Acta Orientalia Academiae Scientiarum Hung.

Volume 72 (1), 33 – 46 (2019)

DOI: 10.1556/062.2019.72.1.3

A BRIEF INTRODUCTION TO TWO MANUSCRIPTS

OF BODHISATTVĀVADĀNAKALPALATĀ FOUND IN TIBET*
LIU Zhen

National Institute for Advanced Humanistic Studies at Fudan University

220 Handan Road, Shanghai, 200433, P.R. China
e-mail: [email protected]

There are two Sanskrit manuscripts of Kṣemendra’s Bodhisattvāvadānakalpalatā preserved at Dre-

pung Monastery in Tibet. The earlier one (according to the colophon) contains all the 108 avadānas
comprising the whole text, while the later one only includes the last 61 avadānas. In this paper, we
compare and analyse two paratextual elements of the extant versions, namely, pallava-endings and
prologue, in order to know how frequently the translator Shong ton Rdo rje rgyal mtshan and the
redactor Zha lu lotsāva Chos skyong bzang po used these two manuscripts.
Key words: Bodhisattvāvadānakalpalatā, Drepung Monastery, Shong ton Rdo rje rgyal mtshan,
Zha lu lotsāva Chos skyong bzang po.

Kṣemendra’s (ca. 980–after 1065 C.E.)1 Bodhisattvāvadānakalpalatā (henceforth:

Av-klp) is a highly influential poetic work within Indo-Tibetan Buddhist circles. His
son, Somendra, composed an epilogue2 according to which Kṣemendra composed the
Av-klp in 1052.
Shong ston Rdo rje rgyal mtshan and Lakṣmīkara translated it into Tibetan be-
tween ca. 1260 and 1280.3 Subsequently, Ta’i si tu Byang chub rgyal mtshan (1302–
1364) included a recension of its translation in the Tanjur edition (van der Kuijp 1994:
139– 142), and Zha lu lotsāva Chos skyong bzang po (1441–1528) prepared a bilingual

*
I would like to express my gratitude to Dragomir Dimitrov, Martin Straube and David
Fiordalis who gave me many valuable suggestions. Without them, this article would not be the same.
1
For the year of his birth and death I follow Formigatti (2019).
2
samvatsare saptaviṃśe vaiśākhasya sitodayo | kṛteyaṃ kalpalatikā jinajanmamahotsave ||
16 ||, referred by Bendall (1992: 18).
3
This dating is suggested by de Jong (1979: 5) and followed by Straube (2006: 41). How-
ever, there are different hypotheses: 1272 (Das 1888 – 1913, Vol. I: iii), ca. 1267 – 1270 (van der
Kuijp 1996: 401), ca. 1270 – 1275 (Mejor 1992: 5, Note 2) and followed by Dimitrov (2002: 45,
Note 185). For a brief summary of dating conjectures, cf. Dimitrov (2002: 37, Note 151).
0001-6446 © 2019 Akadémiai Kiadó, Budapest

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

34 LIU ZHEN

(Sanskrit and Tibetan) version (Mejor 1992: 65ff). During the reign of the Fifth Dalai
Lama, a complete and revised Sanskrit text of the Av-klp, together with its Tibetan
translation, was printed in 1664 and 1665 (cf. Mejor 1992: 64).
Five printing blocks and one manuscript of the Tibetan translation are known
to be preserved. Four printing blocks, Derge, Cone, Narthang and Peking, and one
manuscript, Ganden, are in the Tanjur. There is also the printing block prepared under
the Fifth Dalai Lama. Derge (henceforth: D1 for its transcription of Sanskrit text in
Tibetan script and D2 for the Tibetan translation in Derge-Tanjur, ke 1b –366a and
khe 1b–329a) and Cone include bilingual editions. So does the woodblock prepared
under the Fifth Dalai Lama. These editions belong to Straube’s hypo-archetype δ.
The others, which belong to Straube’s hypo-archetype β, only contain the Tibetan
translation.4
Until now, six Sanskrit manuscripts (henceforth Mss.) are known to be extant:
(1) Ms. A, at Cambridge University Library (Add. 1306, henceforth: A); (2) Ms. B, at
the same library (Add. 913, henceforth: B); (3) Ms. E, at the National Archives of
Nepal (reel No. B 95/5); (4) Ms. H, at the Library of the Asiatic Society of Calcutta
(Hudgson Collection, B. 15); (5) Ms. F, at the Bibliothèque nationale de France
(No. 8);5 (6) and a Ms. at the Kesar Library, Kathmandu (Kesar 519).6
All the above-mentioned Mss. are divided into two parts. Pallavas 1– 49 cor-
respond to the first part, while pallava 50 is the first chapter of the second part. E, H,
F lack the first part. A and B preserve pallavas 41 to 49. B and F preserve the second
part, but are incomplete (Straube 2006: 60–69; 2009: 10–11).
Two other Mss. of Av-klp have been found at Drepung (Tibetan: ’Bras spungs)
Monastery. Luo, in his catalogue, gives us a brief description of one box of Mss. (213
leaves) preserved at that monastery, which includes thirteen texts Luo (1984: 126–
135). However, the Av-klp is not mentioned there. Following Straube’s (2006: 60–
68) model, I provide detailed descriptions of both Mss. below.

