Springe direkt zu Inhalt

Corpus Avesticum Berolinense: Project Description

Overview:

1 State-of-the-art and preliminary work

1.A The loss and regaining of the ritual Avesta within Avestan Studies

1.B The manuscripts

1.C The view of the Avestan transmission and the different goals of the editions

1.D. Preliminary work

2 Objectives and work programme

2.1 Anticipated total duration of the project

2.2 Objectives

2.3 Work programme incl. proposed research methods

2.4 Data handling

3 Bibliography

1       State-of-the-art and preliminary work

Zoroastrianism probably evolved in Eastern Iran from the beginning of the first millennium, spreading to Western Iran and the Near East in Achaemenid and Sasanian times. Up to this very day, most Zoroastrian rituals are celebrated in Avestan, an Old Iranian language. The extant Avestan texts are predominantly those recited in these rituals. They have come down to us in myriad manuscripts copied over the course of the past 700 years that mostly present the texts in their concrete liturgical shape. These manuscripts usually served as a guide for the performance of a ceremony, or were used to teach such performances.

Any edition of these texts would be expected to reflect their function and transmission, but for several and very different reasons (the late antique transmission of the texts; the modern scholarly perspective on the Avesta), none of them does so. As a result, the liturgical nature of the Avestan texts passed on to us is not taken into account by any of the modern Avesta editions. The scholars edited the texts shorn of their ritual context; many of them were even arranged in a different order to the one used in the performances, and the ritual variations are not edited at all, since they were considered merely secondary re-arrangements.

In recent years, our understanding of the Avestan texts has changed drastically, mainly through the work of Jean Kellens. Today, it has become clear that almost all Avestan texts were specially drafted for the purpose of ritual performance. The implications, however, of this crucial turnaround in our understanding of the Avestan texts and their transmission have not yet become apparent in the more recent textual editions. Although the need for a new edition of the Avestan texts that substitutes the 19th century editions has been widely felt since the second half of the 20th century, new editions have only partially contributed to an even more distorted picture of the Avestan texts, and focused solely on linguistic aspects, while still based on the data of the old editions and maintaining their textual arrangement. By contrast, the proposed new edition aims to approach the texts as they are used in the rituals, to put them back into their Sitz im Leben, and present them as they were and are still used by Zoroastrians. The new edition, furthermore, distances itself for the first time from the data of the old editions, and is the result of a previous collection of the available manuscripts. Our systematic search for manuscripts in recent years has indeed brought to light many new manuscripts and a new transmission branch, the Iranian one, whereas all former editions depend almost exclusively on Indian manuscripts. The comparison of both branches has revealed important differences.

Thus, the new edition Corpus Avesticum will present a radical change, namely, a ‘concrete’ picture of the Avesta in which the Avestan texts are conceived as coherent texts used in historical rituals. It will provide an entirely new and solid basis for the study of ancient Iranian ancient Iranian language and religion. Furthermore, it intends to be the reference for all studies related to the history, intellectual and material culture and religion of ancient Iran.

1.A. The loss and regaining of the ritual Avesta within Avestan Studies

As from the Avesta translation by Anquetil-Duperron (1771), it was common practice in the 19th c. to edit the Avestan texts in their entirety. However, some decisions were taken that were never justified or criticised, but which have profoundly shaped our understanding of the Avestan texts:

a)    The Long Liturgy was edited not according to the ritual but to exegetical manuscripts, leading to an abstract modularization of the text and the loss of its ritual variability

b)    All the texts that seemed to be unrelated to the Long Liturgy were brought together under the (modern) designation "Xorde Avesta"

c)    The text of all those rituals that are (shorter) variations of the Long Liturgy (the so-called “Outer Ceremonies”) were totally ignored in the editions/translations.

These decisions meant that the true textual reality, i.e., the ritual structure and performativity of the texts, became obscure. An Avesta was created in the shape of a holy scripture (see for example the title of Geldner’s edition: “die heiligen Schriften der Parsen”), which was seen as the surviving fragment of a great late antique collection of 21 canonized books, the so-called "Greater" or "Sasanian" Avesta, described in the Middle Persian literature. This tendency was reinforced by Haug’s discovery of two Avestan dialects. On the basis of the linguistic distinction of a seemingly older dialect of the Gāϑās, and a seemingly younger dialect of all other Avesta texts, an old textual kernel could be singled out that appeared not only as Zaraϑuštra’s own words, but that seemed to be free from all ritual ‘distortions’. On the basis of Haug’s methodological decisions, the 20th century became a time of Gāϑā philology. In addition, texts in Younger Avesta that were seen as valuable (≈ non-ritualistic) were re-edited according to the Gāϑic model. The Avesta became more or less scattered.

Rejections of this tendency surprisingly emerged within Gāϑā philology. With Humbach’s first Gāϑā translation (1959), which in some way was still consistent with Haug (“Die Gathas des Zarathustra“), scholars again became aware of the link between Old Avesta and Indo-Iranian vocabulary and phrases. This and the structural perspective of M. Molé became the basis for the ritual re-reading of the Gāϑās of J. Kellens and É. Pirart (1988-1991). This was the beginning of a slow process of “re-ritualization” of the other Avestan texts. Nevertheless, while the Avestan texts were henceforth read focusing on their usage of a sacrificial language, the presentation of the texts was not altered. One of the main goals, therefore, of the proposed edition is indeed to present the texts according to their ritual nature.

The first milestone in this process was the discovery of the antiquity of the Long Liturgy. The ritual nature of most Avestan texts is obvious, and has been acknowledged since the very beginning of Avestan studies. However, it was widely considered (and this trend became still more obvious during the 20th century) that their ritual use was secondary. The preserved Avestan texts might have been or still be used in rituals, but they were composed for other purposes. Nevertheless, research over the past twenty years has outrightly challenged this view. The turning point was an article by J. Kellens (1998), in which he pointed out that the extant Avestan texts represent two ritual collections, i.e., the Long Liturgy and a collection of shorter rituals, and that neither of them is of a secondary nature or depends on the Great Avesta (be it as secondary editions as Geldner would have it, or as randomly “saved fragments”, as with Bailey), but that both refer to rituals celebrated at least as early as in Sasanian times. The collections were most probably arranged in Sasanian or post-Sasanian times.

Kellens (2012) has subsequently found textual evidence of the antiquity of the structure of the Long Liturgy. Several Avestan texts describe rituals that share exactly the same structure presented by the Long Liturgy, as described in the manuscripts. Furthermore, Tremblay has revealed undeniable structural parallels between the Zoroastrian Long Liturgy and the Vedic sacrifice (his conclusions were posthumously published by Swennen, 2016). In my own work, I have shown that with the exception of certain phonetic and some minor textual and performative changes the recited Avestan text had adopted its current shape and form even before the composition of the Avestan version of the Nērangestān, i.e., at a time when Avestan was still a living language (cf. below 2.3.1.3).

Whereas the Great Avesta with its 21 books should be viewed today as something of a scholastic construct whose historical reality is questionable, the Long Liturgy is no longer thought of as a secondary or casual composition of surviving fragments, but instead a composition in its own right, and an old one that is the ritual context for the performance of most Avestan texts. It thus occupies an absolutely crucial position among these texts and is a key to a better understanding of the earliest periods of Zoroastrianism. Even yet modern editions have failed to publish the Long Liturgy as a liturgy. Likewise, performative information (accompanying actions, variants depending on the ritual context, etc.) has been completely ignored by scholars. Only the least complex variant of the Long Liturgy, the Yasna, has been edited in its entirety; and not even the Yasna editions reflect the liturgical reality as it appears in the liturgical manuscripts, and as it continues to be celebrated today.

