Early Zoroastrianism and Orality

Oral Tradition, 35/2 (2022):199-210 

Most readers of Oral Tradition may not be overly concerned with the oral transmission of premodern compositions. Those who study the religious texts of the ancient Zoroastrian religion, however, must now take the long period of oral transmission of these texts, and its implications for our understanding of its contents, very seriously indeed. A survey of what is now known about the history and transmission of these texts, therefore, has a rightful place in a volume on orality in Iranian cultures. The present article will offer a brief overview of academic approaches to the study of orality in early Zoroastrianism, Zoroastrianism and its background, the texts and their history, the Middle Persian translation of Avestan texts, the appearance of a written Avesta, the priestly transmitters of religious texts, the process of teaching and learning, and finally some other implications of the oral background of some of our texts.

Academic Approaches to the Study of Orality in the Avesta

Zoroastrianism is the main religion of ancient, pre-Islamic Iran, and still survives in Iran, India, Pakistan, and in the diaspora, mainly in Anglo-Saxon countries. The study of Zoroastrianism traditionally focused on the limited corpus of religious texts that have been preserved in writing. Among these, the Avesta, which is widely thought of as the sacred book of Zoroastrianism, in fact consists of a number of heterogeneous texts in an Old Iranian language (Avestan). Another group of texts is transmitted in a Middle Iranian language, Pahlavi or Middle Persian. Given the relative difficulty of reading texts in these languages, the study of Zoroastrianism was long left mostly to philologists, who treated the extant texts as if they had originated as manuscripts. It is now generally recognized, however, that all Avestan and many Pahlavi texts were transmitted orally for a long time before being written down. This recognition is not yet reflected by the scholarly attention given to this vital fact. Only a handful of contemporary Iranists are studying questions related to the implications of an early oral transmission for our understanding of these texts and, more generally, of premodern Zoroastrianism.

Among these, the Harvard Iranist P. Oktor Skjaervø is an adherent of the school of thought that regards the findings of Milman Parry and Albert Lord, based on a Serbo-Croat epic


tradition, as prescriptive laws informing all oral composition.1 His attempts to apply the “principles of oral literature” (Skjaervø 1994:205) to the Avesta resulted in several articles, one of which, “Hymnic Composition in the Avesta” (Skjaervø 1994), was praised by the French Iranist Jean Kellens because: “It proposes a scheme of the history of the Avestan texts based on the common laws of oral literatures” (1998:490).2These two philologists, in other words, have accepted a generalization of Parry and Lord’s theories and proceeded deductively to apply theories based on the practices of modern Serbo-Croat epic poets to the religious traditions of the ancient Iranians. The problem, in my view, is not only that the Avestan texts are not epic poetry, but also that in the case of Zoroastrian texts we are dealing with questions of transmission rather than composition. Priests were painstakingly trained to recite these religious texts without a single error or deviation from the text as they were taught it. As is well known, even such measures cannot guarantee that a text remains unaltered indefinitely, but the conditions are far removed from those of Serbo-Croat bards. The findings of Parry and Lord, in short, simply do not apply here.

Another approach can be found in the work of Almut Hinze, some of whose work seeks to ascertain to what extent principles or techniques that are found in other oral traditions are applicable to the parts of the Avesta. In her article “On the Literary Structure of the Older Avesta” she considers the question of orality and argues that the Gāthās (on which see below) are instances of ring composition (2002:39-46).

The present writer was trained as an Iranist and historian of religions. Apart from studying ancient Iranian texts, he has led an “oral history” type project on modern Zoroastrianism in India,3 and studied the contemporary, largely oral religious traditions of the Yezidis and Yārsān (Ahl-e Haqq) of Iraqi and Iranian Kurdistan.4 His experience of these modern traditions led him to call into question the relevance of some of Parry and Lord’s theories for “fixed” religious texts whose recitation contains no element of extemporizing, and which cannot by any means be described as epic in character. He advocates a more inductive approach to the study of oral traditions, seeking to learn the conditions governing the transmission and function of religious texts in each culture individually, and challenging commonly accepted views on the genesis and history of certain compositions. Elsewhere (Kreyenbroek 2015) he has shown that one cannot understand the ancient Zoroastrian tradition without assuming the existence of a strong and more or less uninterrupted, “priestly” tradition, informed by, but largely independent of, the transmitted texts. The results of part of his research in this area are presented here.