Ms. C1
Possessor: Drepung Monastery (ZX0675-ZB 38).
Material: Palm leaf.
Size: 33 × 5.5 cm.
Script: Old Newārī.
Number of lines per page: 7.
Number of folios: 296 (complete).
Pagination: On the verso page of each folio, there are three pagination marks. Pagi-
nation mark 1, on the left edge in ‘letter-numerals’,7 is written in the same ink and

4
For his hypo-archetypes, cf. Straube (2006: 90 – 92; 2009: 18).
5
For detailed descriptions of Mss. A, B, E, H, cf. Straube (2006: 60 – 69). For a description
of Ms. F, cf. Filliozat (1941: 4 – 5) and Straube (2009: 10 – 11).
6
According to Martin Straube, this Ms. contains 13 folios of the 108th pallava, and has been
double filmed. One of the films is reel No. C48/9, and the other one is C 104/10 (email commu-
nication on 27 May 2018).
7
I follow Bendall’s terminology (1992: liii).
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

 A BRIEF INTRODUCTION TO TWO MANUSCRIPTS OF BODHISATTVĀVADĀNAKALPALATĀ 35

handwriting style as the main text. Pagination mark 2, on the right edge, reflects
Hindu-Arabic numbering and a similar Old Newārī style. It was probably written by
the same hand. Pagination mark 3, on the right edge beneath pagination 2, written in
New Newārī, was probably added by another person in a later period.
Condition of preservation: Apart from the fact that some akṣaras are beginning to
fade (e.g. 83a, 211a, 220a, 256a, 256b, 267a, 268a, 286b, 287a, 290b), the Ms. is still
in acceptable condition. On several pages, neighbouring pages in particular, the ink
has become blurred. The akṣaras there are hardly legible (e.g. 55a, 76a, 199b, 216a,
220b, 233b, 234a, 264b, 275b, 276a, 278b). Almost all titles of the pallavas are
marked in red.
Glosses/correction (tippana): Sometimes after a verse, the scribe identified figures
of speech (alaṃkāra) according to Alaṃkāraśāstra. Later, scribes or perhaps common
readers marked these terms in red, providing glosses with Tibetan counterparts. These
glosses appear on the blank areas near the Sanskrit terms.8 Occasionally, the words in
the main text also present glosses, such as on 282a and 284a.9 In the upper, lower or
right margin of the Ms., one may find scribal corrections. Also, a later person wrote
some akṣaras either in a script close to Devanāgarī or in Tibetan script. This person
did so in order to transcribe these akṣaras’ old script into a more familiar script. Addi-
tionally, there are some interlinear comments or notes. Assumedly, people who studied
the text produced these notes,10 which reflect four hands,11 being two in red ink,12
and two in black ink.13 Comparing the bilingual Sanskrit–Tibetan glosses on the first
page in a Nepalese Ms. of the Ratnaśrīṭīkā (NAK Acc. No.: 1/468), the first red note
presents a ductus similar to the one in that Ms., i.e., it consists in a possible autograph
by Shong ston Rdo rje rgyal mtshan himself.14
Punctuation: Relatively regular, with one single daṇḍa after each half verse and a
double daṇḍa at the end of the verse. At times, a verse number is added between two
double daṇḍas.

18
For example: (134a7) arthāntaranyāsaḥ is glossed as don gzhan bkod pa below, cf. Kāvyā-
darśa II 166: jñeyaḥ sorthāntaranyāso vastu prastutya kiñcana | tatsādhanasamarthasya nyāso
yonyasya vastunaḥ || and Tibetan translation: gang zhig dngos ’ga’ rab bkod nas | de yi sgrub byed
nus pa can | dngos po gzhan dag ’god pa de | don gzhan bkod par zhes par bya ||.
19
Straube (2009: 22) assumes that Zha lu lotsawa may have only provided glosses in the
margins of a Sanskrit Ms. instead of a new copy of a Sanskrit Ms.
10
For example: (99b, above) dmar chen gnyis po’i bar gyi snyad ’dod bod dpe la mi snang
ngo, ‘between the two big red [marks] the simile in the Tibetan text with wrong intention does not
appear’. Does “the two big red [marks]’ indicate the two terms marked with red in the previous page,
i.e. (99a4) śleṣaḥ and (99a7) upamā?
11
I am not sure if one of them wrote the glosses, too.
12
29a above, 31a above, 58a below, 71a below and 99b above by one hand (red 1), and
1
38a above and 187b below by one hand (red 2).
13
287b below and 293a above by one hand (black 1), and 282a above & below, 283b below
and 284a above by another hand (black 2).
14
The possibility that Shong ston Rdo rje rgyal mtshan left his autograph on the first page
in a Ms. of Ratnaśrīṭīkā was suggested by Dimitrov (2006: 6) who identified the interline notes in
red with another ‘autograph’ by Shong ston Rdo rje rgyal mtshan (personal communication, in No-
vember 2017).
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

36 LIU ZHEN

Dating: 1169 C.E. or shortly before/after 1269 C.E.

Content: (1) Av-klp 1–107 (1b1–281b7); (2) Somendra’s prologue (281b7– 282b6);
(3) Av-klp 108 (282b6–291b1; (4) Somendra’s epilogue (291b1–292a4); (5) So-
mendra’s list of contents (292a4–294a1); (6) praṇidhāna (294a1– 5); (7) colophon
(294a5–7).
Beginning:
(1b1)namaḥ śrīlokanāthāya || cittaṃ yasya sphaṭikavimalaṃ naiva gṛhṇāti rāgaṃ
kāruṇyārdre manasi nikhilāḥ śoṣitā yena doṣāḥ | akrodhena svayam abhihato yena
saṃsāraśatruḥ sarvvajño sau bhavatu bhavatāṃ śreyase ni(b2)ścalāya ||
Colophon:
(294a5) deyadharmmo (a6) ’yaṃ pravaramahāyānayāyinaḥ sakalatathāga-
t{{ā}}akṛtānugrahasya mahācīnadeśīyaśākyabhikṣuśrīmatirājasya yad atra || ||
puṇyaṃ tat* bhavatv ācāryopadhyāyamātāpitṛpūrvvagaṃmaṃ kṛtvā sakala-
sa(a7)tvarāśe{{ḥ śukha}}r anuttarajñānaphalāvāptaya iti || śūnyaṃ grahaś ca
bāhuś ca samvad yayau tadā {{dya}} māghasitetarapañcamadavase (odavase is
scribal error for odivase) yad alekhi mādhavo nityaṃ sarvvasatvārthahetunā || ||
namo buddhāya namo dharmāya namo saṅghāya.
“‘This is Śākyabhikṣu Śrīmatirāja’s religious gift; he is from mahācīna15, the
follower of the eminent Mahāyāna [, and has received] favors from all Tathā-
gatas. Having placed the master, teacher, mother and father in the foreground,
the entire group of living beings will obtain the merit, which [is created] here,
in order to obtain the fruit of the highest knowledge.’ On the fifth day of the
black [half] of the winter month in the lapsed year 290 (= 1169 C.E.)16, this
manuscript, which Mādhava has copied, is [dedicated] to the eternal wellbeing
of all living beings. Pay homage to the Buddha, the Dharma and the monk
community!”
On the penultimate inscribed page (294b) there is a line written in minute Tibetan
dbu med script:
paṇṭia (scribal error for paṇḍita) dge ba’i dbang pos le’u brgya dang bdun
mdzad do | dpag bsam ’khri shing le’u brgya dang brgyad yod pa las le’u
brgyad pa ’di zla ba’i dbang pos mdzad do |
‘Paṇḍita Kṣemendra composed 107 chapters. Since the Av-klp has 108 chap-
ters, the 108th chapter is composed by Somendra.’
Furthermore, after a blank leaf (295), there is a leaf (296) whose recto side is inscribed
while the verso is blank. On the recto page lines 1 and 3, there are three and a half
verses from Daṇḍin’s Kāvyādarśa (II 4–7ab) transcribed in Tibetan script, with the
corresponding Tibetan translation in lines 2 and 4. Besides these, there are occasional
glosses in red ink. All these are in dbu med. The words in the Tibetan translation are

