On the Sukhavativyuhasutra of the Ulan Bator Manuscript Kangyur

【以下の文は2005年11月に開催された国際学会「大乗仏教:歴史と文化」(ニューデリー・チベットハウス主催)に於ける発表の要旨です】

ON THE SUKHAVATIVYUHASUTRA OF ULAN BATOR MANUSCRIPT KANGYUR

Shunzo ONODA

(Bukkyo University, Kyoto JAPAN)

The research works by Helmut Eimer, Paul Harrison and Peter Skilling have contributed enormously to the development of our understanding of the relations between the different versions of the Kangyur. The common recognition made by the above mentioned scholars' joint research is that the versions of each Kangyur text appear in one of two groups or recensions. But the historical map still remains containing unclear parts. It shows a complex family tree with complicated lines of descent, but in which two principal branches can be distinguished. Those are the Tshal pa Kangyur, which was completed in 1351 at Tshal Gung thang monastery in dBus, and the Them spangs ma Kangyur, which was produced in 1431 at rGyal rtse in gTsang by the rGyal rtse ruler Rab brtan Kun bzang 'phags pa(1389-1442). Those two Kangyurs have their common source, the so-called the old sNar thang Kangyur of the early 14th century. So, the old sNar thang Kangyur is regarded as the grandmother or archetype of the two great recensions.

At the beginning of the 14th century, a collection of various copies of all the Tibetan translations of the mDo and bsTan bcos was made at the bKa' gdams pa monastery of sNar thang by dBus pa Blo gsal Byang chub ye shes and other monk scholars under the supervision of their teacher bCom ldan rig pa'i ral gri. That is the so-called the old sNar thang edition of the Kangyur. It is thought that the old sNar thang Kangyur was a raw collection probably containing many duplicates and further work would have been required to edit a "real collection". The above two recensions are thought to be the witnesses of that work.

The first branch, Tshal pa Kangyur side, contains the xylograph Kangyurs of Peking, Li thang, and Co ne. The second branch, Them spangs ma Kangyur side, contains the London MS (L), Stog Palace MS (S) and Kawaguchi or Tokyo MS (T) Kangyur, and the rather inaccessible Ulan Bator MS (U). It is considered that all three accessible editions on this side (London MS, Stog Palace MS and Tokyo Kawaguchi MS) are indeed descendants of the same MS. These are all descended from the Them spangs ma Kangyur.

The fourth one, Ulan Bator MS Kangyur, is also said to be a copy of the Them spangs ma Kangyur. This MS Kangyur was presented in 1671 to the first Rje btsun dam pa Khutuktu Blo bzang Bstan pa'i rgyal mtshan (1635-1723) by the Fifth Dalai Lama (1617-1682) and now kept in the State Library in Ulan Bator.

About this MS Kangyur, there is an interesting tradition which was reported by Geza Bethlenfalvy, and according to which this MS Kangyur is nothing other than the original rGyal rtse Them spangs ma Kangyur itself. Bethlenfalvy analyzes the meaning of title of the Them spangs ma as "that which is forbidden to cross a threshold". Actually we can find two red circles marking the string holes on the surface of the MS. This might be the reason behind such story, that this is the original Them spangs ma. (Dr. Eimer talks about the title of the MS saying "There a manuscript named Them spang ma, i.e. that one of authentic arrangement, served as examplar".) The size of the leaves of the MS is 25 x 72 cms, 8 lines cover an area of about 14.5 x 59 cms. According to Bathlenfalvy, 3 volumes are missing. MS are kept in a storehouse of an underground floor in the State Library of Mongolia. We can see one volume in the exhibition room on the second floor of the same bilding.

