AVESTA ii. Middle Persian Translations



ii. Middle Persian Translations 

The manuscripts transmitting the preserved Avestan texts often also include their translation into Pahlavi [PT], the Zoroastrian Middle Persian language. The ritual Avestan texts belonging to the great rituals (Yasna, Vīsparad, and Vīdēvdād) are transmitted through two different kinds of manuscripts: the Sāde manuscripts, containing only the Avestan text, and the so-called Pahlavi manuscripts, which include the Pahlavi translations. The ritual texts of the minor ritual are transmitted in a more complicated way, with manuscripts containing different series of texts with or without Pahlavi translation. These are limited to only some of the Yašts (Y. 1, 3, 6, 7 11, 14, 20, 21) and to the rest of the texts known as Xwardag Abestāg (Niyāyišn, Gāh, Āfrīnagān, Sīrōzag). Finally, the preserved texts belonging to the Sasanian great Avesta (Hērbedestān, Nērangestān, Hādōxt, Vaēθā) that do not prescribe ritual activity are all transmitted together with the Pahlavi translation.

We know that the Sasanian great Avesta (as described in Dēnkard; see also AVESTA i) included not only the Avestan text, but also a translation into Pahlavi, because its description in the Dēnkard is based on the PT and because the manuscripts of texts deriving directly from the great Avesta always included the PT. On the other hand, the collection of the ritual Avestan texts originally did not include the Pahlavi translation, since the translation plays no role in the Zoroastrian ritual. Nevertheless, part of the extant manuscripts of the ritual Avesta also include the PT. So at a certain time a PT was also added to the ritual manuscripts. In most of the cases the PT of the great Avesta was adapted to that of the ritual texts with only minor changes. In the case of the only text belonging to the great ritual that was at the same time one Nask of the great Avesta, Vīdēvdād, this was quite easy, as most of the labor was already done. In the case of the Yasna collection the translation process was more difficult. To a certain extent, the PT could be taken from the PT of the great Avesta: Stōd Yasn (Y. 14-16, 22-27, 28-54, 56), Baγ (Y. 19-21), Hādōxt (Y. 58), Bayān Yasn (Y. 9-11, 57). As for the rest of texts included in the Yasna collection, we are not sure whether they were included in the great Avesta, and thus we cannot know if their PTs were taken from the great Avesta or newly created. The date of the adaptation process of the PT of the great Avesta to the ritual Avesta is uncertain. For the Yasna, the colophons of Mf4 and Pt4 inform us that the adaptation took place before 1020 (Cantera and De Vaan, 2005). For the rest of the texts we do not have evidence of such an early date. Since the extant manuscripts of the Pahlavi—Vīdēvdād and Vīdēvdād Sāde—go back to the same written tradition (Humbach, 1973), the joining of the Avestan text and its PT must have taken place before the 13th century. The same is true for the Vīsparad transmission. Therefore, we can conclude that this process started in the 10th century at the latest and concluded not later than the 13th century. The situation is more complicated for the Xwardag Abestāg and the Yašts. The different translations were added to the Avestan texts probably between the 12th and 15th century. Some translations were taken from the great Avesta (Ohrmazd Yašt, Sīrōzag, Āfrīnagān) or from later translations (Hušbām, Niyāyišn), while others had to be newly composed at this very late date (Yts. 3, 14; on Yt. 14, see König, 2012). 

