Mahāratnakū A' Scriptures in Khotan. A Quotation From The Samantamukhaparivarta in The Book of Zambasta
Mahāratnakū A' Scriptures in Khotan. A Quotation From The Samantamukhaparivarta in The Book of Zambasta
Mahāratnakū A' Scriptures in Khotan. A Quotation From The Samantamukhaparivarta in The Book of Zambasta
ISSN 1343-8980
25
Annual Report
of
The International Research Institute for Advanced Buddhology
at Soka University
Volume XVII
2014
The International Research Institute for Advanced Buddhology
Soka University
Tokyo2014
PDF Version: ARIRIAB XVII (2014), 337347
2517
CONTENTS
IRIAB BULLETIN:
IRIAB Activities 519
List of Publications of the IRIAB Fellows 522
Books Received 523
Journals Received 526
EDITORIALS:
Contributors to this Issue / Editorial Postscript 529
Gilgit Manuscripts in the National Archives of India Facsimile Edition, Vol. I
PLATES:
1 Harry FALK: The rst-century Copper-plates of Helagupta from Gandhra hailing Maitreya PLATES 16
2 GUAN Di: Three Sanskrit Fragments Preserved in Arthur M. Sackler Museum of Peking University PLATES 78
3 Seishi KARASHIMA: New Research on the Buddhist Sanskrit Manuscripts from Central Asia PLATES 910
4 Akira YUYAMA: Reviewing Rgs XIV 2 & 7: How One Can Rescue Oneself ... PLATES 1112
5 Peter ZIEME: Collecting of the Buddhist Scriptures: Notes on Old Uigur annals PLATES 13-14
6 Tatsushi TAMAI: Tocharian Syllabary with Uigur Explanations: M34.4 of the Mannerheim CollectionPLATE 15
7 Isao KURITA: Gandhran Art (Part 2) PLATES 16-32
Mahratnaka Scriptures in Khotan:
A quotation from the Samantamukhaparivarta
in the Book of Zambasta*
In a previous volume of the Annual Report of the International Research Institute for
Advanced Buddhology at Soka University (2011), I studied the transmission of the Ratnaka/
Kyapaparivarta in Khotan, on the basis of fragments of the Khotanese translation of this
stra and of sourced and unsourced citations of the same identied in the mid-fth century
Khotanese composition known as the Book of Zambasta.1 This time I once again follow the
thread of Mahratnaka scriptures in Khotan, with a note on a quotation attributed to the
Samantamukhaparivarta in the fourth chapter of the same Book of Zambasta.
The Samantamukhaparivarta is no longer extant in its Indian original, but is preserved in
Chinese and Tibetan in the Mahratnaka canonical collections (Da baoji jing , T
310; Dam chos dkon mchog brtsegs pa, abbreviated dKon brtsegs, Derge/Thoku 87 etc.), as
well as in an earlier, individually transmitted Chinese translation (T 315).
I start with a brief overview of Mahratnaka scriptures featured in the Book of
Zambasta (I). I then discuss the Samantamukhaparivarta citation in more detail (II), and,
nally, I make a few observations on the signicance of these scriptural threads for the
reconstruction of the textual history of early Khotanese Buddhism (III).
* In part this article is a revised version of a chapter of my doctoral dissertation (Martini 2010). I thank Bhikkhu
Anlayo, Mauro Maggi, Peter Skilling and Vincent Tournier for their suggestions. In quoting the text from Emmericks
edition of the Book of Zambasta (Emmerick 1968) I have simplied and omitted some information such as stanza numbers
given in the manuscript. For all text editions, on occurrence, I have adjusted the sandhi, punctuation, capitalisations etc.
1.
