Sogdian documents from Khotan, I: four economic documents.

Bo, Bi ; Sims-Williams, Nicholas

INTRODUCTION

In the spring of 2010 the Museum of Renmin University of China acquired a collection of documents from Hetian (Ancient Khotan), Xinjiang. The main part of the collection is a group of manuscripts written in Chinese, Khotanese, Sogdian, and Tibetan, including both religious texts and secular documents such as official documents and economic texts. The present authors were allocated the task of editing and studying the Sogdian part of the collection, which amounts to twelve items, making it one of the more significant Sogdian finds to come from China in recent years. Four of these texts are economic documents and are published in the present article; the rest, apart from a few fragments which are too short or unclear to be identified, are all letters. (1)

Only a few Sogdian documents from the Khotan area have previously been made known. Most of these are now in the British Library and were published in Sims-Williams 1976 with facsimiles in Sims-Williams and Hamilton 1990 (henceforth DTS). Those referred to in those publications as Fragments 12, 15, 16, 23, 27, 30, and 33 belong to the Stein collection and come from Mazar Tagh. Fragment 36 (H.143 MNS 18 = IOL Khot 158/5) belongs to the Hoernle collection. There seems to be no evidence about the exact findspot of this text, but it is certain that it comes from the Khotan region; (2) moreover, according to Yutaka Yoshida's recent reading, Khotan is even mentioned in the text. (3) Yet another Sogdian fragment from this area survives only in the form of a photo in the Trinkler collection, Bremen. This fragment was published by Gropp, (4) who tentatively suggested that it may come from Mazar Tagh. (5) However, since Trinkler only mentions Brahmi and Tibetan documents from Mazar Tagh, it is more likely that the Sogdian fragment is to be identified with the so-called "Uighur" fragment found by Trinkler at Dandan Uiliq. (6) Finally, there is also a small Sogdian fragment amongst a collection of more than five hundred documents and other items from the Khotan area recently acquired by the National Library of China in Beijing. (7)

[FIGURE 1 OMITTED]

Amongst these texts, the most relevant to the new documents published here is British Library Fragment 33 (see the appendix at the end of this article), a fragment of a commercial document similar to document no. 4 below. The surviving part consists largely of numerals and the repeated word pr'[delta] t 'sold'.

Unfortunately the findspots of the new documents are not recorded and the texts themselves provide no clear indications of their place or date of writing. However, a significant Sogdian presence at sites such as Mazar Tagh, Dandan Uiliq, and Old Domoko is clear from references in the Khotanese, Chinese, Tibetan, and Judeo-Persian documents from these sites, (8) as well as from the earlier discoveries of Sogdian documents cited above. It seems likely that the newly acquired texts have a similar provenance and are roughly contemporary with the Chinese and Khotanese documents; that is, they may be attributed to around the eighth to ninth centuries.

THE TEXTS

No. 1 (Fig. 1)

Catalogue No.: GXW 0116. A fragment of paper bearing writing on one side, with a large hole in the centre. Judging from the creases, the paper was apparently used as a makeshift wrapper for a small quantity of some substance, as was a small Sogdian fragment found at Niya. (9)

1. traces only

2. [[beta] M] N [beta] mwt't ct[beta]'r (l00)[pny]

3. [[beta] (yrt) MN nwrtwt'[delta] w100 pny [

4. ([beta] MN sit' [delta] w 100 pny

5. ([beta] MN nmtkr'y 100 30 pny

rest blank

"[Received from] Vamutat: 400 [pny]. -- Received from Nurtut: 200pny. -- Received from Sart: 200 pny. -- Received from the felt-maker: 130 pny"

Commentary

This is a very simple document listing receipts of sums of money ranging from 130 to 400 pny. The word pny, which ultimately derives from Skt. pana-, refers to the Chinese copper coins with a square hole (Chinese qian [TEXT NOT REPRODUCIBLE IN ASCII], Khotanese mura-) that were the normal currency of the "Western Regions" from the middle of the seventh century C.E. (10) The verb [beta] yrt, which introduces each record, is formally a 3rd person form of the intransitive preterite: 'has/have been received'. The recipient is not indicated but is presumably the writer of the text, so that the formula is equivalent to 'I have received'.