Kellens’s recent Études avestiques et mazdéennes (2006-2011, and 2013 with Redard) are the sole exception. In these volumes, Kellens accepts the consequences of his conclusions and has sought to produce an alternative edition of the Long Liturgy. He does not confine himself to editing the Yasna ceremony, as is usually the case, but adds the Visperad variants wherever they correspond. He makes a pioneering contribution in many respects. It is the first systematic analysis of the structure of the Long Liturgy (see also the parallel attempt by Skjaervø 2007a, 2007b). Nevertheless, his editorial work has certain shortcomings that I have discussed at length elsewhere (Cantera, 2016a), all of them are attributable to Kellens’s dependence on Geldner’s edition, and hence, albeit unintentionally, on the exegetical manuscripts. He presents only two variants of the Long Liturgy. One of them, the Visperad ceremony as edited by Kellens is therefore very different to the ceremony described in the liturgical manuscripts. Furthermore, he does not reflect at all either the specifying or the contextual-determined variability (see below §2.3.1.1). It is also unfortunate that Kellens completely forgoes any information other than that recorded in the Avestan language.

A further consequence of the new understanding of the Avestan texts has involved the several attempts made to bridge the gap between the Avestan texts as they appear in the editions and the ethnographical description of the modern ritual practice of Zoroastrians. The description of the modern ritual practice for Western academia was initiated by Modi (1922). This line has been continued in the 20th and 21st centuries with descriptions of the modern performance of concrete rituals (e.g., Kotwal-Boyd’s description of the Zoroastrian ritual language), and of the totality of modern ritual practices by Stausberg (2004). Modi classifies the ceremonies into the following groups:

a. Socio-religious ceremonies and customs;

b. Purification ceremonies;

c. Initiation ceremonies and customs;

d. Inner liturgical ceremonies; and

e. Outer liturgical ceremonies.

By contrast, the traditional division divides the manuscripts in just two types, according to the rituals they contain: 1. Manuscripts of the Long Liturgy; and 2. Manuscripts containing collections of the short liturgies. Recent studies have shown that this philological perspective on the Avesta is partly the result of historical accidents. First, the proposed general distinction of the manuscripts and rituals is a scholarly hypostatization of a (late) historical situation (see König 2012; 2016a). The texts that are not celebrated within the Long Liturgy have very a different origin and usage. Their collection in manuscripts called Tamām Xorde Avesta (“the whole Xorde Avesta”) does not begin before the 16th century. The Yašts seem to be relicts from a past ritual activity that had already disappeared when the manuscripts were first produced. In fact, an internal analysis of the Yašts and the meta-ritual literature reveals different ritual contexts in which the Yašt were performed (Kreyenbroek 2008, Cantera 2013). Secondly, the (generally proposed) splitting of the manuscripts into two categories is mainly the result of the manuscripts’ status in Europe in the 18th/19th centuries. Manuscripts that did not belong either to the Long Liturgy or to the Xorde Avesta/Tamām Xorde Avesta were more or less unknown or ignored (e.g., Ms. K17 (1681), a Bāj-Ms.; Supp. Pers. 983, a Drōn Sade). Since Westergaard’s edition (1852-1854), the general arrangement of the Avestan texts (of the Long Liturgy/Xorde Avesta) has remained unaltered; the texts of all those so-called “Outer Ceremonies” (also “Shorter Liturgies“) are still unedited. The few differences between Geldner’s and Westergaard’s editions indeed reflect a sharpening of this view. Thus, Geldner eliminates from his edition the fragments of Westergaard that are texts recited in different ritual and complex ceremonies, such as the Paragṇā.

In fact, the manuscripts both present and group the rituals in a different way that is closer to the descriptions of contemporary ritual practice. Despite previous descriptions, our analysis of the available manuscripts covers the following types:

a)    Manuscripts describing the different types of the Long Liturgy, and usually only one liturgy in each manuscript. They correspond to Modi’s inner liturgical ceremonies

b)    Manuscripts including minor variants of the Long Liturgy or of parts of it, e.g., collections of Drōns or manuscripts containing different variants of the Farroxši. They correspond basically to Modi’s Outer Ceremonies.

c)    Manuscripts of the Yašt (only one example, F1).    

d)    Manuscripts of the short liturgies, which can be divided in two groups after the inclusion, or not, of the complete collection of the Yašts. This includes many texts celebrated in Modi’s groups a to c.

With the exception of the Yašts, which are a series of ritual texts with common features that no longer had a clear ritual function at the time the manuscripts were produced, but were used earlier in different kinds of rituals, the manuscripts seem to make a classification of the texts that is very close to Modi’s description of ritual practice.

The type b manuscripts, corresponding to Modi’s Outer Ceremonies, have been completely ignored in the editions of the Avestan texts. Only recently, R. Karanjia (2010), an erudite Zoroastrian priest in India, has edited the Drōn ceremony. Despite its philological shortcomings, this is the first edition of an Avestan text that takes into account the ritual nature of the texts. Karanjia presents all possible variations, and provides essential performative information. In recent lectures, König and Redard have started the analysis of the manuscripts containing Outer Ceremonies. Andrés-Toledo (2015) has found textual evidence in old manuscripts of one variant of the modern Farroxši ceremonies.

Furthermore, recent research has pointed out that the boundaries between the Long Liturgy, Outer Ceremonies and Xorde Avesta are somewhat fluid: a) the structure of the Long Liturgy might have once integrated (nearly) all Avestan texts (see König 2016b; on the inclusion of the Yašts in complex variants of the Long Liturgy, see Kreyenbroek 2004, 2008; Cantera 2013); b) the Long Liturgy is again a montage of ritual structures that could be celebrated as Outer Ceremonies (Cantera 2016a).

1.B. The manuscripts

1.B.1 The number and origin of the manuscripts

The first critical editions by Spiegel (1853, 1858) and Westergaard (1852-1854) were prepared almost simultaneously on the basis of the limited number of manuscripts available in European collections. The standard edition was made by K. F. Geldner between 1885 and 1895, and was based mainly on Westergaard’s. Geldner’s most important contribution was the greater number of manuscripts he could use. He had access to a large quantity of manuscripts from Indian collections. He used 135 manuscripts. For a long time, it has long been considered that it is no longer possible to assemble these manuscripts, as many of them have since been lost. After Geldner, only partial editions have been made, and all of them on the basis of Geldner’s data, with occasional use of a few manuscripts that are readily available. In the past ten years, the Avestan Digital Archive (ADA, led by me) has begun an intensive search for Avestan manuscripts, with astonishing results (see the description of the submitter’s previous work). With very few exceptions, all the manuscripts used by Geldner have been recovered, and many new ones have also been discovered. The number of known manuscripts has at least doubled. Most important still is the fact that a new branch of the transmission of the Avestan texts has come to light. All former editions rely almost exclusively on manuscripts written in India. A few Iranian manuscripts were known, but due to their isolation could not be properly evaluated, playing almost no role in the editions. Since Westergaard’s short visit to Iran, the extended belief was that the Iranian manuscripts had been destroyed or lost.

After a seminal article published by Mazdapour (2008-2009), which described twelve Avestan manuscripts found in Iran, ADA decided to intensify the search for surviving Avestan manuscripts in Iran. Thanks to the co-operation of certain Iranian colleagues led by K. Mazdapour (Institute for Humanities and Cultural Studies, Tehran), S. Gholami (Universität Frankfurt) and collectors such as Vahid Zolfeghari, we have succeeded in identifying and digitising more than 70 manuscripts surviving in public libraries, temples, private collections and priestly families. Today, the liturgical Iranian transmission of the Long Liturgy is better known for the early period (until the 18th century) than the Indian one. The same is true for the Outer Ceremonies. In both cases the Iranian manuscripts have brought to light ritual variants that were unknown until only recently, e.g., the Drōn ī Frawarḏin Yašt (Andrés-Toledo 2015). In the case of the short liturgies, the oldest Indian manuscripts are earlier than the Iranian ones; however, they are influenced by an innovation that did not affect the Iranian transmission: the creation of the Tamām Xorde Avesta. As for the Yašts, the Iranian manuscripts have not made any significant contribution for the time being. Our picture of the transmission of the Long Liturgy, the Outer Ceremonies and, partially, the Xorde Avesta, has been profoundly changed.