15
Mahācīna is rendered as dbus gtsang in Tibetan, cf. Roesler (2018: 361 and Note 37).
16
I am in debt to Christopher Minkowski (Oxford) and Michael Witzel (Harvard) who
helped me to calculate the date.
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

 A BRIEF INTRODUCTION TO TWO MANUSCRIPTS OF BODHISATTVĀVADĀNAKALPALATĀ 37

divided with numbering beneath into word unities, which are linked with red dotted
lines to their respective counterparts in the Sanskrit text. According to the extant ver-
sions, both texts, Sanskrit and Tibetan, have scribal errors and a little different word-
ing.17 However, the 7th verse stops at Pāda-b; the rest of it cannot be found on the
verso page. It is difficult to say whether this folio was inserted into the Av-klp by mis-
take.

Ms. C2
Possessor: Drepung Monastery (ZX0650-ZB 22).
Material: palm leaf.
Size: 33.5 × 6.1 cm.
Script: Old Newārī.18
Number of lines per page: 5.19
Number of folios: 230.
Pagination: The right edge of the Ms shows only Hindu-Arabic numbering. It is in
the same style as that in the main text and is similar to the numbering 2 in C1. How-
ever, on 132b, the pagination mark repeats the number ‘131’. Therefore, from folio
132 to 230, the page number is always wrong, since the corresponding number is one
unit smaller than the correct number.
Condition of preservation: The Ms. shows an excellent state of preservation. The
akṣaras are clear. Almost all titles of the pallavas are marked in red.
Glosses/correction (tippana): The scribe made some corrections in both the upper
and the lower margins of the Ms. However, a later person also wrote some akṣaras in
a script close to Devanāgarī in order to decipher the corresponding old styled akṣaras
in the text. Unlike C1, this text has no annotations. The text presents almost no glosses,
except for 209b–210b, which has many glosses and a verse in Tibetan in the form of
notes.20
Punctuation: Relatively regular, with one single daṇḍa after each half verse and
a double daṇḍa at the end of the verse. Only at times, a verse number is added between
two double daṇḍas.

17
The Tibetan text here has the same peculiar readings as in Cone- (C) and Derge-Tanjur
(D), against these in Ganden- (G), Narthang- (N) and Peking-Tanjur (P), like, ldog pa can dang srid
pa can CD vs. ldog pa can dang srid pa dang GNP (II 4d) and lhan cig brjod dang yongs brjes shis
CD and lhan cig brjod dang yongs brjes smon GNP (II 7a). For these five Tanjur editions, cf.
https://www.tbrc.org/.
18
It has the same ductus as that of C1.
19
In each above and below margin of 5b, below margin of 153a, and below margin of 170a,
a line is added.
20
It is noteworthy that there is a verse-citation from the Tibetan translation of the 108th ava-
dāna on the upper margin of 210*a, skabs der tsandan ’khri shing rlung gzhon gyis | bskyod cing
shugs ring glal ba bsten rnams la | ’dod pa’i grogs por gyur pa dpyid kyi ni | mngon par ’dod pa’i
spyod tshul rab tu bstan |. This corresponds to the 14th verse of the Sanskrit version (210*a3):
atrāntare candanavallarīṇāṃ dide◯śa vālānilalolitānāṃ | ucchrasinīnām abhilāṣavṛttaṃ jṛṃbhā-
juṣāṃ kāmasuhṛdvasantaḥ ||. The latter is marked with a cross in the Ms. This means a reader
checked or compared this verse in the Ms. with its translation.
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

38 LIU ZHEN

Dating: 1269 C.E.

