakrāntau”).

Furthermore, together with Vāgbhaṭa’s Aṣṭāṅgahṛdaya, Garbhāvakrāntisūtra is one of the most important Indian texts for Tibetan embryology (Garrett 23-31). In particular, it is the main source for Tibetan theories of winds as a primary factor in fetal 131 PDF Version: ARIRIAB XV (2012), 131-145development (Garrett 146-148). Unfortunately, there are only a few Sanskrit fragments of the sūtra available in the form of quotations in abhidharma and Yogācāra texts, but there are six translations, three in Chinese and three in Tibetan. De Jong has pointed out that the various translations must be carefully compared in order to clarify diﬀerences among the versions of the text, and this paper is a preliminary attempt to do just that.

Table 1—Translations of Garbhāvakrāntisūtra Title Translator Date Pao-t’ai ching 胞胎經 (T. 317) Dharmarakṣa (Chu fa-hu 竺法護)

Fo wei a-nan shuo ch’u-t’ai hui 佛爲阿難説處胎會 (Ratnakūṭasūtra [T. 310, no. 13]) Bodhiruci (P’u-t’i-liu-chih 菩提流志) 703-713

Tshe daṅ ldan pa dga’ bo la mṅal du ’jug pa bstan pa (Translation of Ch’u-t’ai hui) Tohoku 58

Chos grub (Chinese name Fa-ch’eng 法成) ninth century

Fo shuo ju-t’ai-tsang hui 佛説入胎藏會 (Ratnakūṭasūtra [T. 310, no. 14])(also found in the Mūlasarvāstivādavinayakṣudrakavastu [T. 1451: 251a14-262a19])

I-ching 義淨 710 Dga’ bo la mṅal na gnas pa bstan pa (Translation of Ju-t’ai-tsang hui) Tohoku 57 Unknown (perhaps Chos grub) ninth century

Mṅal du ’jug pa źes bya ba’i chos kyi rnam graṅs (found in the Tibetan Mūlasarvāstivādavinayakṣudrakavastu) Tohoku 6

Vidyākaraprabha, Dharmaśribhadra, and Dpal ’byor ninth century

Pao-t’ai ching is by far the earliest version of the sūtra that we have. It is very diﬀerent from the other versions, and the Chinese is diﬃcult to read and probably corrupt in places. Although it is very important for the study of the sūtra, I will not discuss it in this paper.

Bodhiruci’s translation, Ch’u-t’ai hui, is said to be a translation of the same Sanskrit text that underlies Pao-t’ai ching.1 However, there are many diﬀerences between the two.

For example, while Pao-t’ai ching calls the interlocutor Nanda four times at the beginning of the text before changing to Ānanda, Ch’u-t’ai hui is consistent in using Ānanda throughout. Furthermore, Ch’u-t’ai hui includes a pair of verses in its account of Week 13 of gestation that are not found in Pao-t’ai ching. These and other diﬀerences

See, for example, K’ai-yüan shih-chiao lun 開元釋教錄 (T. 2154: 586c20-12).

PDF Version: ARIRIAB XV (2012), 131-145make it seem unlikely that the two are translations of the same text.2 In many respects Ch’u-t’ai hui is more similar to I-ching’s translation, Ju-t’ai ching.3 Chos grub or Fa-ch’eng, the famous Chinese translator based for much of his career in Tunhuang, translated Bodhiruci’s Chinese text into Tibetan (Tohoku 58).4 All the editions of this

translation contain the phrase mṅal du ’jug pa bstan pa (“the teaching about entering into the womb”) in their titles,5 but this phrase does not correspond to the Chinese title, which can be translated as “the chapter on dwelling in the womb.” Conversely, the title of the Tibetan translation (Tohoku 57) of I-ching’s Chinese translation (“the chapter on entering the womb”) generally

contains the element mṅal na gnas pa bstan pa (“the teaching on dwelling in the womb”).6 Thus, it seems likely that the Tibetan compilers of the Ratnakūṭa mistakenly switched the titles of the two texts. Because of this confusion of titles, I refer for the sake of clarity to the two Tibetan translations from Chinese by their Tohoku numbers: Tohoku 58 is the translation of Ch’u-t’ai hui, while Tohoku 57 is the translation of Ju-ta’i ching.

I-ching’s translation of the sūtra was originally part of his translation of Mūlasarvāstivādavinaya. Bodhiruci also inserted it into his compilation of Ratnakūṭa, and these two versions are almost, but not completely, identical. A comparison of the Tibetan translation (Tohoku 57) with the two Chinese versions shows that Tohoku 57 was based on the version in Ratnakūṭa. According to

Ueyama Daishun, the Tibetan translation of Ju-t’ai ching also may well have been Chos grub’s work (“Dai Bankoku,” 178). It seems as though this is the version of the sūtra best known in Tibet. Finally, the Tibetan translation that is found in Vinayakṣudrakavastu (Tohoku 6) is the only one made directly from the Sanskrit. I

refer to this text, too, by its Tohoku number. Currently I am preparing a critical edition and annotated translation of this version, which is the longest extant version in either Tibetan or Chinese. The language of this translation is more diﬃcult to

read than that of the other two, and it contains a considerable number of archaic words and words not found in any dictionaries. An examination of the extant texts suggests that there are two basic versions (or groups of versions) of the sūtra, a longer and a shorter one.

