I met Chiara Barbati long ago in Italy, because we studied at the same University (“Sapienza” University of Rome), but it is only once we had both moved to Vienna that we became friends. She is now a researcher at the Institute of Iranian Studies at the Austrian Academy of Sciences and works on Sogdian, which is probably well-known to many of us because of the many Buddhist texts being translated from Sanskrit (or, in a less amount, from Chinese) into Sogdian. All the others will be perhaps surprised to know that Sogdian exists (almost) only as a corpus of translations, from Syriac (in the case of Christian texts), from Sanskrit (Buddhist texts) or from Middle Persian (Manichean texts).

Q1. EF: What is your current project about?

CB: The project is aimed at identifying and contextualizing the emergence and the development of a Christian Iranian book culture as a result of the cultural-religious activities carried out by the Christian Iranian communities in the Turfan oasis (present-day Xinjian, Uyghur Autonomous Region, China), during late Antiquity and the early Middle Ages. Some background information: About 30–35,000 fragments have been found in Turfan, most of which extremely short [see Q2], in about twenty different languages and scripts. The Christian community in Turfan used Sogdian as a language of cultural exchanges among Christian communities. Thus, they never adopted Middle Persian (at the same age of the Sogdian documents, one already finds Neo Persian ones, so that it appears that the Middle Persian was already outdated as a spoken language and its choice by other communities was due to other reasons), whereas the Manichean (and, back in Iran, the Zoroastrian) church used Middle Persian. The project focuses on a corpus of nearly five hundred fragments in Sogdian language in East Syriac script and circa fifty fragments in Sogdian secular script. Furthermore, it examines also five hundred fragments in Easy Syriac script (out of which 400 are kept in Berlin, Turfan Collection and 100 are in Saint Petersburg).

Q2. EF: How do you feel about having to work with such a restricted corpus? Don’t you miss whole texts?

CB: Every single day. If I happen to find a fragment with 20 lines I have to be more than happy and this is surely frustrating. One works on some pages of the Bible in Middle Persian from the Turfan oasis, some inscriptions on the so-called Nestorian crosses in Pahlavi,
a short rock inscription…No other manuscripts were found, apart from Turfan, although we know —through Syriac witnesses— that there were Christian communities in Iran and that the Bible has been translated in Middle Persian around the 3rd or 4th century. One cannot avoid hoping that an archaeological expedition not focusing on Persepolis will finally uncover some documents even in Iran. Apart from the middle age zoroastrian literature and the Neo Persian one, all we know about Iran is due to fragments and we scholars need to be able to fruitfully compare small indications, e.g., a Middle Persian Psalter found in Turfan, although in the X c. Middle Persian was neither a liturgical language (the liturgical language was Syriac, although out of Syriac sources we know that local languages were admitted for some parts of the Mass) nor a spoken one (Neo Persian was already in use). Thus, one can speculate that the Psalter has been brought to Turfan by traders on the Silk Road or by missionaries accompanying them. The ductus of the Psalter resembles that of a cross found in today’s Afghanistan, so that this could be its origin…It is very frustrating, but it is also fun to reconstruct the puzzle.

Q3. EF: How many languages does one need to know in order to (try to) reconstruct the puzzle?

CB: Theoretically many: Middle Persian, Parthian, Bactrian, Sogdian, Khotanese… Since I focus on the phenomena of interculturation shown through Syriac elements in the Sogdian book culture, I also had to learn Syriac. People who focus on Buddhist Sogdian will have to know Sanskrit and in some case Chinese, too. In fact, the Sogdian corpus consists of translations and, thus, the resulting Sogdian is very much influenced by the source language. This is proved by the fact that the same language, at about the same time and in the same area looks very different according to the source language it translates. The Manichean Sogdian is easier, because the source language (Middle Persian or Parthian) is closer, since it is also a middle persian language of about the same period. Mani used Middle Persian and other middlepersian languages to spread his religion, because he understood himself as the “seal between Jesus, the Buddha and Zarathustra” and thus wanted his religion to be understood. It is not clear whether the invention of the so-called Manichean script (a script based on the Aramaic one) is also due to Mani’s circle as a part of this project (perhaps in order to characterise the new religion with a clear identity marker —-although other scripts were also used for missionary purposes— this topic is currently investigated in a PhD research project in Berlin). The Christian Sogdian has also a similar origin, since Christians also wanted their texts to be spread and understood. Within the Iranian area, the situation of the Zoroastrian religion is, instead, quite different, since Pahlavi texts are deliberately conservative, since one can only be born a Zoroastrian and conversions are not allowed.

Q4. EF: Until now, we have spoken of languages and of the scripts used to represent them. However, you also focus on other elements in your fragments, such as punctuation and the like. Which role do they play in your cultural reconstruction?

