THE SOGDIAN FRAGMENTS OF LENINGRAD The Leningrad branch of the Oriental Institute of the Academy of Sciences of the U.S.S.R. possesses a collection of nearly 150 Sogdian fragments recovered from various sites in Central Asia between 1893 and 1915. During the first three decades of this century a few texts were published by F. Rosenberg and C. Salemann. Thereafter the collection was for a long time ignored, while Russian Sogdianists turned their attention to the important archive of Sogdian documents discovered in 1933 at Mt. Nugh in Tajikistan. Everyone concerned with the languages and culture of medieval Central Asia will be deeply grateful to A. N. Ragoza, whose efforts have finally resulted in the complete publication of these interesting fragments. Her edition, reviewed here, includes texts, translations, commentary, glossary, and 69 pages of facsimi1es.l Although the Leningrad collection does not equal in importance the large collections of Sogdian manuscripts in Paris, London, and Berlin, it is far more varied and extensive than one would have expected from the samples published up to now. Indeed, the true interest and value of these texts seems to have escaped their editor, who has in general been content to transcribe and translate whatever is reasonably legible and comprehensible, with little attempt to obtain a coherent sense from each fragment as a whole or to identify the literary genre to which it belongs. The editing of broken fragments such as these is an exacting task, requiring not only a thorough knowledge of the language and scripts in which they are written but also a certain amount of ingenuity in identifying a word from the outlines of a few ill-preserved letters, in capturing the sense of a passage from a series of disconnected phrases, or in recognizing the reciprocity of outline which may indicate that the left edge of one fragment can be rejoined to the right-hand edge of another. It must be said, with regret, that such qualities are rarely to be seen in the present edition. Many of the editor's interpretations are incompatible with well-established rules of Sogdian grammar and orthography. For instance, 'zucy (13.V1) cannot he compared with C. Sogd. 'wc', which derives from 'wrts'r, but is merely the end of some such word as /3ry'zucy ; 'skzu'tskwn (29.6) must be read 'skwntskzo~z, since -skwn is never used with the subjunctive ; w'p'y~t( 57.3) is not plural but elative; prm'y (57.8) is not 3 sg. impf. but 2 sg. impv. ' please ' ; Byr'nk'r" (58.1) is not a variant spelling of Byr'nk'r'k but its vocative; nyWrn' (58.2) cannot be 3 sg. impf. of 'ny'm (which would be mny'm) but should be read zy"m' (2 sg. subj.) ; M. qndyyt (74.V3) is not the pl. of qn8 'city ' but of qndy ' dug '. A recurrent quirk is the apparent assumption that pronouns can be Akademija Nauk SSSR. Institut vostokovedenija. Sogdijskie fragntenty cent~al'noaziatskogo sobranija Instituta vostohvedenija. Faksintile. Izdanie tekstov, Ctenie, perevod, predislovie, primehnija i glossarij A. AT. Ragoza. 183 pp. Moscow : Izdatel'stvo ' Nauka ', 1980. Rbls. 2.30. Although the edition is intended to be complete, at least two texts seem to have been over- looked : ' S 40 '.published in facsimile by Salemann, 'Manichaica I11', Izv. Imp. Akad. Nauk, 1912, 28-9, and 'Kr. IV Soghd. 3 ', from which Rosenberg, Izv. Akad. Nauk SSSR, 1927, 1398, cited the word ptz-m'ntyh. The reverses of two fragments, Nos. 63 and 72a, have been omitted from the facsimiles. Ragoza mostly refers to the fragments by their complex and inconvenient signatures. In the glossary she uses instead the running numbers 1-120, which represent their sequence in the edition, preceded by the abbreviation ' Fr.' (= ' Fragment '). Here the latter system has been preferred, but the insufficiently distinctive ' Fr.' has been replaced by ' L ' (= ' Leningrad '). This prefix is omitted for brevity where no loss of clarity is likely to result. suffixed to any word, regardless of its function or position in the sentence, 13') e.g. Pr'k-pn, 40.16 (read pr'k ; Prmn', 89.3 (read Prm'n) ; 'krty-in, 119.1 (read 'Arty' s'). The system of transliteration employed by Ragoza is somewhat old- fashioned, with NLPW for ILPtu, RNYW for 'N YW, rm for 'M, and no distinction between y and x, even in final position.2 The diacritic points under z and the disjoined z- (cf. IIJ, xx, 1978, 258) are only sporadically noted. In transliterating from Manichaean script the editor mostly writes t for _t (and vice versa), f for j, thus introducing new