1.B.2. The manuscripts on the basis of the editions: exegetical vs. liturgical manuscripts

Concerning the manuscripts, all former editions are, furthermore, depending on a concrete class of manuscripts that hides the intimate connection between the Avestan texts and the ritual performance. At least for the Long Liturgy, we clearly distinguish between two types of manuscripts: the liturgical and the exegetical manuscripts. Both lines of transmission, the liturgical and the exegetical one, bear witness to the texts recited in the Long Liturgy, but in a different way. The liturgical manuscripts offer full descriptions of a variety of ceremonies, i.e., in addition to the Avestan recitative they include ritual instructions for the performance written in Middle Persian (in later witnesses in New Persian as well) in the case of Iranian sources, and in Gujarati in the case of Indian sources. The exegetical manuscripts, on the other hand, omit the ritual instructions, and merely pass on the Avestan recitative with a translation into Middle Persian or Sanskrit (and later on also into Persian and Gujarati).

All editions (with the sole exception of Brockhaus’ Vidēvdād, 1850) depend mainly on the exegetical manuscripts. This choice was based on different grounds that were rarely made explicit. First, the exegetical manuscripts were considered closer to the “original” non-ritual Sasanian Avesta, and better represent the “real” nature of the texts. Secondly, the exegetical manuscripts reproduce the texts in a simplified manner that avoid, among others, repetitions and minor variations, whereby the text was considered easier to edit. Lastly, they present (in most of the cases) the Avestan texts without additional ritual instructions in Pahlavi or Gujarati. Since the editors were only interested in the texts in the Avestan language and the deciphering of their contents, this presentation seemed more appropriate. Nevertheless, the changes concerning the primary nature of the texts and the antiquity of the rituals and of their descriptions as they appear in the manuscripts have led us to a drastic shift in the procedure and to the assumption of the liturgical manuscripts as the appropriate basis for the edition of the Avestan texts.

1.C The view of the Avestan transmission and the different goals of the editions

The 19th century editions (Westergaard and Geldner) were produced under a concrete view of the Avestan transmission: the Avesta was compiled in Sasanian times, but our manuscripts do not go back directly to this Sasanian archetype, but rather through a “Yazd-original” (in the words of Westergaard), written long after the end of the Sasanian period, and from which all the extant manuscripts derive. Although the ideal goal would have been the reconstruction of the Sasanian version of the text, we are often able to reproduce solely the Yazd-original.

At the beginning of the 20th century, F. C. Andreas (1903) under the influence of Spiegel, the other editor of the Avesta, postulated that the existing texts in Avestan script are simply mistranscriptions of an Arsacid archetype written in a Semitic alphabet. Accordingly, the single goal of any edition was the reconstruction of the Arsacid archetype. It can only rely on linguistic considerations (mainly a comparison with the Vedic language), and the manuscripts are of no use. Nevertheless, during the Second World War, Morgenstierne (1942), Henning (1942) and Bailey (1943) simultaneously refuted this theory, showing that 1) the transmitted Avestan texts reproduced a linguistic reality; 2) the Avestan script had been developed as a phonetic alphabet designed for the accurate and unambiguous transcription of an orally transmitted text; and 3) there is direct continuity between the first Sasanian copy and the manuscripts: the latter were “saved fragments” (Bailey) of the Sasanian collection. These three findings provided the starting point for a careful analysis and methodological revolution, as initiated by K. Hoffmann.

Hoffmann (1971; 1986; K. Hoffmann & Narten, 1989) ultimately traces our manuscripts back to a Sasanian archetype transcribed on the basis of an oral transmission and in a script specially created for this purpose. Following Islamisation, most of the manuscripts descending from this archetype disappeared, and many texts were lost, so that by around the 10th century there was only one existing specimen for each class of manuscripts, viz. the hyparchetypes to which all our manuscripts allegedly go back. The purpose of a text edition would thus consist in reconstructing the phonetic shape of the Sasanian archetype. In fact, however, textual criticism will take us back no further than the hyparchetypes. Therefore, the methodological bases for the reconstruction are, first, an analysis of the original values of the letters of the alphabet invented for writing down the Avestan texts, and second, the philological analysis complemented by the comparison with the related Vedic language.

Hoffmann’s work led to renewed and intensified editorial activity. Hintze (2012) has published a survey of existing and ongoing editions. All of them generally rely on Geldner’s information, although they occasionally consider a number of other manuscripts. They essentially reproduce Geldner’s apparatus (differently arranged) and – with some minor and almost exclusively phonetic deviations – even his text.

Hoffmann’s view has been challenged from different perspectives over the past twenty years. First, Kellens (1998; see above) broke the direct link between the hyparchetypes and the Sasanian archetype: the texts represented in our manuscripts are not fragments of the Sasanian Avesta of 21 nasks (books) described in the Pahlavi literature that was allegedly transcribed for the first time in Sasanian times. They represent an independent collection of ritual texts. The Sasanian archetype of the Great Avesta has thus become a pure entelechy. The idea of the Sasanian archetype, as conceived by Hoffman, was indeed subsidiary to the conception of the ritual use of the Avestan texts as purely secondary. Secondly, I have questioned the theoretical necessity and methodological feasibility of reconstructing the hyparchetypes (Cantera 2012a; 2014: 325 ff.). The discovery of the Iranian branch of the transmission shows, on the one hand, that different ways of reciting the Avestan text were developed in each branch of the transmission, and these differences are reflected in the manuscripts. Although the Avestan text that has usually been edited aims to represent the alleged Sasanian archetype, it represents rather the Indian version of the Avestan texts. The Iranian manuscripts, however, preserve a different version that is mostly more archaic than the Indian one. Furthermore, as the discovery of the ritual diversity (cf. § 2.3.1.1) reveals, producing all the manuscripts needed for the most diverse variants of the Long Liturgy is a much more complex process than simply copying and slightly varying sources. It is therefore fairly unlikely that there would have been a one and only archetype for all the variants of the Long Liturgy. Even if there had only been one, the extant manuscripts’ existence could not be explained by a simple process of copying.

On the other hand, there is clearly a profound interaction between the written sources and the performance of the rituals. The performance of rituals changes and develops in different ways at different places, and the manuscripts are involved in this process in many ways: sometimes as models for spreading changes, sometimes merely as attempts to prevent changes, and sometimes even reflecting a given ritual reality. This also concerns both the way of reciting the text and the way of performing it. The Avestan transmission is thus a fluid and dynamic transmission under the influence of ritual practice. Accordingly, the reconstruction of the Sasanian archetype is no longer the goal, and the reconstruction of the hyparchetypes is methodologically unfeasible (s. § 2.3.1.3).

1. D. Preliminary work

In 2008, I launched the Avestan Digital Archive (ADA) (http://www.avesta-archive.com) with the aim being to locate, digitise and publish online all extant Avestan manuscripts. The project has been running at the University of Salamanca with funding of different national research institutions for 8 years until my move to the Freie Universität Berlin (Mai 2016). Initially, ADA focused on the manuscripts of the Long Liturgy, whose search can be considered as concluded. We have located 293 manuscripts, digitized around 157 and publish online 80 of them. These numbers considerably exceed Geldner’s roughly 100 manuscripts of the Long Liturgy. For updated lists of the manuscripts known to us, see Cantera (2014: 406-417). Furthermore, we have identified around 150 of the rest of manuscript classes, but many of them, especially, in European libraries have still to be digitized. Nevertheless, the manuscripts of the most important available collections as well in India as in Iran have already been digitized, but not yet indexed and published online. The ADA has become an indispensable tool for Avestan research, whereby according to a recent statement made by Skjaervø: “The instantaneous availability of a large number of manuscripts at the ADA is a boon to Avestan and Pahlavi studies that cannot be underestimated. Today, I consider citing Geldner without consulting the manuscripts indefensible.” (Skjaervø 2016: 178).