1 Skjaervø 2012:4, fn. 1: “The beginning of modern study of oral literature is often placed in 1928, which is when Milman Parry presented his two doctoral theses on Homeric style to the Faculty of Letters of the University of Paris. A couple of years later, Parry moved to Harvard University, where he continued his studies, but died of an accident six years later. By then he had acquired and trained Albert Lord as his student, who also became a professor at Harvard and continued Parry’s work. Among Harvard scholars who have explored orality over the last decades is Lord’s student Gregory Nagy, to whom I owe my own introduction to the field.”

2 Translation from French by the present author.

3 See Kreyenbroek 2001, which was based on fieldwork in India during frequent visits, 1994-96.

4 See, for example, Kreyenbroek 1995a, which was partly based on field research in Northern Iraq in 1992.



The common ancestors of the speakers of Iranian and North-Indian languages are thought to have moved away from the wider group of Indo-Europeans perhaps around 3000 BCE, and to have developed a distinctive shared religious culture which continued to evolve when the group split up into “Proto-Indians” and “Proto-Iranians,” perhaps around 2000 BCE. The Iranians reached their present homeland (modern Iran and most of the Kurdish-speaking regions to the west, and modern Afghanistan, Tajikistan, Turkmenistan, and Uzbekistan in the east) in the course of the second and early-first millennia BCE.

Zoroastrians regard themselves as followers of Zarathustra (Avestan: Zarathushtra,Greek: Zoroastèr), who probably lived in the eastern Iranian regions some time before 1000 BCE. Zarathustra was a learned priest in the Indo-Iranian tradition, who reinterpreted the meaning of traditional religious lore and preached that the iniquitous social conditions of his time were the result of a cosmic struggle between the powers of Good and Evil. He taught that it was the task of human beings, and the purpose of their existence, to align themselves with the powers of Good in order to fight Evil. In the course of time this message led to the genesis of a new “religion” (in the sense of a formal bond between men based upon their beliefs), which first spread and developed in the Eastern Iranian lands. Eventually Zoroastrianism also became prominent in the West under the Achaemenid dynasty (550-330 BCE), from the time of Darius I (ruled 522-486 BCE) onwards. Western Iran gradually became the center of Zoroastrian culture, and the place where the tradition that is represented by our sources developed further. After the defeat of the Achaemenids by Alexander the Great (330 BCE), Zoroastrianism lived on. It became politically influential again under the Sasanian dynasty (224-651 CE). With the coming of Islam in the seventh century CE, Zoroastrianism gradually lost territory, and by the tenth century it had been marginalized and reduced to minority status. A considerable community fled to India, where they eventually thrived and became known as Parsis (“People from Persia”).

Texts and Traditions

The (oral) religious texts of the Indo-Iranians probably included hymns to praise the gods, and “mantic” poetry: by formulating a hidden truth in precisely the right way, it was thought, the poet might acquire power, more or less compelling the divine beings to do as he wished. Both these genres are represented in the Avesta. The Yashts (whose authors are not named; see below) are hymns of praise to the divinities, whilst Zarathustra’s own “songs,” the Gāthās, clearly represent the “mantic” genre.5 Zarathustra formulated the Truth as he understood it, and thus sought to compel or persuade the divine beings to grant his wishes because of the truth of his words.

Over time, the Gāthās, together with another composition, the Yasna Haptanghāiti, became the core of the liturgy of the central ritual of Zoroastrianism, the Yasna. Other texts were

5 On the Zoroastrian Yashts as heroic poetry and the Gāthās as magical formulations of truth, see Thieme 1957, especially p. 95.


added to this liturgy which, under the name Yasna, became part of the extant Avesta. The Yashts form another major component of this work, as does the Vendidād or Vīdēvdād, a long composition of heterogeneous texts which originally appear to have belonged to a learned tradition, but later came to be used as part of an extended liturgy. Another text that can be added to the Yasna liturgy is the Vispered, a somewhat repetitive, formulaic text.