A detailed discussion of these diﬀerences will be included in the introduction to my forthcoming translation of the version of the sūtra in Tohoku 6.

Although Fo shuo ju-t’ai-tsang hui 佛説入胎藏會 is the title of the chapter of Ratnakūṭa that contains the sūtra, the sūtra proper is referred to in both versions of I-ching’s translation as Ju mu-t’ai ching 入母胎經 (T. 1451: 253a21; T. 310 no. 14: 328a9) and Ju-t’ai ching 入胎經 (T. 1451: 260b29; T. 310 no. 14: 335b1). For convenience, I use the short title Ju-t’ai ching in this article.

Tohoku 58 and Tohoku 57 are catalogued as separate texts in some Kanjurs but as chapters of Ratnakūṭa in others. For details, see bibliography.

The full titles are as follows: NFS dga’ bo mṅal du ’jug pa bstan pa źes bya ba theg pa chen po’i mdo; P ’phags pa tshe daṅ ldan pa / dga’ bo mṅal du ’jug pa bstan pa źes bya ba theg pa chen po’i mdo; D ’phags pa tshe daṅ ldan pa dga’ bo la mṅal du ’jug pa bstan pa źes bya ba theg pa chen po’i mdo.

The full titles are as follows: P ’phags pa dga’ bo mṅal na gnas pa bstan pa źes bya ba theg pa chen po’i mdo; D ’phags pa dga’ bo la mṅal na gnas pa bstan pa źes bya ba theg pa chen po’i mdo; N bcuṅ mo’u dga’ bo źes bya ba theg pa chen po’i mdo; S gcuṅ mo’u dga’ bo źes bya ba theg pa chen po’i mdo; F tshe ldan pa dga’ bo mṅal na gnas pa bstan pa źes bya ba / theg pa chen po’i mdo.

PDF Version: ARIRIAB XV (2012), 131-145Table 2—Versions of the sūtra

Short version: 1. Pao-t’ai ching (about 4.6 Taishō pages) 2. Ch’u-t’ai hui (about 4.3 Taishō pages) 2a. Mṅal du ’jug pa bstan pa (Tohoku 58: about 11 folios [22 sides] in Derge edition) Long version: 1. Tohoku 6 (about 28.6 folios) 2. Ju-t’ai ching (about 7.3 Taishō pages) 2a. Mṅal na gnas pa bstan pa (Tohoku 57: about 21.3 folios)

As usual, we do not know the dates of the Sanskrit texts on which these translations are based. Pao-t’ai ching and Ch’u-t’ai hui are said to be based on the same Sanskrit text and seem to represent an earlier version or versions. Ju-t’ai ching and Tohoku 6 are deﬁnitely translations of somewhat diﬀerent Sanskrit texts. We do not know which Sanskrit text was earlier, but on the principle of “longer equals later,” we might suppose that Ju-t’ai ching was based on an earlier text than Tohoku 6.

To give some idea of the diﬀerences among the versions, I provide the text of the explanations of the second week of gestation in the three Tibetan translations. The diﬀerent lengths of the passage are typical of the diﬀerent translations. The wordiness and repetitiveness of Tohoku 6 is evident at a glance: it uses more than twice as many near-synonyms as the other two translations to describe the “cooking” of the embryo in the womb.

Table 3—Account of Gestation, Week 2

Tohoku 6 dga’ bo źag bdun pa gñis pa la ni / mṅal gyi zug rṅu mi gtsaṅ ba rul pa rab tu ’bar ba’i ’dam gyi naṅ na ’dug pa lus kyi dbaṅ po thams cad rab tu tsha ba / śin tu mi bzad pa daṅ / dog pa daṅ / ñam ṅa bar gyur pa sdug bsṅal bar ro gcig pa’i rnam par śes pa can de la ma’i ltor naṅ gi rluṅ sṅon gyi las kyi rnam par smin pa las skyes pa kun sdud ces bya ba dag ldaṅ ste / de’i reg pas mṅal gyi zug rṅu de mer mer por ’gyur ro // dga’ bo de la mer mer po’i raṅ bźin ji lta bu źe na / dga’ bo dper na / źo ’am mar mkhraṅ po lta bu ste / mer mer po ni de lta bu’i raṅ bźin yin no // de der źag bdun du gnas śiṅ / de źag bdun po der gduṅ ba daṅ / kun tu gduṅ ba daṅ / tsha ba daṅ / kun tu tsha ba daṅ / yoṅs su tsha bar ’gyur te / de la sa’i khams ni sra ba ñid du / chu’i khams ni gśer ba ñid du / me’i khams ni dro ba ñid du / rluṅ gi khams ni yaṅ źiṅ g.yo ba ñid du ñe bar gnas so //