Ever since the launch of ADA, I have been deeply engaged not only in collecting, but also in assessing the Avestan manuscripts of the Zoroastrian Long Liturgy, as well as in reflecting upon our image of the Avestan transmission and on the conclusions to be drawn accordingly for an edition of Avestan texts, especially of such texts as those belonging to the Long Liturgy. In 2012, I published a monograph (The transmission of the Avesta, Wiesbaden, 2012) in which I gathered contributions from different authors on the Avestan transmission, the collections of manuscripts, the analysis of the manuscripts, and the editorial principles in use in Avestan philology. For my part, I wrote two methodological chapters: an in-depth analysis of the shortcomings of Geldner’s method for the analysis of the genealogical relations between manuscripts, with a description of new electronic tools and methods (based on the Coherence Based Genealogical Method, Mink 2004; Wachtel 2008) I have developed for this analysis (“Building Trees: Genealogical Relations Between the Manuscripts of Wīdēwdād”, pp. 279-346); the second presents the reasons a new edition of the Zoroastrian Long Liturgy is necessary (“Why do we really need a new edition of the Zoroastrian long liturgy?”, pp. 439-478).

Over the interim years I have continued working on the analysis of the relationship between the different manuscripts of the Avestan liturgy (Cantera, 2012a, 2015). I have also supervised works on palaeography by my student Juanjo Ferrer and on orthographic uses in the Avestan manuscripts by Jaime Martínez Porro (as a PhD thesis).

The focus of my research in recent years has been the development of a new methodology for the edition of Avestan texts that takes into account their ritual nature. In fact, this was the difficulties for indexing the manuscripts on the basis of the text arrangement used in the standard editions. The differences were so deep in the arrangement, and the missing texts so numerous, that it became clear that the available editions do not provide an exact picture of the reality of the Avestan texts as presented in the manuscripts. Furthermore, I have reviewed the methodology for the selection of a single reading among the vast quantity of readings for each single word used in the 19th century editions and the new method advanced by the Indo-Europeanist Karl Hoffmann.

At J. Kellens’s invitation, I presented my findings in a series of papers read at the Collège de France in May-June 2013, and published them in a revised monographic form under the title Vers une édition de la liturgie longue zoroastrienne: pensées et travaux préliminaires (2014). They constitute the prolegomena for a new edition of the Avestan texts from a completely new approach. The edition is to address the Avestan texts as performed in ritual practice, and therefore on the basis of the liturgy, and not as has hitherto been assumed, on the exegetical manuscripts. Furthermore, the transmission of the Avestan texts is seen as a dynamic process, conditioned by constant interferences between the production of manuscripts and ritual practice, whereby the existence and reconstruction of a Sasanian archetype as postulated by Hoffmann seems dubious and methodologically untenable. The editor’s objectives have to be redefined accordingly.

1.1       Project-related publications

1.1.1      Articles published by outlets with scientific quality assurance, book publications, and works accepted for publication but not yet published.

  1. Vers une édition de la liturgie longue zoroastrienne: pensées et travaux préliminaires, Cahiers de Studia Iranica nr. 51, Leuven: Peeters (2014)
  2. (Ed.)The transmission of the Avesta, Wiesbaden: Harrassowitz (2012)
  3. The transmission of the Pahlavi Vīdēvdād in India after 1700 (I): Jamasp's visit from Iran and the rise of a new exegetical movement in Surat, Journal of the Cama Oriental Institute 2008, 81-142 (with M. A. Andrés Toledo)
  4. Talking with god: The Zoroastrian ham.paršti or intercalation ceremonies, Journal Asiatique 301 (2013), 85-138
  5. A substantial change in the approach to the Zoroastrian long liturgy. About J. Kellens's Études Avestiques et Mazdéennes, Indo-Iranian Journal 59 (2016) 139-185
  6. Building trees: Genealogical relationship between the manuscripts of Widewdad, in The transmission of the Avesta, Wiesbaden: Harrassowitz (2012), pp. 279-346
  7. Why do we really need a new edition of the Zoroastrian long liturgy? in The transmission of the Avesta, Wiesbaden: Harrassowitz (2012), pp. 439-477
  8. Die Manuskriptologie der Avesta-Handschriften, Handbuch der Iranistik, ed. L. Paul, Wiesbaden: Reichert, 2013, pp. 345-351
  9. The problems of the transmission of the Avesta and the tools for Avestan Text Criticism (TATEC), Analysis of ancient and medieval texts and manuscripts: Digital Approaches, Lectio 1, Leuven: Brepols (2015), pp. 95-116
  10. On Avestan text criticism (2): the accusative singular of the ū̆- and u̯a- stems in the long liturgy, E. Pirart (ed.), Etudes de linguistique iranienne in memoriam Xavier Tremblay (Acta Iranica 57), Leuven: Peeters (2016), pp. 89-156

 

2       Objectives and work programme

2.1       Anticipated total duration of the project

The project is expected to run for 12 years divided into four periods of three years. The edition of the Long Liturgy will run parallel to the edition of all the other ceremonies. The former will take up the first 10 years (three years for Y0-21; four years for Y22-59; and three years for VrS32 [~ Y60]-Y72). The other ceremonies require a first year for preliminary works (digitising certain collections, prior analyses of the manuscripts, etc.) that have already been carried out for the Long Liturgy. The Xorde Avesta will then be edited between the second and fourth years; the Yašts, between the fifth and the seventh; and the Outer Ceremonies between the eighth and tenth. The intercalated texts (Vidēvdād and Vīštāsp Yašt) will be edited in the final stages of the project. So by the end of the proposed timeframe we will have completed the edition of all the Avestan texts used in Zoroastrian rituals, which basically means all the Avestan texts. For a complete chronogram see below.

 

Year

LL

OC

PhD

1st

Edition of Yasna and Wisperad

Preliminary work

Pahlavi Zoroastrian literature

2nd

Edition of Xōrdē Avesta

3rd

4th

Persian Zoroastrian literature

5th

Edition of Yašts

6th

7th

Analysis of “Outer ceremonies”

8th

Edition of “Outer ceremonies”

9th

10th

Edition of Wištāsp Yašt

11th

Edition of Widēvdād

 

12th

 

 

2.2       Objectives

Our aim is to edit all the Avestan texts used in the Zoroastrian rituals, insofar as there is manuscript evidence. Accordingly, this will include all known Avestan texts, with the exception of the Nērangestān and Hērbadestān (as they are not ritual, but meta-ritual texts), and a few minimal fragments included in Middle Persian works whose ritual contexts are not clear. The new edition considers the fresh insights and tendencies in Avestan studies since the 1950s. It will be the first edition to present all the texts belonging to the Avesta in its concrete shape, i.e., as ritual texts, and strictly according to the manuscripts. This textual concreteness establishes new requirements for editorial practice:

  • the representation of the diversity of the rituals, a) regarding the inner variability of the liturgies, and b) regarding the contexts and purposes of the rituals
  • the representation of the performative nature of the texts, which means a) the observation of the structural performativity of the Avestan texts, and b) the observation of their contextual performativity
  • the representation of a historical-topical concreteness of the texts. While the older editions sought to eliminate all historical-topical elements from the text and its representation, the new edition will integrate the historical and local dimensions of the edited texts

The edition is firstly a representation of the ritual reality represented by the manuscripts, and does not therefore aim to provide a complete picture of the ritual life of Zoroastrians. The manuscripts, however, provide a much more complete description of the ritual life of Zoroastrians than the standard editions of the Avestan texts ever did. Accordingly, Corpus Avesticum will include a good number of rituals that were not included in any of the older editions. While secondly, by means of various apparatuses, the edition will mediate this ritual reality of the manuscripts with results from historical and ethnographical research.