In the Biography of the rJe btsun dam pa written by his disciple, the Dzaya pandita Blo bzang 'phrin las (1642-1715), we can find a sentence concerning the MS Kangyur.

vol.4, 66a-6, bkra shis lhun po sogs dbus gtsang la bsdad bas de bar gyi rnam thar mdzad bzang rnams nges pa cig ma byung/ 'gal byed ces pa lcags mo phag lor gnyer pa bin thu nang so (66b-1) dbus gtsang du gtong ba gnang nas rgyal ba yab sras la sku 'tsam 'dri zhu ba'i rten 'bul dang chos sde che chung rnams su mang 'gyed gnang/ gzhung nas rgyal rtse them spangs lugs kyi bka' 'gyur rin po che sbul shin tu legs pa cig gdan (99b-2) drangs/ (He stayed in dBus gTsang at Tashilhunpo etc., During this time we do not have well made records, so we do not know well. 'Gal byed year, i.e. the year lcags mo phag (1671), when a certain Bin thu nang so was sent to dBus gtsang. He gave many presents to the Governor and his family showing his respect, and he made offering a lot to various big and small monasteries. From the Government of Tibet he received a holy Kangyur of very good quality corresponding to the rGyal rtse Them spangs ma. )

The author, Blo bzang 'phrin las a the person who made his own library and collected a lots of important and rare materials. We may believe he had correct knowledge about literature. He mentioned this MS as "Them spangs lugs kyi bKa' 'gyur". He wrote:

vol.4, 197a-1, ...lo pan rnams ki bka' drin la brten nas byang phyogs kha ba can gyi ljongs 'dir dar rgyas su mdzad pa'i bka' 'gyur rin po che las lung ji snyed cig thob pa'i chos kyi rnam grangs thor bkod pa la/ tshal pa bka' 'gyur dang/ rgyal rtse (a-2) them spangs ma sogs su chos tshan bzhugs tshul gyi go rims mi mthun na 'ang da lam rgyal rtse them spangs ma'i steng nas lung thob pas de'i go rims ltar dri bar bya'o//(By the endeavors of great translators, here in the northern snow land of Tibet, we have had a lot of literature as precious jewels of the Kangyur which are spread and they have properly got lungs . Among those there are two Kangyurs: Tshal pa Kangur and rGyal rtse Them spangs ma, and each of two has different ways of classification and ordering. Let me explain here by the system of rGyal rtse Them spangs ma because I have gotten lungs through the line of the Them spang ma.)

The similarity between the ordering system of the Ulan Bator MS, Kangyur and the Tokyo Kawaguchi Ms was already reported in 1982 by Bethlenfalvy. And in the introduction for the catalogue of the Stog Palace Kanjur, Skorupski (1985) added the report that Stog Palace MS as well as Tokyo Kawaguchi MS and Ulan Bator MS contain some twelve works which in the Derge or Peking editions are placed in the Tengyur. Also they contain some fifteen works which do not seem to be listed in the Derge or Peking Kangyurs. But this does not prove the similarity in the level of words or variants of each text, since no scholar could access individual texts until recently.

Tokyo Kawaguchi collection of the Manuscript Kangyur now kept in the Toyo Bunko was handwritten in 19th century (between 1858 and 1878) and brought to Japan from the dPal 'khor chos sde at Gyantse by Kawaguchi Ekai. It is one of several manuscript Kangyurs which are known to be copies of the Them spangs ma recension, another being the manuscript kept in the State Library in Ulan Bator. A recent copy is the Tokyo Kawaguchi MS Kangyur, and an earlier copy is the Ulan Bator MS Kangyur.

It was December in 1931 Kawaguchi published the critical edition of the Tibetan Larger Sukhavativyuhasutra. Kawaguchi used the Gyantse MS when he edited it. After while (about a decade), the set of the so-called Gyantse Manuscript Kangyur which was used for his editorial works was transferred to Toyobunko Library, Tokyo. That is the Tokyo MS or Kawaguchi MS as we call it.

"The contract to aid the editing fee of the Tibetan-Japanese Dictionary" was agreed on in May, 1940 between Kawaguchi and the Toyobunko. We can guess that the transfer was a reward of that aid. But, the end of volume Ka in the dKon brTsegs section of this MS Kangyur, that is, the text of Larger Sukhavativyuhasutra was actually lacking. Toyobunko doesn't own it at present, either. When Prof. Kojyun Saito had checked it in 1977 also, he reported it as "missing". Kawaguchi published the critical edition of the Smaller Sukhavativyuhasutra also at about the same time, but he did not use the Gyantse MS in his editing work, and there remained the MS of the Smaller Sukhavativyuhasutra in the collection.