An important difference between the Gāthās, the Yasna Haptanghāiti, and certain prayers on the one hand, and the other Avestan texts on the other, is the form of the language of the extant texts. The language of the first group appears to be much older than that of the second, which suggests that these texts came to be memorized very precisely(perhaps word-for-word or even syllable by syllable) at an early stage of their existence. A possible explanation for this is that they were regarded as particularly holy or powerful, so that the need was felt to repeat them exactly as Zarathustra may have uttered them. The language of the other texts, on the other hand, reflects a much later stage in the development of Avestan, and may have evolved alongside the natural spoken language for centuries. Still, the language of these Old Iranian texts stopped well short of evolving to a “Middle Iranian” stage, which suggests that they became “fixed” a long time before they were committed to writing in their present form (Kreyenbroek 1996:224-26). Until the emergence of a special “Avestan” alphabet some time under the Sasanians, none of the scripts commonly used by Iranians were capable of rendering the precise sounds of Avestan, so that any hypothetical early written text could not have been read from the page by those who did not already have the text by heart.

This state of affairs begs the question as to how and when the sacred texts, which must have been transmitted orally in a vast region over a considerable length of time and may therefore have shown a range of local variants, could come to be “fixed” in one particular form for further oral transmission. Given that children of priestly families began learning the texts by heart (following their own family tradition) at a very early age (see below), and given the poor communications at the time, it would have been practically impossible to make an existing priesthood give up the tradition they had memorized in favor of a newly instituted, unified “received version.” The only way such a process is imaginable is a scenario where a small, authoritative group of priests taught these texts to another group of transmitters who had no prior knowledge of them. This would have been the case when Zoroastrianism first became influential in Western Iran, under Darius I.

At that time, the religious tradition of the Western Iranians, deriving from the Indo-Iranian one, had its own professional priestly caste, the Magush or Magi, while the earlier inhabitants of their land, the Elamites, also had their own professional priesthood, the Shatin. When Zoroastrianism became prominent in western Iran, its liturgical language remained Avestan, whilst the inhabitants spoke cognate but quite different languages (Old Persian and Median). Τhe ancient Iranians, it seems, did not have the theoretical knowledge that is required for teaching a new language, so that it would have been very difficult for inhabitants of Western Iran to acquire an active knowledge of Avestan. At an early stage, the priests celebrating Zoroastrian rituals in Western Iran may have been immigrants who were native or at least fluent speakers of Avestan. When Western Iranian priests accepted Zoroastrianism and sought to continue their role as priests, they were presumably taught the texts by a small group of Avestan-speaking teachers (Avestan: aēϑrapaiti,Pahlavi: hērbed) and learned the texts by heart exactly as


they were taught. This means that, whilst they could officiate at rituals and recite the Avestan liturgy, they no longer had the capacity to add to or alter the memorized texts to any significant degree. As a result, the “Young Avestan” texts reflect a stage of linguistic development that is similar to that of Old Persian of the Achaemenid Inscriptions (fifth century BCE), which would be commensurate with a “fixation” around that time (Kreyenbroek 1996).

The Zand and the Written Avesta

In the course of time comprehension of Avestan became increasingly difficult for speakers of Western Iranian languages. The priesthood solved this problem by developing a simple way of translating the Avestan texts by using one standard translation for each Avestan word. Although not quite automatic, this relatively simple system may have been informed by the requirements of oral transmission. This translation, known as Zand or “exegesis,” was taught as the basis of “advanced priestly studies (hērbedestān).”6 Eventually, commentaries by prominent teachers became a fixed part of the transmission and were memorized along with the actual translation.

The Zand probably played a key role in the development of a considerable corpus of orally transmitted religious literature in Middle Persian. Some texts that had no known Avestan equivalent were nevertheless thought of as being based on lost Avestan originals (Boyce 1979:136-38). Others show a very limited percentage of Avestan texts with their Pahlavi versions, and consist mainly of priestly comments in Pahlavi. Only a few Pahlavi texts are attributed to a historical author; many were probably transmitted orally for some considerable time, with their language evolving with the natural language of their times. An exception, it seems, is the actual Zand of the Avesta, whose language has been shown to be more archaic in character than that of other Pahlavi works (Cantera 1999). This suggests that the text was either committed to writing earlier than other texts or, more probably, that it was taught with particular care. A final redaction of many Pahlavi texts probably ensued in the ninth and tenth centuries CE.

As was said earlier, in Sasanian times a system was developed for writing Avestan. This key development resulted in a written version of the sacred texts, which came to be thought of as a single “book,” known as Abestāg, a word that may have meant “testament” (Sundermann 2001; other proposed etymologies also exist), and whose use may have been inspired by the Christian term “Testament,” since no such book had existed in Zoroastrianism until then, and a name had to be found for it.