types of ambiguity into an already confused situation. Misreadings are very numerous. Typical examples are tmync for tmylz (8.R4), 'prtync for '(k)rtyh (8.R6), ]ry'n(c) for ](p)rynh (18.R2), pt'yrs'r for pt'ycs'r (24.2), yw'y for xwny (29.12)' w'nkw for wy'kw (40.14), nmy for rmy (43.5)' [plcm'k [my]'m'[n]ty for ](p)cnz'k(zt') cym'nty (51.6)' m'ts'r for mrts'r (52.2), pk'rn for pz'rn (52.6), zmnyyh for zmnwyh (52.9)' ywt'wn for xwt'w' (52.10)' cym'n8 for cym'y8 (82a.21)' Ing'n'k for n]y$'n'k (100b.3). Ragoza does not appear to be aware of the existence of special signs for numerals in Sogd. script. Thus, nw-100 '900 ', and (p)nc-100 ' 500 ' are both read p'c'yn (33.3, 8), 20 is read w (35b.4), 12 is read 8y (84a.4, 5)' and 24 is read [/// /// ///]y (84a.6). Easily recognizable words and phrases are often left unrestored, e.g. syltm'n (3510.2)' p[zoty] r'6wh (38.4-5)' [wzl'rkt myapt' (38.25)' "Iz'wn "z'yth (59.1), qltyprykt (74.R5), (p)r?[y]'m (84a.2), 'wyn my'zun '8(r)y i[k'n](p) [pnc "z'tun w't8'r c'wn] 'pygyh wx'rit k'm (84b.4-5, cf. VJ 1188-90, etc.), wylx Fyz'k (85.10)' mrltxm'k (lOl.Rl), w'tS['r (109.3). This may be symptomatic of a lack of interest in understanding the meaning (as opposed to the linguistic features) of the texts. Such an attitude is evinced also by the imprecision with which the editor identifies the contents of the individual fragments merely as ' Buddhist ' or ' Manichaean '-minimal definitions which are often assigned without any discernible reason and sometimes clearly erroneously. Many of these texts are far less fragmentary than would appear froin Ragoza's presentation. For one thing, she often places points or square brackets at the beginnings or ends of the lines, even where part of the margin is in fact preserved (e.g. L23, 26, 32, 40, 51). If parts of both side margins are preserved, as in the case of L32, this careless habit results in the dismemberment of a piece of continuous text and the loss of the opportunity to interpret it as a whole. For another, she appears to have made no attempt to create larger units of text by rejoining fragments belonging to the same MS. In many cases it is quite easy to see how this should be done, even with only the facsimiles to work from, and access to the originals would make the task still easier. For .. example, L95 and L98a can be joined in such a way that 98a.2 = 95.3 (. p't'k 's[kw](y) ZY MN 'tsy'kyh ZKw), 98a.3 =95.4 (. . . pwrtl'rmykw ZY ZKwy . . x'n'kyh 'z-y'm't w'Py8), 98a.4 =95.5 (. ZY ZKw yrp'kw Byrxwz-'y). L99a and 99c combine to give the following text : (1)] X N [ ] . .[ ] .. [ (2) ]h 'krt-'sp's [ (3) (?)p']tp'B 'zt'yr'yw (y)[ (4) ](r)ty 'prtm[. Here *p'tp'i 'respectful ' (otherwise unattested) m-ould correspond to p'i-p'y'k (attested in a similar context in P2.1180-81) in the same way as 'krt'sp's ' obedient ' (cf. Dhy. 7) to *'sp'skryy (cf. GMS $1044). L91, 97b, and 97c can also be joined, as the 2 My own preference is to distinguish y and x in all positions, cf. my appendix to W. Sunder- mann, JIitteliraniache Texte kirchengeschichtlichen Inhalts der Hanichaer (Berliner Turfantexte, forthcoming). THE SOGDIAK FRAGMENTS OF LEXINGRAU 233 . . . following excerpts illustrate : (91.5) zwxiky (91.7-8) i'tw(x) Byr-/3(r'n) /3wt + cywys py6'r 6'tPr 'yi o . . . (91.9 97c.0) 6/3(t)[y](k) [zwx](i)kt' . . . (97c.l) . . . + 'wr'm wrm[yh] (97b.2) wrcy' 'skwnt (97c.2) 'nc'yt knpy-(97b.3 97c.3)-str ,6wt MN tw' wz-p[yh] cywy8 py6'r tyw . . . '. . . the disciple . . . becomes happy and contented. For this reason you are a judge. . . . Secondly, the disciples . . . are calm and peaceful in quietness (cf. M178.42-3, Henning, BSOAS, XII, 2, 1948, 307) . . . ceases and becomes less from awe of you. For this reason you . . .,. In view of the deficiencies of the editing it is fortunate indeed that facsimiles have been provided by means of which it is possible to check the editor's work to a certain extent. However, the quality of the facsimiles is in general rather poor, and many of the larger fragments have been so drastically reduced in scale as to become virtually illegible. Plentiful substantiation of all these various criticisms will be found implicit in what follows. It would be ungenerous, however, to continue to draw attention to every defect of Ragoza's edition. The remainder of this article is intended to provide a summary description of the most important and interesting texts in the Leningrad collection, principally those whose contents can be defined with some precision. Most of the information to be given here is not to be found in the edition, but derives from an independent examination of the facsimiles. Nevertheless, the most essential corrections to Ragoza's readings and interpretations will be signalled in passing. Of the Buddhist texts now edited by Ragoza, three were first published by Rosenberg in Izv. Akad. Nauk, 1918, 817-42 ; 1920, 399-420 and 455-74; 1927, 1375-98. The first two (01 =L92 and O2 =L93 respectively) contain parts of the tale of the farmer and the fisherman, a further fragment of which is found in the Stein collection in the British Library (Fragment IIa), and of the Sukas~tra. The third (Kr. IV Soghd. 