Abstract
The Buddhist Sanskrit Saṃghāṭa-sūtra includes several longer or shorter passages in verse, mostly ślokas. Many though not all of these verse passages also appear in metrical form in the Khotanese version, which makes use of all three of the metres known from the longest Old Khotanese poem, the Book of Zambasta. The aim of the present article is to analyse these metrical passages in order to determine to what extent the treatment of the metres conforms to the practice of the Book of Zambasta. The relevant passages are therefore presented with a detailed metrical analysis as well as an English translation and brief commentary.
Introduction
The Saṃghāṭa-sūtra (henceforth: “Sgh”), a popular Mahāyāna devotional text originally composed in Sanskrit, was translated into many of the languages used by Buddhists in East and Central Asia, including Chinese, Tibetan, Sogdian and Khotanese. The popularity of the Khotanese version is clear from the fact that it is attested by about 150 folios or fragments belonging to more than 25 different manuscripts. In 1993 Giotto Canevascini, building on earlier work by Ernst Leumann, Sten Konow, H.W. Bailey and Oskar von Hinüber, incorporated almost all of this material into his excellent edition The Khotanese Saṅghāṭasūtra (henceforth: “KS”).1 A few additional fragments have since been identified by Duan Qing, Mauro Maggi and Prods Oktor Skjærvø.2 A preliminary edition of the underlying Sanskrit text, supplied by von Hinüber, is included in Canevascini’s edition, but this is now superseded by von Hinüber’s editio maior of 2021.
While most of the Sanskrit Sgh is composed in prose, several sections are in verse, almost all ślokas. Many of these verse passages also appear in metrical form in the Khotanese version, where they tend to be much more freely translated than the surrounding prose—sometimes so freely that the correspondence between the two texts is hardly discernible.3 The verse passages are presented in all manuscripts as continuous text, making it difficult to distinguish verse from prose in broken passages.4 Some of the Sanskrit verses seem to be translated into Khotanese prose, though it cannot be excluded a priori that some of them may attest a metre or metrical system otherwise unknown.5 Those parts of the Khotanese translation which are clearly in verse make use of the same three metres which are known from the most substantial Old Khotanese poem which has come down to us, the Book of Zambasta (= Z). In a recent book (Sims-Williams 2022), I have given a new analysis of these three metres, conventionally referred to as A, B and C, as they are attested in Z. The existence of verses in the same metres in another source provides an opportunity both to test this analysis and to discover whether these metres were treated differently by different poets. In this connection it is worth noting that the oldest surviving fragments both of Z and of the Khotanese Sgh are written in what Lore Sander has defined as “Early Turkestan Brāhmī, type b”, which suggests that the composition of both works took place not later than the fifth century.6 Any differences between Sgh and Z in the treatment of the metre, therefore, are unlikely to be due to chronological factors.7
Verses in Metre C
Sgh 99 includes a series of verses in metre C, which I have re-edited and studied in a forthcoming article.8 Some verses, all of which are attested only in one manuscript (“MS 6” according to Canevascini’s numbering), appear to contain a few copying errors. Thus, in two pādas (3d, 7a) a 3-mora sequence is lacking and in another (5c) it appears necessary to delete a superfluous negative ne. Similar errors are not uncommon in Z. The only detail in this metre C passage which seems seriously to contravene any of the metrical “rules” deduced from Z is the occurrence of stōrĕ ˈHL “strong” at the end of the hemistich 8ab, where a word of the shape ˈLL is expected. Rather than being a mistake for sturĕ “heavy”, as I first suspected, this cadence may indicate that the first syllable of stora- < *stabra- “strong” can be read with either a long or a short vowel, as in the well-established case of jsei’ṇa-/jsäṇa- ˈHL/ˈLL < *ǰašna- “small, fine”, which occurs several times at the end of a metre C hemistich.9
Apart from the details mentioned in the last paragraph, the metre C verses in Sgh 99 do not contain anything which would be regarded as irregular if it were attested in Z. They do however attest at least one feature which can plausibly be attributed to the fact that the Sgh and Z are works by two different authors, namely, the remarkable frequency of 7-mora cadences ending with ˈLLL (⏑́⏑⏑) instead of the more usual ˈHL (–́ ⏑).10 The difference between the Sgh and Z in this respect is quite striking and clearly statistically significant. The Sgh passage also includes two rare 9-mora cadences, HLLˈLLLLL (or perhaps HLˈLLLLLL) in 9c and LHˈLLLHL in 10c, which happen not to be attested in metre C in Z. However, since they do occur there in metre A, and the two metres seem to employ essentially the same range of 7-, 9- and 10-mora cadences, their occurrence in metre C in Sgh cannot be regarded as having any particular significance.
Verses in Metres A and B
In the following pages I will discuss all the remaining passages of the Khotanese Sgh which correspond to metrical passages in the Sanskrit original, both those in metres A and B and those which appear to be non-metrical. The metrical passages will be edited, translated and provided with a metrical analysis. The notation of the texts is designed, like that of Z in Sims-Williams (2022), to show unambiguously both the readings of the manuscripts and the proposed metrical interpretation. Here however, I will retain the punctuation and verse numbers of the manuscripts, which are omitted from my text of Z. Other differences from the scanned text of Z in that book include a systematic marking of what I take to be the position of the ictus, both in the text (e.g. karma yäˈḍāṃdä) and in the interlinear analysis (e.g. HLLˈHL).
In my analysis of the metre of Z, I have attempted to show that the metrical system was indeed moraic (as Ernst Leumann and others have previously argued) but that it is also characterized by a compulsory ictus at a specific point in most or all of the cadences. In the case of metre A, each hemistich attests one of the following patterns of morae:
5 + 7 |
‖ 5 + 7 ‖ |
5 + 9 |
‖ 3 + 7 ‖ |
5 + 10 |
‖ 2 + 7 ‖ |
The equivalent patterns in metre B are as follows:
5 + 3 |
‖ 3 + 7 ‖ |
5 + 4 |
‖ 2 + 7 ‖ |
In both metres, the hemistich consists of two pādas, each beginning with a non-cadential segment and ending with a cadence. In principle,11 the first non-cadential segment consists of 5 morae, while the second cadence always contains 7 morae. The first cadence is of varying length: in metre A it contains (usually) 7, (rarely) 9 or (even more rarely) 10 morae, while in metre B it consists of 3 or 4 morae. In each case the length of the segment following the first caesura varies with that of the preceding cadence in such a way that the two together add up to 12 morae in metre A and to 6 morae in metre B.
The non-cadential segments are unstructured, containing any combination of light and heavy syllables which adds up to the required number of morae. Each of the cadences, apart from the shortest two, has a specific structure requiring at least a syllable-boundary and an ictus at a particular point. The cadences used in metres A and B are more precisely defined in table 1 above.
Table 1
The cadences of metres A and B
3-mora cadence (metre B) |
|
4-mora cadence (metre B) |
|
7-mora cadence (metres A and B) |
|
9-mora cadence, type 1 (metre A) |
|
10-mora cadence, type 1 (metre A) |
|
9-mora cadence, type 2 (metre A) |
|
10-mora cadence, type 2 (metre A) |
|
The above description of metres A and B is a summary of that in Sims-Williams (2022), which is exclusively based on an analysis of Z but which proves to apply equally well to the metrical passages in the Sgh.12 In particular it is worth noting that nothing in the Sgh contradicts my argument, which some may regard as controversial, that there is a caesura (or at the very least a fixed word-boundary) in the middle of the metre B hemistich.13 The treatment of the non-cadential segments shows the same “irregularities” as in Z: sometimes the segment has only 4 morae instead of the expected 5 (e.g. jaḍīna -LHL 9.3c, similarly 3.1a, 3.1d, 4.2b, 6.1c, 7.2a(?), 7.4a, 9.10a);14 sometimes it appears to have 6 morae, in which case it is likely that a heavy syllable must be lightened (e.g. ysamaśśaṃdiya LLLLL 8.1a, similarly 6.2a, 6.8c, 7.5c).15
Table 2
7-mora final cadences (metres A and B, pādas b and d) in Z and the Sgh16
Z |
Sgh |
||
---|---|---|---|
HLLˈHL |
2,635 (= 74 %) |
58 (= 69 %) |
|
HH|ˈHL |
392 (= 11 %) |
5 (= 6 %) |
|
LLLLˈHL |
359 (= 10 %) |
16 (= 19 %) |
|
LHLˈHL |
97 (= 3 %) |
3 (= 4 %) |
|
LLH|ˈHL |
29 (< 1 %) |
0 |
|
|
29 (< 1 %) |
2 (= 2 %) |
|
Total |
3,541 |
84 |
Some metrical features found in Z are absent from the surviving fragments of the Sgh,17 but this may be due to nothing more than the limited amount of material (about 100 fairly complete hemistichs in these two metres, as opposed to nearly 3,500 in Z).
While the definition of metres A and B based on Z does not require any significant modification in order to accommodate data from the Sgh, it seems that, as in the case of metre C, the two poets display slightly different preferences in their use of the metres.
In the first place, the frequency of the various forms of the 7-mora cadence which occur at the end of each hemistich in these two metres is notably different in the two texts. In particular, as can be seen from Table 2, the final cadence LLLLˈHL is found in the Sgh almost twice as frequently as in Z.
