Bactrian is an East Iranian language written in Greek script, and served as a principal administrative language of what is now Afghanistan during the first millennium ce. Until the 1990s when more than 100 Bactrian documents appeared on the market, our knowledge of the language was based on only one inscription from the second century ce, generally known as Surkh Kotal Inscription. Since then N. Sims-Williams has been the only scholar who has provided us with texts and translations of the documents and other inscriptions written in this language.
Of the two books under review, hereafter BD I2 and BD III, the second concludes the series of publication announced by Sims-Williams in the original version of BD I2 (BD I published in 2001, cf. ibid., p. 7); the second of the series, BD II, appeared in 2007 and contains letters and Buddhist texts. Since its publication in 2001, many points of BD I have been corrected and revised mainly by Sims-Williams himself and partly by the reviewers. As BD I is now out of print Sims-Williams took this opportunity to produce the revised version, incorporating all the corrections and improvements as well as editions of as many as ten legal or economic documents which have appeared since 2001 and have been published in separate articles. However, the “Bibliography”, “Notes on orthography and grammar”, “Glossary”, etc. have been omitted, since they have been superseded by the corresponding sections of BD II. As a result BD I2 is a slimmer book at 177 pages (as compared with 255 pages of BD I) comprising “Preface” (7–8), “List of documents” (9–21), “A note on the presentation of the texts” (22–23), and “Texts and translation” (25–171).
In BD III, 230 plates of monochrome photographs of the documents are preceded by such sections as “Preface” (7), “Catalogue of documents” (9–31), and “List of plates” (33–37; second pl.66a on p. 34 is a misprint of pl.66b), etc. Obviously when BD III was edited, BD I2 did not exist and the references to page numbers of the text editions found in BD III, pp. 12–19 are those of BD I rather than BD I2. As stated by Sims-Williams (BD III, p. 7), the photographs other than those of the Khalili collection are very varied, because they were not prepared for this publication but were provided by the owners. In less satisfactory cases one finds it very difficult to check the readings due to poor quality of photographs (e.g. document bd) and sometimes because the parts containing texts are hidden under creases and bents, which are not well smoothed away, e.g. document cj, line 11. In one case the size of a manuscript (document je) seems to be unknown. In view of the present situation in which the manuscripts are scattered all over the world and the present whereabouts of some manuscripts are not known, the author's effort to make the manuscripts as accessible and readable as possible must be highly appreciated. While colour plates would enhance the readability of manuscripts considerably, the price would make the volume far less accessible to those who are interested. A case in point is plate 229; its colour photograph published in S. Fogg, Manuscripts of the Silk Road, London 2004, no. 14 looks much better. For the colour plates of some other documents, see also Comptes rendus de l'Académie des Inscriptions et Belles-Lettres 2002, p. 1051 (document O) and O. Bopearachchi et al. (eds.), De l'Indus à l'Oxus. Archéologie de l'Asie Centrale, Lattes 2003, p. 393 (document A).
The publication of the facsimile volume marks not only the end of a series of studies inaugurated by Sims-Williams himself in 1997 (New Light on Ancient Afghanistan: the Decipherment of Bactrian, London) but also the beginning of the new epoch of Bactrian studies. He writes in one of his recent articles: “Until the 1990s the Bactrian language was known from a very limited range of sources … however, the discovery of more than 150 Bactrian documents … has transformed our knowledge of the language, as well as providing an invaluable new source for the history of Afghanistan and adjacent regions”, cf. “Turks and other peoples in the Bactrian documents”, in From Ötüken to Istanbul, 1290 Years of Turkish (720–2010), Istanbul 2011, p. 15. We now have texts, translation, glossary, and plates for almost all the manuscripts as well as full-fledged description of the grammar of Bactrian. A monograph on Bactrian onomastics is also to be mentioned as a companion volume: Bactrian Personal Names, Iranisches Personennamenbuch II/7, Vienna 2010 [= IPNB II/7]. However, one must still wait for the work by J. Lerner, who will publish all the sealings accompanying the documents. Before the publication of BD III, not many documents were reproduced and very few specialists knew how letters and ligatures of the Bactrian cursive script, previously referred to as Hephthalite script, are to be read. Acquaintance with these genuine documents will certainly enable the detection of some fakes being sold in the markets, which could mislead even a most competent scholar, cf. G. D. Davary's inclusion of a fake document in his article published in H. Eichner et al. (eds.), Iranistik in Europa – Gestern, heute, morgen, Vienna 2006, p. 59, Fig. 14.
In what follows I draw attention to some minor points of BD I2 which seem to merit discussion. Since BD III allows one to check the texts published in BD II, I also include a few comments on BD II here.
BD I2, p. 22: One may expect the repetition of the explanation of : (colon) found in το:γδο (Tt25) published in Bulletin of the Asia Institute [= BAI] 15, 2001 [2005], p. 27b.
p. 26: After the publication of BD II Sims-Williams changed his mind and proposed to translate πιδοναμγο as “copy”, cf. BAI 19, 2005 [2009], p. 196. Here, his old translation is repeated and translations of the other instances (K16, Nn6, Tt6, and aa5) are not revised either. This seems to imply that Sims-Williams abandoned the proposal and returned to his old idea, but the translation of “copy” may be supported by Sogdian pyšπn'myk “id.” attested in a Mug document.
p. 31, Aa1: Although Sims-Williams transcribes [χ]Pονο, the facsimile (BD III, plate 4a) clearly shows the initial χ.