4 =L71) belongs to a work which is referred to in the text itself as snk'swtr. Rosenberg assumed, not unreasonably, that this spelling stands for ' Sanghasiitra ', but was unable to trace such a work in Indian or Chinese sources. In this case, the newly-published fragments are capable of providing the decisive solution to a long-standing problem. In addition to L71, Ragoza has recognized two more fragments (L8 and 39) as containing the title snk'swtr. L8 is part of a pothi, of which little is preserved beyond the marginal title and foliation. (The last word is to be restored as ptt[r] ' folio(s)', but the preceding numeral is not legible from the published facsimile.) L39, which belongs to a scroll, is more informative, since it contains part of a conversation between the Buddha and p't[sy](n) p(wty)[stp] ' the Bodhisattva P'tsyn ' (line 9). This name (on which see below) recurs in line 5 of L38, a fragment which not only belongs to the same MS as L39 but can be rejoined to it so as to form a continuous text of 40 lines, many of them complete or nearly complete. (39.11 and 38.1 together make up a single line : . . (ZKzu ,8)[y'n](,6xtm pwty K)ZNH (ptyikwy)[ .) Another substantial portion (73 lines) of this ATS can be obtained by joining L41, 84a, 82b, 86, 83, 71, 84b, and 82a. (41.7 =84a.l ; 84a.7-10 =8213.1-4 ; 84a.11-12 = 86.1-2 ; 53.1-3 -=86.7+-9 ; the tips of the final n-s in 71.1 are visible on the bottom edge of 86 ; 71.14 84b.l together form a line which may tentatively be read as 'wyn xzut7tuy m'6 (zu'p'nt) (tyzu xwt'w' ")6(prm )L(')[y](r,6)'(y)[ . . . ; 84b.6 = 82a.l.) The text thus reconstituted continues the story begun at 38.6, which tells of a man whose newly-planted tree grew roots, branches, leaves, flowers, and fruit within the space of a single day. 234 NICHOLAS SIMY-WILLIblfS Fronl nly account of the contents of these fragments, Professor Emmerick was able to recognize immediately that they belong to a Sogd. version of the SailghEtaszitra. This work is already known in Sanskrit and Tibetan (both edited by R. A. Gunatilaka, An edition zuith translation of the Buddhist Sanskrit text ' Sa?ngh@a-siitra', unpublished Ph.D. thesis, Cambridge, 1967), in Khotanese (see R. E. Emmerick, A guide to the literature of Khotan, Tolryo, 1979, 28-9), and in txo Chinese versions (l'aishii Issaikyii, XIII, Pios. 423 and 424). The Sogdian does not agree closely with any of these texts and must be regarded either as a free adaptation or as a translation from a source as yet unknown. In general, the fragments described above correspond to the Slit. + and Chin. texts as follows: L39 38 = Slit. [Gilgit MS 371 ff. 74-6 =Chin. [T. 1.4231 pp. 970a, [T. 1.4241 989b-990a ; L41, etc. = Skt. ff. 76-81 = Chin. P~J9.70b-c, 990a-991a. The lack of the syllable -?a- in the Sogd. transcription of Sanghtltaszitra could be explained in two different ways. Either sanghtita was understood as equivalent to saitgha (cf. F. Edgerton, Buddhist Hybrid Sanskrit dictionary, 549 ; Gunatilalia, op. cit., xlvi-li) or "snk'tswtr was simplified in Sogdian for purely phonetic reasons. In the latter case one may compare the name of the Bodhisattva p'tsyn, a transcription which now proves to represent Skt. Bhaisajyasena. An intermediate form is attested in P8.11, where p'8tsyl& ' Bhaisajyasena ' is named immediately after sr,68zor ' Sarvaiiira ', these two Rodhisatt~yasb eing the Buddha's chief interlocutors in the SafighEtasCtra. It is worth noting in passing that these fragments enable one to correct the translation of p'n'w6'k in Dhu. 39, 164, hitherto wrongly interpreted as custom, manner ' on the basis of the (still obscure) gloss p'nza6yi~ "yfz ' N (Henning, Sogdica, a 17). Spelled p7nw6'y or p7nw6'k, the word occurs four times in the SaizghcT?as4tra (L38.11, 18, 20 ; 53.1), clearly denoting some part of a tree. In Dhzl. 39, p'n'zct'3'k is still