Secondly, and even more strikingly, the proportion of 9- and 10-mora cadences (as opposed to 7-mora cadences) in the first pāda of each metre A hemistich is almost twice as great in the Sgh as in Z. The contrast is even greater if one takes into consideration only the 9- and 10-mora cadences of type 1 (those with the ictus on the 5th mora), as can be seen from Table 3:
Table 3
Metre A first cadences (pādas a and c) in Z and Sgh
Z |
Sgh |
|||
---|---|---|---|---|
7-mora |
1,911 (= 75 %) |
19 (= 53 %) |
||
9-mora (type 1) |
374 (= 15 %) |
11 (= 31 %) |
||
10-mora (type 1) |
57 (= 2 %) |
4 (= 11 %) |
||
9- and 10-mora, type 1, total |
431 (= 17 %) |
15 (= 42 %) |
||
9-mora (type 2) |
186 (= 7 %) |
2 (= 6 %) |
||
10-mora (type 2) |
20 (< 1 %) |
0 |
||
9- and 10-mora, type 2, total |
206 (= 8 %) |
2 (= 6 %) |
||
Total |
2,548 |
36 |
Finally, though it is not strictly a metrical issue, it is worth noting an apparent difference between the preferences of the two poets in respect of the two alternative forms of the 1 sg. m. tr. perfect, one formed with īmä “I am”, e.g. yäˈḍaimä LˈHL “I did” < ˈyäḍĕ + ˈīmä, and one formed with the enclitic equivalent mä, e.g. ˈyäḍĕ mä ˈLLL “id.”.18 In Z the form in -aimä HL occurs more than 4 times as often as that in -ĕ mä LL, whereas in the Sgh the two forms seem to be equally frequent.19
As a matter of convenience I have numbered the passages discussed or edited below from 1 to 12 and refer to them by these numbers. Note that 5.1–40 is the passage in metre C edited in Sims-Williams (forthcoming a). My text, like Canevascini’s, is eclectic. Differences between the copies in grammatical forms or spelling are not uncommon, but in general are only recorded here if they have a bearing on the understanding of the text or its metrical analysis. The texts are printed in Roman type, uncertain readings being indicated by italics. Where text is quoted in italics in the discussion, the usage is reversed, Roman type being used if necessary for uncertain readings.
Photos of manuscripts with the prefix “SI P” can be found in Emmerick and Vorob’ëva-Desjatovskaja (1993) and of those with the prefix “IOL Khot” on the website of the International Dunhuang Project.20 For fragments in the Francke-Körber collection in Munich, of which no photos are available to me, I have relied on the readings of Skjærvø (2002).
1 Saṃghāṭa-sūtra 43
Diplomatic edition and translation in KS 20, based on three manuscripts:
-
MS 19, transcribed in KS 236–237 (IOL Khot 35/4, v2–5, cf. KT5: 175, #337; Skjærvø 2002: 252);
-
MS 22, transcribed in KS 239–240 (IOL Khot 5/1, r3–v5, cf. Leumann 1920, 13–15; Konow 1932: 70; KT5: 388, #24; Skjærvø 2002: 171);
-
MS 24, transcribed in KS 244–245 (IOL Khot 32/2, v1–6, cf. KT5: 126, #227; Skjærvø 2002: 245).
The equivalent Sanskrit text is edited in von Hinüber (2021): 17 (verses 1–7), with English summary of the contents on pp. lx–lxi.
Metre A, cf. the summary metrical analysis by M. Leumann apud E. Leumann (1933–1936): xxvi n. 1.
1.1ab cu mānau nĕ usahyäta ˈpyūṣṭä ttätĕ puva’ṃṇavīya saˈlāva ·
LHH LLLLˈHL ‖ LLLLL HLLˈHL ‖
1.1cd haṃggūjsu nariyŏ duˈkhyau jsa cu ttä vaśīvīndä uysˈnaura 1
HHL LLLLˈHL ‖ LLLH HLLˈHL ‖
1.2ab kyĕ haḍä śśäru yīndä śśäˈruī bī[ḍ]ä [ky]ĕ vā dīru [ya]ˈnīṃdä
LLLLL HLLˈHHL ‖ LH HLLˈHL ‖
1.2cd dukha vaśīvīdä ˈnariyŏ ṣṭāna crrāma karma yäˈḍāṃdä 2
LLLH HLˈLLLHL ‖ HL HLLˈHL ‖
1.3ab ysiyāmatu maraṇu u ˈdukha kā[ṣk]yĕ [tca]mana basta uysˈnaura :
LHLL LLLLˈLLHL ‖ LLL HLLˈHL ‖
1.3cd jaḍa rrŏ vā ōṣku duˈkhautta suhi ju nyāṃnartha nĕ ˈīndä
LLLH HLLˈHL ‖ LLLH HLLˈHL ‖
1.4ab hajuva vā ōṣku suˈhautta balysä vätĕ haurä nä ˈbyāta
LLLH HLLˈHL ‖ HLLL HLLˈHL ‖
1.4cd ṣada nä rrŏ aśtä maˈhāyāña ttä nĕ avāyä nĕ ˈtsīdi 4
HLLL HLLˈHHL ‖ LLL HLLˈHL ‖
1.5ab tvī padī karma ˈsarvaśśūra paḍāṃjsiya yanu ˈbyāta :
HLH HLˈHLHL ‖ LH LLLLˈHL ‖
1.5cd batäna kiḍägānäna ˈsatva mästu vīvāgä baˈrīndä 5
LLLLL HLLˈHL ‖ HLH HLLˈHL ‖
1.6ab bataku juvĕ ttīma päˈrēndä pharu yĕ vīvāgu paˈjāysdĕ ·
LLLLL HLLˈHL ‖ LLLH HLLˈHL ‖
1.6cd ku yĕ ttu mäṣa byēhätĕ ˈbalysāna ku karä ttīma nĕ ˈjīyĕ 6
LLLLL HLLˈHHL ‖ LLL HLLˈHL ‖
1.7ab hajuvānu ṣä suhä cu ˈhayärīṃdi ōṣku balysi paˈrauya :
LLHL LLLLˈLLHL ‖ HL HLLˈHL ‖
1.7cd cu rŏ parēhārĕ häˈryau dīryau ōṣkä [śśäru yaˈnīndä]
LLLH HLLˈHHH ‖ H LLLLˈHL ‖
“1Although you cannot bear to hear these fearful words, the encounter with sorrows in hell which those beings experience, 2yet (he) who does good, to him it brings good; but (those) who do evil experience sorrows in hell (according to) whatever acts they committed: 3birth, death and sorrows, the anxieties by which beings are bound. Fools, too, (being) always afflicted, are not informed about happiness, 4but the wise (are) always happy. They remember to give to the Buddha, they also have faith in the Mahāyāna, (so) they do not undergo an evil rebirth. 5In this way, Sarvaśūra, remember (your) former acts: as a result of a little (evil) action, beings bear a great retribution. 6(When) one sows a tiny seed, one gains much reward, so that one reaches the Buddha-field where the seed does not perish at all. 7This is happiness for the wise, that they always delight in the Buddha’s command, (and) also that they restrain themselves from evil things (and) always [do good].”
1.1. On the expression cu mānau see Emmerick (1970): 74–75.
The ending of usahyäta is only attested in MS 22. In SGS 16, Emmerick tentatively takes the form as 3 sg. injunctive, but the construction seems to require a second person form (see KS 126 n. 2). Leumann (1920): 14 assumes that the scribe has written 2 pl. pres. by mistake for 2 sg., as the Buddha here addresses Sarvaśūra alone. A purely graphic explanation is another possibility, “intrusive -t-” being a fairly common feature of the manuscripts. For further certain or possible examples in the Sgh see 5.10 and 5.36 in Sims-Williams (forthcoming a), 5.44 and 12.3 below; cf. also KS 140 n. 17. At any rate, the metre here requires the suppression of one mora.
1.2. MS 22 has naryo (as in the preceding verse), while MS 19 has the alternative loc. sg. form narya. From the 7-mora cadence ōṣku mä ˈnariya HLLˈLLL in 5.7c (Sims-Williams forthcoming a), it is clear that this word is stressed on the first syllable and hence that the present cadence is to be read vaśī)vīdä ˈnariyV̆ ṣṭāna HLˈLLLHL,21 a “type 2” 9-mora cadence (rather than “type 1”, which would require the stress *naˈriyV̆).22
1.3. u, which is required for the metre, is found in MS 22 but omitted in MS 24.
Leumann read the last word of the verse (only attested in MS 22) as tsīndä. Most subsequent editors have followed him, but Skjærvø’s [ī]ndä is more likely.
1.6. As already noted by Leumann (1920): 14, jve (i.e. juvĕ) stands for ju yĕ. See further Maggi (2017a): 279–281, where this example is cited with several others.
On ttu mäṣa see KS 127 n. 12 and Maggi in SVK3: 69–70 s.v. †ttumäṣa.
2 Saṃghāṭa-sūtra 47
Edition and translation in KS 22, based on MS 17, transcribed in KS 227–228 (IOL Khot 161/2, r4–v4, cf. Leumann 1920: 15–17; Konow 1932: 71; KT5: 70, #155; Skjærvø 2002: 360). The equivalent Sanskrit text is edited in von Hinüber (2021): 20 (verses 10–13), with English summary of the contents on p. lxi.
Most of the text is too poorly preserved to be interpreted metrically. However, Leumann was probably right to recognize bal)ysūśtu vaˈrālstŏ HLLˈHL “towards enlightenment” (cf. 7.1 below) as the 7-mora cadence ending a verse in metre A or B.