Line 2. The personal name [beta] mwt't or [beta] mwtnt, less likely [beta] mwtyt, is unknown. It does not appear to be Sogdian or Chinese.

Line 3 contains another unknown personal name, probably to be read as nwrtwt/zwrtwt or ywrtwt. -- '[delta] w100 '200' here looks like [delta] w100, but '[delta] r is hardly possible as a form of [delta] ry 'three'.

Line 4. srt, less likely srt, could be compared with the Bactrian personal name oapxo (sarto), for which see Sims-Williams 2010: 126 (no. 421). In addition to the possible etymologies suggested there, one could also consider deriving this name from Old Iranian * sarta-'cold', Sogdian srt.

Line 5. The first two letters of nmtkr'y are not joined, which naturally suggests a reading Z-mtkr'y. However, some irregularity in the joining of the letters seems to be a characteristic of the writer's hand without significance for the reading; cf. the unjoined initial [beta] - of fimwt't in line 2 and of [beta] yrt in line 5. If nmtkr'y is indeed the correct reading, this may be not a name but an occupational designation 'felt-maker'. The word for 'felt' itself is attested as nmty in P3, line 229, (11) and as nmfk in the Mug document [??]20, line 3.(12) -- Finally, it is worth noting that 100 is here written in a more elaborate form than in the preceding lines, with an extra initial stroke resembling an aleph.

[FIGURE 2 OMITTED]

No. 2 (Fig. 2)

Catalogue No.: GXW 0438. A fragment of paper with writing on one side.

1. traces only

2. [beta] ntk(sy)r 'yw kpc(k s)y'swn pnc snk

3. ct [beta]'r kpc('kk)[

4. ] * '[delta] (w)[

" ... (to/from) Vandesir, I kapcakk (of) mustard(?), 5 sang (and) 4 kapcakk (of) ... 2(?) ... *

Commentary

This scrap seems to be part of a list of persons and amounts of certain commodities assigned to them (or received from them). The script, evidently written with a brush rather than a pen, is extremely irregular and ambiguous.

Line 2. Although only a small part of each side margin is preserved, this line seems to be complete. The first word must be a Sogdian name beginning with [beta] ntk 'slave' (more common as the final component of compound names) and a second component which is illegible apart from the final-r. The reading/restoration [beta] ntk(sy)r is based on the occurrence of a similar name in the Mug document [??]9, line 1, where the editors actually read fintksyr. (13) Recently this reading has been corrected to ([beta] ntysyr, (14) but this would merely be a later variant of the same name, for which (Sntksyr would in fact be the normal and expected spelling.

Both kpc 'bushel' and kpc'kk iittle kapc' are well attested in the Sogdian documents from Mt. Mug, chiefly as measures for grain (wheat, barley, millet) and wine. Phrases such as '[delta] w kpc ZY 'yw kpc'kk yntm ZY '[delta] w kpc ZY '[delta] ry kpc'kk H'MRH "two kapc and one kapcakk of wheat and two kapc and three kapcakk of wine" (15) show that the kapcakk is a smaller unit than the kapc. The form kpc is apparently an abbreviated spelling of a neuter light stem kpc-, ace. sg. kpew, (16) abl. sg. Christian Sogd. qpcy', (17) nom.-acc. numv. kpcy. (18) The exact form written here is unclear, but kpc(k), a possible though unattested variant spelling of kpc'kk, seems a more likely reading than any form of kpc-.

The following sy'swn (or xy'swn?) is most likely the designation for a particular commodity. No such word is known in Sogdian, but in view of the provenance of the text it seems possible that it may be a transcription of the local Khotanese word sasvam, earlier ssasvana-'mustard', of which the genuine Sogdian equivalent is sywsp [delta] n. (19) The alternative that s/xy'swn is another personal name is less attractive, since this would require one to assume that the commodity measured is left unnamed.

snk is a loanword from Chinese sheng [TEXT NOT REPRODUCIBLE IN ASCII] 'pint'. (20) While this measure is used in Chinese for both solids and liquids, in Sogdian it is so far attested only as a measure for oil.

Line 3. kpc('kk)[ is graphically unclear though the context makes this reading plausible. The word clearly begins kpc- or kwc-, but the following letters are hardly identifiable.

Line 4. Read '[delta](w)[ '2' ? Less likely lL(P)[w '1000'.