Content: (1) Av-klp 48–107 (1b1–208*a1); (2) Somendra’s prologue (208*a1–
209*a5); (3) Av-klp 108 (209*a5–222b*4); (4) Somendra’s epilogue (222b*4–
224*a4); (5) Somendra’s list of contents (224a*4– 226b*5; (6) praṇidhāna (226b*5–
227*a5);21 (7) colophon (227*a5– b2).
Beginning:
(1b1) namo lokanāthāya || svārthapravṛttau vigataspṛhrāṇāṃ paropakāre sata-
todyatānāṃ | kleśeṣv abhītāvyasanair anītā vighnair apīḍākaram eti siddhiḥ ||
Colophon:
(227*a5) *deyadharmmo ’yaṃ pra(b1)varamahāyānayāyinaḥ sakalatathāgata-
śāsanāmṛtasārasya bhoṭadeśīyaśākyabhikṣuśrīvajradhvajasya yad atra puṇyaṃ
tat* bhavatv ācāryopādhyāyamātāpitṛpūrvvaṃ gamaṃ kṛtvā sakala(b2)satva-
raśer anuttaraphalāvāptaya iti || samvat* 389 caitrapañcamyāṃ tithau likhi-
tam idam mādhaveneti || śubham astu sarvvajagatām iti ||
‘This is a religious gift from Śākyabhikṣu Śrīvajradhvaja22 from Bhoṭa,23 who
is a follower of the eminent Mahāyāna [, and has received] the essence of am-
brosia from the instruction of all the Tathāgatas! Having placed the master,
teacher, mother and father in the foreground, the entire group of living beings,
will have the merit, which [is created] here, in order to obtain the highest fruit.
On the 5th day of the second spring month in the year 389 (= 1269 C.E.), this
is copied by Mādhava. May all beings have auspiciousness.’
On 230*a, there is a line in Tibetan dbu med script:
’di ni rgya dpe cig shos mchog tu dag pa’i khyad ’phags |
‘This is the other Indian manuscript, which [contains] the excellence of fore-
most correctness.’
According to the above information, Mādhava copied two Mss. One of these (C1)
had a template whose dating can be traced back to 1169 C.E. Mādhava probably did
not alter the text, but he did replace the previous scribe’s name with his own (cf.
Bendall 1992: xviii–xix). In 1269 C.E., a Tibetan monk—or even more likely, Shong
ston Rdo rje rgyal mtshan himself—commissioned the other Ms. (C2). The colophon
of the second version was carefully prepared and does not reveal any information
about its template. Both versions are earlier than the oldest Ms. (A), found outside
Tibet.

21
There is one more verse: anena saddharmmarasāmṛtena sarvvajñavaktrād bhavaśītalena
| kleśānalaprajvālitāturasya lokasya duḥkhapraśamo ’stu nityam iti ||.
22
This name can be reconstructed into Tibetan as ‘Rdo rje rgyal mtshan’, which is identical
to one of the translators, Shong ston Rdo rje rgyal mtshan. Coincidently, in the colophon of another
Sanskrit Ms., Catuḥstotravivaraṇa, his Sanskrit name is also attested. According to Ye (2011: 175),
this Ms. was copied between 1260 and 1269 C.E. However, the Tibetan translation of this text has
left no trace in Tibet (cf. Luo 1985: 41 – 42; Ye 2011: 175 – 176).
23
In medieval Nepal (8th – 15th century), Bhoṭa can indicate either Tibet or Banepa (cf.
Petech 1984: 27 – 28). According to Roesler (2018: 361), ‘bhoṭa was an umbrella term for ethnically
Tibetan people’.
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

 A BRIEF INTRODUCTION TO TWO MANUSCRIPTS OF BODHISATTVĀVADĀNAKALPALATĀ 39

A Ms. of the Av-klp could have either been copied in its entirety,24 or divided
into two parts. In the latter case, the division may have taken place either before the
50th pallava, as Straube (2006: 60 –69) saw, or elsewhere. For instance, it could take
place at the beginning of the 48th pallava, since the short homage phrase25 is only
found there. As in the Tibetan transcription of the Sanskrit text (D1) and the Tibetan
translation (D2), C1 contains pallava 10, Garbhāvakrāntideśanā(vadāna),26 while
we do not find the Ṣaḍdantāvadāna in either of the two Mss.27 Based on the dating of
the Mss., it is possible that C1 and C2 could have been used in translation, redactions
and transcription. Due to its better condition of legibility, C2 may have been used
more frequently than C1 during some period. This might explain how the first part of
C2 became lost.28
Comparing C1 and C2 with the Tibetan translation and transcription in the
framework of the pallava-endings and the prologue can help us to determine whether
one or both Mss. were used in the translation and transcription process.
As we know, each pallava ends with a formulaic sentence, like iti kṣemendra-
viracitāyāṃ bodhisatvāvadānakalpalatāyāṃ [a title]-avadānaṃ [an ordinal number]
pallavaḥ. Here are alternative expressions, which trace back to scribal error or to a
scribe’s arbitrary decision.

6. Badaradvīpayātra
C1: (29b3) iti kṣemendraviracitāyāṃ bodhisatvāvadānamālāyāṃ29 kalpalatāyāṃ
badaradvīpayātrā ṣaṣṭaḥ pallavaḥ
D1:30 (ke 70b1) iti kṣe(b3)mendraviracitāyāṃ bodhisatvāvadānakalpalatāyāṃ badara-
dvīpayātrā ṣaṣṭaḥ pallavaḥ
D2: (70b2) zhes pa dge ba’i (b4) dbang pos byas pa’i byang chub sems dpa’i rtogs
pa brjod pa dpag bsam gyi ’khri shing las ba da ra’i gling du ’gro ba’i yal ’dab
ste drug pa’o

13. Hārītikadamana
C1: (56a2) hārītakādamanāvadānaṃ
D1: (132a5) hārītakādamanāvadānaṃ
D2: (132a6) ’phrog ma’i rtogs pa brjod gro ba’i

24
Until the colophon, we do not find any other short beginning homage phrase.
25
1b1: namo lokanāthāya.
26
40b4, I am not convinced by de Jong’s (1977: 28 ff) hypothesis concerning why this pal-
lava has been placed after pallava 9.
27
However, as in the other codices, Ṣaḍdantāvadāna as pallava 49 exists in Somendra’s list
of C1 and C2, and Garbhāvakrāntideśanāvadāna does not.
28
We cannot rule out the possibility that C2 is only a partial copy of the whole Av-klp com-
missioned by Shong ston Rdo rje rgyal mtshan.
29
This designation, avadānamālā, appears twice in C1 (here and 15b6 – 7, Av-klp 3) and
three times in C2 (32a5, Av-klp 54 , 40b1, Av-klp 58 and 189*a3, Av-klp 95).
30
Instead of Romanising the Tibetan transcription, I transliterate it to make the comparison
distinct.
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