3 Saṃghāṭa-sūtra 49
Diplomatic edition and translation in KS 23, based on two manuscripts:
-
MS 10, transcribed in KS 197 (IOL Khot 178/3, r1–2, cf. Konow 1932: 72; KT5: 330, fol. 20; Skjærvø 2002: 397);
-
MS 18, transcribed in KS 233 (SI P 67.19, v1–2, cf. SDTV3: 77).
The equivalent Sanskrit text is edited in von Hinüber (2021): 21 (verses 14–15), with English summary of the contents on p. lxii.
What remains of the Khotanese text corresponding to Sanskrit verse 14 does not seem to be metrical: hämätĕ · pyū’vīru uhu nyaṇḍyau “… will be … Listen, you Nirgranthas!”. At least the beginning of the following Sanskrit verse 15 appears to be represented by a verse in metre A, though it is hard to discern a metrical form in the poorly preserved passage which follows. This begins with the words aysu pandāyä näjsätā’kä “I (am) the shower of the way”, a plausible metre A pāda (-LLH HLLˈHL), but also contains the phrase [balysā]nyau tcē’mañyau HHHHLH “with Buddha-eyes” (~ Sanskrit buddha-cakṣuṣā), which is difficult to fit into a regular metrical structure.
3.1ab jaḍyau ku suhū ju ˈniśtä tta ku ṣṭau purrauṣca hä[ˈmātĕ]
-LHL LHLˈHL ‖ LLHL HLLˈHL ‖
3.1cd [rraṣṭŏ pa]ndō nĕ paˈysānda tta ku ṣṭau purrauṣca häˈm[ā]tĕ
HLH HLLˈHL ‖ LLHL HLLˈHL ‖
“Fools! Since you have no happiness, where then [will] be your victory? [Since] you do not recognize the [right] path, where will be your victory?”
My restoration is based on similar contexts such as Z22.275b: cau rraṣṭŏ pandŏ nijsaṣḍĕ “who shows you the right path”. There, however, as in all the other six instances of the acc. sg. pando in Z, the metre clearly requires the reading pandŏ HL. The word for “path” shows the regular inflections of a stem pandāa- in all cases except the nom. and acc. sg., where Old Khotanese attests forms from a stem pandā-: nom. pandĕ < *pantāh (Avestan paṇtā̊), acc. pandŏ < *pantām (Avestan paṇtąm). The unique inflection of this word was clearly unstable: as noted by Emmerick, SGS 308, Late Khotanese tends to replace these two forms by the regular āa-stem nom.-acc. sg. ending -ā, and the acc. sg. paṃdā is already found in Sgh 253 in MS 17 (SI P 53.11 + 18, v2–3, in SDTV3: 58), a manuscript in “late” script, but one where the influence of Late Khotanese orthography is “not strong” (KS 220). If the present passage is indeed metrical, the metre requires the reading pandō HH. Such a form, beside that in -ŏ, might have been created on the analogy of the word for “earth”, for which the two acc./loc. sg. forms śśandō, śśandau (stem śśandaā-) and śśandŏ, śśaṃdu (stem śśandā-) are both unambiguously attested in Z (see Sims-Williams 2022: 32). Unfortunately paṃdau in Sgh 244 (10.1 below), which looks like a variant spelling of pandō HH, does not provide any confirmation, since, if the passage is metrical at all, it must be read paṃdaù = pandŏ HL.
4 Saṃghāṭa-sūtra 51–52
Diplomatic edition and translation in KS 24, based on two manuscripts:
-
MS 10, transcribed in KS 198 (IOL Khot 178/3, v1–5, cf. Konow 1932: 73; KT5: 330; Skjærvø 2002: 397);
-
MS 18, transcribed in KS 233 (IOL Khot 26/8 + 24/6, r1–5, cf. KT5: 157, #295; Skjærvø 2002: 222).
The equivalent Sanskrit text is edited in von Hinüber (2021): 21–22 (verses 16–19), with English summary of the contents on p. lxii.
The translation of the verse in Sgh 52 does not seem to be metrical, but the two verses in Sgh 51 are in metre A.
4.1ab nĕ nĕ ju muhu vaṃña ttaˈttīka māta pätĕ trāstu yaˈnīndä ·
LLLLL HLLˈHL ‖ HLLL HLLˈHL ‖
4.1cd sa[mu] śśū[kĕ . . diˈyā]mä ttuśśĕ k[y]ĕ nĕ ju nyāpätä ˈnīrä ·
LLHL
μμμ LˈHL ‖ LLLLL HLLˈHL ‖4.2ab nĕ nĕ ju vara gyasta nĕ ˈhva’ndä nĕ banhya ō vā ˈkṛṅga
LLLLL HLLˈHL ‖ -LHL HH|ˈHL ‖
4.2cd vara ttŏ diśŏ daindä anˈāsp[ētŏ] [ku nĕ nĕ dä]stä śtä ˈbalysä ‖
LLLLL HLLˈHHL ‖ LLL HLLˈHL ‖
“1Mother (and) father cannot protect me here now. We [see] merely lonely empty [forests], in which no water is found. 2No gods (are) there, no men, neither trees nor cocks.” They see that place there (as) providing no refuge [since] the Buddha did [not appear].
4.1. With some hesitation I have followed Bailey’s reading of the unclear akṣara after ttuśśe as k[y]e (rather than ne with Canevascini and Skjærvø). Here kye would be used as gen. sg. of cu, lit. “of which”.
4.2. The 9-mora cadence of 2c is probably “type 1” in the classification of Sims-Williams (2022): 55–59, with stress on the 5th mora, i.e. anˈāsp[ētŏ]. Since there is no other evidence for the stress in this word a “type 2” cadence with stress on the 4th mora, i.e. ˈanāsp[ētŏ], cannot be ruled out absolutely, though the resulting pattern HLˈLHHL would in fact be unique.
5 Saṃghāṭa-sūtra 99
Diplomatic edition and translation in KS 44–47. The equivalent Sanskrit text is edited in von Hinüber (2021): 39–42 (verses 20–43), with English summary of the contents on pp. lxvii–lxviii.
Of the Sanskrit verses, 28–34 are in triṣṭubh/jagatī metre,23 whereas verses 20–27 and 35–43 are ślokas, like virtually all the other verses in the Sgh. The Khotanese version of Sanskrit verses 20–27, 35–41 (ślokas) and verse 34 (triṣṭubh/jagatī) is in metre C. This was probably also the case with the translation of Sanskrit verses 28–33, but at this point the Khotanese text is too badly preserved for any certainty. At any rate, it seems that the changes of metre in the Khotanese version do not correlate with those in the Sanskrit text. The last two Sanskrit ślokas, 42–43, are represented by four Khotanese verses in metre B.
Since I have already given a metrical restoration of the verses in metre C (numbered as 1–40 in Sims-Williams forthcoming a), only the final verses in metre B are considered here. My numbering of the verses continues from that of the preceding verses in metre C. The relevant passage is partially preserved in two manuscripts:
-
MS 12, transcribed in KS 216 (IOL Khot 162/6, r1–3, cf. KT5: 48, #123; Skjærvø 2002: 363);
-
MS 17, transcribed in KS 229 (SI P 53.9, r2–v1, cf. SDTV3: 55).
5.41ab [cu] mä [īndä dāruṇa] [käḍäyānĕ nĕ] ˈpyūṃjĕ
LLHL HLL ‖ LL HLLˈHL ‖
5.41cd cu buraù aysu yäḍä mä u parstēmä sĕ ˈyanda ·
LLLLL LLL ‖ LH HLLˈHL ‖
5.42ab ttiyĕtĕ k[äḍäyānä ma vīvā]tu baˈrīmä
LLLLL HL ‖ LH HLLˈHL ‖
5.42cd ma nĕ ttä dukha vīyanĕ mamä bērä häˈmāndĕ
LLLLL HLL ‖ LL HLLˈHL ‖
5.43ab räṣayä ttrāy[ākä] [mamä hämātĕ] duˈ[khy]au [jsa]
LLLH HL ‖ LLL HLLˈHL ‖
5.43cd [cu mamä] vätĕ dāruṇa käḍäyānĕ jiˈyāndĕ
LLLLL HLL ‖ LL HLLˈHL ‖
5.44ab [.]ä × yau mä aysmiyata jina ka m[......]
LLHL HLL ‖ LL
μμμμ ˈμμμ ‖5.44cd [aysu tä] hämĕ bīsä yāvajīvī ˈbraucä ·
LLLLL HL ‖ HL HH|ˈHL ‖
“41[Whatever are] my [terrible deeds], I [do not] deny (them), whatever I did and (whatever) I ordered, saying: ‘Do (it)!’ 42May I [not] bear [the result] of this [deed]! May these sorrows (and) pains not have to be borne by me! 43[May] the ṛṣi [be my] deliverer from sorrows! [Whatever] were [my] terrible deeds, may they be removed! 44… in my mind, destroy …! [I] will be [your] servant so long as I live.”
5.41. For the restoration dāruṇa cf. 43cd. Canevascini’s dīra is too short by one mora.
5.42. ttyete is a variant spelling of ttätäye, gen. sg. m. of ṣäta-. This reading (as in SDTV3) is to be preferred to ttye ne (as in other editions of both manuscripts).