[FIGURE 3 OMITTED]

No. 3 (Fig. 3)

Catalogue No.: GXW 04320 A fragment of paper with writing on one side.

1. [[delta]'rt ***] * [*] **** [delta] s lLPw pny [beta] yrt ct [beta]'r (lLP)[w](' [beta])t(100)[ pny]

2. [[delta] 'rt 'ytxw msy [delta] r ' [delta] w lLPw ' [delta] wy 100 pny

3. [delta] 'rt [beta] wy [delta]'t 24 lLPw pny o

blank space

4. [delta] ['r](t) (ypyw)" tryc pnc lLPw '[beta] t100 pn(y)[

"[NN has] 10,000 pny. Received: 4,700 [pny]. - Itkhu (?) the priest has 2,200 pny. -- Vogh-dhat has 24,000 pny. -- The yabghu Atarich has 5,700 pny."

Commentary

This document seems to be a list of the capital possessed by various persons.

Line 1. On the form [beta] yrt see above, commentary to no. 1. Here it is not clear whether the recipient is the person named at the beginning of the line, who has received 4,700 pny in addition to the 10,000 which he already had, or whether it is the writer who has received 4,700 pny out of the 10,000 just mentioned.

Line 2. The title msy [delta] r 'elder, presbyter, priest' was used by both Christians and Mani-cheans. The presbyter's personal name 'ytxw/'ytyw or '[beta] txw/'[beta] tyw, less likely 'ntxwf'ntyw or 'ztxw/'ztyw, is unfortunately unclear.

Line 3. [beta] wy[delta]'t (which could theoretically be read in various other ways, e.g., [beta] wx [delta]'t, fiwsd't, IcwyS't) seems to be a genuine though unattested Sogdian personal name consisting of the elements [deta]wy 'salvation' and - [delta]'t 'given'. The formation of the name, apparently 'Given by Salvation', suggests that the abstract noun [deta] wy may have been used as the designation of a particular deity, which is plausible in view of the use of the Bactrian expression [deta]WYO [sigma][tau][omicron][rho][gamma][omicron] 'the Great Salvation' as an honorific epithet of the Kushan ruler. (21) Cf. also the name of the ninth Sogdian month, [deta] wyyc, which may be named after a divinity like the names of several other months.

Line 4. The reading ypyw is uncertain, since only the tops of the letters are preserved, but fits the traces extremely well. The title yabghu has not previously been found in any Sogdian document, though it occurs in exactly this spelling in the Sogdian version of the Karabal-gasun inscription. It is a very ancient title which is first attested amongst the Wusun in the second century b.c.e. and was later used by the Kushans, Hephthalites, Western Turks, and other peoples. (22) Although the title does not seem to be attested in the Khotanese documents, Chinese records indicates that in the Khotan kingdom a person of very high rank, such as the king's son or brother, could bear this title. (23)

The name "tryc (which could also be read 'ntryc, "trpc, etc.) appears to derive from the name of the god Atar 'Fire' with the hypocoristic suffix * -ica-. Such a name could have originated in Sogdian, Bactrian, or almost any other Iranian language.

No. 4

Catalogue No.: GXW 0434. A fragment of paper with writing on both sides.

Recto(?) (Fig. 4)

1. wx(wsw)[ wys'yny ](100)[ pny

2. <'[deta]t20 syty mr[t]s'r>

3. tym pr'[delta]t wxwsw wys'yny (pn)cl00 (p)[ny

4. tym pr'[delta] t '[delta]r[y wyg'yny

5. pr't [delta]w wysyny ('[delta])[

6. '[delta] w wys'yny ct [deta]'r100 pn'k o tym pr'[delta]t [

7. <yw20 syty mrts'r wys'(?) pr ... >

8. pr'8t pncw wys'yny '[delta]w ptkwk pn(y)[

9. (') [delta] ry wyg'yny pr lLPw ct[deta]'rl00 pny pr't[

10. lLPw (') [delta] ryl00 pny o tym pr't '5ry wysfyny

"... [I sold] six [pieces of cotton cloth for ... ] hundred [pny].

[FIGURE 4 OMITTED]

From the 27th day on.

Moreover, (I) sold six pieces of cotton cloth (for) 500 [pny]. Moreover, (I) sold three [pieces of cotton cloth for ... Moreover,] (I) sold two pieces of cotton cloth [for ... Moreover, I sold] two pieces of cotton cloth (for) 400 pn 'k. Moreover (I) sold ...