PDF Version: ARIRIAB XV (2012), 131-145Tohoku 57 dga’ bo źag bdun phrag gñis pa la ma’i mṅal na mi gtsaṅ ba’i gnas su ’di lta ste / dper na zaṅs kyi naṅ du btsos śiṅ brṅos pa bźin du lus kyi dbaṅ po daṅ rnam par śes pa lhan cig tu sdug bsṅal mi bzad pa myoṅ ba na ma’i mṅal nas kun tu reg pa źes bya ba’i rluṅ sṅon kyi las kyi dbaṅ gis raṅ byuṅ nas mṅal de la reg pa gaṅ yin pa de ni nur nur po źes bya’o // de’i dbyis ni źo ska ba’am / mar ’khyags pa ’dra ste / źag bdun gyi bar du btsos rab tu btsos pas na khams bźi po dag mṅon du snaṅ bar ’gyur ro //

Tohoku 58 źag bdun phrag gñis pa la ma’i mṅal du gnas pa’i tshe las kyi rluṅ mṅon par grub pa gaṅ yin pa de ni kun sdud ces bya ste rluṅ śin tu phra mo des ma’i rtsib logs g.yon pa daṅ / g.yas pa bus te mer mer po’i lus kyi mtshan ma rim gyis snaṅ bar byed pas de’i dbyibs źo ska ba’am / mar ’khyags pa ’dra ste / mṅal gyi naṅ du btsos / rab tu btsos pas nur nur po’i lus su gyur nas ’di ltar ’byuṅ ba chen po bźi po dag mthar gyis mṅon par grub par byed do //

The remainder of this article is essentially a comparison of several passages in the two Tibetan translations from the Chinese and the two Chinese texts that are the bases of the translations. Scholarly interest in Tibetan translations from Chinese has been focused for the most part on the fact that the Tibetan used in the translations is often quite diﬀerent from the Tibetan (broadly

speaking, the vocabulary prescribed by Mahāvyutpatti) used in translations from Sanskrit. Stein analyzes in detail some of the translations from the Chinese, and he identiﬁes examples of what he calls “Chinese vocabulary” in contrast to the “Indian vocabulary” found in texts translated after Khri lde sroṅ btsan’s edict of 814, in which the standardization of the translation vocabulary was decreed (152). For example, the standard Tibetan translation of saṃsāra is ’khor ba, while in a number of texts

translated from Chinese, the word skye śi, a literal translation of the Chinese sheng-ssu 生死 (“birth and death”) appears (Stein 174-175). In the case of the Garbhāvakrāntisūtra, neither of the translations from Chinese seems to contain what Stein refers to as the “Chinese” vocabulary.7 However, Tohoku 57 and Tohoku 58 are of interest for what they may tell us about Tibetan translation practices in Tunhuang in the early ninth century.

In many cases, I believe, diﬀerences between Tohoku 57 and 58 and the Chinese translations on which they are based reﬂect the translator’s reliance on another text, not the one he is supposedly translating. There are many possibilities, including: the other Chinese text (e.g., the translator of Ju-t’ai ching may have been looking at Ch’u-t’ai hui and translated what he found there rather than in Ju-t’ai ching); one of the other Tibetan translations (e.g., the translator of Ju-t’ai ching may have preferred a reading from Tohoku 58 or from Tohoku 6); a Sanskrit manuscript, of the text from which the Chinese translation had been made, the text from which Tohoku 6 had been translated, or yet

I have not read Tohoku 57 and Tohoku 58 in their entirety, but the portions that I have read seem to conform to Stein’s observation that Chos-grub generally follows the dictates of Mahāvyutpatti (149).

PDF Version: ARIRIAB XV (2012), 131-145 another Sanskrit version of the text, unknown to us. Of course, there are also diﬀerences that are mere mistakes or that simply reﬂect the translator’s idiosyncratic understanding of the Chinese.

However, I believe that in most cases when a diﬀerence between a Chinese translation and its corresponding Tibetan is attributable to reliance on a diﬀerent text, the Tibetan translator has deliberately accepted either the wording or the general meaning of the corresponding passage in Tohoku 6. In the following, I discuss a number of passages in Tohoku 58 and Tohoku 57 that appear to rely on another text.

Tohoku

Diﬀerences between Tohoku 58 and Ch’u-t’ai hui can be more or less convincingly explained as reﬂecting inﬂuence from Tohoku 6. When the reading in Tohoku 58 disagrees with Ch’u-t’ai hui but agrees with several other translations in addition to Tohoku 6, the inﬂuence of Tohoku 6 is less deﬁnite. When Tohoku 58 agrees only with Tohoku 6, the inﬂuence is fairly clear.