For ma + 1 sg. pres. indicative see Emmerick (1991): 105 n. 22.
dukha vīyanĕ “sorrows (and) pains”. Although the underlying Sanskrit compound duḥkhavedanā is most naturally interpreted as a tatpuruṣa “unpleasant sensations”, it seems that Khotanese dukha- is only used as a noun (see Canevascini, KS 126, note 5 to § 42).
5.43. It is difficult to see how väte can be the postposition here. Canevascini, KS 256a s.v. ah-, hesitantly takes it as 3 pl. m. intr. perfect of the verb “to be” (the expected form being väta).
5.44. aysmyata is another example of “intrusive -t-” (see above on 1.1).
6 Saṃghāṭa-sūtra 199
Diplomatic edition and translation in KS 80–81, based on two manuscripts which only overlap for one hemistich (6.6cd):
-
MS 7, transcribed in KS 183–184 (IOL Khot 170/4,24 r5–v5, cf. KT3: 141; Skjærvø 2002: 381);
-
MS 10, transcribed in KS 208 (IOL Khot 182/3, r1–4, cf. Konow 1932: 91–92; KT5: 339, fol. 77; Skjærvø 2002: 402).
The equivalent Sanskrit text is edited in von Hinüber (2021): 80–81 (verses 54–63), with English summary of the contents on p. lxxxiv. The Khotanese version is in metre A.
6.1ab [. .]ṣvajsei’ āchei ˈōśä ācheina puva’ṇa häˈmārĕ ·
Η LH HH|ˈHL ‖ HHL LLLLˈHL ‖6.1cd naro haḍĕ hīśtä maˈraṇä pūryau āchei ma puˈva’ta’
-LLLL HLLˈLLHH ‖ H HLLˈHL ‖
6.2ab tta parrìyū hämät[ĕ] u ˈmaṃ pūryau āchyau puva’ˈṇyau jsa
LLLH LLLLˈHHH ‖ H HLLˈHL ‖
6.2cd ttārä haḍĕ daidä sĕ ˈmāvīyau nĕ maraṇä diyĕ ˈātĕ | |
HLLL HLLˈHHH ‖ LL LLLLˈHL ‖
6.3ab biśä nä [...] aṃṅga maˈraṇu dyāmu uysyāniyĕ nĕ nĕ ˈvaṃña
LLL
μμ HLLˈLLHL ‖ LH LLLLˈHL ‖6.3cd bajāṣa {nĕ} pyùvā’manĕ ˈguvyau jsa tcēmanyō rūva nĕ {nĕ} ˈdyāmä 3
LHLL HLLˈHL ‖ HLH HLLˈHL ‖
6.4ab [...... bu]śä nĕ buˈvā[r]ĕ [.............]
μμμμμ LLLLˈHL ‖μμμμμ μμμμ ˈμμμ ‖6.4cd [....]ī × hamara ˈgūsīndä samu khaù cakalä ⟨... .⟩
μμμμμ LLLLˈHHL ‖ LLL LLLμ ˈμμμ ‖6.5ab ⟨ttiyĕ⟩ ttaraṃdarä mā ˈttrā[yākä] [niśtä ... ... . .]
LLLH LLH|ˈHHL ‖ HL
μμμμ ˈμμμ ‖6.5cd [...............] ttau hvatĕ mä kō nĕ puˈvā’ta
μμμμμ μμμμ ˈμμμ ‖ HLLL HLLˈHL ‖6.6ab maraṇū ju karä na[rŏ ˈātä] [.............]
LLHL LLLLˈHL ‖
μμμμμ μμμμ ˈμμμ ‖6.6cd nĕ ju ttavai āchai ˈyīndä mā dukha maraṇä nä ˈātä 6
LLLH HH|ˈHL ‖ HHL LLLLˈHL ‖
6.7ab nuṣṭhuru ṇä ātä maˈraṇä patana cĕ mā jīvätä ˈnāstĕ
HLLL HLLˈLLLLL ‖ LH HLLˈHL ‖
6.7cd ttaraṃdarä ṣṭānä nä ˈhaṃthrīśtä ttä dukha harbiśśä ˈdyāmä 7
LHLL HLLˈHHL ‖ LLL HLLˈHL ‖
6.8ab cĕ nä rrŏ trāyākä u ˈāspāta hämätĕ vaṃña ttaˈttīka
LLLH HLLˈHHL ‖ LLL HLLˈHL ‖
6.8cd ttiyĕ māràpätara tta ˈhvāñīndä aśka auṣṭai ˈgyasta 8
LLHL LLLLˈHHL ‖ HL HH|ˈHL ‖
6.9ab gyastānu gyaysnä [... .] [... .]ru hämätĕ ×[. .]
HHL HLLˈ
μμμ ‖μμμμ L LLμμ ˈμμμ ‖6.9cd mara[ta] yanda khŏ ttĕrä ˈīyä ka haḍĕ mamä tsāṣṭä häˈmātĕ 9 | |
LLLH LLLLˈHL ‖ LLLLL HLLˈHL ‖
“1The … illness (is) bad (and) fears arise from the illness, but death is not yet coming, sons. Do not fear the illness! 2So there will be deliverance from illnesses (and) fears for you, sons.” But those (sons) see (rightly, saying): “… death has come to be seen by us. 3All our limbs [ache]; we see death for (our)self. Now we do not hear sounds with (our) ears, we do not see shapes with (our) eyes. 4[Our nostrils do] not perceive odours … [our] joints are being loosened/broken just like a ⟨senseless?⟩ piece of wood. 5For ⟨this?⟩ body we [have no] protec[tor] …” [The father says:] “Thus I said to you: ‘You should not be afraid!’ 6Death [has] not [yet come] for you.” [The sons say:] “The fever illness is not causing us sorrows: death has come to us. 7Harshly has death come before us, which will take our life. Our body is oppressed; we see all these sorrows. 8Who will be our protector and refuge now here?” His parents say thus: “Perhaps you have angered the gods. 9A sacrifice to the gods … will be …” “Do (it) here so that it may be thus. If only it might be easy for me!”
6.1. Canevascini (KS 272) draws attention to the spelling of 2 pl. impv. puva’ta’ with t rather than the expected tt.
6.2. u maṃ “and for me” cannot be correct, since the mother and father are both speaking (hvāñīndä “they speak”). Canevascini’s emendation to umā “for you” is probably correct (though this word is redundant beside its encl. equivalent -ū).
māvīyau nĕ maraṇä diyĕ ātĕ is problematic. Canevascini left the sequence vī × ne uninterpreted and tentatively understood the rest as: “Our (mā) … have come (āte) to-be-seen (dye < däte) in death (maraṇu)”, comparing the expression in Sgh 85.4: [… gya]sta balysa däte tsutānda “[these] Lord Buddhas have come to-be-seen [by you]” translating ete … tathāgatās … upasaṃkrāntā darśanāya. I have adopted this interpretation of diyĕ ātĕ, but take nĕ (for nä) to be the 1 pl. encl. pronoun and the subject of ātĕ (for ātä, 3 sg. m. intr. perfect) to be “death”, cf. 6cd below: nĕ ju ttavai āchai yīndä mā dukha maraṇä nä ātä “the fever illness is not causing us sorrows: death has come to us”. A possible but very hypothetical interpretation of māvīyau might be “O (parents of) ours”, voc. pl. of an adj. māvīya- formed from māvu, gen. pl. of the 1 pl. pronoun, here used as a noun like French “les nôtres”.
6.3. As Canevascini implies, uysyānye stands for uysānye “self”, translating Sanskrit ātmanaḥ. The treatment of the first syllable as light is therefore as expected.
On the writing pyūv’- in place of pyŭv’-, the pre-vocalic form of the verb pyūṣ- “to hear”, see Sims-Williams (2022): 42. Even if one reads pyùv- here as in other such cases, pāda c is still one mora overlong. The simplest correction is to delete the negative nĕ (which is redundant, as the sentence already has a double negative nĕ nĕ in pāda b). As it stands in the manuscript, pāda d is also overlong by one mora, which can similarly be corrected by deleting one of the two negatives.
6.4. The words hamara gūsīndä samu khau cakalä are discussed by Emmerick in SVK1: 125–126, where the meaning of hamara “joints” is established. The equivalent Sanskrit text (verse 58) was previously read as aṅga-m-aṃgāni mucyaṃti kāṣṭhā iva acetanāḥ “my limbs are loosened like an unconscious piece of wood” (see ibid., 126 n. 1, and KS 80), on the basis of which gūsīndä was interpreted as belonging to ggūs- “to escape, be delivered, set loose”. However, the correct reading of the verb is now known to be bhidyaṃti (see von Hinüber 2021: 81 and lxxxiv n. 208). While bhid- can have a meaning compatible with that of ggūs- (cf. PW s.v. bhid- 5) “lösen, entwirren; Pass. sich lösen, aufgehen”, as Ruixuan Chen points out to me, the literal meaning of bhidyaṃti “are split” seems to make better sense of the simile comparing the limbs with a piece of wood. It is therefore worth considering the alternative possibility that gūsīndä stands for gŭsīndä, 3 pl. pres. act. of an otherwise unattested verb *gusid- “to be split” < *wi-sid(y)a-, cf. MP wisinn-/wisist- “to break, split”, Old Indian vi-chid-. Phonologically, *gusīndä would be exactly parallel to bīndä, 3 pl. pres. act. of bid- “to pierce” < *bida- (cf. SGS 96). Both interpretations allow a metrically regular reconstruction of the cadence of pāda c and the following segment, either 9 ‖ 3:
[....]ī × hamara ˈgūsīndä samu khaù cakalä ⟨....⟩
μμμ H LLLLˈHHL ‖ LLL LLLμ ˈμμμ ‖
or 7 ‖ 5:
[. .]ī × hamara gùˈsīndä samu khaù cakalä ⟨......⟩
μμ Hμ LLLLˈHL ‖ LLLLL Lμμμ ˈμμμ ‖
In each case one must assume that there is an omission in the manuscript after the words samu khaù cakalä “like a piece of wood”, where ttaraṃdarä “body” appears to be neither meaningful nor metrically acceptable.