Kawaguchi brought this MS when he returned from Tibet after his second Journey to Tibet. During the Journey he asked to 13th Dalai Lama to give a MS Kangyur set, and H.H. accepted his request. He received the MS which had been kept in Gyantse dPal skor chos sde monastery. That was in January of 1915.

Prof. Kojyun Saito carefully read the documents and prayers which are included in each section of the MS Kangyur and concluded that the set was a compilation of the manuscripts which had been prepared for a long time. Some inscriptions can be dated 1861 or 1878. Prof. Saito also writes that there is a possibility that some of the manuscripts had connections to the old Narthang Kangyur, but it wasn't proved there. He wrote it in 1977 when we didn't have got enough data or materials yet.

Being led by Bethlenfalvy's and Skorupski's description above, my first interest in the Ulan Bator MS is like this: "The Tokyo MS and the Ulan Bator MS belong together stemming from one version. It might be possible that the Ulan Bator MS can cover the missing of the Larger Sukhavativyuha text in the Tokyo Kawaguchi MS set". I felt such expectations. In other words, I had simply believed that a lot of the same variants can be found in both the Ulan Bator and Tokyo MSs. The Tokyo MS of the Larger Sukhavativyuha Text itself is unfortunatly lost, but we have Kawaguchi's description in the footnotes of his edition. So, this time, I first tried to check it the variants, which can be found in Kawaguchi's notes appear similar in the Ulan Bator MS's.

In the opposite direction to which I expected, there are many places where the Tokyo MS doesn't correspond to the Ulan Bator

TYPE I (the reading of Ulan Bator MS does not share with other the three of accessible Them spang ma recensions.)

a) U 319a5

tshe dang ldan pa ba glang bdag U (N)

tshe dang ldan pa ba lang bdag T / L

tshe dang ldan pa glang bdag S

b) U 320a6

rgyal bar gnas pa dang/ U (N)

rgyal po'i gnas pa dang/ T

rgyal ba'i gnas pa dang/ L / S

c) U 320b4

glang po chen po'i gnas pas... U (N)

glang po che chen po'i gnas pas T / L / S

d) U 323b3

skyes bu 'dul ba'i kha lo bsgyur ba/ U (N)

skyes bu 'dul ba'i kha lo sgyur ba/ T / L / S

e) U 324b1

bde la dga' bar gyur// U (N)

bde la dga' bar 'gyur// T / L / S

f) U 323b8

nyi zla nor bu sna tshogs kyi 'od rnams ni// U 

nyi zla nor bu tshogs kyi 'od rnams ni// T / L / S

nyi zla nor bu sna tshogs 'od rnams ni// (N)

g) U 322b5

mngon par dges pa zhes bya ba dang/ U (N)

mngon par dgyes pa zhes bya ba dang/ T / L / S

h) U 319b6

bcom ldan 'das ga la ba de logs su U / S (N)

bcom ldan 'das ga la ba der logs su T / L

For example, (example a) Ulan Bator MS, 319a 5-6 introduces a name of Ayusmat "tshe dang ldan pa ba glang bdag ". Upon this name the reading of the Tokyo MS is "tshe dang ldan pa ba langs bdag" which shares the variant readings with London MS. The reading of Stog Palace MS seems to have some confusion: "tshe dang ldan pa glang bdag".

(example b) Ulan Bator MS, 320a6. The sentence of Ulan Bator MS goes "rgyal bar gnas pa dang/". Tokyo Kawaguchi MS reads "rgyal po'i gnas pa" against "rgyal ba'i gnas pa dang/" of London MS and Stog Palace MS.

(example c) Ulan Bator MS, 320b4. The sentence goes "glang po chen po'i gnas pas...". Tokyo Kawaguchi MS maintains the reading of "glang po che chen po'i gnas pas", which shares with readings of Stog Palace MS and London MS.