From the 21st day on. Cotton(?) ...

(I) sold five pieces of cotton cloth (for) two strings of pny. [Moreover, I sold] three pieces of cotton cloth for 1,400 pny. (I) sold [ ... pieces of cotton cloth] (for) 1,300 pny. Moreover, (I) sold three pieces of cotton cloth ... "

Verso (?) (Fig. 5)

1. tym pr'[delta]t [delta]s' wys'yny '[delta]ry100 65(?) [pny

2. pr'[delta]t 19 wys'yny ct[beta]' r100 20 pn'(k)[

3. (obliterate by lines 1a-2a, which are added here the opposite way up)

4. ...[delta](r)y wys'yny(?) p(r't)(?) '[

5. pr'yw(?) ct[beta]'r wysyny 1LPw * [

6. [beta]yrt 'sp' syckk(?) s'r *[

7. 'nyw 'st100 pny MN [

[FIGURE 5 OMITTED]

8. pr'tpkn'w(p)[

9. [deta] cw [deta]rp' ** [delta] yrf * [(?) ] p(r'yw)(?) [delta]s' pt(kwk) ct [deta] ['r

10. 'nyw ct [delta]'r ny's *** (?)pr'[delta] (t)(?) ct [delta]'rl00(?) p(ny)(?)

" ... Moreover, (I) sold ten pieces of cotton cloth (for) 365(?) [pny. Moreover], (I) sold 19 pieces of cotton cloth (for) 420 pn'k. ... three pieces of cotton cloth (I) sold(?) ... [I exchanged](?) with [NN] four pieces of cotton cloth (for) 1,000 [pny ...].

(I) received [...] for General Sijaka(?) [...and] another 400 pny from [NN ... ]

(I) sold Ito](?) Pak Nop[...] ... is ... Cho Varpa[...]. (I) received with(?) ... 10 strings (and) 4[00 pny ...] another 4 ... (I) sold(?) (for) 400(?) pny(?) ..."

Verso(?), two lines written the opposite way up to the rest of the text (Fig. 6)

la ](p)yyym(?)'sty lLPw ptk(w)k pny s'[delta] (?) pny pr

2a ]pc ... (?) ZY syr'krtyh sky(?) zy'mt k'n

" ... we are ready(?). There are 1000 strings of pny ... pny for ... and piety ... he will spend ..."

[FIGURE 6 OMITTED]

Commentary

The main part of this document is an account of sales of wyS'yny or wysyny 'cotton cloth'. (24) Like [deta]ert in no. 1, the constantly repeated verb pr'[delta]t/pr't 'has/have been sold' is formally a 3rd person intransitive preterite, but a first person agent is again implied: '(I) sold'. Each record of a sale is followed by the mention of a sum in piny or pn'k, evidently the price, though the figure is only once introduced by the expected preposition pr 'for' (R9).

Lines R2 and R7 each contain a date naming a day of the month by means of the formula: numeral + syty 'elapsed' (25) + mrts'r, which can be taken either as an adverb 'onwards' or as a postposition 'from'. (26) As their spacing indicates, both lines are secondary additions to the original text. It is strange that the dates seem to appear in reverse chronological order; but the main problem is the numerals themselves, which are written in a very unconventional fashion. Normally, in Sogdian numerals written wholly or partly in figures, the tens precede the units, e.g., 30 wxwsw '36'. (27) Here, however, the writing follows the order of the spoken forms, in which the unit precedes, e.g., Christian Sogdian yw-wystmyq '21st'. (28)

R5. The form pr'r (also in R9, R10, V4 (?), V8) is obviously a variant of pr'[delta] t [pra [delta] (a) t] 'sold'. Cf. the loss of [0] and [5] before [t] in the Christian Sogdian forms of the conjunctions m't 'that' < m'[delta] ZY [ma [theta] - [??]ti] and qt 'if; that' < k [delta] wty [ka[delta] utij. (29) One might also consider the possible influence of the cognate Khotanese form parata 'sold', but if we are right to interpret 'spr in V6 as a transcription of the Khotanese title spa < spata 'general' (see below), we would expect Khotanese parata, too, to appear in its later form para.