6.5. The partial restoration of 5ab is of course very hypothetical and is intended only to show that the surviving words could fit into a meaningful and metrically correct sentence. In the next hemistich Bailey’s “-tau” which I have completed to ttau “thus to you”, is preferable to the graphically and metrically impossible [pū]ryau “O sons” of the other editions. The 3 sg. m. form hvatĕ mä “I said” implies that the father is speaking rather than the mother as in the Sanskrit. An emendation to the equivalent f. form *hvatāmä, as tentatively suggested by Canevascini (KS 149 n. 14), is metrically impossible. This discrepancy is similar to that between the consistent sg. “son” in the Sanskrit and the varying sg./pl. “son(s)” in the Khotanese version of the story (see below on 6.8). In any case, the content of the speech cannot be renconciled with that of the Sanskrit text: vaktuṃ nārhasi putraivaṃ mā me trāsaparāṃ kuru “You should not say thus, son! Do not cause me excessive fear!”.
I owe the correct reading puvā’ta (2 pl. subj., rather than puva’ta, 2 pl. impv.) to Mauro Maggi.
6.6. MS 10 has mā dukha, the reading adopted by Canevascini. Since kh in this word is sometimes treated as a double consonant (Sims-Williams 2022: 85), one can read mā dukha HHL as a regular 5-mora segment. Before dukha MS 7 inserts tä ttä, suggesting a possible alternative reconstruction of the segment as mā ttä dukha HLLL.
6.7. ṇä ātä maraṇä patana “death has come before us”. I owe the excellent reading patana (for *patäna) to Ruixuan Chen. Canevascini read pvatana (taking this for *pvaṇä “fear”, with “intrusive -t-” and irregular -n- for -ṇ-).
6.8. It seems that, except where is is used as a proclitic article, the gen. sg. form ttye is always realized as disyllabic ttiyĕ (see Sims-Williams 2022: 33 with n. 54). This implies that either the first or the second syllable of the following mārāpätara “parents” must be subject to metrical lightening in order to obtain a regular 5-mora segment. In the text above I have placed the reading mārà°, but the choice of this solution rather than màrā° is essentially arbitrary (cf. Sims-Williams 2022: 71 (iv)).
As noted by Canevascini, KS 148, note 2 to Sgh 198, by referring to “his parents” the Khotanese version here abruptly switches to referring to one son (as in the Sanskrit text) rather than several. The verb in 8d has so far been read as 3 sg. m. tr. perfect auṣṭĕ “he has angered”, but a reading 2 sg. m. auṣṭai “you have angered” seems equally possible from the traces visible and is to be preferred as giving a regular 7-mora cadence HH|HL.
6.9. The reading and interpretation of this verse is largely guesswork. Only the last pāda is really clear.
7 Saṃghāṭa-sūtra 213
Diplomatic edition and translation in KS 88. The passage is partially preserved in MS 10 (whole passage) and MS 11 (verse 1 only):
-
MS 10, transcribed in KS 210 (IOL Khot 183/3, r4–v1, cf. Konow 1932: 99–100; KT5: 341, fol. 83; Skjærvø 2002: 404);
-
MS 11, transcribed in KS 214 (IOL Khot 169/2 + FK. 913 Kha. 28,25 v5, cf. Skjærvø 2002: 379; for the London fragment only cf. Konow 1932: 98 n. 1; KT5: 264, #556).
The equivalent Sanskrit text is edited in von Hinüber (2021): 88–89 (verses 66cd–69), with English summary of the contents on p. lxxxvii.
The Khotanese text is in metre B. Manu Leumann, apud Leumann (1933–1936): xxvi n. 2, also gives a metrically restored version of this passage based on Konow’s edition.
7.1ab nĕ ma śtā śtä rruśtä ttagatä · vasväutä yäˈḍaimä
LLHL HL ‖ LLL LLLLˈHL ‖
7.1cd praṇähānä hastamu balysūśtu vaˈrālstŏ
LLHL HLL ‖ H HLLˈHL ‖
7.2ab [ku] ṣṭä nä[r]vā[nä tsāṣṭä] balysa tsuˈtāndä
-LLH HL ‖ HL HLLˈHL ‖
7.2cd hālstŏ aysu tsīñä u balysūśtu buˈvānĕ ·
HLLL HL ‖ LH HLLˈHL ‖
7.3ab umā hvāñīñä dātu tcamna yĕ ˈhīśtä
LHH HL ‖ HL HLLˈHL ‖
7.3cd närvā[nä kī]ntha ku biśśä puva’ṇa jiˈyārĕ ·
HHL HL ‖ LLL LLLLˈHL ‖
7.4ab palaṃgu bastĕ ṣä naḍĕ tta hvatĕ yäˈḍaimä
-LHL HL ‖ LLL LLLLˈHL ‖
7.4cd paḍā käḍätānĕ cu banŏ kaṣṭumä ˈrruīya ·
LHLL HL ‖ LLL HLLˈHL ‖
7.5ab kyĕ bu[r]ŏ ttätä pyuvā’ndĕ cu aysu hvataì mä saˈlāva
LLLLL HL ‖ LLL LLLLˈHL ‖
7.5cd käḍätànai harbiśśä härṣṭāyä jäˈtāndĕ ‖
LLLH HLL ‖ H HLLˈHL ‖
“1The wealth of royalty is not necessary for me. I have made a pure vow towards best enlightenment. 2Where the buddhas have gone [to peaceful] nirvāṇa—thither may I go and realize enlightenment. 3May I tell you the Law by which one comes to the city of nirvāṇa, where all fears are removed.” 4The man adopted the paryaṅka position. He said: “Formerly I committed (evil) deeds whereby I fell into the royal prison. 5Whoever may hear these words which I have spoken—may all his (evil) deeds really be removed.”
7.1. vasvätä “pure” (MS 10) seems to be a compromise between Old Khotanese vasutä and later forms such as vasvä, i.e. vasuvä. The same spelling is attested in Z2.167, see Sims-Williams (2022): 41 n. 83, referring to the explanation of this form by Leumann and Emmerick.
hastamu “very excellent” is the reading of MS 10, in place of which MS 11 has the equally possible variant aysmiya “in (my) mind” (though this is preceded by mä “my”, which would have to be deleted as unmetrical).
7.2. The restoration tsāṣṭä “peaceful” (~ Sanskrit śānte) is suggested by Ruixuan Chen.
7.5. käḍätānai is one mora overlong. M. Leumann proposes lightening the final diphthong, for which there is no parallel. A reduction of the long ā is more plausible, cf. 5-mora segments such as paramàṇavV̆ (Sims-Williams 2022: 70 (ii) with n. 142).
8 Saṃghāṭa-sūtra 214
Diplomatic edition and translation in KS 89, based on two manuscripts:
-
MS 8, transcribed in KS 185 (IOL Khot 177/5 + two fragments from the Francke-Körber collection, Munich, r2–v1, cf. Skjærvø 2002: 394; for the London fragment only see also KT5: 353, #721);
-
MS 10, transcribed in KS 211 (IOL Khot 183/3, v5–6, followed by IOL Khot 184/1, r1–4, cf. Konow 1932: 100–101; KT5: 342, fol. 83–84; Skjærvø 2002: 404).
The equivalent Sanskrit text is edited in von Hinüber (2021): 89–90 (verses 70–75), with English summary of the contents on p. lxxxvii.
The Khotanese version of verses 70–72 does not seem to be metrical, and the king’s second question (verse 74) is not translated. However, the last speech of the birds (corresponding approximately to Sanskrit verses 73 and 75), which is described as being ggāhäna “in verse” (IOL Khot 184/1, r2), is represented by three verses in metre B.
8.1ab ysamaśśaṃdiya hastamä rrūndētĕ yaˈnākä
LLLLL HLL ‖ H HLLˈHL ‖
8.1cd hämätĕ balys[ū]śtu bustä cĕ baṃhyä ˈrr[u]stä ·
LLLH HL ‖ HL LHLˈHL ‖
8.2ab cĕ nä rrustä baṃhyä ṣä vā śätĕ naḍĕ {tta} ˈdyāñä
LLHL HL ‖ LH LLLLˈHL ‖
8.2cd ṣä vā dīvatattä [.]e[...]ttaru khŏ ˈbalysä ·
LHH LLL ‖
μμμ LLLLˈHL ‖8.3ab ttänau śäru ⟨tcērä⟩ cĕ vīvāgä kṣaˈmīyä
LHLL HL ‖ LH HLLˈHL ‖
8.3cd ku yĕ nĕ puña yandĕ ṣä vīvātä ku ˈnāstĕ ‖
LLLLL HL ‖ LH HLLˈHL ‖
“1He will be the best enlightener in the world. (He) whose tree grew has realized enlightenment. 2(He) whose tree did not grow, the second man, (is) to be viewed thus: He (is) Devadatta … rather than the Buddha! 3Therefore (he) to whom a reward would be pleasing ⟨should act⟩ well. When one does not perform meritorious acts, where will one obtain the reward?”