(example d) Ulan Bator MS, 323b3. The sentence goes "skyes bu 'dul ba'i kha lo bsgyur ba/". The readings of Tokyo Kawaguchi MS, London MS and Stog Palace MS are "skyes bu 'dul ba'i kha lo sgyur ba/".

(example e) In Ulan Bator MS, 324b1 there is a sentence "bde la dga' bar gyur//". The readings of Tokyo Kawaguchi MS, London MS and Stog Palace MS are "bde la dga' bar 'gyur//".

(example f) In Ulan Bator MS, 323b8 there is a sentence "nyi zla nor bu sna tshogs kyi 'od rnams...". The readings of Tokyo MS, London MS and Stog Palace MS are "nyi zla nor bu tshogs kyi 'od rnams...".

(example g) In Ulan Bator MS, 322b5 there is a sentence "mngon par dges pa zhes bya ba dang/". The readings of Tokyo MS, London MS and Stog Palace MS are "mngon par dgyes pa zhes bya ba dang/".

(example h) In Ulan Bator MS, 319b6 there is a sentence "bcom ldan 'das ga la ba de logs su". This reading of Ulan Bator MS shares with Stog Palace MS. The readings of Tokyo MS, and London MS are "bcom ldan 'das ga la ba der logs su".

TYPE II (the reading of Ulan Bator MS does share with London MS, but not Tokyo Kawaguchi MS)

i) U 320b8-321a1

de bzhin gshegs pa'i ye shes... U / L (N)

de bzhin gshegs pa ye shes T / S

j) U 321b7

'od dri ma med pa zhes bya... U / L (N)

'od dri ma med ces bya... T / S

(example i) Ulan Bator MS, 320b8 to 321a1 includes the sentence "mthong ba sdud kyang de bzhin gshegs pa'i ye shes...". The readings of Stog Palace MS and Tokyo Kawaguchi MS are "...de bzhin gshegs pa ye shes ". The London MS agrees with Ulan Bator MS's reading.

(example j) Ulan Bator MS, 321b7 contains the sentence as "de bzhin gshegs pa 'od dri ma med pa zhes bya ba dang/". For this name of tathagata "dri ma med pa zhes bya" is in Tokyo MS spelled as "dri ma med ces bya" which is shared with Stog Palace MS. The reading in Ulan Bator MS is shared with London MS.

TYPE III (the reading of Ulan Bator MS does share with Tokyo MS)

k) U 322a2

de bzhin gshegs pa bai Du r-ya U / T (N)

de bzhin gshegs pa bai Du r-yar L / S

l) U 321a2

de bzhin gshegs pa bzhed na/ bsod snyoms... U / T (N)

de bzhin gshegs pa bzhed na bsod snyoms... L / S

m) U 318b8

nyon mongs pa med pa gnas pa dang ldang pa/ U / T (N)

nyon mongs pa med pa/ gnas pa dang ldang pa/ L / S

n) U 320a4 etc.

lpags pa U /T/ L (N)

pags pa S

o) U 320b1-2

rnyad pa yang ma mchis/ U /T/ L (N)

rnyad pa'ang ma mchis/ S

p) U 327b3

grangs rnyed pa gyur na yang U / T (N)

grangs rnyed par gyur na yang/ L

grangs rnyed par gyur na'ang S

q) U 319a2

cang shes pa glang po chen po U / T (N)

cang shes pa glang po chen po// L

cang shes pa/ glang glang chen po/ S

From the above examples, we see that there are many places where the Tokyo MS doesn't correspond to the Ulan Bator MS. But we also find cases where the Tokyo MS coreesponds to the Ulan Bator MS.

(example k) In Ulan Bator MS, 322a2, there introduces a tathagata's name "de bzhin gshegs pa bai Du r-ya snang ba". This "bai du r-ya" is spelled as "bai Du r-yar" in London MS and Stog Palace MS. The reading of Tokyo Kawaguchi MS shares with Ulan Bator MS. Using the spell as r-ya (not rya of ya-rtags) is maybe the common characteristic of Them spangs ma recensions.