At the end of the line, ('8)[ could be restored to '[delta] w '2' '[delta] ry '3', or ILPw '1000'.

R6. For pn'k as a variant spelling of pny cf. V2.

R7. The words following the date formula (cf, above on R2) are very faintly written. If we may read wys' this could be interpreted as the noun 'cotton' from which the substantivized adjective wys'yny '(cloth) made of cotton' is derived. The underlying noun has been reconstructed as * w(y)s(w)9 but wyS' could be a later form as in the case of *kynpw 'flax', later kynp'. (30)

R8. '[delta] w ptkwkpny 'two ptkwk of pny' evidently represents a substantial sum. The context suggests that ptkwk, which also occurs in V9 and Via, translates Chin, guan [TEXT NOT REPRODUCIBLE IN ASCII], which means both 1. 'to run a thread through, string together' and 2. 'a string of 1,000 coins' and which is attested in the latter sense in contemporary Chinese documents from the Khotan area. (31) By etymology ptkwk is no doubt cognate with the verb ptkwc, which translates Chin, chuan [TEXT NOT REPRODUCIBLE IN ASCII] 'to pierce, thread, string together' (cf. MacKenzie 1970: 35; Gershevitch 1970: 306), a verb which is synonymous with guan (in its first sense). The word ptkw 'string of 1,000 coins' also occurs, presumably as a Sogdian loanword, in the two Judeo-Persian documents from the Khotan region, which refer to '30 ptkw' and '100ptkw' respectively. (32)

In the present text, ptkwk seems to be used only for numbers from 2,000 upwards, while for smaller numbers (1,000; 1,300; 1,400) the ordinary Sogdian word lLPw '1,000' is preferred (R9, 10, V5, Via). Somewhat similar is the use of the term ysa'ca 'thousander' in Khotanese documents for referring to amounts above 10,000 mura- (= 10 ysa'ca) as opposed to the ordinary ysara '1,000', which is the only term used to refer to smaller sums (and sometimes larger sums too). (33) The Khotanese usage seems to have been influenced by Chinese, which typically uses guan [TEXT NOT REPRODUCIBLE IN ASCII] 'string of 1,000 coins' only for amounts over 10,000 coins, while the simple qian [TEXT NOT REPRODUCIBLE IN ASCII]' 1,000' (along with its two alteration-proof forms [TEXT NOT REPRODUCIBLE IN ASCII] and [TEXT NOT REPRODUCIBLE IN ASCII]) is used for smaller sums and sometimes also, as an alternative, for sums over 10,000.

V4-5. These lines seem to contain a slightly different formula from the preceding part of the text. The postposition pr'yw 'with' in line 5 was presumably preceded by the name (or other designation) of a person with whom the four pieces of cotton cloth were exchanged or deposited. One may compare the use of the synonymous postposition nfi'nt in a "Turco-Sogdian" document from Dunhuang, e.g.: 'yw "n cwrt'n n'm swimy ninny nw"spyty pr wxwsw krmyr s'r "with a certain An Chortan from Solmi (I exchanged) 9 white (rayzi) for 6 red." (34)

V6. The verb [beta] yrt 'received' (cf. no. 1 above) signals another change of formula. Together with the postposition s'r 'for' it seems to imply receipt on behalf of another person. Here the person in question seems to be named as 'sp' syckk(l), where 'sp' [aspa] may represent the local Khotanese title spata 'general', later spa, Tibetan spa, with the characteristic Sogdian prothetic vowel before the initial consonant cluster. The general's name is not clearly written (apart from the initial s-) but may possibly be interpreted as syckk. If so, it may be compared with the name of the spata Sijaka, which is attested on a wooden stick from Mazar Tagh. (35)

V8-9. pk (or p) n'w(p')[ and cw [beta] rp' ** may be Chinese names consisting of a monosyllabic family name (perhaps bai [TEXT NOT REPRODUCIBLE IN ASCII] and zhou [TEXT NOT REPRODUCIBLE IN ASCII], Late Middle Chinese pfia:jk and t[section]iw respectively (36)) and a two-syllable personal name.

V10. Instead of 'nyw 'another' (as in line 7) one could read fxw (definite article). -- ny's ** (name of a commodity?) and all the following words are more or less uncertain.