8.2. Pāda b, as it stands in both manuscripts, is overlong by one mora. This can most easily be corrected by suppressing tta “so”. Alternatively, one might delete either śätĕ “second” or naḍĕ “man” and read diyāñä LˈHL in place of dyāñä ˈHL.
ṣä vā, the first two words of pāda c, are clear in MS 8, where they are followed by a lacuna. The following words are found in MS 10, where the akṣaras (apart from the superscript vowels) are partially obscured by mirrored writing transferred from another folio. Here Canevascini cautiously has only -ī v- ta -ä -e, while Konow has dīvata [väte] and Bailey and Skjærvø read dīvata väte. However, nothing in the Sanskrit text explains the mention of a “god” (dīvatā-), while Sanskrit verse 75 clearly identifies the “fool whose tree did not grow” as Devadatta.
8.3. The first pāda, found only in MS 8, seems to be lacking three morae. The sense requires a form of the verb “to do”, perhaps yīndä “he does” or tcērä “(it is) to be done”.
9 Saṃghāṭa-sūtra 243
Edition and translation in KS 101–102 (with transcription of the single MS 17, i.e. SI P 53.10, in KS 229–230, and Emmerick 1995: 163–164) and in SDTV3: 56–57. The equivalent Sanskrit text is edited in von Hinüber (2021): 106–107 (verses 111–119), with English summary of the contents on p. xcv.
The Khotanese version of these verses, which is perfectly preserved, is in Metre B.
9.1ab ku vā bihŏ buḍĕ mä jsatĕ mä pūrnyau ˈgūysna
LHLL LLL ‖ LLL HH|ˈHL ‖
9.1cd tta vā pharu haṃndara datĕ tvīṣṣĕ yäˈḍaimä
LHLL HLL ‖ LL HLLˈHL ‖
9.2ab īṇāta lauva ttāma gūśtŏ hvaˈḍāndä
HHL HL ‖ HL HLLˈHL ‖
9.2cd dukha ju vīvātu vaṃña śūkä baˈrīmä ·
LLLH HL ‖ HL HLLˈHL ‖
9.3ab nĕ ju yiḍĕ mä āysda maraṇu sĕ rraysgu ˈhīśtä
LLLLL HL ‖ LLL LHLˈHL ‖
9.3cd jaḍīna pāḍä mä ttätä atäraña ˈaṃga
-LHL HLL ‖ LL LLLLˈHL ‖
9.4ab ku mä maraṇä ātä karä mä trāṇä nĕ ˈvätä śtä
LLLLL HL ‖ LLL HLLˈLLL ‖
9.4cd ysanĕ ma samu śāmiña spāśīrŏ ttu ˈkālu
LLLLL HLL ‖ H HLLˈHL ‖
9.5ab gūnĕ bärätāndä u bärṣṭāndä prraˈhōṇu ·
HLLL HL ‖ LH HLLˈHL ‖
9.5cd haryāsa auśa cu pamätāndä ttu ˈkālu
HHL HL ‖ LLL HLLˈHL ‖
9.6ab kamalu vätä phānä dukhäna pärja duˈvīrĕ
LLLLL HL ‖ LLL HLLˈHL ‖
9.6cd nĕ ma ju haḍĕ ciṃdäku hanu trāstu yaˈnārŏ ·
LLLLL HLL ‖ LL HLLˈHL ‖
9.7ab bärga rrūvāsa ṣuṃndä śvānū ˈsuṭhṭha
HLH HL ‖ HL HH|ˈHL ‖
9.7cd ājäväṣä bihara kyĕ samu gūśtŏ hvaˈrīndä
HLLL LLL ‖ LLL HLLˈHL ‖
9.8ab ttaraṃndarä ṣṭāna ttä vaṃña mamä hvaˈrīndä
LHLL HL ‖ LH LLLLˈHL ‖
9.8cd kyĕ nĕ ju dukhä hämätĕ kyĕ ttätä yiḍĕ käḍäˈtānĕ
LLLLL LLL ‖ LLL LLLLˈHL ‖
9.9ab nĕ ju aruva’ vījä nä haṃndara saṃˈbāra
LLLLL HL ‖ LH LLHˈHL! ‖
9.9cd muhŏ maraṇu vīrä ttaṃdu trāstu yaˈnārŏ
LLLLL HL ‖ HL HLLˈHL ‖
9.10ab kyĕ ttä yiḍĕ dīra karaṇa vara śätĕ ˈvaṃña
-LLLL HL ‖ LLL LLLLˈHL ‖
9.10cd ṣä ju nĕ d[u]khu buttĕ u maṃ vīyanĕ ˈvaṃña
LLLLL HL ‖ LH HLLˈHL ‖
“1When I rode out to the hunt I killed deer with arrows. Thus I destroyed many other animals. 2Strangers then ate the meat, (but) now I alone bear sorrows (and) retribution. 3I did not pay attention to death, (thinking): It will come swiftly! Out of foolishness I nourished these ungrateful limbs. 4When my death came I had no protection at all. (My) kinsmen would just look me in the face at that time; 5they dishevelled (their) hair and in anger they tore the black clothes which they put on at that time; 6(there was) dust on (their) head; out of sorrow they would beat (their) breast. But they will not be able to protect me, (not) even a little. 7Wolves, jackals, ravens, dogs and birds of prey, snakes (and) …, which eat only meat—8they are now eating my body, which has no sorrow (but) which did these (evil) deeds. 9No medicine, (no) doctor, no other equipment will be able to protect me (even) a little at (the time of) death. 10(The body) which did these evil acts lies there now. It does not know sorrow and now the pains (are) mine!”
9.2. The Sanskrit text says nothing about the meat being eaten by others. The words īṇāta lauva “foreign people, strangers” seem to derive from the phrase paraloka ajānatā “not knowing the other world”, though it is hard to imagine that the translator could have misunderstood such a commonplace expression if that was the reading of his source.
9.5. On bärätāndä “they dishevelled” see Emmerick (1995).
9.6. On pärjā- “breast” see Maggi (2022): 326–327.
duvīrĕ is an unusual spelling for 3 pl. opt. *duvīrŏ, see KS 157. For the use of the optative cf. spāśīrŏ in verse 4.
9.7. bihara- is an unknown word, which evidently denotes some carnivorous animal.
9.9. haṃn)dara saṃˈbāra LLHˈHL! seems to be an exceptional 7-mora cadence without what seems in general to be a compulsory word- or compound-boundary between the two heavy syllables: LLH|ˈHL. In the whole of Z there is only one possible example of such an irregularity, namely, marä kauˈśāmä “here in Kauśāmbī” Z24.466b. However, as noted in Sims-Williams (2022): 50 n. 101, the hemistich 466ab is problematic in other respects, and it is likely that the cadence should be emended, perhaps to maratä kaùˈśāmä LLLLˈHL, with lightening of the syllable preceding the ictus by the “uysnora-effect”. A similar solution of the problem here can be achieved by emending to *haṃn)dāra saṃˈbāra HLLˈHL. Alternatively, it may be that LLHˈHL, without the usual word- or compound-boundary, may have been regarded by some poets, perhaps including the author of the present text, as an acceptable 7-mora cadence. Possible examples in other texts in metre A include khu ji urˈmaysdi “like the sun” in the Suvarṇabhāsottama-sūtra26 and sar)bätä urˈmaysd[ĕ] “the sun rises” in the Bhaiṣajyaguruvaiḍūryaprabharāja-sūtra,27 in both of which it appears that the last word scans as HˈHL, without the lightening to urˈmaysdĕ LˈHL which is systematically attested in this context in Z.
10 Saṃghāṭa-sūtra 244
Edition and translation in KS 103 (with transcription of the single MS 23, i.e. IOL Khot 159/9, r1, in KS 243); cf. also Leumann (1920): 41; Konow (1932): 105; KT5: 80, #167; Skjærvø (2002): 357. The equivalent Sanskrit text is presumably verse 140 as edited in von Hinüber (2021): 109, with English summary of the contents on p. xcvi, though the correspondence is quite vague.
The text as restored by Canevascini can be scanned as part of a verse in metre B, but since only a few words are preserved it is not quite certain that the passage is metrical.
10.1ab [. . hutsutu] paṃdaù jsātä pīrmŏ bäˈśśā[nu]
μμ LLL HL ‖ HL HLLˈHL ‖10.1cd [rraṣṭŏ balysūśtu] [buttĕ .......]
HLH HL ‖ HL
μμμμ ˈμμμ ‖
… he will tread the [well-trodden] path, [he will realize true enlightenment], foremost of all (things).