Martini 2011a. A new folio of the Khotanese translation of the Ratnaka (= Kyapaparivarta) is identied and
studied in Maggi forthcoming b.
surprising, given that the extant manuscripts of the Book of Zambasta do not include colo-
phons containing titles or summaries of contents for chapter two (or for any other chapters,
with the exception of chapter nineteen) whether at the beginning and end of the chapters or of
the work itself. As a result, one cannot be certain that the Bhadramykravykaraa was
known to the Khotanese by the Sanskrit title that we know today or by a Khotanese
equivalent, although this may well have been the case. Chapter two, rst studied in relation to
the Tibetan version by K. Rgamey (1938),2 is a rather free adaptation of the Bhadramy-
kravykaraa as we know it from the Chinese and Tibetan translations belonging to the
Mahratnaka collections.3 The Khotanese version seems to be substantially a paraphrase
except for a few stanzas that correspond closely to the Tibetan version (and thereby to its
underlying Indian original). Rgameys pioneering work which has, inter alia, laid the
methodological foundation for subsequent philological studies of Mahyna stras in general
as well as for another Mahratnaka scripture in particular, the *Ugraparipcch (Nattier
2003)4 may, seventy-ve years after its publication, deserve reconsideration in the light of
the multiple model of oral, aural and written formation that in my reading best explains the
redaction, transmission, function, features and overall textual and literary patterns of the Book
of Zambasta. In other words, the adaptive and paraphrasing character of the Khotanese
witness vis--vis the text preserved in the parallel versions of the Bhadramykravykaraa
may be better appreciated when we take into account such a complex model of transmission.
Out of twenty-three extant chapters of the Book of Zambasta, chapter two is one of only ve
chapters for which complete sources or parallels have been identied.5 Thus, in the light of
ndings made since the publication of Rgameys work, a new study would be especially
helpful for the evaluation of the mode or modes of incorporation of textual sources into the
Book of Zambasta.
The other Mahratnaka works attested in the Book of Zambasta (nos. 2 to 5 as listed
above) are all mentioned by their own titles, including the Samantamukhaparivarta, which I
will discuss in the next section:
Samantamukhaparivarta: samntamukha-parivartto, Z 4.39;
Ratnaka (= Kyapaparivarta): ratnaklna vt, Z 8.39; ratnaklna, Z 13.42;6
Vinayavinicaya-Upliparipcch: vinayavinicayo sttro, Z 13.33;7
2.
Cf. also Leumann 19331936: 361367 and Emmerick 1968: 1011.
3.
In addition to T 310.21, translated by Bodhiruci, there is an early individual translation, T 324,
attributed to Dharmaraka ( ) in the Taish edition. A translation of T 310.21 is available in Chang
1983: 326; see Rgamey 1938/1990 for the edition, study and translation of the Tibetan version
(corresponding to Derge/Thoku 65 etc.).
4.
Cf. the remarks in Nattier 2003: 203 with note 4.
5.
The chapters with identied complete parallels are chapters two, three, sixteen, twenty-three and
twenty-four. On there being twenty-three instead of twenty-four extant chapters see Maggi and Martini
2014.
6.
For a detailed study of the sourced (Z 8.38-39, Z 7.29, Z 9.19 and 13.42) and unsourced citations (Z
4.95, Z 8.44 and Z 9.16) of the Ratnaka (= Kyapaparivarta) in the Book of Zambasta, see Martini 2011a
(and cf. also Martini 2008), where I also discuss at length the fact that all extant evidence, in Sanskrit,
Khotanese and other languages, points to Ratnaka rather than Kyapaparivarta as the original title of the
text transmitted throughout India and Central Asia.
7.
On the Vinayavinicaya-Upliparipcch quotation as cited and re-read in the Book of Zambasta see
Martini 2013.
338
PDF Version: ARIRIAB XVII (2014), 337347
8.
The citation of the Daadharmaka-stra (T 310.9, translated by Buddhanta in A.D. 539 according to
T 2154 at T LV 542a26; Derge/Thoku 53 etc.) awaits closer inspection. To complete the overview of
Mahratnaka scriptures in the Book of Zambasta, it may be noted that the meeting of the Buddhas
father with his son, which constitutes the topic of chapter ve of the Book of Zambasta, is the theme of
another Mahratnaka work, the *Pitputrasamgama-stra, a text that contains an important formula-
tion of the doctrine of the two truths, and is quoted in Kamalalas Bhvanakrma I, ntidevas iksamu-
ccaya, Prajkaramatis Bodhicaryvatra-pajik, and Candrakrtis Madhyamakvatra (see Lamotte
2000: 46, Yuyama 2001: lix and Yuyama 2002: 199200). From a cursory reading of the extensive Yab
dang sras mjal ba zhes bya ba theg pa chen poi mdo (Derge/Thoku 60 etc.), its ninth-century Tibetan
translation, I have not been able to nd any direct parallel to the Khotanese text. The text is also available
in Chinese translation, the Pusa jian shi hui ( , T 310.16), translated by Narendrayaas in A.D.