The new edition will appear in printed form (hardcopy and e-book) in the new series Corpus Avesticum by Brill, as well as in an electronic edition (s. § 2.4).

The edition will provide an entirely new and solid basis for the study of ancient Iranian language and religion. Furthermore, it develops a new concept of text and new ways of editing that will be fruitful for other areas of Iranian philology. Historical linguistics will be able to base its research on a reliable Avestan text and have access to abundant new material for the analysis of the evolution and transmission of the Avestan language. Comparative religious history will re-think its traditional view of Iran and its geistesgeschichtlichen Ort in the world of the first millennium BCE, as the expected fallout from the new edition will change our understanding of Old Zoroastrianism. Our edition will for the first time enable a structural analysis of the text that provides the information about the religious views and Weltanschauung of the performers of such rituals. If we could locate the celebration of these rituals in the Achaemenid period in Western Iran, as we assume, the edition will significantly contribute to the reconstruction and understanding of the (historical) Ancient Wold in the Near and Middle East.

2.3       Work program incl. proposed research methods

2.3.1. Methodology

2.3.1.1. The texts in their performative variability and context

The Long Liturgy and other rituals acquired their current shape (or at any rate, a shape very similar to the current one) at a time when the composition of new Avestan texts was actually still possible, i.e., probably before or in Achaemenid times (Cantera, 2014, pp. 187 ff.). The ritual instructions accompanying the Avestan text can lay claim to old age, too, as already conjectured by Darmesteter (Cantera, 2014, pp. 189 ff.; Darmesteter, 1892, p. 1.xcii). The liturgical manuscripts continue a tradition of meta-ritual literature dating back at least to Sasanian times, but probably beginning much earlier, as there are even Avestan versions of some of the ritual instructions. Accordingly, it seems that the ritual continuity stretches as far back as Achaemenid times, and the same applies for the tradition of the meta-ritual descriptions of these rituals.

An important feature of the Long Liturgy and all the other rituals, as described in the meta-ritual literature and in the manuscripts, is the possibility of variation. Variability can be of different types:

  1. Specifying variation: There is a set of variable elements that must be chosen for the celebration of each ceremony or section of a ceremony depending on the place (fire temple or not), time of day, date, special dedicatories (šnūman), or the type and number of officiating priests (wāž gīrišnīh).
  2. Context-determined variation: each ritual is adapted to its performative context. The variation concerns mainly the utensils and offerings used in the same ritual, as well as the celebration’s ritual framework. The set of possible variations is theoretically unlimited.
  3. Typological variation: the same basic structure of a ritual (like the Long Liturgy, but also others) can be adapted, expanded or contracted for creating different variants of the Long Liturgy for serving different purposes (Cantera, 2013). There is a broad range of variants of the long ceremony, including fairly long celebrations with extensive intercalated texts. The evidence suggests that the number of possible variants was larger in Antiquity than attested in the manuscripts, and even more so than in modern practice. The following variants of the Long Liturgy are described by the manuscripts handed down to us (Cantera, 2014, pp. 216 ff.): a) the standard daily ceremony, known as Yasna; b) a simpler variant celebrated at the beginning and end of the summer, called Yasna ī Rapihwin; c) a more solemn variant, the Visperad, celebrated on special occasions, and forming the basis for all the more complex variants; d) the Dō-Hōmāst, a more complex variant of the Visperad containing two hōmāst (i.e., sections for the pressing of the haōma plant), and attested only in manuscript K7; e) the Vidēvdād; and f) the Vištāsp Yašt.

Former editions reproduced the Avestan texts not of the liturgical manuscripts but instead of the exegetical ones (see § 1.B.2). However, only the liturgical manuscripts include the Avestan texts as performed in the rituals. The exegetical manuscripts offer a simplified text, free of contextual variants and performative information. They do not record the specifying and context-determined variations at all, and the typological variations only very limitedly so: Thus, Yasna ­i Rapihwin is not included at all in the editions, Visperad is included in a much simplified version, and the ritual texts framing the Vidēvdād and Vištāsp Yašt are completely omitted. None of the numerous variants of the Zoroastrian Long Liturgy has ever been edited in such a way as to reflect its celebration in ritual practice (Cantera, 2012b, pp. 451 ff.), and some of them have not been edited at all.

The proposed edition of the Long Liturgy will be based, by contrast, on the liturgical manuscripts, and accordingly represent the intrinsic, contextual and typological variety of the Zoroastrian rituals. This choice rests upon three observations:

  1. The demonstrated antiquity of the liturgy as described in the manuscripts, and the antiquity of the description itself that continued a tradition started in the Avestan language, upheld in Sasanian times, and preserved in manuscript production through to the 19th century
  2. Only the liturgical manuscripts include performative information that is essential for understanding the ritual discourse and the recited texts in their performative context. Our edition will therefore reproduce not only the texts in the Avestan language, but also the ritual instructions in Middle Persian (since our basis will be the Iranian manuscripts, see below)
  3. Only the ritual manuscripts suitably record the specifying, contextual-determination and typological variations, and therefore only these ones transmit a suitable picture of ritual life

In fact, the choice between liturgical and exegetical manuscripts concerns only the Long Liturgy. This classification does not hold, however, for either the Xorde Avesta or the Yašts or the Outer Ceremonies. In the case of the Outer Ceremonies, only liturgical manuscripts exist, and have therefore never been edited (with the exception of Karanjia’s edition of the Drōn). The principles of their edition will be the same as for the Long Liturgy. In the case of the Yašts, no Middle Persian translations of the ‘greater’ Yašts have survived. In the manuscript tradition of the Yašts (F1 and then the Mss. of the Tamām Xorde Avesta) the texts appear within a liturgical framework and all liturgical repetitions, although the typical characteristics of the liturgical manuscripts of the Long Liturgy (intercalation; variants; performative instructions) are missing. In the case of the Xorde Avesta, the most important texts were translated into Pahlavi and/or Sanskrit. However, the situation regarding the Avestan text of the Xorde Avesta is similar to the Yašts, and there is no further clear distinction between the exegetical and liturgical manuscripts of the Xorde Avesta. However, because of the fluent and open nature of the Xorde Avesta tradition, an edition of these texts has to deal with a) a diachronic variability, and b) the fact that the Xorde Avesta also includes non-Avestan texts (see § 2.3.1.2).

2.3.1.2. The representation of the dynamism of the ritual practice

Contrary to the view dominating Avestan studies until only recently, whereby the Avestan manuscripts more or less mechanically reproduce an alleged archetype or several hyparchetypes, we assume that the manuscripts both reflect and are an active part in the dynamism of the ritual life of Zoroastrian communities. Thanks to a comparison of a) the internal ritual information to be gleaned from the Avestan text; b) the descriptions given in the Avestan and the Middle Persian versions of the Nērangestān; and c) the liturgical manuscripts, it has become clear that, ritual continuity notwithstanding, changes inevitably occurred in ritual practice. This casts doubt on the wisdom of making the reconstruction of an alleged Sasanian archetype an objective of any edition. Even if it were possible to establish the linguistic form of the Sasanian text, we should never be able to reconstruct the ceremonies as they were celebrated in Sasanian times, but only as they appear in the manuscripts.