more specifically part of the seed. In every case it is the subject of the verb rw6 'grow ' or (L38.20) sn- ' rise '. It follows, I believe, that p'n'zu6'k must mean sprout, shoot '. ' Further fragments of the Sogd. version of the SarighEtasdtra are to be found in the German Turfan collection. The title snk'swtr occurs in the text of Tiiig 313, cited by Hansen, ZD-MG, XCIV, 1940,149 n. 1, 155, and (as TiiiS 313) by Gershevitch, ~ 35 6404l ~and pass in^. Dr. Gershevitch draws my attention also to the following four fragments which contain the name sr/38wr and are therefore likely to belong to this szitra : TiiS 23 : ~iiig23 (2) (two pieces) ; ~iiig23 (3). Some of the remaining Buddhist Sogd. fragments in the Leningrad collection may be described more briefly. L11, in which the Buddha converses with Subhiiti, would appear to be connected in some way with the VajracchedikE. However, since it contains the term 'Pratyekabuddha ', mhich is not found in Vaj., it is more likely to belong to a commentary on that 1%-ork( cf. Sunder- mann, BSOAS, XL. 3, 1977, 634b) than to a translation. LlOOc, 96, and lOOa contain part of a text concerning stiipas ('pw'6'lz, cf. D. N. MacKenzie, The S~troaf tltc causes and efects of actions ' in Sogdian, 37). The three fragments can be joined together in such a way that 100c.12 == 96.1 and 100a.l reads ](p)~~t-'p~'6'l2it1iilk 'np ' Buddha-stiipa (Chin. W f$ <fo t'a, cf. MacKenzie, The Buddhist Sogdian texts of the Byitish Library, 11, 164), eleven storeys ', with the tips of the 6 and final p visible on the bottom edge of L96. 11:ut the heg~nn~nogf L39 scellis to col~cspondtu a niuch cather passage 111 tllc Sht. tcxt (f. 71). indicating a lnajor onlission or ~c-nrrnngemmotf material in tho Sogd. version. THE SOGDIbN FRAGXENTS OF LENISGRAD 235 Another group of fragments belonging to a single NS consists of L40, 49, 50, 52, 81, and 89 (perhaps also 35a and 35b). This text, which should surely be identifiable, seems to recount a disputation between the heretic Upaka ('wp'k', 40.3, 10) and the Buddha, who is several times referred to as ' Gautama the Bramapa (k'ul8'm imny, 49.1, 3 ; 81.9) or as ' g8kyaputra ' (B'k(y-z)t'h., ' 52.16). The setting for this encounter is the city of RLjagrha (r'ckry knsyh, 81.14) in Magadha (rn'kt 'zut'kh, 89.5), whose ruler King Biinbistira (pyfns'r xwt'zci, 40.14 ; 49.8 ; 81.14) also plays a part in the story. Leaving the Buddhist texts, we may next consider L47 and 48, which contain a medical text of Indian type, apparently a recipe for a remedy against piles ('rs'x, 48.7). The tm-o fragments are the left and right portions of the same sheet. They do not join directly, but may be put together so that 48.1 =47.3, etc. This reconstruction, suggested in the first place by the shapes of the fragments (which are almost mirror-images of one another, as a result of the way in which the 1IS has been torn while folded), is confirmed by the convincing sense which the consecutive ends and beginnings of lines + combine to give: '8w nym'k s'nk "ph (47.2 48.1) ' 29 gang (= Khot. ,sa%a, cf. Emmerick, BS4OAS, XLII, 2, 1979, 239 ff.) of water ' ; nlym'k+ ink prxs't (47.3) 'boil until Bang remains ' ; 'yw pty'p nym'b 'xiph (47.4 48.3) ' he sl~oulclt ake one part at midnight ' ; 'yzo (p)ty'p xw(r)[ .. . (47.5 +48.4) ' one part at sun-set(?); 'yw my8 xwr't ZY ['Iyw my8 'nc'yt xwrt (47.6 f 48.5) ' for one day he should take nourishment and for one day he should cease to take nourishment ' ; . . . xwr't L' zwtk xzor't L' . . . (47.7 +48.6) ' he should ?lot consume . .., nor beer, nor . ..' ; x'm'k zrywn (47.8) ' ram vegetables '. This inedical text is the only identifiably secular document in the collection, unless one counts as such the quite numerous fragnients containing the honorific introductory formulae of letters, mostly written as scribal exercises and mostly by Manichaeans (L27, 44, 63, 111 Verso, 118). L27 is written in a particularly repulsive, unpractised cursive. With all due reserve, the following interpretation inay be proposed: . . . ,Byzo xwt'tc R(Bkw 'nwth) ZY (y)r'n pst'(t c)nn (wys)pn7cw" 6'y 'yzo (yzo)'ncykstr ZY 'yto '(xin)kystr pr '(zw)'nty 6'm RBkw 'nzoth ZY y(r'n)(?) pst't cnn '8w cy~z' 'xinky-st(r) . . . ' To the noble lord, (my) great support and mighty defence (= 11. and C. ptst't), most necessary and most excellent of everyone in the living world (cf. P2.565)' great support and mighty defence, more excellent than (the sight of my) two eyes . ..'