Chinese text (Figure 1)

1. 1 大象二年歲

2. 2 次庚子十月

3. 3 癸丑廿日壬

4. 4 申相州商客

5. 5 遊埿々[埿]槃陁

6. 6 妻康紀姜合

7. 7 㘸舊相州城

8. 8 西北六里際

9. 9 陌河所銘記

Figure 1. The Chinese inscription (rubbing). © Yan Yan.

Translation

On the twentieth day (day of renshen 壬申) of the tenth month, the month of guichou 癸丑, in the second year of the era Daxiang 大象 (580), the year of gengzi 庚子, You Nini Pantuo 遊埿埿槃陁 (Nanai-vande), the travelling merchant from Xiangzhou 相州 (i.e. Ye 鄴), and his wife Kang Jijiang 康紀姜 (Kekan) were buried together, six li to the north-west of the city of Old Xiangzhou, in the place Jimo 際陌 where the [Zhang] river ran through. The inscriptions were engraved (on the stone).

Commentary

As in the case of the epitaph of Shi Jun, the Sogdian version is much longer and more informative than the Chinese. Only two details in the Chinese text are missing in the Sogdian and therefore need to be discussed here: Nanai-vande's surname You 遊 (line 5) and the place name Jimo 際陌 (lines 8–9).

5 You 遊 (EMC *juw).Footnote ⁶ Judging from its position in the text, this should be Nanai-vande's surname, which is not given in the Sogdian text. In view of his Sogdian personal name (see below), his merchant background and his marriage to a Sogdian woman, it is natural to assume that Nanai-vande is of Sogdian or Central Asian descent, in which case his surname would normally indicate his family's place of origin. However, since You is not one of the surnames commonly used by the Sogdians in China, we also have to consider the possibility that Nanai-vande is Chinese or half-Chinese. You 遊 is attested as a Chinese surname in a document from Mazar Tagh near Khotan (mid-eighth century)Footnote ⁷ and several documents from Dunhuang (c. late ninth to late tenth centuries).Footnote ⁸ Later on, the surname 遊 seems to occur only in documents from the area to the west of Dunhuang, so it might be a Chinese surname originating in that region, or perhaps just a local corruption of its homophone 游, a well-attested surname in historical sources and epitaphs from central China as well as in documents from Dunhuang and Turfan. According to Lin Bao's Yuanhe Xingzuan 元和姓纂 (Register of the Great Families from the Yuanhe Reign (806–820), completed in 812 ce), juan 5, the surname 游 originates in an area not far from Ye. However, it is difficult to imagine that a powerful Chinese family of Ye would have had marriage connections with a foreign family at the time of our epitaph. It may be more relevant to note that some people also bore the surname 游 in the Hexi Corridor and even further west. In the biography of Qu Yun 麴允 (d. 316 ce), an official under the Western Jin (265–316 ce), it is recorded that he was originally from Jincheng 金城 (Sogdian kmzyn, modern Lanzhou), where his family Qu and the family You 游 had been local great families (haozu 豪族) for generations. See the Jinshu 晉書, juan 89, which even refers to a verse indicating the grandeur and wealth of these two families, which is said to have been widely known in his hometown and the surrounding areas (xizhou 西州, lit. “the Western prefectures”). It therefore seems possible that Nanai-vande's surname indicates a connection with the You 游 family from the Hexi Corridor, whose surname may have been corrupted to You 遊 in the Western regions.

8–9 Jimo 際陌 (EMC *tsiaj^h maɨjk). According to historical sources and epitaphs this place is located five or seven li to the north-west of the city Ye (see Figure 2). The place was originally called Jimo 祭陌, with ji 祭 (EMC *tsiaj^h ) meaning “to offer a sacrifice, hold a memorial ceremony for” and mo 陌 meaning “an east–west path in the fields”. The name refers to an event during the Warring States period, when the local witches and wizards are said to have married young girls to the river god by throwing them into the Zhang River 漳河 to prevent it from flooding. In 345 ce, Shi Hu 石虎, a ruler of the Later Zhao (319–351 ce), built a pontoon over the Zhang River beside the place Jimo and named it Zimo Qiao 紫陌橋 (lit. Zimo Bridge), which indicates that by this time Jimo had been changed to Zimo 紫陌 with the first character 紫 (EMC *tsiəă/tsi’) meaning “purple, violet”. Although Zimo is well-attested in historical texts and epitaphs from the Northern Qi period, the form Jimo is still attested in an epitaph of 546 ce under the Eastern Wei (534–550 ce) and in the epitaph under discussion.