568 (no translations in Western languages are presently available). No Sanskrit manuscript containing a
text corresponding to the *Pitputrasamgama-stra of the Mahratnaka has come to light so far,
although the two fragments from the Northern Silk Road (Kucha) in the Fond Pelliot containing a passage
which is otherwise known from a quotation in the iksamuccaya could in fact belong to a manuscript of
this text (see Pauly 1996: esp. 290291 and Yuyama 2002: 200). At best, the Khotanese could be an
extremely abridged and summarised (thereby simply unrecognisable) rendering of such a long and
articulated text as the *Pitputrasamgama-stra.
9.
On the translation cannot all enter it (ni bi h tranda hmre) with present potential construction
for have not all been able to enter it in Emmerick 1968: 83 see Maggi forthcoming a ( 6).
10.
Since in Old Khotanese the phonological distinction -u-/-- is accurately represented, the form
Praku can be taken to reect a variant Pr rather than Pru, cf. Edgerton 1953: 343, s.v.
11.
On bna music instead of lotuses in Emmerick 1968: 83 see Degener 1989: xxxiii.
339
PDF Version: ARIRIAB XVII (2014), 337347
[specic] mind, just as in a dream one sees what does not [really] exist. Therefore there
is no creator of them. [4.38] There are seven divine classes of gods. They are such as is
imagined. Unhindered they go through walls. He [i.e., a god] has no wound which he
laments, [4.39] because gods [exist] only by virtue of perception. They are ones own
imaginations, not [made up of] material substance. In the Samantamukhaparivarta the
Buddha has shown the meaning of this in detail.12
Among earlier scholars who have dealt with this passage, S. Konow (1939: 26) was the rst
to point to the Samantamukhaparivarta of the Mahratnaka, although without locating
the passage in the text. E. Leumann (19331936: 57), followed by R.E. Emmerick (1968:
83), remarked that the Samantamukhaparivarta constituting chapter twenty-four of the
Saddharmapuarka-stra does not contain any content relevant to the passage in chapter
four of the Book of Zambasta in which the title Samantamukhaparivarta is quoted. I was then
able to locate a parallel to the Khotanese excerpt in the Chinese and Tibetan translations of
the Samantamukhaparivarta of the Mahratnaka.13 A parallel to the Khotanese passage
occurs in the context of a description of the samdhi of divine appearance(s) ( ) or
divine form(s) samdhi (lhai gzugs ting nge dzin).14
With this identication, the Khotanese citation now provides evidence for:
(a) the Indian form of the title of this work which to the best of my knowledge is
otherwise neither attested nor cited in Sanskrit texts; and
(b) the circulation of an Indian (most likely Sanskrit) recension of the Samantamukha-
parivarta in Southern Central Asia from at least the mid-fth century, as an individual
scripture not yet transmitted as part of a larger collection.
As mentioned earlier, two Chinese versions of the Samantamukhaparivarta are extant: an
early translation, the Pumen pin jing (, T 315), carried out by Dharmaraka ()
in A.D. 287, and the Wenshushili pu men hui ( , T 310.10), translated by
Bodhiruci (d. A.D. 527) and contained in the Mahratnaka/Da baoji jing
collection. A detailed comparison of the passages as found in the two Chinese translations
and in relation to the Tibetan version falls outside my present concern; sufce it to note that
Dharmarakas translation is rather idiosyncratic if the original on which it depends was close
to the textual basis of the later Chinese and Tibetan versions.15 In order to provide the parallel
to the Khotanese citation I quote in full the text of the versions included in the Mah-
ratnaka collections. The Chinese, in verse, reads:
, ? :
12.
Ed. Emmerick 1968: 82; translated after ibid.: 83 with modications.
13.
Martini 2010: 150155.
14.
The expression *deva-rpa-samdhi or other equivalents to it appear to be unattested as the name of a
samdhi in Sanskrit sources.
15.
T 315 at T XI 772a25b4: : ? , , ,
, , , , , , , ,
, , , , , .
, , , . , . In
Dharmarakas translation, the name of the samdhi in question mentions bodhisattvas journeying through all
the heavens, .