. L44 is clearly not an exercise but the beginning of a genuine letter sent by a queen (xCit.iin) of Argi to a Manichaean ' Teacher ' (9nwcE). The style of address is exceedingly courteous, with the unfortunate result that in all its 17 lines the writer hardly progresses beyond the conventional preliminaries. The letter describes itself by the old-fashioned word Gykh, and its phraseology strikingly resembles that of the ' Ancient Letters ' and of the more forinal and archaic of the Mugh letters (especially B16). For instance, the phrase ' greatest of all supports ' (line 2) occurs also in B16.1. The sentence (11. 9-11) ' And, sir, if you there are well and at ease, sir, in high fortune and great honour (restore ptpyw), (then) we consider ourselves immortal(?)' inay be compared with B16.4-6 and Nov. 2, R5-6, 1Iost remarkable of all is the preceding sentence (11.8-9) : 'We pay homage (to you) just as (cf. GMS $1592) one pays homage to the Buddhas of the various periods '. This is allnost identical to the formula of the 'Ancient Letters ' (e.g. AL 5.4 : ' (when) I inay pay hoinage froin nearby, as to the gods '), but with Py'nzo (evidently understood a.; ' gods ' rather than as 'His Najestv the King ', cf. Henning, ZDMG, xc. 1936, 198) 236 XICHOLIS SINS-WILLIAMS replaced by ~wnzy-zwrny-cy-kt pzut'yity (cf. 11. zwrnyy z(wr)[nyy plwiyit, Sogdica, p. 19 ; 3-zwmycyki pnc bzoiyi~,H enning, BSOS, VIII, 2-3, 1936, 586 ; S. wyspw Z-wmycyktwp (zu)[t'yit, L106.3, on which see below). Several individual words in this letter are worth of comment. The un- explained zuyikyr in line 3 must be a verbal noun from zuyikyr ' to open (the eyes) '. Its meaning ~villb e similar to that of C. 'zolp'r-cmpy' (cf. IllacKenzie, BSOAS, XXXIII, 1, 1970, 116-17)' which in an unpublished fragment of the JIS C14 translates Syr. prhsy' '.rrappqola, freedom of speech, confidence ', b-prhsy' ' openly, publicly, in person '. The phrase may perhaps be para- phrased : ' seeing you and meeting you face to face is more necessary to me than every other sight(?) '. Similar feelings are expressed, in equally hyperbolical language, in many of the ' Ancient Letters '. For pt,ByScyw in the same line one may read ptp'ys'cnw, cf. ptpy-'icy'n'k ' worthy of honour ' (Dlty. 14).4 In line 6, Byr-"y6ch cannot be the sender's name-this is excluded by the word- order- but is a synonym of the expected *lyr-xwzch ' well-wisher, friend '. The text on the reverse of Llll is obviously the exercise of an unskilled writer : 'tkw [prylzu 0%)-dnkyp 'ic'n pr't i(my)wn s'r MN 6zor z'yh .. . ' To the dear, excellent, respected brother Simoa, from a distant place (cf. L44.7) . . .'. The name Semy6n (Syr. irn'wn), attested in just this spelling in a C. Sogd. text in Sogd. script (IIJ, XVIII, 1976, 63, Fragment 18, line l), is one of the few Semitic names known to have been used by the Manichaeans in Central Ssia (Jywywn, M1.83). Another favourite Manichaean scribal exercise was the copying of lists of names, parts of the body, doctrinal points, etc. (see Henning, Sogdica, Chapter 1, also Fragments 17 and 28 in IIJ, XVIII, 1976, 62-3 and 71). The Leningrad collection includes two such texts, which are easily recognizable by the roughly vertical lines which separate the columns of writing. L105 is a list of disabled persons: kzvr mrty ' blind man ', k(m) mrty ' deaf man ', (k')t mrty ' dumb man ', etc. L33 belongs to a doctrinal text. The surviving fragment includes the enumeration of four (out of an original five) hells corre- sponding to the five ' Dark Elements ' or ' 'CVorlds ' into which, according to Mani, the Realm of Darkness is divided : [cnn] zu'tny tm' 'from the hell of Wind ', cnn ["Itr-(n)'y (t)m' ' from the hell of Fire ', cnn "pny tm' ' from the hell of Water ', cnn myyny tm' ' from the hell of Fog ' (= C. myy, cf. Asia Major, XVIII, 1, 1973, 99, here corresponding to al-dab& ' fog ' in the Arabic sources). What it is that derives from each of these hells cannot at present be deciphered. The Leningrad collection is particularly rich in fragments of tales and parables, most of which seem likely to be of Manichaean origin. L103, which is complete in itself, being an exercise written on both sides of a scrap torn from a Chinese scroll, deserves to be mentioned first, since it attests the previously unknown phrase for ' Once upon a time ' : pr ny'wr zmnw, literally ' au temps d'autrefois '. The story begins : ' Once upon a time there were two thieves, and they were very bold.. .. One day they set out and went to a distant land . ..'