The Zoroastrian tradition soon adapted its narrative to the existence of a written Avesta by claiming that copies of such a book had existed in ancient times but were destroyed or stolen by “Alexander the Accursed,” or Alexander the Great (Boyce 1984:114). Because only a limited number of manuscripts probably existed in Sasanian times and the traditional oral way of teaching and learning continued, the existence of a written Avesta does not seem to have profoundly affected the religious tradition as a whole for several centuries. In oral transmission,

6 See Kotwal and Kreyenbroek 1992:15-18.


the testimony of the various Avestan texts could not easily be studied and compared, and perhaps because of this, all religious “judgments” by highly qualified priests were held to be equally valid even if they contradicted one another. There is no evidence to suggest that this approach ever changed during the Sasanian period.7 Several centuries later, in the work of the ninth-century high priest Manushchihr, we see that theological thinking had advanced to the extent of accepting that only one opinion could be the true one; however, the only criteria cited in such matters was the number of qualified priests supporting the various opinions rather than, for example, an objective comparison of the contents of the Avesta (Kreyenbroek 1994:12).

Transmitters and Traditions

The Zoroastrian priesthood is hereditary. Sons of priestly families were taught to recite Avestan in their home environment (or at least by a family member) from an early age. Presumably, they were taught texts that were most likely to be needed in everyday ritual practice. As to higher priestly education, an Avestan text on this subject with Pahlavi translation and commentary, the Hērbedestān, has been preserved (Kotwal and Kreyenbroek 1992). This work (Herb.12-14)8 states that “advanced priestly studies (hērbedestān)” entailed a three-year-long period of study with three different priests (Kotwal and Kreyenbroek 1992:62-63), a course of study that probably included memorizing more “advanced” texts, exegetical texts such as the Zand, a wider tradition of priestly knowledge (see below), and learning to deal with the religious questions and problems of the laity. There is evidence to suggest that, at an early stage in the history of Zoroastrianism, one of the main parts of the curriculum was a form of “exegesis” of the Gāthās and the most important prayers (Kreyenbroek 2015). This “exegesis,” however, mainly seems to have connected words of the ancient texts with the fundamental teachings of the faith (for example, “God is the Lord”), rather than explaining the actual meaning of the texts. As was shown elsewhere (Kreyenbroek 2015),a more complex and wide-ranging system of priestly teachings must have complemented the information found in such texts.Later, when the Zand had come into existence, it can be shown that this word-for-word translation of Avestan texts, together with the priestly comments that had become part of the text, became the core of hērbedestān teaching (Kreyenbroek 1987), while the transmission of further priestly knowledge presumably continued in this setting.9

When studying the various discourses that constituted the “Zoroastrian Oral Tradition,” one finds that these must have included at least three main categories, each with some sub-categories:

7 It was still adhered to under Khosrow I; see Kreyenbroek 1994:10.

8 The following abbreviations will be used here: Av. for Avestan; Herb. for rbedestān;k. for karde; Ner. for rangestān; v. for verse; Phl. For Pahlavi; Y. for Yasna: Yt. for Yasht.

9 Pace Skjaervø (1994:203 with n. 7), I do not think the later Zand was a direct continuation of the early form of exegesis exemplified by the Bagān Yasn. The ways of explaining the texts are fundamentally different; the Bagān Yasn connects parts of prayers with key beliefs; the later Zand essentially translates the Avestan texts.


1. Religious texts

            — Religious texts memorized verbatim

            — Religious texts in “freer” transmission

2. Priestly knowledge

            — Exegesis of texts

            — Wider religious knowledge

3. Lay discourse on religion

            — Discourse that repeated and confirmed traditional knowledge, myths, stories

            — Discourse on religious questions, prompted by priests

The fixed transmission of the Gāthās, and at a later stage of all Avestan texts, has been discussed above. Before the fixation of the Young Avestan texts, one question is how free the transmission of these texts was. As will be shown below, Avestan texts that were recited as part of a liturgy needed to be pronounced absolutely correctly, and priests were trained to be word-perfect. It is therefore unlikely that the transmission of these texts was very free. Given that the language of the texts evolved along with the natural language of its speakers, however, it seems likely that, instead of the extremely meticulous (syllable-for-syllable?) way of teaching the Gāthās, a somewhat freer method (perhaps line-by-line, or verse-by-verse; see below) was used.