340
PDF Version: ARIRIAB XVII (2014), 337347
.
[The Buddha said:] Then, Majur, what is the so-called samdhi of divine
appearances? Then, he explained in verse:
Because of a mind with serene and pure faith
and by way of manyfold wholesome deeds,
one will receive the excellent reward of [birth as] a god:
[that is,] a body of wondrous beauty.
Treasures and various palaces
will be completed without having been constructed.
The coral tree has wonderful blossoms,
yet no one has planted it.
Like this, without being preconceived,
it all arises through the power of karma,
[which] is able to manifest myriad forms,
just as [images appear] in a clear piece of beryl.
In this way, the awesome bodies
as well as all the varieties of palaces,
all arise from what is unreal.
This is called the samdhi of divine [appearances].16
jam dpal, de la ji ltar na byang chub sems dpas lhai gzugs ting nge dzin du rtogs par
bya zhe17 na rnam rtog dge las byung ba yi las dgei bras bu lha rnams te rab tu dang
bai yid kyis na18 mdzes pai lha rnams mngon par grub. gzhal med khang pa nyams dga
ba, de dag gzhan sus byas pa min. me tog man ta19 ra ba yang de dag gzhan sus bskyed
pa min. sgyu mai las las kun byung bas byung ba rnams kyang bsam du med, bzang
bor20 snang21 bai rya log pas grub pa rnams kyang brdzun. yang dag min pas kun
bskyed pas lha rnams brdzun par shes pas na ting nge22 dzin zhi ba brjod pa ni lhai23
gzugs kyis bstan pa yin. jam dpal, de ltar na lhai gzugs ting nge dzin du rtogs par
byao.
Therefore, Majur, if it is asked how a bodhisattva should understand the divine form
samdhi, [the answer is:] the gods are the fruit of good karma [that is] born of good
16.
Wenshushili pu men hui (), T 310.11 at T XI 159c27160a7; translated after Chang 1983:
138 with modications.
17.
Derge/Thoku 54 reads:zhi.
18.
Derge/Thoku 54 reads: ni.
19.
Derge/Thoku 54 reads:d.
20.
Derge/Thoku 54reads: por.
21.
Derge/Thoku 54adds:bai.
22.
Derge/Thoku 54omits:nge.
23.
Derge/Thoku 54reads:lhayi.
341
PDF Version: ARIRIAB XVII (2014), 337347
thoughts. Handsome gods are created through a mind endowed with serene faith.
Charming aerial pavilions [will manifest]: these have not been produced by anybody
else. Flowers of the coral trees: these have not been produced by anybody else.
Inconceivable objects produced by magic, [such as] beryl of the nest water, are
fabricated and false. Being non-existent (yang dag min pas, Skt. abhta) and entirely
[mentally] fabricated, arisen from falsity, the gods are illusory: the detailed explanation
of the samdhi [practised] through the divine form is expounded. In this way, Majur,
the divine form samdhi should be understood.24
Although the philosophical theme of the passage in the Book of Zambasta relates it beyond
doubt to the passage(s) of the Samantamukhaparivarta in Chinese and Tibetan translated
above, on reading the preceding and subsequent Khotanese stanzas against the parallels, the
excerpt in the Book of Zambasta gives the impression of a loose and at the same time
expanded rendering. Moreover, it is not immediately self-evident where the actual scriptural
quotation begins and ends. Such a loose and non-verbatim correspondence between the
Khotanese stanzas and the verse and prose counterparts in the Chinese and Tibetan versions
is partly explained by the seemingly aural formation and composite textuality of the Book of
Zambasta. However, in the absence of the Sanskrit source(s) which was or were actually used
or consulted at the time and place where the passage was translated, it is not fruitful to
attempt an evaluation of the faithfulness of the Khotanese rendering by comparing it to
parallels translated centuries later in entirely different cultural and intellectual milieux,
almost certainly on the basis of different Indian recensions of the text.
The Samantamukhaparivarta can be classied, in terms of genre, as a meditation text, and
its citation in the Khotanese Book of Zambasta lls in a small blank spot in the obscure
puzzle that is the Central Asian formation and transmission of a number of visualisation and
meditation scriptures.25 It is structured around the description of various kinds of samdhi
that take up different meditative objects and domains of existence. The meditative
instructions are based on the use of visualisations that reveal the ultimately illusory nature of
all mind-fabricated phenomena.