Via. If ](p)yyzym is correctly read, it can be understood as the 1st person plural present indicative or optative of the verb py'yz, Manichean Sogdian pyyh for which the meaning 'to * be ready, * intend' has been proposed. (37) This meaning also suits a so far unnoticed occurrence in a letter from Mt. Mug: rty c'n(k)w 'skwy cw frkh 'krt(w) [delta]'ry cw fpyy-z-y-skwn "How are you? What have you done (and) what do you * intend (to do)?" (38)

s'[delta] (or sn [delta]) is incomprehensible. The writing of snvisvery strange, but it is hard to envisage an alternative reading.

V2a. sky 'above' does not seem to make sense here. In view of the context provided by the preceding Syr'krtyh 'piety, virtue', one may perhaps suspect a mistake for * snky 'for the samgha, for the (monastic) community'. But we must admit that we have failed to make sense of this part of the text.

APPENDIX: BRITISH LIBRARY FRAGMENT 33

This fragment of an economic document from Mazar Tagh (Or. 8212/1763 = M. Tagh 038.d) was previously published in Sims-Williams 1976: 73 with a facsimile in DTS: PI. 36b. (39) One side of the paper is blank; the other side looks the middle of a double folio whose two pages are inscribed the opposite ways up. Probably the paper was folded over to form a single page of double thickness, the inscribed sides being the Recto and Verso of this double-thickness page.

Recto(?)

1. pr'[delta]t [delta] *[

2. 40 p[ny(?)

3. p(r') [delta] [t

Verso(?)

1 (22)[

2 rtms pr'[delta](t) [

3 [ ]130[

4 [ ](k) pr 'L(P)[w(?) blank space

5 [ ] ** [

"(I) sold ... 40 p[ny(?) ... (I) sold ..." "22 ... (I) also sold ... 130 ... for 1000(?) [pny)(?) ... "

(1.) We are grateful to the Museum, to the working group of scholars and experts to whom the study of the new collection has been entrusted, and in particular to the Museum's Director, Prof. Wei Jian, for permission to publish these manuscripts. We should also like to thank Ursula Sims-Williams (British Library) for providing information on the history and findspots of the manuscripts of the Stein, Hoemle, and Trinkler coUections.

(2.) It formed part of a consignment of manuscripts from the Khotan area. At least one of these (H. 143 NS 34 = IOL Khot 160/2) presumably came from Khadaliq, since it is part of the same leaf as the excavated fragment Kha. ix.13 = IOL Khot 39/3 (Skjwrvo 2002: 358).

(3.) Yoshida 2010: 6. In the first part of this article, Yoshida gives an overview of the Sogdian texts found in recent years in Xinjiang, including the Khotan area (pp. 5-7).

(4.) Gropp 1974: 364 with fig. 219a on p. 367. For a corrected reading see Sims-Williams 1979: 337 n. 6, where it is shown that this fragment belongs to a religious text, either Buddhist or (less likely) Manichean.

(5.) Gropp 1974: 362, citing Trinkler 1930: 148.

(6.) Trinkler 1930: 155.

(7.) This document is a slip of paper, with a clay seal and hemp string attached in the middle, on which the name of a s'rtp'w 'caravan-leader' is preserved. It was displayed in Beijing in 2009 in the "Exhibition of the Special and Rare Collections of the National Library of China."

(8.) See Rong 2006: 514-18; Rong 2009: 405-7.

(9.) Sims-Williams and Bi Bo, forthcoming.

(10.) See Henning 1946: 723; Yoshida 1994: 379 with nn. 11-12 on p. 304; Sims-Williams 1996: 50; Yoshida 1996: 71-73. For a thorough survey of the monetary system in the Khotan area see Wang 2004

(11.) Benveniste 1940: 69, 199.

(12.) Bogoljubov and Smirnova 1963: 66.

(13.) Bogoljubov and Smirnova 1963: 32, 92.

(14.) Lurje 2011: 140-41 (no. 296).

(15.) E2, line 3 (Bogoljubov and Smirnova 1963: 29).

(16.) Mug V4, R15 (LivSic 2008: 62).

(17.) Cited by Sundermann 1984: 58.

(18.) E.g., 'dwy 100 kpcy 'two hundred kapc', Mug A18, R9, and B13, line 7 (Livsic 2008: 148, 167); cf. Sims-Williams 1979: 342, on the "numerative" of nouns of this class.