Lay discourse on religion is rarely reflected by the texts, but the way priestly teaching could affect the laity is shown by the history of the movement of Mazdak, a sixth-century priest who pointed out at a time of famine that hoarding food and forms of social inequality ran counter to the teaching of the religion as represented by the Zand. His preaching led to a popular revolution in Iran, which initially caused the king, Kawād I (ruled 488-531 CE), to accept Mazdak’s teachings. The traditional priesthood and the nobility positioned themselves against Mazdak, accusing him of heresy, and deposed the King. Kawād I eventually returned as king and gave up his support of Mazdak. The movement was then defeated by the king’s heir, the later King Khosrow I (ruled 531-79), who had Mazdak killed (c. 524). Khosrow was evidently so alarmed by the potential results of popular religious belief based on the Zand, that he forbade the priesthood to teach the Zand to the laity.10


Whilst early “exegetical” texts seem to be rather perfunctory as regards the meaning of texts (see above), they stress the importance of correct pronunciation. The texts should be pronounced “without insertions or displacements [that is, of sounds, syllables, or words].” When one recites a prayer without error, this will help one to cross into Heaven (Y.19.5-6).The emphasis on proper pronunciation, rather than understanding, points to a reception of these texts as being effective primarily as pronouncements, rather than vehicles of meaning, a development that could be described as “manthrification” (from Av. mąϑra, “holy pronouncement”). This trend continued or increased when the Avestan texts became harder to understand, particularly when the center of Zoroastrianism shifted to Western Iran. In later times at least, priests turned to

10 Zand i Wahman Yasn II.1-4; see Kotwal and Kreyenbroek 1992:17-18. This ban was never revoked and probably played an important role in the developments that eventually led Zoroastrianism to become a mainly orthopractic religion.


the Zand (and presumably to the priestly tradition)for meaning, whilst the Avestan texts were perceived predominantly as holy pronouncements. A further consideration was that mispronunciation and other errors in the recitation were regarded as sinful (Herb.13).

The Process of Teaching and Learning

On the method of advanced teaching and learning in post-Achaemenid times, we find the following in a Pahlavi comment (Herb.13.4): “. . . if he (the student) pronounces it with him [the teacher] three times and he can recite it the fourth time, then it is considered that he has memorized it” (Kotwal and Kreyenbroek 1992:69).

According to an earlier text (Y.19.6), the oral learning process had four stages:

  • First the candidate spent time “concentrating (fra.mar-)” on a sacred text, that is, studying it without uttering it, probably by listening to the teacher’s recitation.
  • Then he recited the holy words softly (drǝṇjaiia-), careful not to make them sound like a formal utterance (where an error would be sinful).
  • After mastering this stage, he went on to recite (srāuuāiia-) the text aloud, formally but mainly in a ritual context.
  • Finally, he was allowed to perform (yaz-) it during a formal ritual.

Another relevant point is that the Nērangestān (26.1, 27.1) refers to the failure to recite the Gāthā texts for half a year as a punishable sin. This may be connected, in an oral tradition, to the need to keep repeating these difficult texts in order to avoid lapses. In contemporary (traditional) Zoroastrian practice the Yasna liturgy is still recited without reference to a written text, and some priests recite the Gāthās together informally several times a week so as to avoid errors.

Units of Text

Avestan texts (Y.57.8; Ner.5.1, 2) state that the Gāthās were recited “in verse-lines, in stanzas (afsmaniuuąn vacastaštiuuat̰),”11which seems to reflect the way the divisions of these texts were perceived and formally recited (even though they might initially have been memorized syllable-by-syllable or word-for-word). This division into verse-lines and stanzas applies to other poetic Avestan texts than the Gāthās (that is, those that originally had a syllabic meter),12 even though later changes in meter (perhaps the shift to a stress meter) obscured the original meter and led to additions which no longer followed the original poetic rules.

In many cases, a number of stanzas belonging together form a karde (Av. yasnō.kǝrǝti, Phl. kardag). A karde generally has characteristic initial and closing formulae (for some

11 See Kreyenbroek 1985:80.

12 So Ner.6.1, referring to the “act of worship (Yasna)” generally.


exceptions see Hinze 2014). A collection of kardes, or other longer sequences of poetic texts, could be known by various names depending on the text they belonged to: in the Yasna such sections are known as hāiti or ;in the case of the hymns to the Divinities (yazata), each of these is now referred to as a Yasht.13 In prose works such as the Vendidād and the Nērangestān, a “chapter” is called a Fragard.