Figure 2. Map of Ye and area. Adapted from Wang Shiduo Reference Wang1861.

In the present epitaph, the character following Jimo is he 河 “river”. It is not easy to decide whether we should read these words together as “Jimo River” or separately as “Jimo and the river”. Besides the epitaph of the Sogdian couple, the above-mentioned epitaph of 546 ce also contains the combination 際陌河, and the similar expression 紫陌河 appears in three epitaphs from the Northern Qi period. In these four epitaphs, the burial places are all indicated by the phrase “to the north of the river”. This might seem to imply that there is a Zimo (or Jimo) River to the west of Ye. However, judging from the available literary and archaeological sources, the river referred to in these texts must be understood as the Zhang River. Since Zimo was a fixed place well known to the local residents of Ye, it may be that they tended to give the name “Zimo River” to this part of the Zhang River, using it as a landmark to record the precise locations of the burial places of their family members.

Sogdian text (Figures 3–5)

1. 1 ʾwyn tyntwʾn δwʾ srδ kʾs srδ 10my

2. 2 mʾxw 23yh nnyβntk xwʾcʾk^a ZK cynʾʾkk

3. 3 xwʾcʾk^a BRY ʾnkpcʾnʾk ʾḤRZY cnn βnδʾ^b

4. 4 ʾpwrʾsty^c rty ʾwyn mwpʾyn δwʾ srδ xrγ-ʾwšʾ(k)^d

5. 5 srδ 11my m(ʾ)xw pnc sγ-th tʾrʾ cnn ʾBYʾ

6. 6 nyzβrt kʾw ʾnkpcʾnʾk k(n)δh nšmy pʾš kwsy

7. 7 wxwšw ʾβsʾx kʾw R(B)k xwʾcʾk cynʾʾkk

8. 8 txmw nšmy pʾš kyrʾn txmw wʾ(s)t^e ʾḤRZY tʾ(z)ʾy(n)^f

9. 9 ʾyw srδ kʾs (srδ) δβty m(ʾ)xw ʾwyh 19yh

10. 10 kykʾʾnh δβʾnmʾh^c ZKh wnxrʾk δwxth^g

11. 11 ʾnkpcʾnch nnyβntk xwʾcʾk wδwh cnn βnδʾ^h

12. 12 ʾpwrʾsth^c tʾzʾʾnⁱ δwʾ srδ 10my mʾxw 20 sγ-th

13. 13 tʾrʾ (c)[n]n mʾmh^j nyzβrt kʾw ʾBYʾ^k sʾr

14. 14 ʾnkpcʾnʾk knδh nšmy pʾš kwsy wxw(š)w ʾβsʾ(x)^l kʾw

15. 15 txmw pʾw mʾmh kʾw sʾcy wyʾʾk ʾyw zʾyh wʾ(sn)t^m

1. ^a Here could be xwʾcyk, but -ʾk is sure in lines 7, 11. Hardly xwycʾk.

2. ^b Hardly βʾδʾ. Or βγ-ʾ?—cf. γ immediately below. But cf. line 11.

3. ^c Sic.

4. ^d Last letter at edge of stone, rather unclear, but -k is more likely than -y.

5. ^e Or wʾxt? But cf. line 15.

6. ^f The last letter is a rather long vertical stroke at the very edge of the stone.

7. ^g Looks like δwsth.

8. ^h The reading seems certain here.

9. ⁱ Or tʾzyʾn? The writing cannot be reconciled with that in line 8.