Furthermore, a comparison of the manuscripts has brought to light differences in the performance of the rituals, not only between the Indian and the Iranian community, but also chronological differences within the same community (Cantera, 2014, pp. 259 ff.). Although ritual knowledge was imparted through regular contacts between the Indian and the Iranian communities from the 15th century on, a process which implied sending liturgical manuscripts from Iran to India, the communities’ separation ultimately generated two independent lines in the development of the Long Liturgy celebration. Thus, the idea is brought home to us of a dynamic transmission of the Long Liturgy, with repeated attempts at harmonisation being denied by the emergence of ever new variants of recitation and ritual practice. The tension between homogenising tendencies and unstoppable local developments, as well as between written testimonies and ritual practice, has powerfully impregnated the history and transmission of the Long Liturgy. A relevant edition must take all this into account.

Therefore, the edition is compelled not only to represent the rituals as they were performed at a concrete historical moment, but also to reflect the historical variation. The first is ensured by choosing as the basis for the edition a coherent group of manuscripts produced in the same area (Yazd-Kerman) in Safavid times (between the end of the 16th century and the beginning of the 18th century). Our analyses have indeed shown that they reflect the broadest ritual variety and, from a philological and linguistic point of view, represent the most archaic stage that can be reached.

The geographical and chronological changes would be represented through two apparatuses: 1. the apparatus of variants will chronologically and geographically collect and organize the textual variants, as well as the variants of the performative instructions; 2. an apparatus that gathers all the additional ritual information that is not included in the manuscripts, including the internal information provided by the Avestan texts, and the ritual information in the Middle Persian literature and in the meta-ritual literature in Middle and New Persian.

In the case of the Xorde Avesta, the situation is different. We also note an intense dynamism; however, it concerns not merely differences in the performance of each single ritual, but mainly which rituals are included in the collection at different historical periods. Accordingly, the main challenge of an edition of a very moveable text is its definition and the range of texts to be included. It seems reasonable to prelude the edition with three studies: 1) a reconstruction of the Xorde Avesta in the first millennium CE on the basis of the information from Pahlavi and classical sources; 2) a definition of the Xorde Avesta according to the history of its translation; and 3) a definition of the Xorde Avesta according to the statistics and sequences of the texts in the manuscripts. Based on these studies, a certain kernel of the Xorde Avesta may well be defined. This kernel seems to be clearly mirrored in certain Iranian Xorde Avesta manuscripts from the early 18th century. It seems that the (slightly older) Indian tradition was especially productive regarding the addition of further texts (e.g., the tradition of the Tamām Xorde Avesta seems to be unknown in Iran). The edition of the Xorde Avesta will react to this dynamism by splitting the bulk of the Xorde Avesta texts into a) those texts that constitute a historical kernel with a stable sequential structure; and b1) those texts (Avestan and non-Avestan) that are later additions; and b2) those Yašts that are not an integral part of the Xorde Avesta.

2.3.1.3. Diplomatic edition or reconstructed text?

Rituals in Avestan have been performed over millennia in very different regions, so the homogeneity of the recitation has been (and still is) seriously challenged. The invention of the Avestan script in Sasanian times (probably around the 6th century CE) was an attempt at establishing a canonical recitation to be propagated in all Zoroastrian regions. This canonical recitation forms the basis of the preserved manuscripts. As the discovery of a Sogdian Ašəm Vohū with a completely different phonetic shape goes to show, alternative local ways of recitation survive, and new ones arose despite the introduction of the script. A systematic comparison between the Indian and the Iranian manuscripts indicates that although both transmissions originate from the official Sasanian recitation, important and systematic variations attributable to dialectal influence evolved after part of the community had emigrated (Cantera 2014a: 133 ff.; 2016). The manuscripts are guides for learning the correct performance of the rituals, and as such maintain a difficult equilibrium, different in each manuscript, between the canonical recitation and the real recitation, although all of them are under the influence of the actual recitation. Accordingly, the Avestan manuscripts reveal an enormous diversity of readings for each single word: for instance, a single proper name such as θraētaōna records a different evolution of the recitation in both India and Iran, as the readings of the manuscripts show. Here just a selection of readings in just one passage (Y9.7):

Iranian manuscripts (ordered chronologically)

Indian manuscripts (also chronologically)

θraētaōnō 4000, 4010 (first quarter 17th century)

θraētōnō 4065 (middle 17th century)

θrītōnō 2035 (18th century)

θraētaonō 4200, 4220 (17th century)

θraetanō in 4240, 4250, 4280 (middle 17th century)

θaritanō in 231 (undated, strong oral influence)

 

The same word is rarely written in exactly the same way even in the same manuscript. For example, in 2030, an accurate Iranian manuscript by Rostam Goštāsp (1716 CE), we find different attestations for caŋraŋhāc- “following the belt”: caŋvraŋhācąm in VrS1.1, caṇgraŋvhāca in VrS2.1, caṇgraŋhāca in VrS7.1, caŋraŋvhācąm in VrS41.1 and caṇgraŋhācasca in Y71.9. The variation in this case concerns the group ŋr and ŋh, finding ŋr, ṇgr, ŋvr, ŋh and ŋvh in different combinations; however, a similar or bigger variation is found in almost every word. On the issues involved in editing Avestan texts regarding the number and nature of the variants, see Cantera (2014: 273 ff.; 2015).

This challenging situation explains the success of the Andreas theory, according to which the transmitted Avestan is simply a mistaken vocalization of an older Arsacid archetype written according to the Semitic system (without indicating the short vowels), and so the Avestan text can be reconstructed only on the basis of its comparison with the Vedic. The manuscripts were of no use. According to Hoffmann, however, the Avestan texts were written for the first time around the 4th Century CE, with a phonetic script invented ad hoc. The editor’s aim should be to reconstruct this archetype on the basis of the attested readings and a comparison with the Vedic evidence.

However, the existence of the archetype is questionable, as is the methodology used for its reconstruction. According to Hoffmann, systematic changes have to be attributed to the archetype, but the occasional ones are later developments. However, we have been able to show that this distinction does not hold. We find systematic changes that happened as late as the 17th century, or solely in one of the branches. For instance, ū systematically becomes ī in the Iranian manuscripts after the middle of the 17th century. If similar changes occurred before the two transmission branches separated, they would have appeared in all the manuscripts, and would be erroneously ascribed to the archetype. Furthermore, Hoffmann understood he could reconstruct the shape of the Sasanian archetype on the basis of the analysis of the original use of certain letters. However, this is a pipe dream, as we are not sure, on the one hand, of the original use of some letters (e.g., ġ), and on the other, of the exact date of certain evolutions, e.g., the different waves of epenthesis before a nasal and the loss of ń in the Indian manuscripts.

Moreover, as J. Ferrer has been able to ascertain in his study of Avestan palaeography, conducted within the context of ADA, the catalogue of letters of the Avestan alphabet was longer than Hoffmann had assumed (Ferrer Losilla, 2016). For example, the Indian transmission has preserved different letters for the pre-consonantal nasals. These appear in a small group of manuscripts, where they are used arbitrarily. A palaeographic and typological investigation suggests that different values were originally attached to them. It was only by chance that we learnt about the existence of these letters. By no means can it therefore be ruled out that even more letters were lost before the emergence of our earliest manuscripts. Hence, from a methodological point of view it must be considered well-nigh impossible to reconstruct the supposed Sasanian archetype.

Indeed, while aiming to reconstruct an alleged Sasanian archetype, all past editions (the old complete ones and the modern partial ones) in fact reflect the Indian way of reciting the Avestan texts. The discovery of the Iranian manuscripts has shown, however, first that there were important differences in the recitation at different places, and second, that these differences were reflected in the manuscripts, although in different degrees depending on the type of manuscript, the inclination of the scribe, and the manuscript’s purpose, etc. (see the analysis of the ac. sg. of –ṷa- stems in Cantera 2016b).