The PT of the great Avesta used for the adaptation was the result of a previous canonization process. During the Sasanian period there were different exegetical schools with different translations and commentaries of some Avestan texts. The translations and opinions of these schools were transmitted mainly orally, but the existence of written versions even before the canonization process is not to be ruled out completely. Some of them, such as the Vīdēvdād version of Medōmāh, were able to survive even after the canonization process. Although a canonical translation was chosen from among several translations and commentaries, alternative interpretations and exegetical debates of past and contemporary authorities were, nevertheless, also incorporated. Past authorities were alluded to by mentioning the name of the pōryōtkēšān (authorities of the former times) responsible for the interpretation. Most probably, the final result of the canonization process was written in order to establish it in fixed form, and it is likely that the invention of the Avesta script (see AVESTAN LANGUAGE i) and the writing down of the Avesta was closely related to the process of canonization of the PT. We do not have a sure date for this process, but it could have taken place throughout the 6th century, since historical facts and persons such as Ādurbād, Mazdak (see IRAN ix. RELIGIONS IN IRAN (1) Pre-Islamic (1.1) Overview and SASANIAN DYNASTY), and Zurwandād mentioned in the commentaries are never later than this date, and the same is true for the different exegetes alluded to (although here not all scholars agree; cf. Gignoux, 1995, Cantera, 2004, pp. 207 ff.; Secunda, 2012). 

The canonized translations were not homogeneous. The translations of the different texts were produced at different times, and only minor changes and commentaries were introduced later. In fact, a linguistic and philological analysis reveals chronological differences even among the translations included in the great Avesta. We distinguish at least three chronological levels: (1) old translations like the Vīdēvdād, Nērangestān, and Hērbedestān; (2) new translations of texts partially integrated in the great Avesta, but with a definitive writing around the 9th century, probably in order to adapt the translation of the ritual Avesta to the Avestan text (such as the Yasna, Ohrmazd Yašt, Sīrōzag, and Āfrīnagān); (3) a further level includes all the translations of ritual texts composed after the Arab conquest without counterpart in the great Avesta. Among the old translations, Vīdēvdād is probably the oldest one. The language of the old translations and especially the PT of Vīdēvdād shares several archaic traces with the Manichean Middle Persian and with the Middle Persian of the inscriptions, which allows one to place the date of the translation around the 3rd to 5th centuries (Cantera, 1999). 

The general organization of the PT is similar in all the translations. The translators divided the Avestan text into sentences with complete meaning and embedded the PT between the Avestan sentences. To each Avestan word there corresponds a word in the PT, and the word order of the Avestan text is mostly preserved, even though it is not always the conventional one in Middle Persian. Very often glosses explaining some words are included, words necessary to make the resulting Pahlavi text understandable (such as auxiliary verbs). Besides the glosses, short explanatory commentaries are also added, mostly introduced by the conjunction or ēd kū. These are especially frequent in the old translations and in others like the Yasna, Sīrōzag or Āfrīnagān. Moreover, in the PT going (almost) directly back to the great Avesta (mainly Vīdēvdād, Nērangestān, and Hērbedestān), long commentaries dealing mostly (but not always) with practical matters (ritual or legal) appear. In these commentaries, very often exegetical disputes among different opinions and schools are reproduced, so that they are our principal source of information about the exegetical traditions of the Avesta in Sasanian times. The distribution of the long commentaries is not totally clear. Although they appear predominantly in the translations stemming from the great Avesta, several facts remain unclear, for example, why long commentaries are very frequent in Vīdēvdād 1-9, but very rare in 10-22 (with the exception of Vd. 15 and 16). 

The technique of the translation is very peculiar. It is a word-by-word translation. To each Avestan word a Pahlavi equivalent must correspond. The word order of the Avestan is also kept. The translation of every word is mostly consistent, even in different texts. Therefore the existence of an academic school of translators must be assumed. Despite the tendency to homogeneity in the translation, polysemy is often recognized. There is a conscious endeavor to use, when possible, etymological correspondences of the Avestan words, and a considerable number of loanwords from the Avesta are found in the translation. Nevertheless, termini technici of the religious language are mostly translated, even if the loanword was usual in the religious language.  Since the vocabulary used in the translation is not always usual in the Pahlavi literature, very often explanatory glosses are necessary. Actually, the PT is more than just a word-by-word translation; it is an attempt to understand and translate the real meaning of an Avestan text. This implies in many cases a reinterpretation of the Avestan text divergent from its original intention. This is especially clear in the case of the translation of the Yasna (Josephson, 2003). 