In the ﬁrst group of passages, Tohoku 58 diﬀers from Ch’u-t’ai hui but agrees with all other versions. Table 4—Passages in Tohoku 58 that disagree with Ch’u-t’ai hui (1) Diﬀerences probably explainable by the reliance of Tohoku 58 on Tohoku

Tohoku 58 Ch’u-ta’i hui
(處胎會)
Ju-t’ai ching
(入胎經)

Tohoku 57 Tohoku 6
interlocutor Nanda
(dga’ bo)
Ānanda
(阿難)

Nanda
(難陀)
Nanda
(dga’ bo)

Nanda (dga’ bo)
fault of the womb
oppressed by phlegm

(bad kan...gis ñe bar non par
gyur—237b1 [all references are to the Derge edition unless otherwise noted])
(missing) having phlegm

(或有黄病痰
癊—T.
1451:

253b16; T. 310 no. 14: 328b4)
aﬀected by phlegm
(bad kan gyis yoṅs su zin par
gyur—211b7)

oppressed by phlegm
(bad kan gyis bgags
pa—126a1)

PDF Version: ARIRIAB XV (2012), 131-145fault of the womb
ﬁlled with medicine
(sman gyis gaṅ

ba—237b1-2)
has salty disease (有醎8病—T.
310 no. 13: 322b1)

taking medicine (服藥—T.
1451:

253b17; T. 310 no. 14: 328b5)
taking medicine
(sman ’thungs—
212a1)

altered by medicine
(sman gyis bsgyur
ba—126a1)

Week 15 the 20,000 channels in the front of the body are called
sa ga or sa ka (lus kyi mdun na yod pa ñi khri ni sa ga
[FPS ka] źes bya’o— 241a5)
{sa ga and sa

ka probably are transliterations of Sanskrit
sakha

(Companion)}
the 20,000 channels in the front of the body are called Conch
(shang ch’ia)
(身前二萬名
曰商佉—T.

310 no. 13: 323c12)
{shang ch’ia
probably is a transliteration of Sanskrit
śaṅkha}
(the ﬁve types of channels in the front of the body) are called Companion (或名伴 T.
1451:

254c28; T. 310 no. 14: 329c9)
(the ﬁve types of channels in the front of the body) are called sa ga or
sa ka (miṅ ni sa ga
[PS ka] źes
kyaṅ bya—
216a2)

{sa ga and sa
ka probably are transliterations of Sanskrit
sakha

(Companion)}
there are channels called Companion
(grogs źes bya ba yod—
132a3)

In these cases, it is clear that Tohoku 58 does not correspond with the Chinese text that it is supposed to be translating. It is possible that Chos grub relied on either Ju-t’ai ching or its Tibetan translation. My working hypothesis is that Chos grub

translated Tohoku 57 as well, in which case, even if he had translated Tohoku 58 ﬁrst, he would most likely have already been familiar with the readings of Ju-t’ai ching, and if he had translated Tohoku 57 ﬁrst, he would have had his own translation to refer to. So it is not impossible that Ju-t’ai ching or Tohoku 57 is the source of his translation in Tohoku 58.

However, as we can see below, there are other cases in which the reading in Tohoku 58 can only be explained with reference to Tohoku 6, which strongly suggests that it, too, was available to Chos grub. It is possible that the greater length of Tohoku 6, and the fact that it was already translated into Tibetan, prompted Chos grub to give it more authority than the Chinese texts he was

translating. Therefore in these passages, Tohoku 6 is probably the ultimate source of the reading in Tohoku 58. Thus, in the third example, my supposition is that Chos grub must have seen “salty disease” but been aware that, according to Tohoku 6, the condition had something

The Koryō Tripiṭaka reads 鹹 for 醎 (K 22 [13]: 455a20.) I am indebted to Karashima Seishi for this information.

PDF Version: ARIRIAB XV (2012), 131-145to do with medicine. As we can see, the translation “ﬁlled with medicine” does not correspond precisely with “altered by medicine” in Tohoku 6. This could mean that the source is a translation or manuscript unknown to us. However, the “fault of the womb” mentioned immediately before this one is described as “the space is ﬁlled with ﬂesh” (mtshams śas gaṅ ba) in both Tohoku 58 and Tohoku 6, and the word “ﬁlled with” (gaṅ ba) may have been repeated in Tohoku 58 under the inﬂuence of the previous item.