40 LIU ZHEN

45. Ajātaśatrupitṛdroha
C1: (146b7) ajātaśatrupitṛdohao
D1: (342a5) ajātaśatrupitṛdrohao
D2: (342a6) ma skye dgras pha bsad pa

54. Sattvauṣadha
C1: (172a6) satvauṣadho nāmāvadānaṃ
C2: (khe 32a5) satvauṣadho nāmāvadānaṃ
D1: (36b5) satvauṣadho (b7) nāmāvadānaṃ
D2: (36b6) sems (b8) can sman zhes bya ba’i rtogs pa brjod pa

56. Gopālanāgadamana
C1: (175b6) gopālanāgadamanaṃ ṣaṭ*paṃcāśaḥ pallavaḥ
C2: (37b1) go(b2)pālanāgadamanāvadānaṃ ṣaṭ*paṃcāśaḥ pallavaḥ
D1: (45a5) gopālanāgadamanaṃ ṣaṣṭpañcaśaḥ pallavaḥ
D2: (45a6) klu ba lang skyon btul ba’i rtogs pa brjod pa’i yal ’dab ste lnga bcu rtsa
drug pa’o

58. Puṇyabala
C1: (177b5) puṇyavalāvadānaṃ nāmao
C2: (40b1) puṇyavalāvadānaṃ nāmao
D1: (47a1) puṇyavalāvadānaṃ nāmao
D2: (47a2) bsod nams stobs kyi rtogs pa brjod pa

74. pṛthivīpradāna
C1: (229b3) pṛthivīpradānaṃ (b4) catuḥsaptatitamaḥ pallavaḥ
C2: (121b2) pṛthivīpradānāvadānaṃ catuḥsaptatitamaḥ pallavaḥ
D1: (167b3) pṛthivīpradānaṃ catuḥsaptatitamaḥ pallavaḥ
D2: (167b4) sa gzhi phul ba’i rtogs pa brjod pa’i yal ’dab ste bdun cu tsa bzhi pa’o

82. Nārakapūrvika
C1: (241b2) pallavaḥ
C2: (145*a3) without pallavaḥ
D1: (196a1) pallavaḥ
D2: (196a2) yal ’dab ste

83. Rāhulakarmapluti
C1: (242b7) rāhulakarmmapluty
C2: (147*a) rāhulapūrvvakarmmapluty

Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

 A BRIEF INTRODUCTION TO TWO MANUSCRIPTS OF BODHISATTVĀVADĀNAKALPALATĀ 41

D1: (199a1) rāhulakarmmapluty

D2: (199a2) sgra gcan ’dzin31

89. Dharmaruci
C1: (258b7) dharmmarucyavadānam ekān navatitamaḥ pallavaḥ samāptaḥ
C2: (171*b1) dharmmarucyavadānam ekān navatitamaḥ pallavaḥ
D1: (236b5) dharmmarucyavadānam ekān navatitamaḥ pallavaḥ
D2: (236b6) chos sred kyi rtogs pa brjod pa’i yal ’dab ste brgyad cu rtsa dgu ba’o

92. Maitrakanyakā
C1: (264a1) maitryakanyakā
C2: (186*b) maitrakanyakā
D1: (248b1) maitryakanyakā
D2: (248b2) mdza’ bo’i bu mo

100. Puṇyaprabhāsa
C1: (273b1) punaḥ prabhāsāvadānaṃ
C2: (194*b3) puṇyaprabhāsāvadānaṃ
D1: (270b5) puṇyaḥ prabhāsāvadāna
D2: (270b6) bsod nams rab gsal gyi rtogs pa brjod pa

105. Raivata
C1: (279b5) rivatāvadānaṃ
C2: (204*b2) raivatāvadānaṃ
D1: (285b3) raivatāvadānaṃ
D2: (285b4) rai ba ta’i rtogs pa brjod pa

106. Kanakavarmā
C1: (280b6) kanakavarmmāvadānaṃ
C2: (206*a2) kanakavarmmāvadānaṃ
D1: (288a1) kanakavarmmāvadānaṃ
D2: (288a2) gser gyi go cha’i rtogs pa brjod pa

Here, the Tibetan translation largely follows C2, only two of its passages conform to
C1 (pallava 6 and 54). The Tibetan transcription is mostly drawn from C1. It is safe to
argue that the Tibetan translation is based on one or more Sanskrit Ms(s)., but there
are already several redactions made after the translation had been finished. Therefore,
almost all the variations in such formulaic expressions have been eliminated. There

31
It can be reconstructed as rāhulaka, i.e. the 83rd title in Somendra’s list of contents.
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

42 LIU ZHEN

are several examples reflecting the detailed relationship between C1, C2, the Tibetan
transcription, and the translation.