The quality of the PT and the advantages of its use as a tool for the understanding of the Avesta has long been a subject of discussion. At present, this discussion has been definitely settled. For a long time it has been known that the quality of the different translations is not homogeneous: the old translations are more correct than the recent ones. Concerning the old translations we can state that, although the exegetes’ knowledge of Avestan grammar was far from being as good as ours, they were not at all ignorant of important aspects of it. However, their grammatical analyses are informative about their awareness, or lack of it, of Avestan grammar, but not about Avestan grammar itself. The PT is more useful in the field of lexicography. Since the translators very often give the meaning of Avestan words accurately, the PT is a useful tool, especially when comparative and philological methods do not allow sure results. As for the interpretation of passages, it is evident that the PT is the result of an exegetical process in Sasanian times that involves an adaptation of the contents of the Avesta to a broadly new theological situation, so that the PT can hardly be adduced as an argument for interpretation of a passage. 

The PT of the Avesta seems to be the result of the combination of two different exegetical traditions: (1) the Indo-Iranian exegesis of the sacred texts, and (2) the Jewish Bible exegesis, especially the Babylonian Talmud. On the one hand, the Pahlavi translation continues the Indo-Iranian tradition of linguistic and rudimentary philological analysis of the sacred texts. This tradition was highly developed in India and only scarcely in Iran, but several common practices of linguistic analysis reveal their common ancestral origin (Cantera, 2004; 2004a). The linguistic analysis was part of the oral learning of a text, and in Iran, when the Avestan language was no longer spoken or even understood, this analysis was substituted or complemented by a translation into the vernacular language. In fact, it seems likely that other translations in other Iranian languages existed. On the other hand, the similarities and parallelisms between the Avesta exegesis and the Bible exegesis, especially the Babylonian Talmud, are very significant. Both traditions share a main focus on legal debates and on the exegesis of the legal parts of the sacred texts, as well as the same dialectical pattern of reproducing the opinions of different exegetes, whether anonymous or mentioned by name. Also from the chronological point of view there are significant coincidences: in both cases there is a long oral exegetical tradition that was fixed around the end of the 5th century (Babylonian Talmud) and the 6th century (Avesta exegesis). The interrelation of these exegetical traditions has received little attention in the past, despite their similarities. However, in May, 2007, an international conference on “The Talmud in its Iranian context” was held at the University of California, Los Angeles (see Bakhos and Shayegan; see also ZAND Ī FRAGARD Ī JUD-DĒW-DĀD).  


B. T. Anklesaria and D. Kapadia, Pahlavi Vendidâd (Zand-î jvît-dêv-dât). Transliteration and Translation in English, Bombay, 1949. 

Carol Bakhos and Rahim Shayegan, eds.,  The Talmud in Its Iranian Context, Tübingen, 2010.

A. Cantera, “Die Stellung der Sprache der Pahlavi-Übersetzung des Avesta innerhalb des Mittelpersischen,” Studia Iranica 28, 1999, pp. 173-204. 

Idem, Studien zur Pahlavi-Übersetzung des Avesta, Wiesbaden, 2004a. 

Idem, “El análisis del lenguaje en la tradición oral indoirania,” Actas del V Congreso de Lingüística General. León, 5-8 de marzo de 2002, ed. M. Villayandre Llamazares, Madrid, 2004b, pp. 513-22. 

Idem and M. de Vaan, “The Colophons of the Avestan Manuscripts Pt4 and Mf4,” Studia Iranica 34/1, 2005, pp. 31-42. 

K. Dehghan, Der Awesta-Text Srōš Yašt (Yasna 57) mit Pahlavi- und Sanskritübersetzung, Munich, 1982. 

B. N. Dhabhar, Zand-i Khurtak Avistak, Bombay, 1927. 

Idem, Pahlavi Yasna and Visperad. Bombay, 1949. 

Idem, Translation of Zand-i Khåurtak Aviståak, Bombay, 1963. 