(example l) Ulan Bator MS, 319a2. The sentence goes "de bzhin gshegs pa bzhed na/ bsod snyoms...", which has shad after bshed na. This reading of Ulan Bator MS agree with Tokyo Kawaguchi MS, but the Stog Palace MS and most of other editions has no shad.

(example m) In Ulan Bator MS, 318b8 we see the sentence "nyon mongs pa med pa gnas pa dang ldang pa/". But most of other editions have shad as "nyon mongs pa med pa/ gnas pa dang ldang pa/" which means that two names are mentioned here. Ulan Bator MS considers it as one long name.

(example n) As for the spelling of the tibetan word for "skin" both of pags pa and lpags pa can be found in dictionaries. For example, the Ulan Bator MS, 320a3 and 321a3 use the spelling of "lpags pa" whereas many xylograph Kangyur editions and the Stog Palace MS use the spelling as "pags pa". Tokyo Kawaguchi MS and London MS share the reading of the Ulan Bator MS.

(example o) In Ulan Bator MS, 320b1-2, there is a sentence as "rnyed pa yang ma mchis/". In this case only Stog Palace MS appears "rnyed pa'ang ma mchis/". All of the other editions use "yang". Stog Palace MS usually uses "na'ang" instead of "na yang". But when we carefully compared them paying attention to tseg and shad, an interesting point does appear.

(example p) Ulan Bator MS, 327b3 includes the sentence as "grangs rnyed pa gyur na yang". The reading of Stog Palace MS is "par gyur na'ang". The reading of London MS is "par gyur na yang/"(with shad). Ulan Bator MS uses "na yang" but without shad.

(example q) Ulan Bator MS, 319a2 introduces one of the excellent characteristics of bhikkhu as "cang shes pa glang po chen po/" This reading shares the variant with Tokyo Kawaguchi MS. But the Stog Palace MS has shad after shes pa as two characteristics:"cang shes pa/ glang po chen po/". The reading of London MS is "cang shes pa glang po chen po//"(with gnyis shad).

In every cases of which I showed above, i.e. examples (a) to (q), The readings of the Ulan Bator MS agree with the variant readings of the Narthang edition(N), except the example (f). The reading of the Ulan Bator MS in the example (f) is obviously not recensional, but transmissional variants. This is simply an error. This sentence stands as a stanza of verse, so here the sentence must have 9 syllables. The reading of the Narthang edition shows correctly 9 syllables. We can guess a recension of the Ulan Bator MS was same as the reading of the Narthang edition.

The upshot of the foregoing preliminary examination on the manuscript of the Lager sukhavativyuhasutra which is included in the Ulan Bator manuscript Kangyur set, is the following fact : the Ulan Bator MS and Tokyo Kawaguchi MS are not simply brother recensions. And the Ulan Bator MS has some characteristics which are closest to the Nartang edition of the Kangyur published in 1730-1732.

The Ulan Bator manuscript Kangyur itself is a copy from the Them spangs ma side manuscript in 1671. In other words, the Ulan Bator manuscript Kangyur might have the oldest contents in those sets of so called copies from Them spangs ma manuscript. And if we believe that the original of this Ulan Bator manuscript Kangyur was a the earliest copy of the Them spang ma manuscript, we may conclude that the contents which are shown in this Ulan Bator manuscript are the closest to old Narthang Kangyur.

As for the formation process of Narthang zylograph edition, it is not clear up to now. We do not have any concrete opinions. The hypothesis that "the original text of the Narthang edition was the old Narthang Kangyur" also exists. If we think so, some approximations with the Ulan Bator manuscript and Narthang edition is easy to explain.

This hypothesis does not corespond with the opinion of Dr. Skilling who checked a few sutras and says that the origin of the Narthang edition must be a text close to the Chimpataktse manuscript of the branch of the Tshal-pa recension. We have to consider more on this point.