CB: A very important one. I have been, e.g., working on the small crosses drawn on some folia and have found them in Syriac and Sogdian texts. Where do they come from? Some scholars believe that the smaller ones mark the versum of manuscripts, but if so, why do they occur only on some folia? Perhaps they marked the end of a quire, but it is difficult to ascertain it, since we only have small fragments. Do they depend on the scribe? On the scriptorium? Are they the imprimatur sign of a certain monastery? These are the questions I will try to answer in my current project. As for punctuation, Christian Sogdian texts tend to follow the Syriac punctuation. Because it was more authoritative or because there was no Sogdian one? Some elements are found also in Sogdian Buddhist texts, e.g., four points building a square at the end of a text to indicare a pause. The opposite corners are, respectively, black and red and I am currently working on the meaning of such an alternation. Also the small points in Manichean texts probably had a meaning, but it is not clear what is the direction of the borrowings (Manichean and Christian texts are more or less contemporary). Similarly shared is the number of pages per quire and this time it seems that a certain usage has been brought to Turfan by the Christians.

Q5. EF: What needs to be done now in your area? Which priorities would you set?

CB: We can now build upon over one hundred years of decipherment works. The decipherment of the fragments, due to their paucity is thus almost completed and the critical editions are accurate. It is now time to reconsider the whole corpus and investigate on
1. which phenomena are due to the influence of the language one is translating,
2. which ones are due to contacts (as I said before, we must remember that those were bi- or trilingual communities)
3. which ones are common developments in the Iranian area.
For instance, certain phenomena are found in the Eastern Iranian area, but also in Iran, centuries before. This proves that the Eastern Iranian area remained for centuries more conservative. But how long? Other elements are found in some rather conservative dialects of central Iran (see my forthcoming article in the proceedings of the 7th International Conference of the Societas Iranologica Europaea). Such comparisons would enable the distinction between phenomena due to interferences (contact or translation) and ones due to inner developments of the Iranian Asia. As for 1, the influence of the language one was translating has often been underestimated. For instance, W.B. Henning has studied the sociolinguistic of Sogdian and decided that the Manichean Sogdian corresponded to a higher sociolinguistic level and the Christian one to a lower one. Among the elements he considered are the frequency of periphrastic constructions in Christian Sogdian. However, these differences might be explained also through the fact that the Manichean religion had its own Iranian terminology, whereas the Christian one needed to create one and thus used periphrastic verbs (e.g., “cross” + “to put” to translate “crucify”) and could thus avoid Syriac loanwords. Further periphrastic constructions (such as the periphrastic preterit of some verbs) could be explained diachronically rather than diastritically. It is also important to remember that these varieties of Sogdian were never spoken languages and that in this sense sociolinguistic considerations need to be supplemented with a translation-studies analysis. But in order to do that, one needs to master many languages and to be aware of the religious and historical background.

EF: One also needs to master many scripts…

CB: The various scripts might be of help, since for instance the Manichean and the Sogdian secular scripts are based on the Aramaic one and, thus, do not record vowels. By contrast, whenever Sogdian is transcribed in Syriac script (a script also derived from the Aramaic one, but with added diacritics), we can gather more information about the vowels through the diacritic signs.

Q6. EF: It is fascinating to see how you start from concrete elements in languages and scripts but interpret them in order to reconstruct a wider cultural scenario.

CB: Yes, through the study of linguistic phenomena and of scripts one can reconstruct the relations between communities and their nature. Were they due to trade? Did they imply bilingualism? An instance is the Sogdian Psalter written not in the eastern Syriac script (the standard one for Christian texts), but in Sogdian secular script [see above, Q 2]. Why writing it in a secular script, usually employed for letters, bills, etc.? Perhaps because the text was meant for a wider diffusion outside the monastic communities, e.g., for traders? This is confirmed also by the fact that the Psalter is the text more translated in the various languages of Turfan (Uighur, Neopersian and other spoken languages of that time). And also the most translated ones among Mani’s Psalms have been written using the Sogdian secular script instead of the Manichean one. Thus, through the script one can understand the reception and diffusion of a text. Through a linguistic study, one can further reconstruct that there were trilingual communities around Turfan, since after a certain point of time, turkish was also spoken there. In fact, one finds Turkish-Sogdian documents, bilingual letters in which the main language is Sogdian, but the morphology is halfway between the two and one notices phenomena of code-copying. Going back to the small crosses referred to above [Q 4], after some of them one finds a sign I interpret as the Syriac word hayé ‘life’ (I discuss this topic in my forthcoming book, The Sogdian Lectionary in E5). Why does the word occur only in some cases? Does this have to do with the kind of text (this does not seem to be the case)? Or with the scribe? The presence of the word hayé is anyway not out of place because the cross has a different meaning in the Oriental Churches, where it is never depicted with the dying Jesus on it and it is rather a symbol of life and glory.

Readers who work on manuscript fragments from Gandhāra and the like are also encouraged to answer Question 2!
For more on scriptoria and their cultural role, see also this interview.

This post is part of my series of Interviews. If there are additional questions you would like to ask or if there is someone, either a specific person (i.e., yourself), or a representative of a given category (e.g., “A scholar of Nyāya”) you would like me to interview, please let me know.