10.1. Regarding paṃdaù HL see above on 3.1.
11 Saṃghāṭa-sūtra 253
Diplomatic edition and translation in KS 107–112, based on four manuscripts, of which only the first two overlap:
-
MS 24, transcribed in KS 251 (FK 210,2 Do.14, otherwise unpublished), verses [38]–[46]28 = Sanskrit verses 188–196;
-
MS 17, transcribed in KS 231 (SI P 53.11 + 18 and 53.12, cf. SDTV3: 57–60), verses [43]–[59] = Sanskrit verses 193–209;
-
MS 22, transcribed in KS 240–241 (IOL Khot 6/1, cf. Leumann 1920: 34–37; Konow 1932: 106–108; KT5: 77, #163; Skjærvø 2002: 172), verses [71]–[80] = Sanskrit verses 221–230;
-
MS 2, transcribed in KS 177–178 (IOL Khot 141/1, r1–v2, cf. Leumann 1920: 37–39; Konow 1932: 108–109; KT5: 78, #164; Skjærvø 2002: 322), verses [83]–[90] = Sanskrit verses 233–240.
In addition, Skjærvø (2002): 217 and 313 identifies IOL Khot 23/2 and IOL Khot 115/6 (not in Canevascini’s edition) as containing parts of verses [65]–[71] and [86]–[87] respectively, corresponding to Sanskrit verses 215–221 and 236–237.
The equivalent Sanskrit text is edited in von Hinüber (2021): 117–123, with English summary of the contents on pp. c–cii.
Some of these verses are too fragmentary for analysis. Those that are better preserved do not appear to be metrical. Canevascini, KS xvii, apparently considered that some of these verses attest variant forms of the usual metres, for instance describing the phrase ysojsa khāysañä –⏑–⏑⏑ in his verse [56] as a metre B “cadence 9” (according to Emmerick’s system), here appearing irregularly at the end of the verse. In the following verse [57] he sees a possible trace of “an older metrical system based on the morae”, suggesting that it could be scanned thus “as an Indian gaṇacchandas”:
u kṣamätä mājai ttarandarä ttonäka śärka skomata skute
⏑ ⏑ ⏑ ⏑ | – – | ⏑ – ⏑ | ⏑ – ⏑ | ⏑ – ⏑ | – | ⏑ ⏑ ⏑ ⏑
Leumann (1920): 34–35 regarded the text of the well-preserved folio IOL Khot 6/1 as a kind of free verse or rhythmical prose,29 noting that several phrases can be interpreted as pādas in metre A,30 or even as complete hemistichs in metre A or metre B:
sarvvasatvānu uysˈnaurānu vaska hāvaṃ-ˈgārä [75], metre A
HLH HLLˈHHL ‖ HL HH|ˈHL ‖
lakṣaṇyau haṃphutĕ kvī pajsamu yaˈnāmä [77], metre B
HLH HLL ‖ H LLLLˈHL ‖
It would be possible to reach more than one conclusion on the basis of such observations. One is that the text was composed in the standard Khotanese metres, but that the original metrical structure of the verses has been obscured by later revisions or interpolations. The other is that it was originally composed in a sort of semi-verse or a mixture of verse and prose as Leumann implies.
12 Saṃghāṭa-sūtra, Unlocated
Transcribed in KS 230, based on MS 17, i.e. SI P 53.16, edited with translation in SDTV3: 63. Not definitely located, but Canevascini draws attention to parallels in Sgh 253. The numbering indicates that the underlying Sanskrit verses were numbered 22–29.
Metre B.
12.1ab [..........] [......]tä ˈvä[tä] śtä
μμμμμ μμμ ‖μμμ μμμ LˈLLL ‖12.1cd yanā ku nä niśtä satvä cĕ hāva ˈpaśtä 2
LHLL HL ‖ HL LHLˈHL ‖
12.2ab [.]ä[ .......] [..........]
[
12.2cd [cu] burŏ yäḍĕ īyä ggarkha dāruṇa ˈkarma ‧
LLLLL HL ‖ HL HLLˈHL ‖
12.3ab pātcu mara tta [...] [.........]yä
HLLL
μμμ ‖μμμ μμμμ ˈμμμ ‖12.3cd viyata paru × ttĕrä gyasta balysa häˈmāru 23
LLLLL LLL ‖ HL HLLˈHL ‖
12.4ab cĕrä ttā[mu ....] [..........]
LLHL
μμμ ‖μμμ μμμμ ˈμμμ ‖12.4cd [....]ä × × [.]āña [.]ĕ ttū dātu pyuˈvā’tĕ 4
μμμμμ HL ‖ LH HLLˈHL ‖12.5ab ttäna ka pharu [...] [..........]
LLLLL
μμμ ‖μμμ μμμμ ˈμμμ ‖12.5cd [.....]dä hämätĕ ttōlstŏ ustamu ˈbāḍä ·
μμμμ L LLL ‖ HL HLLˈHL ‖12.6ab ttāmu [ha]späśtä ku [.........]
HLH HL ‖
μμμ μμμμ ˈμμμ ‖12.6cd [.........]śtĕ ttärä puñōndä khŏ ˈbalysä 25
μμμμμ μμ L ‖ LLL HLLˈHL ‖12.7ab nūvarā satva panata ma[.......]
HLH HL ‖ LLL
μμμμ ˈμμμ ‖12.7cd [..........] [...] mara [nĕ] kaˈsīndä
μμμμμ μμμ ‖μμμ LLLLˈHL ‖12.8ab ka vā mā käḍäna balysä närmätĕ ˈī[yä]
LHH LLL ‖ HL HLLˈHL ‖
12.8cd [..........] [.....]ndä duˈkhyau jsa ‖
μμμμμ μμμ ‖μμμ HLLˈHL ‖12.9ab jaḍa hāḍĕ hva’ṃdä cĕ bata bvāmata ˈhau[ta]
LLHL HL ‖ LLL HLLˈHL ‖
12.9cd [..........] [....]× mā ˈnäśtä ·
μμμμμ μμμ ‖μμμ μμ H|ˈHL ‖12.10ab ṣṣai prrārätä satvä paiya hvāñätä ˈdrūjŏ
HHL HL ‖ HL HLLˈHL ‖
12.10cd [..........] [....]× × nda × 8
μμμμμ μμμ ‖μμμ μμμμ ˈμμμ ‖12.11ab ṣä biśśu tta tta rraṣṭu khŏ burŏ mara hvatĕ [ˈyīndä]
LLLLL HL ‖ LLL LLLLˈHL ‖
1… was … May you make (it happen) that there is no being amongst them whose benefit matures [2]22[but who is not saved](?), whatever serious, terrible acts he may have committed.3Then here … there will be so many lord buddhas 234As then … he should hear this law. [2]45Therefore if many … he may be … thither at a future time.6Then he will strive so that … as meritorious as the buddha. 257The young beings rose up … here they will [not?] fall.8Would that the buddha may be conjured up for us … from woes!9But foolish men, whose knowledge (and) ability (are) small … is not …10Even a common being sometimes tells a lie … [2]811He [will do] everything rightly, exactly as he said here.
12.1. Since the agent noun yanāku (so read but left untranslated in SDTV3) does not seem to make sense, it seems better to interpret the sequence as yanā 2 sg. subj. mid. (a form attested in Z12.53) + ku.
The interpretation of satvä cĕ hāva paśtä as “being … whose benefit matures” implies two minor irregularities: hāva for nom. sg. hāvä and paśtä for paśtĕ, 3 sg. pres. mid. of pach- “to be cooked or refined, to become mature”. As an alternative Mauro Maggi suggests that hāva may be nom.-acc. pl., the object of paśtä, 3 sg. pres. act. of pajs- in the sense “to bring to maturity”.
12.3. Bailey, DKS 129b (s.v. ttiranda-) and 393b, lists vyata as a form of the verb “to be”. Skjærvø in SVK1: 50–51, shows that it is a misreading in the passage cited by Bailey. Nevertheless, here it is possible that we may have 3 sg. opt. viya with “intrusive -t-” (cf. above on 1.44). Unfortunately the following akṣaras cannot be interpreted.
12.10. prrārätä “common, ordinary” < Sanskrit prākṛta-. Cf. pratarā-, prrarā- “nature”, no doubt a loanword < Sanskrit prakṛti- (with Degener 1989: 317, 318) rather than a native Khotanese word as contemplated by Bailey, DKS 253–254.
See Canevascini (1993) and the earlier publications listed below in the bibliography. I take this opportunity to thank Ruixuan Chen and Mauro Maggi for a number of helpful suggestions, the most important of which will be acknowledged at the appropriate point.
Duan (2011); Maggi (2017b); Skjærvø (2002): 119–120 (Or. 12452D/14), 217 (IOL Khot 23/2), 231–232 (IOL Khot 28/9, which supplements 28/6), 232 (IOL Khot 28/10), 243 (IOL Khot 31/11), 313 (IOL Khot 115/6), 395–396 (IOL Khot 177/8, which supplements 177/7), 408–409 (IOL Khot 186/6), 427 (IOL Khot 192/1), 429 (IOL Khot 193/4), and probably some others listed in the index ibid., 608, under the heading “Saṅghāṭa-sūtra(?)”.
See for instance below on 6.5 and 9.2.
One Sanskrit verse is often translated by two in Khotanese. The occasional verse numbers in the Khotanese text often seem to refer to the underlying Sanskrit verses, see KS xvii.
See discussion below of passage no. 11 (Sgh 253).
See Maggi (2004), especially p. 186.
Note the following conventions employed below: H = heavy syllable, L = light syllable, L = heavy syllable with metrical lightening,
Sims-Williams (forthcoming a), text no. 4. Originally the series probably consisted of about 40 verses, of which some 24 are reasonably well preserved.
Sims-Williams (2022): 40, 45.