10. ^j From the rubbing mʾth seems possible, but mʾmh (as in line 15) is clear on a negative photo.

11. ^k The B seems to be altered (from x?).

12. ^l The end of the word is clumsily written.

13. ^m The middle part of the word is unclear, but neither wʾxt nor wʾst (as in line 8) seems likely here. One may either read wʾ(š)t (with a complex š similar to that in wxwšw, line 7), or wʾ(sn)t.

Figure 3. The Sogdian inscription (rubbing). © Yan Yan.

Figure 4. The Sogdian inscription, right side of lines 11–15 (negative photograph). © Bi Bo.

Figure 5. The Sogdian inscription, left side of lines 11–15 (negative photograph). © Bi Bo.

Translation

In year two of (the era) tiantong, a pig year, on the 23rd of the 10th month, the merchant Nanai-vande, the son of the merchant Chinakk, (resident) of Ye – then he departed from the *world; and in the year two of (the era) wuping, a hare year, on day five of the 11th month, (his) *body was taken away from (his) father; (at) six parasangs (distance) to the north-west side of the city of Ye, in a north-west direction to (the graves of) the great merchant Chinakk family, the family laid (him).

Then in the year one of (the era) daxiang, a pig year, on the 19th of the second month, the lady Kekan, the daughter of Wankharakk, (resident) of Ye, the wife of the merchant Nanai-vande, departed from the *world; in the year two of (the era) daxiang, on day 20 of the tenth month, (her) *body was taken away from (her) *mother to (her) father; (at) six parasangs (distance) to the north-west side of the city of Ye, in the family (grave), without (her) *mother, in a suitable place (in) one (piece of) ground, (they) laid (her).

Commentary

1–2. The dating formula is similar to that at the beginning of the Sogdian inscription on the tomb of Shi Jun. The month is indicated by an ordinal number. The form 23yh “on the 23rd day” is the oblique of a special feminine form of the numeral which indicates the day of the month, confirming Yoshida's reading of exactly the same form in line 3 of the Shi Jun inscription (Yoshida Reference Yoshida, de la Vaissière and Trombert2005: 63). Similarly ʾwyh 19yh “on the 19th day” in line 9 of our text.

The year is identified by the transcribed Chinese era-name (here tyntwʾn = tiantong 天統, EMC *t^hεn t^hawŋ^h , an era of the Northern Qi dynasty), as well as by its place in the animal cycle. It seems that the composer of the epitaph made a mistake in calculating this date, which of course is some years earlier than that of the inscription itself. Since the second year of tiantong was in fact a dog year, either “pig year” is a mistake for “dog year”, in which case the date indicated would correspond to 20 November 566 ce, or “second year of tiantong” is a mistake for “third year”, in which case the date would correspond to 9 December 567 ce.

2–3. The deceased Nanai-vande bears one of the commonest of all Sogdian names (Lurje Reference Lurje2011: 271–3, no. 787), reproduced in the Chinese version as Ninipantuo 埿埿槃陁. His father's name cynʾʾkk, which does not seem to be attested elsewhere, is probably a hypocoristic from a name containing the noun cynʾ, cynʾkh “desire”, cf. such names as Avestan xšaθrō.cinah-, Old Persian Aspacanah-. Connection with cyn “Chinese” seems less likely, especially as a suffix -ʾʾkk would be hard to explain. Both father and son are described by the previously unattested term xwʾcʾk. It does not seem possible to read the word as xwycʾk “open” (which could potentially be interpreted as “freeman”). On the basis of the Chinese version, which has shangke 商客 (EMC *ɕɨaŋ kʰaɨjk/kʰε:jk), lit. “travelling merchant”, it seems likely that this word means “merchant”. If so, it might be cognate with xwʾkr “id.”, where the element xwʾ- no doubt derives from *wahāka- “trade” as proposed by Henning (Reference Henning1937: 116a). However, the suffix -cʾk would not be easy to explain. Another possibility would be to understand it as a slightly adapted loanword from Chinese huozhu 貨主 (EMC *xwa^h tɕuă') “seller”, lit. “owner of goods”,Footnote ⁹ an expression attested in Chinese since Eastern Han,Footnote ¹⁰ as well as in early Chinese Buddhist scriptures such as the Pratyutpanna-samādhi-sūtra (Banzhou sanmei jing 般舟三昧經) translated by the Yuezhi monk Lokakṣema in 179 ce Footnote ¹¹ and the Ṣaṭ-pāramitā-saṃgraha-sūtra (Liudu ji jing 六度集經) translated by the Sogdian monk Kang Senghui 康僧會 under the Three Kingdoms (third century ce).Footnote ¹²