In Somendra’s prologue:
netrāmṛtapracitalepyamayī babhūva
buddhāvadānavividhādbhutakalpavallī || 1cd
c: netrāmṛtapracitalepyamayī A, B, C1, C2, D1] mig gi bdud rtsi rab skyed yi ge’i
rang bzhin = *netrāmṛtaprabhavalekhyamayī D2, citrāmṛtapracitalekhyamayī
Das.32
somendranāmā tanayo ’tha tasya
kavicchannakkāparanāmadheyaḥ | 4ab
b: kavicchannakkāparaº Apc, C1, C2] kavicchannakkāparāº D1, ming gzhan nā ko …
mkhan D2, kavir nākāparāº de Jong, kavir niruddhāparaº Aac, B, Das.
bandhaḥ keralakāminīkucabharākāraḥ paraṃ saṃhataḥ
kāñcīkāntakapolakomalataraḥ ko ’pi prasādodayaḥ |
kārṇādīnayanacchaṭā paricitā karṇāntarāsaṅginī
bhaṅgiryasya taraṅgiṇī rasanidher vāndyaḥ sa sūktodadhiḥ || 5
a: ºkucaºA, B, C1, Das] kacaº D1, skra D2o, ºkarao C2.
b: kāñcīº A, B, C1] kerīo C2, D2, kāṃrī D1, kāñjīº Das.
uṃkārāt kuṭilatvam eva paramaṃ yaiḥ śikṣitaṃ nāparaṃ
yeṣāṃ svastipadaṃ kadācid api na spaṣṭaṃ mukhān nirgatam | 6ab
a: nāparaṃ A, C2, D1] cāparaṃ B, gzhan pa yin D2, om. C1, līlayā Das.
b: yeṣāṃ A, B, C1, C2, D1] yasya Das, gang gi D2.
saṃsārorupariśramasya dadhataḥ kāmāsavakṣībatāṃ
rohaṃ mohatamo nimīlitadṛśaḥ suptasya luptasmṛteḥ |
saṃnaddhaṃ jagataḥ prabodhanavidhau niḥśeṣadoṣāpahaṃ
bhāsvantaṃ bhagavantam eva satataṃ buddhaṃ prabuddhaṃ namaḥ || 8
a: osavao C1, C2, chang D2] ºvaśaº A, B, ºpavaº D1.
b: rohaṃ mohaº C2] rohanmohaº A, B, rohasmāhaº C1, mohasmāhaº D1, gti mug
rgyas pa D2; luptaº A, C1, C2, D1, nyams pa D2] suptaº B.
c: saṃnaddhaṃ jagataḥ prabodhanavidhau C1, C2, D1] illegible A, saṃbuddhaṃ
kramataḥ prabodhanaparaṃ B, ’gro ba rab tu rtogs par sgrub la chas shing =
*sādhanaṃ jagataḥ prabodhanavidhau D2.
ānandabandhum asakṛt pṛthumānasānāṃ
sūktāṃśubhir vihitasarvamukhopadeśam |
kṣemendram ujjvalanijābhijanābdhicandraṃ
kīrttiprakāśajanakaṃ janakaṃ namāmi || 9
b: sūktāº C1, C2, D1, legs bshad D2] śuddhāº A, B; ºsarvaº A, B, C2, D2] ºsarvveo
C1, D1.
c: kṣemendram ujjvalanijāº C1, C2, D1] illegible A, kṣemendrasaṃjam aniśam B,
dri med … dge ba’i dbang = *kṣemendram aniśam D2.

32
The reference to Ms. A and Ms. B is based on de Jong (1979), but I have also compared
his reported readings with the manuscripts; de Jong for de Jong (1979), Das for Das (1888 – 1913).
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

 A BRIEF INTRODUCTION TO TWO MANUSCRIPTS OF BODHISATTVĀVADĀNAKALPALATĀ 43

vākpākapāvananidhānajināvadāna-
nirmāṇapuṇyakalanākuśalābhiyoge | 10ab
a: ºnidhānaº C1, C2, D1, gter D2] ovidhānao A, B, Das.
sarasvatī tūlikayā vicitra-
varṇakramair ullikhitāvadānaḥ | 12ab
a: sarasvatī A, B, C1, C2, dbyangs can D2] saratī D1.
b: varṇaº A, B, C2, D1, D2] karṇao C1; ullikhitāº A, C1, C2, D1, bris D2] saṃkalitāº
B, Das.
dikṣu pratiṣṭhāpitapustapālī-
sthiraprasaktapratimāgaṇasya || 13cd
c: ºpustapālīº A, B, C1, C2, po ti’ ’phreng ba D2] ºpustakālīº D1, ºpuṣpapālīº Das.
mādhuryadhuryam amṛtaṃ śrutiśuktipeyam
āmodasadmamukhapadmapade dhvanantīm | 15ab
a: śrutiśuktipeyam A, C1, C2, D1] śrutisūktipeyam B; thos pa’i skyogs33 kyis btung
bya’i D2, śrutipātrapeyam Das.
b: āmodasadmaº A, B, C1, dri bzang khang pa D2] ātmadasadmaº C2, āmoda-
padmaº D1.

Some variations are merely caused by misreading, e.g., u to a, c to r, sa to pa, m to s,

etc.
These examples, however, do not allow one to conclude whether the transla-
tors or earlier redactors had access to these two Mss. or to another lost Ms.34
In the colophon of the hypo-archetype δ of the Tibetan translation, there are
two puzzling sentences between the paragraph about the translation by Lakṣmīkara
and Shong ston Rdo rje rgyal and the paragraph about the bilingual edition by Zha lu
lotsawa:
‘ In the glorious monastery of Sa-skya the revision was accomplished,
after [the translation] was carefully compared with a Sanskrit manu-
script.  Since an accurate great revision for the two languages was ac-
complished, [this] is distinctly superior over the other.’35

33
However, śukti, ‘shell’, can also serve as a ladle.
34
In order to draw an ultimate conclusion and add these two Mss. into Straube’s stemma
codicum (Straube 2006: 73 – 87; 2009: 12), we have to go through more avadānas from the whole
codices.
35
Translated from Straube’s German translation, cf. Straube (2009: 21, Note 2). For the Ti-
betan text, cf. Mejor (1992: 65): dpal sa skya’i gtsug lag khang du rgya dpe dang gtugs nas zhu dag
tshar bar byas so || skad gnyis ka la zhu chen zhib pa cig grub pas gzhan las khyad du ’phags so ||.
Zha lu lotsawa’s bilingual edition is mentioned Subsequently: rgyal ba’i rtogs brjod snyan ngag
chen po dpag bsam ’khri shing zhes bya’i bstan bcos skad gnyis zung du sbyar ba ’dir || yi ge pa
dang dag byed pa yi skyon las yig ’bru lhag chad brda nor ’khrul ba ji snyed mchis pa rnams || ’di
yi skad gnyis zung sbyar thog mar byed po ’jig rten mig gyur chos skyong bzang po zhes bya’i dge
slong gis || legs sbyar gzhung dang sdeb sbyor bstan bcos bod skad brda yi bstan bcos dang bstun
blo gros zhib mos dag par byis ||.
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