In my initial study of the Samantamukhaparivarta citation, I found it difcult to detect the
actual beginning of the excerpt, and I was not able to single out which, if any, of the adjacent
stanzas might have originated from a different source, perhaps creatively adapted to illustrate
the theme of the relationship between ones vipka (the karmic effect of previous intentional
action), saj (the perceptual act of conceptual identication and the deriving consolidated
cognition), and parikalpa (the process of proliferative and thus faulty imagination). These are
all doctrinal topics encountered in different places in the Book of Zambasta, and they are
certainly dear to a Yogcra-inspired meditative milieu close in philosophical avour and
24.
Phags pa kun nas sgoi leu shes bya ba theg pa chen poi mdo, Peking/tani 760.10, dKon brtsegs, dsi,
209b6210a2 [= Derge/Thoku 54, dKonbrtsegs, kha, 188a3188a5]; my translation.
25.
Cf. esp. Yamabe 1999, Martini 2011b with references, and now Greene 2012. Most notably, the so-
called Majurnairtmyvatra-stra (ms. P 4009), a Late Khotanese original composition based on Old
Khotanese sources preserved in a tenth-century manuscript from Dunhuang, comes very close to the genre
of visualisation texts, cf. Martini 2011b: 127, note 11.
342
PDF Version: ARIRIAB XVII (2014), 337347
time to the Book of Zambasta.26 Immediately after the description of the heavenly landscape,
the statement that such a beautiful realm is visible in dependence on ones perception, and the
reference to the Samantamukhaparivarta, other examples are given to illustrate the theme of
the illusory nature of perception maintained to be nothing more than a relationship between
cause and effect:
[4.40]
catmahrya-bma gyasta-bhma yaka-vimna
nit ggar sttro tta hvnde ku va yaka-bhavana ne ro
[4.41]
ttnu parikalpe vpk ttna ju m kro ni tsndi
kho ju ha saitto ditte kho ye cya-nrmtu daiy
[4.40]
There are dwellings of the Four Great Kings, dwellings of the gods, palaces of the
yakas. There is no mountain so it is stated in the stra so that there are no palaces of
the yakas. [4.41] Their imaginations are [their own] karmic result. Therefore they have no
effect on us. They are just as what seems [to exist] or appears in a dream, [or] as one sees
what has been conjured up by magic.27
Now, although this doctrinal elaboration is compatible with the visionary and
philosophical content of several passages in the Samantumkhaparivarta as well as with that
of, broadly speaking, other Yogcra-inspired meditation texts, I was not able to pin it down
to a specic passage in the Samantamukhaparivarta nor to any other possible source or
parallel.
Fortunately, among the annotations made by the late R.E. Emmerick on his personal copy
of the Book of Zambasta recently edited by M. Maggi, a reference indicates that the example
provided in Z 4.4041 does not come from the Samantumukhaparivarta but from a different
source. We learn from Emmericks note that stanzas Z 4.4041 were identied by G.
Schopen as a loose and in itself cryptic but expanded reference to a passage in the Su-
khvatvyha.28 In particular, as remarked by Maggi, the identication of the source of the
passage makes clear that the unnamed ggar mountain in Z 4.40 is Mount Sumeru and that
sttro in the stra refers to the Sukhvatvyha and not to the Samantamukhaparivarta
quoted in Z 4.39. The passage in the Sukhvatvyha takes up the mechanisms of karma and
rebirth in the Land of Bliss (Sukhvat) where, contrary to normal world systems, the
cosmographic model lacks a Mount Sumeru on which gods of the lower realms live.29 The
26.
Martini 2010: 150155. The status and validity of perception is taken up in several other places of the
Book of Zambasta; especially relevant in this respect is the text of the Dharma sermon given by the Buddha
to his father at the time of his rst return to Kapilavastu, found in chapter ve (cf. Martini 2014). The
cross-referencing between dierent chapters and the by and large homogeneity of the philosophical
discourse of the Book of Zambasta seems to me to be a noteworthy element in support of the hypothesis of
the unitary redaction of this work, that has been discussed by Maggi 2004 and Martini 2014.
27.
Ed. Emmerick 1968: 82; translated after ibid.: 83 with modications.