As we are dealing with a fluid tradition, we might consider the convenience of a diplomatic edition. However, the high degree of instability even in a single manuscript advises against such a procedure. In order to represent a nearby historical text, as well as maintain the highest possible degree of coherence, we have decided to investigate the orthographical uses of each manuscript, manuscript class, place and time of production, etc. This is the subject of the doctoral thesis being prepared by Jaime Martínez Porro, proposed as PD1, which is now in its final stages. We will edit the text according to the orthographical uses dominating the production of manuscripts in the Yazd-Kerman region in Safavid times. These features have been defined by Martíneyz porro in the mentioned work. We will therefore choose the same space-temporal frame for the linguistic aspects of the text as for the shape of the ceremonies. When framework evolutions are stated in the Iranian manuscripts of this time, we will choose the most ancient spelling, even if it is not statistically the most frequent (see, for instance, the analysis of ŋhr in Martínez Porro 2015). Although in many manuscripts ŋr is very frequent, if not more frequent than ŋhr, the different statistical spelling for hazaŋra- and other words containing the same group shows that the original spelling is ŋr and that there is a dissimilation in the case of hazaŋra- < *hazaŋhra-). However, only uses actually attested in the manuscripts of this period will be accepted as editorial criteria.

2.3.1.4. Digital platform for the editorial work and for the display of the digital edition

The digital component at the centre of this long-term project is the web-based collaborative research tool “Virtual Manuscript Room Collaborative Research Environment” or VMR CRE (www.vmrcre.org), which has been used by many German and European projects involving digital critical editions. Based on the award-winning open-source portal “Liferay” (www.liferay.com), VMR CRE supports researchers at all stages in the production of a digital or printed edition. This comprises the tasks associated with witness or manuscript management, including cataloguing, imaging and indexing manuscripts; transcribing and encoding of texts in TEI; regularisation; automated collation; editing of an apparatus, and finally the publication of the critical edition.

 

Witness

Management

Imaging

Indexing

Manuscripts

Transcription &

TEI Encoding

Digital or Print

Edition

Critical

Apparatus

Automated

Collation

Regularisation

 

 

 

 

 

The VMR CRE was originally developed for the edition of the Greek New Testament, which faces a similar problem as to our project: the need for transliterating and collating a high number of witnesses. In this respect, VMR CRE is very mature and reliable software. Another core feature of the VMR CRE is its component-based or modular architecture. On the one hand, this feature enables teams to customise their software suite by selecting the right tools for their project. On the other, the modular architecture provides a framework for further extensions to be added to the core of VMR CRE. As part of this long-term project, we plan on adding two core functionalities to VMR CRE.

The VMR CRE was originally funded by the DFG for the development of the “Virtual Manuscript Room” software at the “Institut für neutestamentliche Textforschung” at the University of Münster. Since then, a number of other scholarly projects have contributed to the development of the VMR CRE as a platform. The following are some of the projects that use the VMR CRE: New Testament Virtual Manuscript Room, Coptic-Sahidic Old Testament Project, The Canons of Apa John the Archimandrite, Erstellung einer kritischen Edition der Johannesapokalypse; Museum of the Bible, Greek Paul Project, Greek Psalter Project, Syriac Climacus' Ladder of Divine Ascent Project, Τομέας Μελέτης Χειρόγραφης Παράδοσης - Byzantine Project, Paratexts of the Bible. More recently, the KELLIA (Koptische/Coptic Electronic Language and Literature International Alliance) project, which will also be using the VMR CRE, has received funding from the DFG and from the NEH.

ADA has been working on the adaptation of this platform for the concrete needs of Avestan philology, and a trial version is already up and running (http://adavmr.usal.es; user: DFG@fu-berlin.de; password: Antrag2017). We are also working on it within the context of the ERC advanced Grant MUYA (see § 5.4.1), led by Almut Hintze at the SOAS, and we intend to continue to expand this platform and develop further modules for a fully functional online digital edition that can at the same time serve as a basis for future works (such as a lexicographical project and grammatical studies). The first feature to be implemented is a new module for the digital edition that we want to be highly customizable by the end user, and that addresses the needs of displaying ritual texts. The second core feature that we would like to see in the VMR CRE is a tighter integration of TEI or XML encoded texts into the output layer of the VMR CRE (s. § 2.3.2.6).

 

2.3.2. Tasks

2.3.2.1. Task 1: Digitisation, indexing and analysis of manuscripts

ADA’s databases are now being integrated with those of VMR CRE, so that all the manuscripts available in ADA will also be available in VMR CRE, and vice versa, before the launch of the project. For the Long Liturgy, all available manuscripts have already been digitised, but around 80 manuscripts have still to be indexed and integrated in VMR CRE. The situation is different for the other classes of manuscripts (Outer Ceremonies, Xorde Awesta, Yašts). All Iranian manuscripts have already been digitised, as have all the manuscripts in a number of Indian libraries (Meherjirana Library, Navsari; Bhandarkar Oriental Institute, Poona; and Bombay University Library). Thus, around 50 manuscripts of the Yašts, Xorde Awesta and Outer Ceremonies have already been digitised. Nevertheless, a systematic search still needs to be conducted in other Indian libraries, especially at the Cama Oriental Institute (Mumbai).

The manuscripts of these classes still need to be indexed and published at ADA. Since we lack detailed analyses of the contents of the Xorde Awesta manuscripts and the manuscripts of the Outer Ceremonies, and they are by far less constant in their contents than the manuscripts of the Long Liturgy or the Yašts, the indexing task is much more difficult and requires an analysis that is still pending. Accordingly, whereas in the case of the Long Liturgy we can start directly with the edition (since all the preliminary work has already been done in recent years), preliminary work involving the analysis of the three other classes of manuscripts will be carried on during the first months of the project. Furthermore, as there are no appropriate guidelines for indexing the manuscripts of these classes, PD2 will be responsible for creating similar ones to those I have developed for the Long Liturgy (cf. http://ada.usal.es/pages/completeceremonies).

Last but not least, it would seem expedient to cultivate contacts that have been established over the past years and that have led to the discovery of numerous Iranian manuscripts. The emergence of new manuscripts is still a probability. Accordingly, the team plans to make regular trips to Iran: 2 in the first period and then one trip every three years.

 

2.3.2.2. Task 2: Transcription and digital publication of each manuscript

Faithful transliterations of every single manuscript lie at the heart of our editorial work. The transliterations will be preserved and made accessible together with the manuscripts’ images in VMR CRE. In the transliteration phase, we also encode the texts, following the guidelines defined by the “Text Encoding Initiative” (TEI) to ensure that documents can be shared on a long-term basis. For transliterations, we use the “Online Transcription Editor” (OTE), developed at Trier University, and integrated into VMR CRE. The OTE allows transcriptions to be encoded according to the TEI standard via a graphical user interface, thus eliminating the need for an in-depth knowledge of TEI. The OTE also facilitates the revision process. Having the transliterations checked by a second person is a crucial step in the quality control process.

When libraries do not permit manuscripts to be digitised, a prolonged stay at these libraries will be necessary. This is the case, for example, with the highly important collection of the Cama Oriental Institute (COI). A four-month visit to the COI by PD 1 and 2, accompanied by the PI for some of the time, is therefore planned. Likewise, a few manuscripts in Iran remain in private hands, and have to be transliterated on site.

Transliterations are most time-consuming, yet their accuracy is essential, as they form the basis of any reliable collation. They will be performed by specially trained students. We have many years of experience with this kind of work, given that several complete manuscripts, as well as certain parts of all the manuscripts (Y9, Y28-29, Vr1-2, V19), have already been transliterated. We have recently organised an internship at the Institute for Iranian Studies at FU Berlin, in which several students have been trained for this task. The time needed for transliteration, calculated on a basis of five hours per 1,000 words, is as follows:

Texts

Working hours

1st period (1st to 3rd year): Y0-Y21.5

6 004.5

2nd period (4th to 7th year): Y22.1-Y59.33

10 092

3rd period (8th to 9th year): VrS32.1/Y60.1 to the end

4 619

4th period (10th to 12th year): Widēwdād (22 books)

9 625

Wīštāsp Yašt

104.5

Xorde Awesta (1st to 4th year)

1 125

Yašts (5th to 6th year)

1 000

Outer Ceremonies (7th to 8th year)

528

Total

33 098

 

The total number of hours will require three student aides (80 h/month) throughout the life of the project (except the final year). A student aide’s total number of hours is 880 per year, which means a total availability of 2658 h/year for the three aides. For a more detailed analysis of the working hours for transliterations, see appendix 1.