. L120, on the other hand, contains the end of a tale, with the caption (line 6) : ' This is the parable ("2-'nt) concerning the greedy Hearers ', followed by the single word c'y6'n. This is probably to be interpreted as the Turkish (originally Chinese) term fayclan ' Bema ' (also attested, as j'yd'nyy, in a semi- Sogdian caption to a Parthian text, cf. Henning, JRAS, 1945, 155), which ]night be taken as an indication of the occasion on which the parable was to 'See further S. Sims-Williams ancl H. Halen, ' The Middle Iranian fragments in Sogdian script from the Rlannerheim collection ', Studia Orientalia LI, 13, Helsinki, 1980, '7. THE SOGDIAN FRAGMENTS OF LENINGRAD 237 be read. (For the liturgical employment of stories or parables cf. Henning, loc. cit., and BBB, p. 47.) Of the parable itself only the last few words are preserved: '. .. they do not obtain knowledge and wisdom (yr/3(')[kyh Z]Y '2-nh), (their) fate becomes (wp't or w/3nt) (to dwell) with the evil carnivorous (y't-xw'rt) beasts continually for eternal ages '. Most of these stories are too incomplete for a coherent narrative to emerge. L32 may belong to a fable in which a cat gives advice : tw' "2-wnt Gykh myn'mk7m ' I shall think about your children's letter ' (?). L29 (and Llll Recto, which is another copy of L29.2-5) contains a story about a gardener (/3wGstnpy, line 20) and a parrot (twty mryy). The extant fragment recounts how the tree in which the parrot has its nest ('t$'r?) becomes deprived of water and sheds its leaves ; as a result the parrot is forced to move with its young to another tree. The significance of the fact that the parrot ' knows the language of men ' would no doubt have become clear as the story progressed. L37 contains part of a folktale of a well-known type, in which an old man (m'sy) overcomes a series of human and animal foes, in each case by sheer luck. Ragoza justly conipares the tale No. 28 in Jagnobskie teksty. Much closer in detail are the South Indian and Sinhalese versions published by S. 11.Kat&Sa Sbstri, Folklore in Southern India, Bombay, 1884-8, 104-15, and by A. E. R. Corea, ' Sinhalese Folklore ', The Orientalist, 11, 1885-6, 102-6, which enable one to reconstruct the plot of the incomplete Sogdian tale. At the beginning of the Tamil version the hero sets out on a journey with a parcel of poisoned cakes which his jealous wife has prepared as a gift for her co-wife. d band of robbers eat the cakes while the hero is asleep and he takes the credit for killing them and for rescuing a princess whom they had abducted. This seems to be the incident recounted in L61 (which clearly belongs to the same ASS as L37), cf. especially ll. 3-5 : ' the thieves (t'gt) . . . at the "sweetmeats (pzuitw?) . .. and when they ate ((xwr')[nt) the "sweetmeats .. . they all died on the spot '. Of the next episode the Sogdian version is comparatively well preserved. The old man, having won a reputation as a warrior, is armed with a two-edged ' sword ' (GPG'r xnyr, 61.16) and sent to dispose of a troublesome lion. The lion forces him to take refuge in a tree, but fortune smiles on him once again : ' (The lion) placed its two (fore)paws on the tree and "opened (wy's or wy'i, cf. Khot. biyEid-2) its mouth upwards ; the old man's hands trembled (Grpt, cf. Henning, BSOAS, XI, 1, 1943, 68, G 6) from fear and he dropped the sword out of his hands ; it fell down (c'Gr-s'r) and entered (tys) the lion's niouth and "throat, and the lion died ' (37.4-8). d third adventure follows which begins with the old nian having himself tied on to the back of a horse and setting out to battle. The Sogd. text breaks off at this point, but the outconie is known from the Tamil and Sinhalese versions, which recount how the horse gallops under a branch, the hero finds himself clutching the uprooted tree, and the enemy army retreats in panic at the sight of a tree approaching. In view of the great distance which divides them, both geographically and chronologically, it is remarkable that the medieval Sogdian and the modern South Indian versions of this tale agree so closely." I Kot certain, possibly k'r'm = C. qrm ' karma ', used as a technical term in an anti-JIani- chaoan polemic, C3 = TiiB 8, R14. This ITS is written in a rather difficult cursive script and therc are many misreadings in the edition. The following corrections (in addition to those mentioned in passing above) may be worth listing : wy'k wy'ky (61.7) ; wy'kw (61.9) ; r(ty)pcp'ty ' suddenly ' (37.1-2) ; trb ' fled ', of. IIJ, xwr, 1976, 58 (37.3); 'tpn (37.10) ; 'xws'nt (37.12); rn]twxcn'k (37.14); c'nkw '2-zu ZKwh s'ny wy'k wyn'n rty "r'k wP('mk)'m ' when I see the ene~np'sp osition I shall go berserk ' (37.16-17). 