28.
Schopens letter to Emmerick dated 26 February 1980, in Maggi forthcoming a (note 7). One
fragmentary but all important Central Asian folio of an Old Khotanese translation of the Sukhvatvyha
survives in the British Library collection (ms. Kha 0013c2), preserving, in translation, a Sanskrit original
somewhat dierent from the Nepalese recension, see Maggi 2009: 381382.
29.
Ed. Fujita 2011: 39715 (= ed. Ashikaga 1965: 3319342 and ed. Kagawa 1984: 20219): evam ukta
yumn nando bhagavantam etad avocat: ye punas te bhagava cturmahrjakyik dev sumeru-
prvanivsinas tryastri v sumerumrdhni nivsinas, te kutra pratihit. bhagavn ha: tat ki
manyase, nanda, ye ta iha sumero parvatarjasyopari ... dev, kutra te pratihit iti. ha: acintyo bhaga-
343
PDF Version: ARIRIAB XVII (2014), 337347
yakas are absent in the Sukhvatvyha, but their imagery ts well with the theme of
apparitional visions to which all dharmas are classically compared.30 Thus, most of stanza Z
4.41 is an amplication required to t the Sukhvatvyha example into the context.
Interestingly, the real cosmography of Sukhvat that is, with the real non-existence,
there, of the very mountain that otherwise serves as the dwelling place of the lower gods
provides a good example of the non-reality of mental and meditative experience according
to the meditation system of the Samantamukhaparivarta. This nding closes the circle while
throwing once again into relief the composite and compilative genesis explanatory,
didactic, locally adaptive and inclusive by nature of the textual material of the Book of
Zambasta, as exemplied by these passages in chapter four.
of the hypothesis of the Central Asian formation of a number of scriptural assemblages, in-
cluding the Mahratnaka itself. Scriptural collections inherently selective and therefore
able to provide foundational frameworks of reference for local communities, functioning like
syllabi and compendia may well have had a major part to play in the transmission of the
Dharma across Greater (Buddhist) Serindia, in times of both scriptural abundance and scrip-
tural scarcity.33
3. In order to better contextualise the Khotanese materials and the signicance of the type
of philological work they require, it may be useful to take into account the general situation
of textuality of fth/sixth-century Khotanese Buddhism.
3.1 In spite of royal patronage, early Khotanese Buddhism appears to have never under-
gone a process of canonisation in the sense of a programmed, centrally sponsored and for-
malised cultural operation of scriptural selection, classication and translation. The Buddhist
communities of Khotan certainly possessed their own implicit and explicit awareness of the
notions of scriptural canonicity, authenticity and normativity. One of the ways such notions
are actualised is the selection of a number of texts that are ideologically and practically
adopted as the frame of reference of religious learning and practice. This in itself renders
especially important the study of the Indian sources that were received and chosen to serve
the purpose of transmission of the Dharma and of the consequent building of religious
identity.
3.2 Further, the limited and fragmentary nature of the surviving Khotanese Buddhist cor-
pus makes even ostensibly minor pieces of evidence historically valuable. Thus, one of the
possible ways of tracing back the canonical scriptures of the Buddhism of Khotan by
reconstructing a theoretical canon of significant and authoritative scriptures is to pay close
attention to its Indian sources, for instance by way of sourced and unsourced citations of
Mahyna works included in Khotanese Buddhist works.
3.3 Even when scriptural quotations can be traced back to their Indian originals (most
often via their Chinese and Tibetan translations), it is in most cases hardly feasible to position
these quotations accurately in terms of the Indian textual recension they witness and on
which they depend. This is not only explained by the common problems posed by the
philology of Mahyna scriptures (lack or fragmentary condition of the Indian originals,
multiple versions circulating in India, diverse recensions of the same text witnessed by dif-
ferent Sanskrit manuscripts as well as the Central Asian, Chinese and Tibetan translations
etc.). Other complicating factors include the notable localising and adaptive tendencies and
literary tastes of the Khotanese transmitters, and the stylistic gaps between originals that were
34.
Cf. the case study of the Vinayavinicaya-Upliparipcch as foundational to the Mahyna identity of
early Khotanese Buddhism in Martini 2013.
346
PDF Version: ARIRIAB XVII (2014), 337347
347