The next passage is somewhat more certainly inﬂuenced by Tohoku 6. Table 5—Passages in Tohoku 58 that disagree with Ch’u-t’ai hui (2) Diﬀerence very probably explainable by the reliance of Tohoku 58 on Tohoku 6

Tohoku 58 Ch’u-t’ai hui
(處胎會)
[Ju-t’ai ching is
identical]
Tohoku 57 Tohoku 6
fault of the womb
center is like barley
(dbus nas ’dra ba)
has barley belly disease
(麥腹病)
center is like barley
(dbus nas ’dra ba)
center is like barley
(dbus nas ’dra ba)

In this case, the two Chinese translations agree. It appears as though both Tohoku 57 and Tohoku 58 have adopted the reading from Tohoku 6 or perhaps from a Sanskrit manuscript on which Tohoku 6 was based. Thus, the translator or translators of both Tohoku 58 and Tohoku 57 rejected “barley belly” disease in favor of a condition in which the center of the womb is like barley. The only extant full text of the sūtra with such a reading is Tohoku 6, although an abbreviated quotation from the sūtra in Yogācārabhūmi also suggests that center of the womb is being compared to a seed or grain.9

saced yonir vātopastabdhā bhavati pitto pastabdhā vā tilamadhyā vā... (Yogācārabhūmi 22.2-3; for a discussion of this passage, the Chinese and Tibetan translations of which mention barley as well as sesame, see Kritzer “Garbhāvakrāntau”).

PDF Version: ARIRIAB XV (2012), 131-145Finally, the following passages deﬁnitely seem to have been inﬂuenced by Tohoku 6.

Table 6—Passages in Tohoku 58 that disagree with Ch’u-t’ai hui (3) Diﬀerences almost certainly explainable by the reliance of Tohoku 58 on Tohoku 6
Tohoku 58
Ch’u-t’ai hui
(處胎會) Tohoku 6
name of wind, Week 2
All-uniting
(kun sdud)
All-completing
(遍滿)

All-uniting
(kun sdud)
Week 36 since the fetus gives rise to a thought of emerging (from the womb), it becomes unhappy
(de ’byuṅ ba’i bsam pa skye bas mṅon par mi dga’ bar ’gyur
ro—244a5)

the fetus gives rise to a thought of weariness and is unhappy
(生厭離心不以爲
樂—T. 310 no. 13: 324c19)
wishing to emerge, the fetus is unhappy
(de ’byuṅ ’dod ciṅ mṅon par mi
dga’o—136b2)

Week 37 the last of ﬁve unmistaken notions is the notion of emerging due to being weary
(yoṅs su skyo nas phyir ’byuṅ ba’i ’du śes skye ba gaṅ yin pa
ste—244a6)

the last of ﬁve unmistaken notions is the notion of loathing (厭惡想—T. 310 no. 13: 324c22)
the last of three unmistaken notions is the notion of emerging
(’byuṅ ba’i ’du
śes—136b3)

explanation of why all activities are suﬀering
“In short, as for this body consisting of the ﬁve upādāna
skandhas...” (mdor na ñe bar len pa’i phuṅ po lṅa’i lus
’di ni—247a5)

“This body consisting of the ﬁve skandhas...” (此五陰身—T. 310
no. 13: 326a2)

“In short, the ﬁve
upādāna skandhas...” (mdor na ñe bar len pa’i phuṅ po
lṅa—141a7)
Here, Ju-t’ai ching and Tohoku 57 do not correspond closely to the other three translations, so the most obvious source for Tohoku 58 is Tohoku 6. For example, in the second passage, Ch’u-t’ai hui mentions the fetus’s thought of weariness, but Chos grub follows Tohoku 6 in saying that it is a thought of emerging. Ju-t’ai ching and Tohoku 57

PDF Version: ARIRIAB XV (2012), 131-145mention neither weariness nor emerging but simply say that the fetus is unhappy.
On the other hand, there are some diﬀerences that can only be explained with reference to either Ju-t’ai ching or Tohoku 57.
Table 7—Passages in Tohoku 58 that disagree with Ch’u-t’ai hui (4) Diﬀerences explainable by the reliance of Tohoku 58 on Tohoku 57 or Ju-t’ai ching
Tohoku 58 Ch’u-t’ai
hui

(處胎會)
Ju-t’ai ching
(入胎經)
Tohoku 57
name of wind, Week 4 Inner Explanation
(naṅ rab tu byed pa)
Gather and Receive
(攝取)
Inner Explanation
(内開)

Inner Explanation
(naṅ rab tu byed pa)
name of worm that lives in the anus
Small Bundle
(po ta ra ka)
{Sanskrit poṭalaka}
Breast Wrinkles
(臆皺)
Small Bundle
(小束)
Small Bundle
(po ta ra ka)
{Sanskrit
poṭalaka}

In these cases, Tohoku 6 is quite diﬀerent and cannot be the source for the diﬀerences between Tohoku 58 and Ch’u-t’ai hui. For example, in Tohoku 6 the name of the wind in Week 4 is “Drawing Inward,” and the name of the anal worm is, aptly, “Possessing the Rotten.” These are clearly not reﬂected in any of the other translations.
Therefore, rejecting the reading in Ch’u-t’ai hui, Chos grub must have based his translation on either Tohoku 57 or Ju-t’ai ching.
Tohoku 57