That this division of poetic texts does indeed reflect the way such divisions were perceived and memorized is suggested by the occasional presence of a verse-line in more than one place; this occurs more often in the case of stanzas,andsometimes the same, or a very similar karde may crop up in more than one Yasht. This indicates that these units represented a reality in the perception of priests (minor differences between versions being due, one imagines, to variations in priestly transmissions in different localities).

The Origin of the Yashts

The extant hymns to individual Divine Beings show many variations. Some are extremely long (Yt.10 has 145 stanzas) and contain material that may go back to the Indo-Iranian past, while others are very short (Yt.21 has only two stanzas). In some of the shorter Yashts (Yt.16, 18), moreover, the text has no obvious relation to the Yazata to whom it is dedicated, and seems to be “borrowed” from hymns to another Yazata.

The sequence of Yashts to various divinities, and also other standard enumerations of divinities (Siroza), correspond to that of the Yazatas to whom the days of the Zoroastrian calendar are dedicated. In all probability, the Zoroastrian calendar was instituted in Achaemenid times (Hinze 2014). It seems plausible, therefore, that the extant Yashts were composed or compiled in Achaemenid times on the basis of existing texts (it was apparently not possible to compose new texts for calendar divinities for whom no such texts existed), and taught to the newly Zoroastrianized Western Iranian priesthood (see above).

In the Yashts the meaning of both verse-lines and stanzas is usually clear and coherent, and the same is true of most—though not all—kardes. The structure of many Yashts (that is, the sequence of kardes), on the other hand, shows a very different picture. There, the same passage may occur twice in the same Yasht, and again in a different hymn.14 Contradictory images occur next to each other,15 while kardes dealing with the same aspect of a Yazata’s nature may be strewn more or less all over a Yasht interspersed with kardes dealing with other aspects of his or her personality.16 It is interesting to note that the passage found in Yt.10.31-34, and again in

13 In the texts themselves, the word yasht may refer to what is now known as a Yasna.

14 For example, Yt.10,k.8.v.31, k.13.v.56; Yt.10.v.55 is partly repeated in Yt.8 and Y.11.

15 For example, Yt.10.31, where Mithra is invoked by name, and v.55 (= v.74), where he complains that men do not call him by name; v.55 is followed by v.56 (= v.31); in Yt.19, the divinity Ardwīsūr is described both as a river and as a statue.

16 For example, Mithra’s warlike aspects are stressed in Yt.10.k.2; k.3 (which repeats part of k.2); k.5, second half; k.9; k.11; k.18; k.19, second half; k.28; k.31, second half; k.33. The kardes between these texts mostly deal with other aspects of Mithra’s being.


v.56-59, may originally have been an independent prayer to Mithra.

It seems legitimate, in view of all this, to wonder whether the original texts used to praise the Yazatas were not much shorter than our extant Yashts, and may in fact have consisted of one or more kardes. This would explain the use of the relatively short (seven stanzas) prayers (niyāyesh) to the sun and moon as Yashts for these beings; presumably the prayers were already well known and were not felt to require lengthening. The contents of longer Yashts, some of whose kardes or stanzas may appear more than once, often in slightly different forms, suggest that extant Avestan texts were collected, possibly from priests representing different lines of transmission, in order to compile the hymns as we have them now.

It may also have been at this time that some simple pious statements, to the effect that the great God, Ahura Mazdā, was superior to the recipient of the Yasht, or repeating other key teachings of Zoroastrianism, were added to compositions whose contents mainly praised the Yazad alone, parts of which may have originated in much earlier times.

Some Further Implications of Early Oral Transmission

It seems important to stress that, given the originally oral transmission of the texts, the state of the grammar of a composition cannot be taken to be an indication of its date of origin, as is often maintained by Iranist philologists. As was pointed out earlier, the language of many texts evolved for a long time before it was “fixed.” Moreover, “learned” texts, such as the Vendidād and the Hērbedestān and Nērangestān, which were long transmitted by a single teacher to his pupils, typically show more grammatical “errors” than texts that were used in the liturgy, in which more than one priest always participated. In the latter case, the priests presumably corrected each other’s mistakes, so that the texts remained grammatically purer.