238 SICHOLAS SI~IS-\VILLIAJI+ The lllost nearly complete of these tales is that of the child who pretends to be deaf and dumb (krn k't), of which at least four fragments survive (L69, 60, 87, and 68, to be read in that order). L69 recounts how a mother on her death- bed admonishes her son (11. 12-13) : ' Make (kw[n']) yourself deaf and dumb, zvithout speech and without hearing ! '. After her death her husband takes another wife, who, after bearing seven sons of her own, begins to hate her deaf and dumb stepson and to plot to kill him. The boy, however, having become aware of her evil intentions (as a result of overhearing an unguarded remark by his stepmother, hence as an indirect result of the ruse enjoined upon him by his mother ?), mixes the poisoned food which she sets before him with that intended for his stepbrothers and takes an uncontaminated portion for himself. The story is continued immediately by LGO-it is probable that not even one line is missing here-which describes the death of the seven brothers and their mother's grief. The last line of L60 begins with the words 'BYzo prw tk'zod "yt ' their father came to look ', from which the bottoms of several letters are visible along the top edge of L87. At 87.3 one reads : rty 'xww w' (z')ky Sstzu (ny)'s ' He took the boy's hand . . .'. At this point a substantial lacuna robs us of the denouement. L68 contains the epirnythion : ' Lord Mar Mani said to the magus (w~w(yw):) I, together " with my disciples and Electi, am like that child (z'kw) who was silent as an expedient (knph [xwp](y)h, cf. line 7, also knph in 69.6, 24), (who) did not . .. speak and did not hear. So we too are silent (dwkw), and we speak with no one and perform good deeds and pious actions as an expedient, (but) that time will come at last when I shall speak before all, like that child (z'kzu), and we shall demand justice for ourselves ". Then the magus said : Now let me go " before the chief magus and tell this (yzony) so great word before the chief magus ! ". Lord Mar Mani said : Listen, 0 magus, let me tell a parable " '. " There follows (68.12 ff.) the beginning of a fable about a wolf who lost his way in the forest (mryy kysny, line 13, cf. P2.811), a sheep (myd, line 14), and a fox. In a case such as this, where the stories are provided with a historical or legendary framework, the genre of tales and parables merges into that of church history. The four fragments just described probably belong to a single scroll (though this is not quite certain in the case of L68). A further fragment, L83a, which cannot (if its dimensions are correctly stated by Ragoza) belong to the same JIS, nevertheless appears to contain part of the same text and to have been copied by the same scribe. A distinctive spelling is xzonwx in 68.8, 9, 69.16 (sic), and 83a.2. The numerous instances of dittography are also characteristic: zwzdkt', 68.3-4 ; ZY ms pwtk'm, 68.7-8 : pyrnm (sic), 68.8-9 ; my', 69.14 ; w7/3ymp ts'r, 83a.8. L83a contains the continuation of hlani's conversation with the magus: ' Why do you say (w'pysk) ' Come to Lord Ptw ' ('ys 'kw pyw pt(w)[ s'r]) ? " (For) you both are my enemies (mn' s'nt 'nsS'), one another's . . . accomplices in falsehood, and you are evildoers (ps'tkr'yt) and false witnesses (~ymz~~ycu".~ yt) Again the magus (mzoyy) said : You(?) ~i~inlolt go to Lord Ptw ; but, come " ('ys), let us go to the palace gate 8 and tell this word (cf. Schwartz, Altorientalisch~ The d~sh1x 1 question is called yrn~'yclz,perhaps = KP r,i (Asadi, Luyat-i furs, etl. 'AbbLs Iclb51, Tehran, 1940, 521, s.v. d?s)ai,d by IqbB1 to be a food made of cooked millet. 8 cf. (yr'k 68r- ' id.', Padm. 3, considered erroneous by MacKenzie. Possibly @r'k(k) is an adjectival derivative of TVRIIr. s'ahr (s'ir) ' realm ', in which case the phrase may be compared with Syriac tr' mlkwt' ' gate of the kingdom ' = ' court, palace ' (a meaning very appropriate to the context in L83a). THE SOGDIAN FRAC-IIENTS OF LEKIKGRSD 239 ". Porschungen, I, 1974, 261) before the S'ni'y Then Lord filar Mani said. " Blessed ("pryty) and praised be the S'ns'y . . . ! " '. The identity of ' Lord P ~ 'T is unfortunately unclear. A possibility would be to read the name as Bat and to equate it with the Coptic baa^, Parth. b't, the name of the vassal king who travelled with Mani on his last journey to Belapat at the command of Bahranl I (cf. Henning, BSOAS, x, 