44 LIU ZHEN

Mejor (1992: 65) builds an independent paragraph with these two sentences and inter-
prets that this revision is that one sponsored by Ta’i si tu Byang chub rgyal mtshan in
the 14th century.36 Since the name of Ta’i si tu does not appear in the hypo-archetype
δ, and the long praise in the hypo-archetype β looks like a later insertion, Mejor’s
interpretation remains difficult to accept.
Straube (2009: 20, Note 4) finds the phrase rgya dpe dang gtugs nas puzzling
and assumes that this Sanskrit Ms. must be another version, not the edition on which
the translation relied. However, he notes that the text in the colophon of his hypo-
archetype δ and dkar chag of Derge-Tanjur37 make the meaning clearer, and distrib-
utes sentence  to the prior paragraph, i.e. record of Lakṣmīkara and Shong ston
Rdo rje rgyal, and sentence  to the following paragraph, i.e. record of Zha lu
lotsawa. For sentence , he assumes that possibly Lakṣmīkara and Shong ston Rdo
rje rgyal used an additional Sanskrit Ms. for their translation and proof-text (Straube
2009: 20). For sentence , he regards that the added sentence component, this, is the
version prepared by Zha lu lotsawa (Straube 2009: 21, Note 2).
Since we have evidence that there are two Mss. and that there were possibly
more than two Mss., one or some of them may have been copied in Nepal under
Shong ton Rdo rje rgyal mtshan’s request38 for the purpose of translation or, possibly,
also for the purpose of the first redaction in the Sa-skya Monastery taking place just
after the translation. It is reasonable that both sentences  and  belong to the prior
paragraph, which is about Lakṣmīkara and Shong ston Rdo rje rgyal’s translation.39
If so, it could support Straube’s assumption, and we could interpret that this manuscript
was superior over the other manuscripts. However, it is not impossible that Shong ton
Rdo rje rgyal mtshan did not use C2, even though he let the scribe put his name in the
colophon.40
According to Deb ther sngon po, Shong ton Rdo rje rgyal mtshan was encour-
aged by ’Phags pa to study Sanskrit poetics and grammar in India, when the latter
returned to Tibet from the Mongolian court. Then he spent five years in Nepal for
study (Roerich 1976: 784–785). Since ’Phags pa himself was at the Sa-skya Monas-
tery from 1265 to 1267 (Petech 1990: 18 and Note 52), their first meeting must have
occurred in this period. It then seems reasonable to conjecture that Shong ton Rdo rje
rgyal mtshan obtained C2 in 1269 or a third Ms., for the purpose of translation or

36
Probably his interpretation was based on the colophon of the hypo-archetype β of the
Tibetan translation, in which sentence  remains and sentence  is replaced by a long praise for
the sponsorship of Ta’i si tu Byang chub rgyal mtshan. For the Tibetan text, cf. Mejor (1992: 86ff).
37
Straube (2009: 21, Note 2): dpal sa skya’i gtsug lag khang du rgya dpe dang legs par
gtugs nas skad gnyis ka’i cha ma la zhu chen zhib pa cig grub pas gzhan las khyad du ’phags pa |
slar yang ’di’i skad gnyis zhung du sbyar ba la | ’jig rten gyi mig tu gyur pa’i zha lu lotsā ba chos
skyong bzang po legs par sbyar ba’i skad yi gzung la sogs pa dang bstun te | zhu chen mdzad pa
bzhugs so ||.
38
Cf. Note 22.
39
The text in dkar chag of Derge-Tanjur has one more word, slar yang ‘afterwards’, and
the paragraph about Zha lu lotsawa can then begin after this word. Cf. Note 37.
40
He commissioned another Ms., Catuḥstotravivaraṇa. This text is regarded as one of the
Sanskrit texts that have never been translated into Tibetan, cf. Note 22.
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

 A BRIEF INTRODUCTION TO TWO MANUSCRIPTS OF BODHISATTVĀVADĀNAKALPALATĀ 45

redaction. From 1276 to 1280, ’Phags pa spent the rest of his life in Tibet at the Sa-
skya Monastery where he saw this translation by Shong ton Rdo rje rgyal mtshan.
According to these historical considerations, the period when this translation task
took place can be narrowed to 1265– 1280.41
Finally, in any case, we cannot connect both sentences  and  to the
paragraph about Zha lu lotsawa. According to the biography of Zha lu lotsawa, he had
never been to the Sa-skya Monastery, but had a good relationship with the Sa-skya
school. So, it is likely that he did not prepare the bilingual version there, but rather
got one or two Ms(s)., a little more likely C1, from the Sa-skya Monastery.42