Dr Pavel Lurje has kindly suggested to us that xwʾcʾk might be the same word as Persian x^vāǰa, which is used as a title for men of importance, including rich merchants. The word is already attested in Rudaki, but apparently not in Middle Persian, and has no obvious etymology,Footnote ¹³ so the possibility must be considered that it is a Sogdian loanword in Persian. If so, the replacement of Sogdian [č] by Persian [ǰ] would be comparable to the case of Persian xāǰ “cross” from Armenian xač̣ (see Hübschmann Reference Hübschmann1895: 227). This explanation would be compatible with either of the etymologies (Sogdian or Chinese) suggested above for xwʾcʾk. It seems less likely that the Persian form is an inherited cognate of the Sogdian, with -ǰ- from *-č- as in a few other Persian words (see Hübschmann Reference Hübschmann1895: 226–7). In the latter case one would have to reconstruct an underlying form such as *xwāčaka-, and the connection with Sogdian xwʾkr “merchant” from *wahāka- “trade” would be excluded.

3 ʾnkpcʾnʾk “of (the city) Ye”, f. ʾnkpcʾnch (line 11). See Henning (Reference Henning1948: 608–9) on the writing of the name of the city Ye 鄴 as ʾnkpʾ in Ancient Letter II, line 13. The Chinese version refers to the city as Xiangzhou 相州 (Xiang prefecture), its official name under the Northern Zhou, which had replaced the Northern Qi in 577 ce (see Zizhi Tongjian 資治通鑑, juan 173). In the Chinese text the expression “Old Xiangzhou” is used to refer to the burial place, probably because the city had been destroyed by Yang Jian 楊堅, the founder of the following Sui Dynasty, in August 580 ce, just before the making of the epitaph in November 580 ce (cf. the note on line 12), and the residents and the seat of the prefecture had been transferred to Anyang 安陽, 45 li to the south of the old city (see the Zhoushu 周書, juan 10).

3–4. The phrase cnn βnδʾ ʾpwrʾst-, also in lines 11–12, must mean “died”, perhaps as a euphemism. Despite the strange spelling, ʾpwrʾsty, f. ʾpwrʾsth, is probably the 3 sg. preterite of ʾpwʾrt “to turn”. The same verb is used in the Bugut inscription in another expression for “to die”: kʾw βγy/βγyšt sʾr pwrsty/pwrst, lit. “turned (= departed) to the god/gods” (Yoshida Reference Yoshida, Takao and Ayudai1999: 123–4). Since it is here construed with cnn “from”, the governed noun βnδʾ might mean “world”, “body” or “life”. Cf. the expressions for “to die” in the Qara-Balgasun inscription: tnpʾr pryc/prʾγt “to abandon the body” (a calque on Turkish ät’öz qod-, see Hansen Reference Hansen1930: 29; Yoshida Reference Yoshida and Sundermann2009: 573), ʾβcʾnpδy xrʾm “to proceed (from) the world” (Yoshida Reference Yoshida1988: 44; wrongly Hansen Reference Hansen1930: 32). It is worth noting that one of the commonest expressions in Chinese for “to die, pass away” is lishi 離世, lit. “to part from the world”. If the phrase here is a calque on the Chinese expression, βnδʾ should mean “world”. According to Benveniste (Reference Benveniste1940: 130, 224), a noun βʾδ is attested in P12, line 6. The form is quite unclear in the published facsimile but it seems likely that it can equally well be read βnδ, of which βnδʾ here could be an ablative form. However, since its meaning is quite unknown (Benveniste conjectured “seat” on the basis of an unacceptable etymology), this does not help.