(19.) See Henning 1965:29,35.

(20.) See DTS: 32-33.

(21.) abatak inscription, line 1 (Sims-Williams 2008: 55).

(22.) See Sims-Williams and de la Vaissifcre 2007.

(23.) See Kin Tangshu [TEXT NOT REPRODUCIBLE IN ASCII] (New Tang History), Chapter 126, 127 (referring to the years 649 and 760 c.e.). 221, Beijing 1975: 6235, 6236; Chavannes 1903: 126, 127 (referring to the years 649 and 760 C.E.).

(24.) See DTS: 56-57.

(25.) TS: 34.

(26.) DTS: 28.

(27.) Gershevitch 1954: [section] 1324.

(28.) Sims-Williams 1985: 104 (55V.29).

(29.) Sims-Williams 1985: 184 n. 7.

(30.) On all these words see DTS: 57.

(31.) The practice of measuring Chinese coins by strings of 1,000 can be traced back to the Han dynasty (Wang 2004: 13).

(32.) See Utas 1968: 133, where ptkw is tentatively understood as 'jar', and Zhang and Shi 2008: 78, 96, where ptkw is defined as a kind of currency unit, without any more specific identification.

(33.) Wang 2004: 95-97 with tables 39-40. Cf. also Yoshida 2005: 47 n. 63.

(34.) Document A, lines 8-9 (DTS: 24).

(35.) IOL Knot Wood 41 (Skjsrv0 2002: 569).

(36.) Pulleyblank 1991:27,411.

(37.) Sundermann 1985:49-50.

(38.) V17, R16-17 (Livsic 2008: 128).

(39.) A digital color image can be found by searching for "Or.8212/1763" on the website of the International Dunhuang Project (http://idp.bl.uk).

REFERENCES

Benveniste, E. 1940. Textes sogdiens, Paris: Librairie Orientaliste Paul Geuthner.

Bogoljubov, M. N., and O. I. Smirnova. 1963. Xozjajstvennye dokumenty. Sogdijskie dokumenty s gory Mug, vol. III. Moscow: Izdatel'stvo Vostocnoj Literatury.

Chavannes, E. 1903. Documents sur les Tou-kiue (Turcs) occidentaux. Sbornik trudov Orkhonskoj fekspedicii, vol. VI. St Petersburg: Iniperatorskaja Akademija Nauk.

Gershevitch, I. 1954. A Grammar of Manichean Sogdian. Oxford: Basil Blackwell.

-- 1970. Review of MacKenzie 1970. Indogermanische Forschungen 75: 303-6.

Gropp, G. 1974. Archdologische Funde am Khotan, Chinesisch-Ostturkestan. Bremen: Verlag Fried-rich Rover.

Henning, W. B. 1946. The Sogdian Texts of Paris. Bulletin of the School of Oriental and African Studies 11/4:713-40.

-- 1965. A Grain of Mustard. Annali delV Istituto Universitario Orientate di Napoli, Sezione linguistica 6: 29-47.

LivSic, V. A. 2008. Sogdijskaja epigrafika Srednej Azii i Semirec 'fa. St. Petersburg: Filologideskij Fakul'tet Sankt-Peterburgskogo Gosudarstvennogo Universiteta.

Lurje, P. B. 2011. Personal Names in Sogdian Texts. Iranisches Personennamenbuch, vol. 11/8. Vienna: Verlag der Osterreichischen Akademie der Wissenschaften.

MacKenzie, D. N. 1970. The "Sutra of the Causes and Effects of Actions" in Sogdian. London: Oxford Univ. Press.

Pulleyblank, E. 1991. Lexicon of Reconstructed Pronunciation in Early Middle Chinese, Late Middle Chinese, and Early Mandarin. Vancouver: UBC Press.

Rong Xinjiang. 2006. Sogdians around the Ancient Tarim Basin. Eran ud Aneran: Studies Presented to Boris ll'ic Marsak on the Occasion of his 70th Birthday, ed. M. Compareti et al. Pp. 513-24. Venice: Libreria Editrice Cafoscarina.

-- 2009. Further Remarks on Sogdians in the Western Regions. Exegisti monumental Fest schrift in Honour of Nicholas Sims-Williams, ed. W. Sundermann et al. Pp. 399-416. Wiesbaden: Harrassowitz Verlag.