References

BANERJEE, Anukul Chandra 1939. Kāvyādarśa, Sanskrit and Tibetan Texts. Calcutta: University of
Calcutta.
BENDALL, Cecil 1992. Catalogue of the Buddhist Sanskrit Manuscripts in the University Library,
Cambridge. [Publications of the Nepal-German Manuscript Preservation Project, 2; Ver-
zeichnis der Orientalischen Handschriften in Deutschland, Supplementband 33.] Stuttgart:
F. Steiner.
DAS, Sarat Chandra 1888 – 1913. Avadāna Kalpalatā. A Collection of Legendary Stories about the
Bodhisattvas by Kshemendra with Its Tibetan Version Called Rtogs brjod dpag bsam hkhri
śiṅ by Śoṅton Lochāva and Paṇḍita Lakshmīkara. Now First Edited from a Xylograph of
Lhasa and Sanskrit Manuscripts of Nepal by Sarat Chandra Dās and Paṇḍit Hari Mohan
Vidyābhūshaṇa (Vol. II: and Satis Chandra Vidyābhūṣaṇa). [Bibliotheca Indica.] Vols. I – II.
Calcutta: Baptist Mission Press.
DIMITROV, Dragomir 2002. Mārgavibhāga. Die Unterscheidung der Stilarten. Kritische Ausgabe
des ersten Kapitels von Daṇḍins Poetik Kāvyādarśa und der tibetischen Übertragung Sñan
ṅag me loṅ nebst einer deutschen Übersetzung des Sanskrittextes. [Indica et Tibetica, Bd.
40.] Marburg: Indica et Tibetica Verlag.
DIMITROV, Dragomir 2006. ‘Bilingual Sanskrit – Tibetan Glosses in a Nepalese Ms of the Ratna-
śrīṭīkā.’ Newsletter of the NGMCP 2: 4 – 7.
FILLIOZAT, Jean 1941. Catalogue du fonds sanscrit. Fas. I, Nr. 1 – 165. [Bibliothèque Nationale.
Department des Manuscripts.] Paris: Libr. d’Amérique et d’Orient Adrien-Maisonneuve.
FORMIGATTI, Camillo 2019. ‘Kṣemendra.’ In: Jonanthan SILK et al. (eds.) Brill’s Encyclopedia of
Buddhism, Vol. II. Leiden: Brill (forthcoming).
DE JONG, Jan W. 1977. ‘The Bodhisattvāvadānakalpalatā and the Ṣaḍantāvadāna.’ In: Leslie S. KA-
WAMURA and Keith SCOTT (eds.) Buddhist Thought and Asian Civilization. Essays in Honor
of Herbert V. Guenther on His Sixtieth Birthday. Emeryville: Dharma Publishing, 27–38.
DE JONG, Jan W. 1979. Textcritical Remarks on the Bodhisattvāvadānakalpalatā (Pallavas 42-108).
[Studia Philologica Buddhica, Monograph Series, II.] Tokyo: Reiyukai Library.
VAN DER KUIJP, Leonard 1994. ‘Fourteenth Century Tibetan Cultural History 1: Ta’i-si-tu-Byang-
chub-rgyal-mtshan as a man of religion.’ Indo-Iranian Journal 37: 139 – 149.

41
Dimitrov’s (2002: 45, Note 185) dating, ca. 1270 – 1275, relies on the same historical re-
cord.
42
Cf. an English summary of his biography at https://treasuryoflives.org/bo/biographies/
view/ChokyongZangpo/11285.
Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC

46 LIU ZHEN

VAN DER KUIJP, Leonard 1996. ‘Tibetan Belles-Lettres: The Influence of Daṇḍin and Kṣemendra.’
In: José Ignacio CABEZÓN and Roger R. JACKSON (eds.) Tibetan Literature. Studies in Genre.
Essays in Honor of Geshe Lhundup Sopa. New York: Snow Lion, 393 – 410.
LUO Zhao 罗炤 1984. Luobulinka suocang beiyejing mulu, fu Zhebangsi xiancang beiyejing gai-
kuang 羅布林卡所藏貝葉經目錄, 附哲蚌寺現藏貝葉經概況 [A catalogue of Sanskrit
manuscripts preserved at the Norbulingka, together with an appendix of a survey of San-
skrit manuscripts preserved at Drepung monastery]. (Unpublished manuscript.)
LUO Zhao 罗炤 1985. Budala gong suocang beiyejing mulu (danzhuer) 布達拉宮所藏貝葉經目錄
(丹珠爾) [A catalogue of Sanskrit manuscripts preserved at Potala-Palace (Tanjur)]. (Un-
published manuscript.)
MEJOR, Marek 1992. Kṣemendra’s Bodhisattvâvadānakalpalatā. Studies and Materials. [Studia
Philologica Buddhica, Monograph Series, VIII.] Tokyo: International Institute for Buddhist
Studies.
PETECH, Luciano 1984. Medieval History of Nepal (c. 750 – 1482). (Second, thoroughly revised ed.)
[Serie Orientale Roma, LIV.] Roma: Istituto Italiano per il Medio ed Estremo Oriente.
PETECH, Luciano 1990. Central Tibet and the Mongols. The Yüan – Sa-skya Period of Tibetan His-
tory. [Rome Oriental Series, LXV.] Roma: Istituto Italiano per il Medio ed Estremo Oriente.
ROERICH, George 1976. The Blue Annals. Parts I & II. Delhi: Motilal Banarsidass.
ROESLER, Ulrike 2018. ‘Rgya gar skad du – “in Sanskrit”?’ In: Oliver VON CRIEGERN, Gudrun
MELZER and Johannes SCHNEIDER (eds.) Saddharmāmṛtam. Festschrift für Jens-Uwe Hart-
mann zum 65. Geburtstag. [Wiener Studien zur Tibetologie und Buddhismuskunde, Heft 93.]
Wien: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien, 351 – 368.
STRAUBE, Martin 2006. Prinz Sudhana und die Kinnarī. Eine buddhistische Liebegeschichte von
Kṣemendra. Texte, Übersetzung, Studie. [Indica et Tibetica, Bd. 46.] Marburg: Indica et Ti-
betica Verlag.
STRAUBE, Martin 2009. Studien zur Bodhisattvāvadānakalpalatā. Texte und Quellen der Parallelen
zu Haribhaṭṭas Jātakamālā. [Veröffentlichungen der Helmuth von Glasenapp-Stiftung, Mo-
nographien 1.] Wiesbaden: Harrossowitz Verlag.
VERHAGEN, Pieter C. 1994. A History of Sanskrit Grammatical Literature in Tibet. Volume One,
Transmission of the Canonical Literature. [Handbuch der Orientalistik, Zweite Abteilung,
Indien, 8. Bd.] Leiden and New York: Brill.
YE Shaoyong 2011.《Zhonglun song》yu《Fohu shi》–– jiyu xin faxian fanwen xieben de wen-
xianxue yanjiu 《中论颂》与《佛护释》 — 基于新发现梵文写本的文献学研究 [Mūla-
madhyamakakārikā and Buddhapālita’s Commentary. A philological study on the basis of
newly identified Sanskrit manuscripts]. [Fanwen Beiyejing yu Fojiao Wenxian Xilie Cong-
shu 梵文贝叶经与佛教文献系列丛书, Series of Sanskrit Manuscripts & Buddhist Litera-
ture 1.] Shanghai: Zhong-Xi Shuju.

Acta Orient. Hung. 72, 2019

Unauthenticated | Downloaded 07/02/20 11:09 PM UTC