4–5. The second date given here must be that of Nanai-vande's burial, four or five years after his death, the fifth day of the eleventh month in the second year of the era mwpʾyn = wuping 武平 (EMC *muə̆’ biajŋ), corresponding to 7 December 571 ce. This was a hare year, as correctly given in the Sogdian text, line 4. The date formula is similar to that in lines 1–2, but this time the day is indicated using another traditional expression with sγ-th, lit. “elapsed” (as also in line 12).

5–6. It is clear from the contexts that the expressions tʾrʾ cnn ʾBYʾ nyzβrt here and tʾrʾ (c)[n]n mʾmh nyzβrt in line 13 refer to the burial of the husband and the wife respectively. For tʾrʾ and nyzβrt we are happy to adopt the suggestions of Professor Yutaka Yoshida, who interprets the former as a noun referring to the remains of the dead persons and the latter as 3 sg. intransitive preterite of a verb nyz βr- < *niž-bara- “to bear off, to take away”. Assuming that mʾmh is a variant of mʾth “mother”, as discussed in the note to line 13 below, the two sentences would mean “(his/her) body was taken away from (his/her) father/mother”.

Although the verb nyz βr- is certainly not common, it seems to be attested in the variant spelling ʾnzβr- in Ancient Letter II, line 45, where it may mean “to withdraw”, with reference to a financial transaction (Sims-Williams Reference Sims-Williams, Schmidt and Bisang2001: 276), and probably survives in Yagnobi žĭvár- “to bring out” (Andreev and Peščereva Reference Andreev and Peščereva1957: 370). Moreover, Avestan niž-bar- and its derivative niž-bərəθi- are specifically used, as would be the case here, with reference to the “removal” of a corpse (Vidēvdād 6, 31ff.). The noun tʾrʾ “body” does not have an obvious etymology, but a connection with Khotanese ttarandara- “id.”, itself of obscure origin, seems worth considering.

6–7 kʾw ʾnkpcʾnʾk k(n)δh nšmy pʾš kwsy wxwšw ʾβsʾx “(at) six parasangs (distance) to the north-west side of the city of Ye”. Similarly (but without kʾw) in line 14. The word-order nšmy pʾš follows that of 西北, lit. “west-north”, in the Chinese text. “Six parasangs” corresponds to 六里 “six li” in the Chinese text, the Sogdian unit of distance ʾβsʾx “parasang” being equated with the Chinese li 里, although the original values of the two units were quite different. Similarly in the Sogdian gospel lectionary E5, fsx is used to translate Syriac mylʾ “(Roman) mile”, without strict regard to the distance indicated by either unit (Sims-Williams Reference Sims-Williams2016: 80).

7–8 kʾw R(B)k xwʾcʾk cynʾʾkk txmw nšmy pʾš kyrʾn “in a north-west direction to (the graves of) the great merchant Chinakk family”. We are again grateful to Professor Yoshida for the suggestion that txmw “family” here implies “family grave”. Thus the text would indicate that a new tomb was built for Nanai-vande beside those of his ancestors. The parallel passage referring to the burial of his wife in lines 14–15 has merely kʾw txmw “in the family (grave)”, i.e. in the same tomb as her husband.