Unsurprisingly, since the texts are longer, there are more diﬀerences between Tohoku 57 and Ju-t’ai ching than there are between Tohoku 58 and Ch’u-t’ai hui. Again, I present several groups of examples.
Table 8—Passages in Tohoku 57 that disagree with Ju-t’ai ching (1) Diﬀerences probably explainable by the reliance of Tohoku 57 on Tohoku 6

Tohoku 57 Ju-t’ai ching
(入胎經)
Ch’u-t’ai hui
(處胎會)

Tohoku 58 Tohoku 6
name of wind, Week 3
Treasury Opening
(mdzod kha)
Sheath Opening
(刀鞘口)
Treasury Opening
(藏口）
Treasury Opening
(mdzod ka)

Treasury Opening
(mdzod sgo)
140
PDF Version: ARIRIAB XV (2012), 131-145name of wind, Week 17
Yak Face
(’bri gdoṅ)
Hair Wipe Mouth
(毛拂口)
Yak Face

(髦牛面)
Yak Face
(’bri gdoṅ)
Yak Face
(’bri’i gdoṅ)
Week 21 (simile)
like a plasterer, or a plasterer’s pupil, who prepares some mud with which to plaster a wall
(dper na źal źal mkhan nam / źal źal mkhan gyi slob mas ’jim pa legs par sbyar nas rtsig ṅos la źal źal byed pa—
217a7-b1)

like a plasterer, who prepares some mud with which to plaster a wall
(譬如泥師先
好調泥泥於
牆壁—T.
310 no. 14: 330a19-12; T. 1451:
255b11-12)
like a plasterer, or a plasterer’s pupil, who prepares some mud with which to plaster a wall
(譬如泥師及
其弟子。能

善調泥泥諸
牆壁—T.
310 no. 13: 324a22-23)
like a plasterer, or a plasterer’s pupil, who prepares some mud with which to plaster a wall
(dper na źal źal mkhan nam / źal źal mkhan gyi slob mas ’jim pa legs par sbyaṅs nas rtsig pa rnams la źal źal byed pa—
242b1-2)

like a plasterer, or a plasterer’s pupil, who prepares some mud with which to plaster a wall
(dga’ bo dper na / źal źal mkhan nam / źal źal mkhan gyi slob ma mkhas pa ’jim pa legs par byas pas rtsig pa’i gźi la źal źal byed do—
134a2)

These, again, are cases in which the Tibetan translation diﬀers from the Chinese text on which it is supposedly based and agrees with the other Chinese and the two other Tibetan translations. In the ﬁrst example, we can imagine the underlying Sanskrit, something like kośamukha. Here, I-ching understands kośa in its meaning of a sheath for a sword. But the translator of Tohoku 57, probably based on Tohoku 6, prefers a translation of the other sense of kośa, a storehouse or treasury.
In the following examples, the probability of reliance on Tohoku 6 is even higher.

PDF Version: ARIRIAB XV (2012), 131-145Table 9—Passages in Tohoku 57 that disagree with Ju-t’ai ching (2) Diﬀerences almost
certainly explainable by the reliance of Tohoku 57 on Tohoku 6
Tohoku 57 Ju-t’ai ching
(入胎經)
Tohoku 6

Week 4 (shape of the fetus)
like a whetstone or millstone
(DP: de’i dbyibs ni bdar ram mchi gu ‘dra ste—
214b6-7 [P 258a4]) [or like a piece of dirt or a millstone (NFS:

de’i dbyibs ni boṅ pa ’am / mchig gu ‘dra
ste—N 441b1; F 383a4-5; S 401a4)]
Like a shoe last or warming stone
(状如鞋楥。或如温
石—T. 310 no. 14: 329b3; T. 1451: 254b16-17)
like a whetstone or millstone
(gdar ram mchi gu lta bu ste—130a6-7)

Week 27 (results of bad karma)
even if the being, when reborn, has the three karmas of body, speech, and mind, when he tells people, they will not believe or remember what he says

(lus daṅ / ṅag daṅ / yid kyi las ji sñed btsal ba dag gźan la smras na yaṅ gźan gyis yid mi ches śiṅ yid la mi ‘dzin pa dag mṅon ’grub par ’gyur ro—218a5)
even if the being, when reborn, has the three karmas, when he tells people, they will not believe or remember what he says
(所有三業向人説時。
他不信受不將在意—T.