p. 69: N'9 αρηγο and N'10 µασο are misprints of μαρηγο and ασο. Incidentally, one finds a sentence “Where … question-marks” printed twice in the preface, pp. 7–8.
pp. 80–81, O2 οτομο λαδδηιο χοησαοο “I have given you an undertaking”: Here and in other places the older translation “I have given an undertaking” has been revised with the recognition of the indirect affectee construction, cf. Sims-Williams “Differential object marking in Bactrian”, in A. Korn et al. (eds.), Topics in Iranian Linguistics, Wiesbaden 2011, pp. 23–38, esp. p. 34 with n. 26. Moreover, one may wonder if some of the other examples listed under the past optative (BD II, pp. 45, 226) also belong to this construction. Thus, δογγο πιτιιαμο οδο μαχο πιταδδηιο could equally well be translated as “we guarantee thus, we have (hereby) guaranteed you that …”.
pp. 92–93, R8: What Sims-Williams reads now in BD I2 is ιθααυασιδο and is the same as that of his edition published in BAI 15, 2001 [2005]. However, in BD II, p. 43 he gives ιαδαυαηιο instead as a reading better than ιθδδηιο of BD I, while one finds this very ιαδαυηιο in BD I2, Tt7. In light of plate 60 of BD III it now becomes clear that ιαδαυηιο of BD II, p. 43 is a simple error for ιθααυασιδο.
pp. 94–95, S5 βονοσογολιγο “Bunchuqlïgh”: In IPNB II/7, p. 87 s.v. no. 249, Sims-Williams follows P. Lurje's suggestion and transcribes the word as Bunsuglig, which refers to a Sogdian settlement, i.e. βονο “land, landed property” + σογολιγο “Sogdian”. It is not clear which reading Sims-Williams prefers now. Nevertheless, “Sogdian hypothesis” is really attractive because the Sogdians' presence further west in Merv is recorded during the period when document S, dated to 26 February 693 ce was written in Guzgan, cf. de la Vaissière, Sogdian Traders, Leiden 2005, pp. 182–83. For the omnipresence of the Sogdian traders or their settlements, one may also refer to sūlī biśa “Sogdian village” attested in a Khotanese document, cf. O. Skjærvø, Khotanese Manuscripts from Chinese Turkestan in the British Library, London 2002, p. 132, b 8. I owe this reference to Prof. Rong Xinjiang. On the Khotanese word biśa translating Chinese cun 村 “village”, see Duan Qing, Journal of Inner Asian Art and Archaeology 3, 2008, pp. 65–70. I differ from Duan in not taking biśa for a locative sg. of bis″- “house” but for a so far unknown independent noun biśā- meaning “village”.
p. 142, Y6: ιωπαριο “altogether” looks more like ιωπαριιο, which seems to be supported also by its etymology *ιω “one” + 2παρο + ιγο.
p. 149, ab20 with n. 195: A flourish preceding α- of αζαδοϕαρδαρο is compared with a similar flourish found in document A, which looks like a numeral κ′ (= 20). However, the shape of the flourish here looks more like the one encountered in aa36; it is curious to note that in both cases the signs precede signatures.
p. 165: Apart from Δραχμο and ηβοΔαλαγγο, those words containing the sequence -δγ- (μορδγο of ll. 3, 11 and αβηδγο of l. 11) appear to show a special δ (delta) form.
BD II, pp. 15, 27: Sims-Williams is of the opinion that document jf is almost certainly a draft, because it lacks an address and bears no sealing, and since the contents suggest that two different recipients are addressed. Cf. also Sims-Williams, in A. Panaino et al. (eds.), Proceedings of the 5th Conference of the Societas Iranologica Europæa … Vol. I, Milan 2006, p. 703. Nevertheless, since parchment cannot have been cheap in those days, one may not expect a draft to be written on such a material. It seems to me that document jf was a real letter put in an envelope bearing the address. There is nothing strange about one single letter addressed to more than one addressee. One good example is Text F edited by N. Sims-Williams and J. Hamilton, Documents turco-sogdiens, London 1990, pp. 51–61. One of the Sogdian Ancient Letters was enclosed in an envelope bearing the address, cf. Encyclopaedia Iranica, II/1, 1985, p. 7. I wonder if a parchment currently housed in the Afrasiab museum which bears a sealing and an address in Sogdian is another example of an envelope, cf. F. Grenet and E. de la Vaissière, Silk Road Art and Archaeology 8, 2002, pp. 179–180 with figures 10–12.
p. 125, ja12: ωγοτο looks more like ωσο το. Cf. ωσο of line 5. What Sims-Williams reads κιρο shows a short vertical stroke between ρ and ο. Can it be ι, i.e. κιριο? For a similar shape of ι see ναμοοι[νδο] in line 1. If this reading is correct, κιριο could be a 2nd sg. optative form, cf. BD II, p. 42.
p. 163, xp16: I prefer to read κοαδοο αζο for Sims-Williams's κοαδο μαζο. My reading would eliminate an exceptional usage of the prohibitive particle μα, cf. BD II, p. 47.
Comparing Sims-Williams's texts with the plates published in BD III, one will no doubt be impressed by how much he was able to read from these faint and damaged documents written in the so-far not well known language. Thus, we feel all the more obliged to him for his enormous effort and his formidable competence which have made the Bactrian materials available and accessible to all those who have interested in the language and history of Pre-Islamic Bactria in such a handsome way and within such a short period of time. As I stated above, with the three volumes of BD I–III we have entered a new epoch of the Bactrian studies.