In Sgh 99 the metre C 7-mora cadences ending ˈLLL make up 20–30 % of the total (depending on how one scans a few ambiguous cases) as opposed to less than 2 % in Z. It is interesting to note that the parts of Sgh in metres A and B contain only two more clear instances of a 7-mora cadence ending in ˈLLL (below, 9.4b and 12.1b, both ending ˈvätä śtä). It seems therefore that the poet’s apparent predilection for such cadences was limited to metre C, perhaps because such cadences echo the usual LˈLL at the end of that metre’s 5- and 6-mora cadences. Cf. the strong association in Z between metre C and 7-mora cadences beginning with LH, for which see Sims-Williams (2022): 49–50.
For exceptions see Sims-Williams (2022): 64–72.
For a possible exception see the commentary to 9.9 below.
On this point see Sims-Williams (2022): 19.
Cf. Sims-Williams (2022): 66–67.
Cf. Sims-Williams (2022): 67–71.
In this table and the next cadences which are unclear or apparently corrupt are ignored. Percentages are given to the nearest whole number.
For example, the types of metrical lightening which I have named the “ttarandara-effect” (Sims-Williams 2022: 52–53) and the “uysnorāṇu-effect” (ibid., 56–58) are absent from the surviving Sgh fragments. However, the more common “uysnora-effect” (ibid., 51, 91–92; cf. Hitch 2014: 15–17) is attested at least twice (in 1.1d and 1.3b).
On the two forms see Emmerick and Maggi (1991): 69 and Sims-Williams (2022): 30.
Z has the HL ending ×32 (spelt -aimä, once -ēmä, Z21a.55, see Sims-Williams forthcoming b), LL ×7 (-ĕ mä, once -aìmä, Z22.293b, see Sims-Williams 2022: 30 n. 42). If one discounts jsatai (jsataì?) mä, 5.35c (where the scansion is ambiguous, see Sims-Williams forthcoming a), the metrical parts of the Sgh have the HL ending ×8 (-aimä, once -ēmä, 5.41d), LL ×8 (-ĕ mä or -ä mä, once -aìmä, 7.5b).
Here the parenthesis is used to mark off part of a word which does not form part of the segment under discussion. In Sims-Williams (2022) I used a square bracket for this purpose, but this can lead to ambiguity.
On the two types see Sims-Williams (2022): 55–59.
See von Hinüber (2021): lxviii with n. 151.
This is the number given in Skjærvø (2002): 381. The photo is found in the IDP database under the number IOL Khot 170/3.
A fragment from Khadalik in the Francke-Körber collection, Munich.
Suvarṇabhāsottama-sūtra 3.2a, see Leumann (1920): 57; Skjærvø (2004, vol. 1): 36.
IOL Khot 147/3, r3, see Leumann (1920): 107; Skjærvø (2002): 332.
Verse-numbers in [square brackets] are those of Canevascini’s edition.
Though he does not actually use these terms, referring rather to the “wechselvolle Rhythmengruppierung” resembling “eine rhythmisch ähnliche Variabilität der griechischen Lyrik”.
Leumann’s examples (here transposed into modern orthography and accompanied by Canevascini’s verse-numbers) include: ku yĕ ttuśāttētu dyĕ ˈyīndi LLLH HLLˈHL [71], ku yĕ thatō balysä paˈnamätĕ LLLH HLLˈLLL [76], sĕ āysda yanīru usˈkyāstu LHLL HLLˈHL [78], paṃjsa satĕ ysārĕ kuˈlāra HLLL HLLˈHL [79].
References
Bailey, Harold W. 1956. Indo-Scythian Studies, being Khotanese Texts, Volume III. Cambridge: Cambridge University Press.
Bailey, Harold W. 1963. Indo-Scythian Studies, being Khotanese Texts, Volume V. Cambridge: Cambridge University Press.
Bailey, Harold W. 1979. Dictionary of Khotan Saka. Cambridge: Cambridge University Press [= DKS].
Canevascini, Giotto. 1993. The Khotanese Saṅghāṭasūtra: A critical edition. Wiesbaden: Reichert [= KS].
Degener, Almut. 1989. Khotanische Suffixe. Stuttgart: Steiner.
Duan Qing. 2011. “Some fragments of the Saṅghāṭa-sūtra from the Xinjiang Museum, Urumqi.” Annual Report of the International Research Institute for Advanced Buddhology at Soka University 14: 127–134.
Emmerick, Ronald E. 1968. Saka Grammatical Studies. London: Oxford University Press [= SGS].
Emmerick, Ronald E. 1968a. The book of Zambasta. London: Oxford University Press [= Z].
Emmerick, Ronald E. 1970. The Khotanese Śūraṅgamasamādhisūtra. London: Oxford University Press.
Emmerick, Ronald E. 1991. “Khotanese ma ‘not’.” In Proceedings of the First European Conference of Iranian Studies held in Turin, September 7th–11th, 1987 by the Societas Iranologica Europaea, Part 1, edited by Gherardo Gnoli and Antonio Panaino, 95–113. Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1990 [1991].
Emmerick, Ronald E. 1995. “Khotanese bärätāndä.” In Proceedings of the Second European Conference of Iranian Studies held in Bamberg, 30th September to 4th October 1991, by the Societas Iranologica Europaea, edited by Bert G. Fragner et al., 163–167. Rome: Istituto Italiano per il Medio ed Estremo Oriente.
Emmerick, Ronald E., and Mauro Maggi. 1991. “Thoughts on Khotanese e and o.” In Corolla Iranica. Papers in honour of Prof. Dr. David Neil MacKenzie, edited by Ronald E. Emmerick and Dieter Weber, 67–73. Frankfurt: Peter Lang.
Emmerick, Ronald E., and Prods Oktor Skjærvø. 1982, 1987, 1997. Studies in the vocabulary of Khotanese, I, II, III. Vienna: Verlag der Österreichischen Akademie der Wissenschaften [= SVK1, 2, 3].
Emmerick, Ronald E., and Margarita I. Vorob’ëva-Desjatovskaja. 1993. Saka Documents Volume VII: The St. Petersburg collections. London: School of Oriental and African Studies.
Emmerick, Ronald E., and Margarita I. Vorob’ëva-Desjatovskaja. 1995. Saka Documents Text Volume III: The St. Petersburg collections. London: School of Oriental and African Studies [= SDTV3].
von Hinüber, Oskar. 1973. Das Saṃghāṭasūtra. Ausgabe und kommentierte Übersetzung eines nordbuddhistischen Lehrtextes in Sanskrit und Sakisch. Mainz, unpublished Habilitation thesis.
von Hinüber, Oskar. 2021. The Saṃghāṭasūtra. A Popular Devotional Buddhist Sanskrit Text, editio maior (Supplement to the Annual Report of The International Research Institute for Advanced Buddhology at Soka University 24). Tokyo.
Hitch, Doug. 2014. “Meter in the Old Khotanese Book of Zambasta.” Ars metrica 2014/11: 1–43.
Konow, Sten. 1932. Saka Studies. Oslo: Oslo Etnografiske Museum.
Leumann, Ernst. 1920. Buddhistische Literatur, nordarisch und deutsch, I: Nebenstücke. Leipzig: Brockhaus.
Leumann, Ernst. 1933–1936. Das nordarische (sakische) Lehrgedicht des Buddhismus. Aus dem Nachlaß herausgegeben von Manu Leumann. Leipzig: Deutsche morgenländische Gesellschaft, in Kommission bei F.A. Brockhaus.
Maggi, Mauro. 2004. “The manuscript T III S 16: Its importance for the history of Khotanese literature.” In Turfan revisited. The first century of research into the arts and culture of the Silk Road, edited by Desmond Durkin-Meisterernst et al., 184–190. Berlin: Dietrich Reimer Verlag.
Maggi, Mauro. 2017a. “Annotations on the Book of Zambasta, IV: Ronald E. Emmerick’s notes.” In Studia Philologica Iranica. Gherardo Gnoli memorial volume, edited by E. Morano et al., 273–291. Rome: Scienze e lettere.
Maggi, Mauro. 2017b. “Two fragments of the Khotanese Saṅghāṭasūtra in the Turfan Collection in Berlin.” In Zur lichten Heimat. Studien zu Manichäismus, Iranistik und Zentralasienkunde im Gedenken an Werner Sundermann, 371–379. Wiesbaden: Harrassowitz.
Maggi, Mauro. 2022. “Two Khotanese etymologies: parrama- and pūhei’tä”. In The Reward of the Righteous. Festschrift in honour of Almut Hintze, edited by A. Cantera et al., 321–333. Wiesbaden: Harrassowitz.
Sims-Williams, Nicholas. 2022. The Book of Zambasta: Metre and Stress in Old Khotanese. Wiesbaden: Reichert.
Sims-Williams, Nicholas. Forthcoming a. “Further texts in the Old Khotanese ‘metre C’.”
Sims-Williams, Nicholas. Forthcoming b. “A missing chapter of the Book of Zambasta.”
Skjærvø, Prods Oktor. 2002. Khotanese manuscripts from Chinese Turkestan in the British Library. A complete catalogue with texts and translations, by Prods Oktor Skjærvø with contributions by Ursula Sims-Williams. Reprinted with corrections 2003. London: British Library.
Skjærvø, Prods Oktor. 2004. This most excellent shine of gold, king of kings of sutras: The Khotanese Suvarṇabhāsottamasūtra, 2 vols. Cambridge MA: Department of Near Eastern Languages and Civilizations, Harvard University.