310 no. 14: 330b13-14; T. 1451: 255c8-9)
whatever little bit he accomplishes in the world through body, speech, and mind, none of it will be delightful or acceptable in the world
(des ’jig rten la lus daṅ ṅag daṅ / yid kyis gaṅ cuṅ zad bsgrubs pa de thams cad kyaṅ / ’jig rten la bzuṅ bar ’os pa daṅ / mṅon par dga’ bar mi ’gyur ro—
135a1)

ten stages of life: stage 1
weak, he lies on his back
(daṅ po ni byis pa’i gnas skabs ste ñam chuṅ źiṅ gan rkyal du ñal bar
byed do—222b5-6)
he lies in swaddling clothes
(初謂嬰兒位臥於襁
褓—T. 310 no. 14: 332a9; T. 1451: 257b7)
weak, he lies on his back
(gnas skabs daṅ po la ni byis pa dman pa gan rkyal du ñal bar ‘gyur
ro—140a1-2)

the desirability and diﬃculty of being reborn as a human
a “good world” refers to birth as a human
(bde ’gro’i ’jig rten gaṅ źe na / ’di lta ste / de ni mi’i ’gro ba gaṅ yin
pa’o—227a7)
a “good world” refers to birth as a human or god
(云何世間善趣。謂是

人天—T. 310 no. 14: 333b10-12; T. 1451: 258c8-10)
a “good world” of the gods is being born as a human
(dga’ bo lha rnams kyi bde ’gro’i ’jig rten gaṅ źe na / mir skye ba yin
te—147b2)

PDF Version: ARIRIAB XV (2012), 131-145These passages are clear evidence of the agreement of Tohoku 57 with Tohoku 6 rather than Ju-t’ai ching. In these cases, Tohoku 58 and Ch’u-t’ai hui do not correspond to the other translations. Taking the third passage as an example, we can see that Tohoku 57 could not possibly be translating what is in Ju-t’ai ching, while it very closely matches Tohoku 6.

Finally, in some cases the reading in Tohoku 57 can best be explained by reliance on Tohoku 58.
Table 10—Passages in Tohoku 57 that disagree with Ju-t’ai ching (3) Diﬀerences probably explainable by the reliance of Tohoku 57 on Tohoku 58
Tohoku 57 Ju-t’ai ching
(入胎經)
Ch’u-ta’i hui
(處胎會)

Tohoku 58 Tohoku 6
fault of the womb
the womb is triangular like the joint of a carriage
(śiṅ rta’i tshigs bźin sogs kar ’dug pa—
212a1)
the womb is like a carriage shaft
(或如車轅—
T. 310 no. 14: 328b7; T. 1451:
253b19)

the womb is like the bent wood of a carriage shaft
(或如車轅曲
木—T. 310
no. 13: 322b2)
the womb is triangular like the joint of a carriage
(śiṅ rta’i tshigs pa bźin sogs kar ’dug
pa—237b2)

the opening of the womb is like a carriage
(sgo śiṅ rta lta
bu—126a2)
Week 27 (results of bad karma)
the body of the fetus will be ugly and resemble a
preta (lus mi sdug ciṅ yi dags daṅ ’dra bas—
218a4-5)

the shape of the fetus will resemble a
preta
(形如餓鬼—
T. 310 no. 14: 330b12-13; T. 1451:
255c7-8)
the body of the fetus will be ugly and resemble a
preta
(其身醜陋猶
如餓鬼—T.

310 no. 13: 324b16)
the body of the fetus will be ugly and resemble a
preta (lus mi sdug ciṅ yi dags daṅ ’dra
ba—243a4)
the fetus will resemble
pretas (yi dags rnams daṅ ’dra bar ’gyur
gyi—134b7)
In the ﬁrst case, there is no other translation that agrees with Tohoku 57 and Tohoku 58. In the second case, Ch’u-t’ai hui agrees with Tohoku 57 and Tohoku 58 against Ju-t’ai ching and Tohoku 6. Perhaps this suggests that Tohoku 58 was translated before Tohoku

PDF Version: ARIRIAB XV (2012), 131-145Conclusion
The textual history of Garbhāvakrāntisūtra is very complicated, and the six extant texts are probably only pieces of the whole puzzle. Therefore, any conclusions about the relative age of the diﬀerent Tibetan translations must remain tentative and subject to revision if new information becomes available.

One thing is deﬁnite: the Tibetan translator or translators of Ch’u-t’ai hui and Ju-t’ai ching must have used other texts of the sūtra, either other translations or Sanskrit manuscripts. My own speculation is that Tohoku 6 preceded both Tohoku 58 and Tohoku 57. However, it is diﬃcult to establish the chronological order of Tohoku 58 and Tohoku 57. It seems possible, especially if both translations were indeed the work of Chos grub, that there was two-way contamination, either while the texts were being translated or during the period in which they were being revised and edited. Since the texts were found side-by-side in the Chinese Ratnakūṭa, it is very likely that the translator or translators were familiar with the contents of both texts before they ﬁnished translating either one.