4,1942, 941 ff.). In that case, the present fragments must be taken to recount the events which led up to Bat's conversion, which probably took place about three years before Mani's final audience with Bahram (cf. H. J. Polotsky, Manich~ischeH o~nilien. Stuttgart, 1934, 46, line 13). The title S'ni'y, which was previously taken to be a personal name, has been identified by Dr. JV. Sundermann in a forthcoming article I\-hose contents I do not wish to anticipate. It occurs again in L59, which also mentions the Light Elements (mr8'sp'nt py'ydt), the Father of Greatness (zrzr' pyy), and the Primal Man ("]66[j31y, cf. Sundermann, Altorientalische Porschu~zyen,T I, 1979. 129, n. 217). Ragoza misleadingly translates these last two terms as ' Brahmii ' and ' 6akra ', as if this were a Buddhist text. Incidentally, L59 is an interesting example of a clearly Manichaean text written in the script often referred to as ' Buddhist ' or ' Sfitra script '. Its existence lends support to the preference expressed in IIJ, X~III1, 976, 44, for the neutral term ' formal script '. The texts in Nanichaean script (L3, 72-5, 77, and 117) include three scraps (72a, 72c-both published previously by Salemann-and 74) which appear to relate to calendars and calendrical calculations. The most interesting of these is L74, which contains the new word 'xdnyt (V6), possibly ' moments ' (borrowed via Parthian from Skt. ksana-). In the preceding line Ragoza reads 6's'-zny'il. assuming that 8's' is a variant of 6s' ' 10 '. Although this is obviously impossible, a numeral is clearly required. From the very poor facsimile the most likely readings would seem to be 'MP'I ' 526 ' or 'MPP'I ' 546 '. If the latter should happen to be correct, it may not be coincidental that 546 ii the number of years which elapsed between the birth of Mani and the conversion to 3Iani- chaeism of the Uyghur Khan (cf. F. TV. K. Killer, Ein Doppelblatt aus ein~nl manichciischen Hymnenbuch (Mahrndmag), APATV, 1912, 36-7). Amongst the Manichaean fragments in Sogd. script are at least four (L34, 65, 80, and 106) which belong to confessional texts, some of them to the Sogdian original of the famous Uyghur Xwiistwiinift. L34 contains the beginning of Xw. X and is thus a variant to the fragment published in Soyclica, 66-7 ; in line 1the chapter is introduced by the phrase ](m)stym (6s?n)[yk ' Furthermore, tenthly '. LSO corresponds to Xw. XIVB-XVB, the only substantial divergence between the Turkish and Sogdian versions being in line 7 (line 8 may be read ]'(/3)z-'ykh zoy'kh pr['yst in full accord with the Uyghur yaj3laq yiryririi barir ' goes to an evil place '). L106 is too small to be identified with certainty, but may possibly correspond to Xw. IIIC-IVB. If so, the phrase zoyspw z-zurnycgktw p(w)[t7yJt'the Buddhas of all periods ' (cf. abore, pp. 235-6) clarifies the ineani~lg of soki in the equivalent Turk. phrase soki tuyri yala/3aEi' burxanlar (on r~hich see W. Bang, Le Musion, xxxvr, 1923, 192-3, and J. P. Asmussen, X71dstciinvt: Studies in Manichaeism, Copenhagen, 1965, 216). As the last but by no means the least of the Manichaean texts of the Leningrad collection one may mention L67, which contains several stanzas of a hymn, each beginning with the words zzukyh /3't ZKn Sln'xzo ' Hail be to you '. The hymn in question is in fact part of the hymn-cycle GGzviSn ;iy Grizu-zindag ' The utterance of the Living Self ', originally composed in Middle Persian, of which the German Turfan collection includes numerous fragments. 24 0 THE SOGDIAN FRAGMESTS OF LENINGRAD The Sogd . version is best preserved in a scroll to which belong the sizeable pieces So. 13399, 13401 [=TiiD], and 14790-92 [=TiiT 10. 1-31 . To this group one may now add the Leningrad fragment, which not only forms part of the same 31s but joins the top of So . 13401 and helps to complete its first verse : [ztu](ky)h,6't ZKn Bm7xtu . . [ Iy7ntyo o ky ZY c('m)'k p(y)G7r p(...y)G (Z)Y '13z.y/3r7nt 'skw8' 'Hail be to you, 0 ...,who ...and suffer for my sake '.g IXDEXOF FRAGNENTS DISCUSSED L~OOlo~o.c......... p.234 L103 ...............p.236 LLL111001561 R.. .........................................ppp...222333697 LL1l1l8lV .. ..........................pp..223365 L120 ............pp.236-7 . So 13401 is kilown to me from a photograph in the collection of Hamburg University. Q Seminar fiir Geschichte und Kultur des Vorderen Orients. Abteilung 1ranistik.-Further correc- tions to the printed text of L67 : prpyrnyfy (line 3). prxiynyt oo (line 6). prxi8'skwn (line 7).