8 txmw wʾ(s)t “the family laid (him)”. The noun txm- “family”, which is usually masculine, here seems to be treated as a neuter. This is perhaps an archaism, as the Avestan cognate taoxman- is likewise neuter. Cf. also the (metathesized?) form txwm (Sims-Williams and Durkin-Meisterernst Reference Sims-Williams and Durkin-Meisterernst2012: 195a)? For the use of the verb ʾwst (y) “to lay”, imperfect wʾst, with reference to the deposition of a corpse, one may compare the Mug document B-8, line 15 (Livshits Reference Livshits2015: 40). See also below on the awkward form wʾ(sn)t in line 15.

tʾ(z)ʾy(n) may be a mistake for tʾzʾʾn or tʾzyʾn, as the Northern Zhou era-name 大象 daxiang (EMC *da’/daj^h-zɨaŋ’) seems to be written in line 12. The same era is named in the epitaph of Shi Jun, where it is written tʾy zʾnw (with silent final -w, see Yoshida Reference Yoshida, de la Vaissière and Trombert2005: 63).

8–9. The date given here is that of the death of the wife. The nineteenth of the second month in the first year of daxiang, a pig year, corresponds to 1 April 579 ce.

10. kykʾʾn, the name of the deceased wife, is otherwise unknown and has no obvious Sogdian etymology. The Chinese equivalent is Jijiang 紀姜 (EMC *kɨ’/ki’ kɨaŋ), which looks like a genuine Chinese name consisting of two common surnames. The Chinese text also gives her surname Kang 康, which indicates that her family originally came from Samarkand.

δβʾnmʾh is presumably a mistake for δβʾmʾnh “lady” or a similar form.

For wnxrʾk, the name of Kekan's father, see Lurje (Reference Lurje2011: 413, no. 1331).

12. This is the only date which is also given in the Chinese version. It corresponds to 12 November 580 ce. According to the Chinese text, this is the date when the couple were buried together. The long gap between death and burial – in Kekan's case more than a year, four or five years in that of her husband – suggests that the Sogdians in China followed the Chinese tradition of selecting an auspicious date for the burial.

13 tʾrʾ (c)[n]n mʾmh nyzβrt “(her) body was taken away from (her) *mother”. Comparison with lines 5–6, tʾrʾ cnn ʾBYʾ nyzβrt “(his) body was taken away from (his) father”, suggests that mʾmh, which occurs again in line 15, must be a variant of mʾth “mother”, probably a childish form with reduplication of the first syllable: “mama”. One could imagine that it was merely conventional to mention the father in this idiom in the case of a man and the mother in the case of a woman. However, the expressions surely imply that the parents mentioned – Nanai-vande's father and Kekan's mother – were still alive at the time of the burial of their children. The addition here of the phrase kʾw ʾBYʾ sʾr “to (her) father” seems to indicate that Kekan's father had predeceased her, while the fact that Kekan's mother survived her daughter is further emphasized by the phrase pʾw mʾmh “without (her) *mother” in line 15. Possibly the couple both died young; it may be significant that, unlike the case of the Shi Jun inscription, no children are mentioned as having been involved in the construction of the tomb for their parents.

15 kʾw sʾcy wyʾʾk “in a suitable place”: cf. sʾcʾw wyʾʾkh “id.” at the end of the Shi Jun inscription.Footnote ¹⁴ Since there seems to be no equivalent expression in contemporary Chinese epitaphs, it is possible that this phrase is connected with the religious beliefs and customs of the Sogdians. In Zoroastrianism it is extremely important that a dead body should not come into contact with earth, fire or water but must be laid in a “suitable place” such as the stone couch which is typical of the Sogdian burials in China. Professor Almut Hintze has kindly drawn our attention to certain Avestan passages which refer to taking the corpse to the “lawful room” (dāitiia- kata-, Vidēvdād 5.11) or “lawful (place)” (dāitiia-, scil. gātu-? Vidēvdād 5.40).

ʾyw zʾyh “(in) one (piece of) ground”, i.e. in the same place as her husband. The Chinese text also mentions that the couple were buried together.

wʾ(sn)t, if this is the right reading, may be a mistake for *wʾstnt “they laid (her)”, or possibly a deliberately abbreviated form due to the lack of space at the end of the inscription. As in line 8, the object is not expressed. Although the verb ʾwšt is the intransitive equivalent of ʾwst, the alternative reading wʾ(š)t, lit. “she stood”, could hardly be understood as having passive meaning “she was laid”, for which a periphrastic form would be expected.