References

Chinese texts Ch’u-tai hui=Fo wei a-nan shuo ch’u-tai hui 佛爲阿難説處胎會 (= Garbhāvakrāntisūtra), translated by
Bodhiruci (P’u-t’i-liu-chih 菩提流志), T. 310 no. 13.
Ju-t’ai ching=Fo-shuo ju-t’ai tsang hui 佛説入胎藏會 (= Garbhāvakrāntisūtra), translated by I-ching 義淨,
T. 310 no. 14.
Ju-t’ai ching= Mūlasarvāstivādavinayakṣudrakavastu (Ken-pen shuo i-ch’ieh-yu pu p’i-nai-yeh tsa-shih 根本
説一切有部毘奈耶雜事), translated by I-ching, T. 1451.
Pao-t’ai ching 胞胎經 (= Garbhāvakrāntisūtra), translated by Dharmarakṣa (Chu Fa-hu 竺法護), T. 317.
Tibetan texts Tohoku 6= ’Dul ba phran tshegs kyi gźi (=Vinayakṣudrakavastu). Derge ’dul ba tha 124b6-153a1 (sūtra
proper).

Tohoku 57= Dga’ bo la mṅal na gnas pa bstan pa (= Garbhāvakrāntisūtra). Derge dkon brtsegs ga
205b1-236b7 (sūtra proper: 210b5-232a7);. Peking 760 (14) dkon brtsegs wi 248a3-282b4 (sūtra proper: 253b4-277a7); Narthang 45 dkon brtsegs ga 426b6-477a6 (sūtra proper: 434b7-469a2); Phug brag 45 dkon brtsegs kha 369a7-416a8 (sūtra proper: 377a1-409a); Stog 11 (14) dkon brtsegs ga 388a2-435a6 (sūtra proper: 395b1-428a3).
Tohoku 58= Tshe daṅ ldan pa dga’ bo la mṅal du ’jug pa bstan pa (= Garbhāvakrāntisūtra). Derge dkon brtsegs ga 237a1-248a7; Peking 760 (13) dkon brtsegs wi 234a3-248a7; Narthang 44 dkon brtsegs ga 408a1-426b6; Phug brag 44 dkon brtsegs kha 355b7-369a6; Stog 11 (13) dkon brtsegs ga 369a7-388a1.
Sigla for Tibetan editions

D Derge N Narthang F Phug brag P Peking S Stog Palace
Sanskrit text
Yogācārabhūmi. Ed. Vidhushekhara Bhattacharya. Calcutta: University of Calcutta, 1957. Part 1.

PDF Version: ARIRIAB XV (2012), 131-145Secondary sources Garrett, Frances. Religion, Medicine and the Human Embryo in Tibet. London: Routledge, 2008. de Jong, J.W. “The Bodhisattvāvadānakalpalatā and the Ṣaḍdantāvadāna.” In Buddhist Thought and Asian civilization : Essays in Honor of Herbert V. Guenther on his Sixtieth Birthday, ed. Leslie S. Kawamura and Keith Scott. Emeryville: Dharma Press, 1977. 27-38.

Kritzer, Robert. “Childbirth and the Mother’s Body in the Abhidharmakośabhāṣya and Related Texts.” In
Indo tetsugaku bukkyō shisō ron shū: Mikogami Eshō kyōju shōju kinen ronshū (インド哲学佛教
思想論集: 神子上恵生教授頌寿記念論集). Kyoto: Nagatabunshodō, 2004. 1085-1009. _______“Life in the Womb: Conception and Gestation in Buddhist Scripture and Classical Indian Medical

Literature.” In Imagining the Fetus: the Unborn in Myth, Religion, and Culture, ed. Vanessa Sasson
and Jane Marie Law. New York: Oxford University Press, 2009. 73-90.
_______ “Garbhāvakrāntau (‘In the Garbhāvakrānti’): Quotations from the Garbhāvakrāntisūtra in Abhidharma Literature and the Yogācārabhūmi.” In The Yogācārabhūmi and the Yogācāras, ed.

Ulrich Timme Kragh. Cambridge, Mass: Harvard University, Department of Sanskrit and Indian Studies, (forthcoming).
Stein, R.A. “Tibetica antiqua I: Les deux vocabulaires des traductions Indo-tibétaine et Sino-tibétaine dans
les manuscrits de Touen-huoang.” Bulletin de l’École Francaişe de Extrême-Orient 72 (1983):
149-236.

UEYAMA Daishun. “Dai bankoku daitoku sanzō hōshi shamon Hōjō no kenkyū (jō)” 大蕃國大徳三蔵法師沙門法成の研究（上). Tōhō gakuhō 38 (1967) 133-198.
YAMABE, Nobuyoshi. “On the School Aﬃliation of An Shigao: Sarvāstivāda and Yogācāra,” Unpublished draft of a paper given at the international workshop on “The Works of An Shigao” (Leiden, 19-20 Dec., 1996).

Source

Tibetan Texts of Garbhāvakrāntisūtra: Diﬀerences and Borrowings

Source

Navigation menu

Personal tools

Namespaces

Variants

Views

More

Search

Navigation

Categories

Tools