Y. Elman, “Orality and the Redaction of the Babylonian Talmud,” Oral Tradition 14/1, 1999, pp. 52-99. 

Idem, “‘Up to the Ears’ in Horses’ Necks (B.M. 108a): on Sasanian Agricultural Policy And Private ‘Eminent Domain’,” Jewish Studies: an Internet Journal [JSIJ] 3, 2004, pp. 95-149. 

Idem and Daphna Ephrat, “Orality and the Institutionalization of Tradition; the Growth of the Geonic Yeshiva and the Islamic Madrasa,” in Y. Elman and I. Gershuni, eds., Transmitting Jewish Traditions: Orality, Textuality, and Cultural Diffusion, 2000, pp. 107-37. 

P. Gignoux, “La controverse dans le mazdéisme Tardif,” in A. Le Boulluec, ed., La controverse religieuse et ses formes, Paris, 1995, pp. 127-49. 

H. Humbach, “Beobachtungen zur Überlieferungsgeschichte des Awesta,” Münchener Studien zur Sprachwissenschaft 31, 1973, pp. 109-22. 

H. Jamasp, Vendidâd. Avesta with the Pahlavi Translation and Commentary and Glossarial Index, Bombay, 1907. 

J. Josephson, The Pahlavi Translation Technique as Illustrated by Hōm Yašt, Uppsala, 1997. 

Idem, “Remarks on the Pahlavi Translation of the Gāthās,” Studia Iranica 32, 2003, pp. 7-34. 

Idem, “The Importance of the Pahlavi Translation Texts for a Middle Persian Dictionary,” in C. Cereti and M. Maggi, eds., Middle Iranian Lexicography, Roma, 2005, pp. 357-73. 

M. F. Kanga, “A Pahlavi Commentary on the Ašəm Vohu,” Bulletin of the Asia Institute of Pahlavi University 2, 1975, pp. 1-20. 

Idem, “Pahlavi Version of Gatha Spənta Mainyu Yasna Has 47-50,” Bulletin of the Asia Institute of Pahlavi University 1-4, 1976, pp. 45-96. 

J. Kellens, “Considerations sur l’histoire de l’Avesta,” Journal Asiatique 286, 1998, pp. 451-519. 

G. Klingenschmitt, “Die Pahlavi-Version des Avesta,” ZDMG, Supplementa I,3. Deutscher Orientalistentag vom 21. bis 27. Juli 1968 in Würzburg, ed., W. Voigt, Wiesbaden, 1969, pp. 993-97. 

Idem, “Der Beitrag der Pahlavi-Literatur zur Interpretation des Avesta,” Münchener Studien zur Sprachwissenschaft 37, 1978, pp. 93-107. 

Goetz König, “The Pahlavi Translation of Yašt 14,” in A. Cantera, The Transmission of the Avesta, Wiesbaden, 2012, pp. 495-518.

F. M. Kotwal and P. G. Kreyenbroek, The Hērbedestān and Nērangestān, 3 vols., Paris, 1992-2003. 

D. P. Sanjana, The Zand î javît shêda dâd. The Pahlavi Version of the Avesta Vendidâd, Bombay, 1895. 

Shai Secunda, “On the Age of the Zoroastrian Sages of the Zand,” Iranica Antiqua 47, 2002, pp. 317-49.

S. Shaked, “The Traditional Commentary on the Avesta (Zand): Translation, Interpretation, Distortion?” La Persia e l’Asia Centrale da Alessandro al X secolo (Atti dei convegni Lincei, 127), Roma, 1996, pp. 641-56. 

Z. Taraf, Der Awesta-Text Niyāyiš mit Pahlavi- und Sanskritübersetzung, Munich, 1981. 

(Alberto Cantera)

Originally Published: March 9, 2015

Last Updated: March 9, 2015

Cite this entry:

Alberto Cantera, "AVESTA ii. Middle Persian Translations,"Encyclopædia Iranica, online edition, 2015, available at http://www.iranicaonline.org/articles/avesta-02-middle-persian-translations (accessed on 09 March 2015).