Many Iranists have sought to assign a date to the composition of the hymns on the basis of their language. The notion that it is possible to establish a clear point of origin of most oral compositions may well be incorrect; it seems more probable that a core of ancient verses was added to (and perhaps partially omitted) over centuries until it was finally “fixed” in the Achaemenid period.

Oral transmission necessarily implies continuity. The fact that Zarathustra used mantic poetry of a kind that goes back to Indo-Iranian times shows that an unbroken chain of transmitters of that tradition must have existed at the time of the prophet and taught him his craft. In ancient Indian poetry, the name of the composer of a religious poem is often mentioned (Jamison and Brereton 2014:9). It is extremely likely that this was also usual in the Iranian tradition before Zarathustra. The fact that no composers other than Zarathustra are mentioned in the Avesta can therefore be interpreted as a case of “conscious forgetting,” perhaps informed by a strong sense of disruption caused by the exceptional status of the prophet and his “religion.”

Georg-August University Göttingen



Boyce 1979
Mary Boyce. Zoroastrians: Their Religious Beliefs and Practices. London: Routledge & Kegan Paul.

Boyce 1984
______. Textual Sources for the Study of Zoroastrianism. Manchester: Manchester University Press.

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

Hintze 2002
Almut Hintze. “On the Literary Structure of the Old Avesta.” Bulletin of the School of Oriental and African Studies, 65:31-51.

Hintze 2014
______. “Yašts.” In Encyclopædia Iranica. Online ed. http://www.iranicaonline. org/articles/yashts

Jamison and Brereton 2014
Stephanie W. Jamison and Joel P. Brereton. The Rigveda: The Earliest Religious Poetry of India. 3 vols. Oxford: Oxford University Press.

Kellens 1998
Jean Kellens. “Considérations sur l’histoire de l’Avesta.” Journal Asiatique, 286:451-519.

Kotwal and Kreyenbroek 1992-2009
Firoz M. Kotwal and Philip G. Kreyenbroek. The Hērbedestān and Nērangestān. 4 vols. Paris: Association pour l’Avancement des Études Iraniennes.

Kreyenbroek 1987
Philip G. Kreyenbroek. “The Dâdestân î Dênîg on Priests.” Indo-Iranian Journal, 3:185-208.

Kreyenbroek 1994
______. “On the Concept of Spiritual Authority in Zoroastrianism.” Jerusalem Studies in Arabic and Islam, 17:1-15.

Kreyenbroek 1995
______. Yezidism: Its Background, Observances, and Textual Tradition. Lewiston, NY: Edwin Mellen.

Kreyenbroek 1996
______. “The Zoroastrian Tradition from an Oralist’s Point of View.” In K. R. Cama Oriental Institute, Second International Congress Proceedings. Bombay: K. R. Cama Oriental Institute. pp. 221-37.

Kreyenbroek 2001
______. Living Zoroastrianism: Urban Parsis Speak about Their Religious Lives. Richmond, UK: Curzon.

Kreyenbroek 2015
______. “Religious Knowledge in Oral and Written Tradition: The Case of


Zoroastrianism.” In Wissen in Bewegung: Institution—Iteration—Transfer. Ed. by Eva Cancik-Kirschbaum and Anita Traininger. Wiesbaden: Harrassowitz. pp. 93-103.

Skjaervø 1994
Prods Oktor Skjaervø. “Hymnic Compositions of the Avesta.” Die Sprache, 3.2:199-241.

Skjaervø 2012
______. “The Zoroastrian Oral Tradition as Reflected in the Texts.” In The Transmission of the Avesta. Ed. by Alberto Cantera. Iranica, 20. Wiesbaden: Harrassowitz. pp. 3-48.

Sundermann 2001
Werner Sundermann. “Avesta und Neues Testament.” In Philologica et Linguistica: Historia, Pluralitas, Universitas. Festschrift für Helmut Humbach zum 80. Geburtstag am 4. Dezember 2001. Ed. byMaria G. Schmidt and Walter Bisang. Trier: Wissenschaftlicher Verlag Trier. pp. 258-64.

Thieme 1957
Paul Thieme. “Vorzarathustrisches bei den Zarathustriern und bei Zarathustra.” Zeitschrift der Deutschen Morgenländischen Gesellschaft,107.1:67-104.