Sims-Williams, N. 1976. The Sogdian Fragments of the British Library. Indo-Iranian Journal 18: 43-82.

-- 1979. On the Plural and Dual in Sogdian. Bulletin of the School of Oriental and African Studies 42: 337-46.

-- 1985. The Christian Sogdian Manuscript C2. Berliner Turfantexte, vol. XII. Berlin: Akademie Verlag.

-- 1996. The Sogdian Merchants in China and India. In Cina e Iran da Aiessandro Magno alia dinastia Tang, ed. A. Cadonna and L. Lanciotti. Pp. 45-67. Florence: Leo S. Olschki Editore.

-- 2008. The Bactrian Inscription of Rabatak: A New Reading. Bulletin of the Asia Institute 18:53-68.

-- 2010. Bactrian Personal Names. Iranisches Personennamenbuch, vol. 11/7. Vienna: Verlag der Osterreichischen Akademie der Wissenschaften.

Sims-Williams, N., and Bi Bo. Forthcoming. A Sogdian Fragment from Niya. Festschrift for Zhang Guangda.

Sims-Williams, N., and E. de la Vaissiere. 2007. Jabguya, i: Origin and Early History. Encyclopaedia Iranica 14/3: 314-15.

Sims-Williams, N., and J. Hamilton. 1990. Documents turco-sogdiens du IXe--Xc siecle de Touen-houang. Corpus Inscriptionum Iranicarum, Part II, Vol. HI/3. London: School of Oriental and African Studies.

Skj [much less than] erv0, P. O. 2002. Khotanese Manuscripts from Chinese Turkestan in the British Library: A Complete Catalogue with Texts and Translations. London: British Library.

Sundermann, W. 1984. Review of J. Benzing, Chwaresmischer Wortindex. Kratylos 29: 55-59.

-- 1985. Ein manichaisch-soghdisches Parabelhuch. Berliner Turfantexte, vol. XV. Berlin: Akademie Verlag.

Trinkler, E. 1930. Im Land der Stiirme. Leipzig: F. A. Brockhaus.

Utas, B. 1968. The Jewish Persian Fragment from Dandan-Uiliq. Orientalia Suecana 17: 123-36.

Wang, H. 2004. Money on the Silk Road: The Evidence from Eastern Central Asia to c. AD 800. London: British Museum Press.

Yoshida Yutaka. 1994. Sogudogo moji de hyoki sareta kanjion [Chinese in Sogdian Script]. Tohd gakuho [Journal of Oriental Studies] 66: 380-271 [sic].

-- 1996. Additional Notes on Sims-Williams' Article on the Sogdian Merchants in China and India. In Cina e Iran da Alessandro Magno alia dinastia Tang, ed. A. Cadonna and L. Lanciotti. Pp. 69-78. Florence: Leo S. Olschki Editore.

-- 2005. Kotan shutsudo 8-9 seiki no Kotango sezoku monjo ni kansuru ohoegaki [Notes on the Khotanese Documents of the 8th-9th Centuries Unearthed from Khotan]. Monograph Series in Foreign Studies, vol. 38. Kobe: Kobe-shi Gaikokugo Daigaku Gaikokugaku Kenkyujo.

-- .2010. Shinshutsu no Sogudogo shiryo ni tsuite: shinmai shoki no chichi e no tegami kara: Saigonji Tachibana shiryo no shokai o kanete [On Some Recently Discovered Sogdian Materials -- Edition of a Draft Letter Written by an Inexperienced Scribe: A Sample from the So Far Unnoticed Saigonji Tachibana Texts). Kyoto Daigaku Bungakubu kenkyu kiyo [Memoirs of the Faculty of Letters, Kyoto University] 49: 1-24.

Zhang Zhan and Shi Guang. 2008. Yijian xin faxian Youtai Bosi yu xinzha de duandai yu shidu [The Date and Interpretation of a Newly Discovered Judeo-Persian Text]. Dunhuang Tulufan yanjiu [Journal of the Dunhuang and Turfan Studies] 11: 71-99.

BI Bo and NICHOLAS SIMS-WILLIAMS RENMIN UNIVERSITY OF CHINA, BEIJING / SCHOOL OF ORIENTAL AND AFRICAN STUDIES, LONDON