The transliteration of the ritual instructions in Middle Persian involves specific problems, which means it cannot be taken on by students. Instead, PD 1 and 2 will assume the transliteration of the Middle Persian instructions.

 

2.3.2.3. Task 3: Collation of manuscripts and creation of the critical apparatus

Once the manuscripts have been transliterated, they will be collated using CollateX (http://collatex.net/doc/), a tool integrated into VMR CRE and designed to help scholars and philologists working on critical editions with the collation and production of a critical apparatus. It can thus read multiple versions of a text in a series of formats. To make the comparison, CollateX splits the versions of the text into parts or tokens, i.e., smallest meaningful units – a process called tokenisation. It can accommodate XML elements such as those defined by TEI, and may also be combined with other, external tokenisers. In two further steps, CollateX normalises and aligns the tokens. At present, the alignment process still requires some refinement through manual intervention in order to cope with the complexity of critical editions and apparatuses. CollateX offers the result in a variety of formats that can be used to produce the critical apparatus or for the genealogical analysis of manuscripts. To model the data, CollateX relies on Desmond Schmidt’s (2009) idea of a graph-oriented model, where the textual variance of the input texts is represented on a graph.

CollateX renders it possible to be both thorough and flexible at the same time, as readings can be grouped according to the user’s individual criteria. Geldner’s apparatus was not particularly user-friendly, nor did it aim at exhaustiveness. It fails to painstakingly record all the variants of all the manuscripts, simply providing a selection, which is sometimes arbitrary, and grouping them in such a manner that minor deviations are overlooked. Geldner’s apparatus should not be viewed as a collection of readings attested in the manuscripts, but rather as a representation of his understanding of each single word’s transmission. The manuscripts quoted after each reading do not in fact all have the same reading; their being quoted means nothing more than that according to Geldner they contain witnesses to the reading in question. They may show differences, but these are considered unimportant. Consequently, it is not clear from Geldner’s apparatus exactly which textual reading can be found in each manuscript, rendering the apparatus useless for orthographical investigations. Its somewhat chaotic disposition has also often attracted criticism: there is neither a chronological nor a geographical nor a manuscript class order to go by (Andrés-Toledo, 2012). Modern editions drawn up in the wake of Hoffmann have striven to compile a well-organised, convenient apparatus. Nonetheless, they remained dependent on Geldner’s account of the transmission and on his textual-critical decisions.

Our project’s apparatus aspires to offer a faithful and complete reading of all the manuscripts, and thus create a more dependable basis for future research. Featuring such an extensive apparatus electronically will be no problem, or for the e-book. In the hardcopy version, however, the sheer quantity of readings rules out such a presentation. For although the Avestan transmission has only a very small number of variants that are grammatically possible and make sense in a given passage, there is an enormous amount of purely phonetic and orthographical variants because of the strong influence of the actual ritual practice. The hardcopy will include just an apparatus of selected variants. We understand variants in a broad sense. They are all readings that are not to be explained as phonetic or orthographical variants depending on the uses of single manuscripts or manuscript groups. This limitation alone makes the size of the apparatus manageable, and the bulk of the readings of the different manuscripts will be discarded. Furthermore, the readings of manuscripts produced after 1750 will be displayed only if a manuscript is relevant for some reason. Apographs will also be discarded.

Furthermore, readings and variants will be presented according to clear rules: frequent readings will appear first, so it will be easy to discern where the editor parted from them. For the apparatus to be consistently organised, the witnesses for each reading will be grouped according to manuscript classes, being chronologically ordered within each class. It is the same principle as the one governing ADA’s “manuscript sigla”, according to which each sigla denotes the manuscript class and each one’s chronological position within its class (for a key of the sigla used for the Long Liturgy, see Cantera 2014: 406).

In addition, there will be one apparatus containing parallel passages.

2.3.2.4. Task 4: Constitutio textus

Textual arrangement (cf. § 2.3.1.1)

Establishing a reliable text is the ultimate goal of any edition. The constitutio textus is therefore the last and most important step in the process – and, in the case of the Avestan texts, so much more important because it is, to my mind, the weakest point of all earlier editions. Previous editors always made it their primary objective to select the correct spelling, i.e., the phonetic value of each word, from the large number of existing ones. This is no less true of Hoffmann’s revolution, whose idea it was to reconstruct the Avestan text as it had allegedly been written down in Sasanian times. The 19th century, too, saw a lively debate over Avestan’s exact phonetics. The shape of the text, however, i.e., textual variants of greater consequence, has rarely been discussed in the history of Avestan studies. The problem of the major divergences between exegetical and liturgical manuscripts has never been explicitly raised.

Our edition focuses on

a)    the representation of the texts in the ritual arrangement

b)    the representation of the specifying, contextual-determined and typological ritual variation

c)    the performative aspects

Accordingly, and as already mentioned, we will base our constitutio textus on the liturgical manuscripts (more specifically on the oldest Iranian liturgical manuscripts in each class), whereas all former editions were based on the exegetical Indian manuscripts (with the exception of Brockhaus’ Vidēvdād and Kellens’ Études avestiques et mazdéennes). This implies that we are going to edit numerous texts that have never been edited before. Some of them are completely new; others are just different combinations of previously known texts. In order to maintain the traditional numbering as far as possible, we have been forced to introduce a new numbering system for the additional parts. We have been using this new system (which is compatible with the traditional one) for indexing the manuscripts published in ADA, where one may find a guide for the numbering of the new parts (http://ada.usal.es/pages/table).

For the representation of the specifying variation, we will label all variable sections with the following tags: ratu or part of the day, day and month, dedicatory, wāž gīrišnīh. The different contextual-determined possibilities for each label will be conveniently set forth in the introduction to the first volume of the edition. Furthermore, the ritual apparatus will indicate which options are pertinent for each passage, and their source. The digital edition will be accessed by clicking on the tag to all the possibilities available in a certain passage.

The contextual-determined variation appears mostly when a section of the Long Liturgy is celebrated as an Outer Ceremony. In this case, we will follow the use of the manuscripts, and edit Outer Ceremonies separately. In the printed edition, the ritual apparatus will, however, indicate the differences with the parallel sections in the Long Liturgy or in other ceremonies. In the digital edition, the different versions will be made available simultaneously.

The typological variety concerns mainly the Long Liturgy. In the printed edition, the text of each variant will be edited synoptically. In the digital edition, the variants will appear as multiple choice menus. The result will be the first complete edition of all the variants of the Long Liturgy attested in the manuscripts produced in Iran in the 17th century. They represent the richest ritual variety attested, and correspond most faithfully to the descriptions of Sasanian meta-ritual literature. Many of these variants are omitted in the standard editions or only partially included: 1. Yasna ī Rapihwin (not included); 2. Yasna; 3. Visperad (partially included); 4. Dō-Hōmāst (not included; only attested in K7); 5. Vidēvdād (partially included); and 6. Vīštāsp Yašt (partially included). In the case of the intercalation ceremonies attested in the manuscripts (Vidēwdād, Vištāsp Yašt, and perhaps also Hādōxt Nask), the intercalated texts will be edited as separate volumes. The edition of the short but difficult Vištāsp Yašt will be left for the final phase of the project, and is likely be the subject of a PhD thesis.

In the case of the texts of the Xorde Awesta and the Tamām Xorde Awesta, the textual arrangement will follow the principles mentioned under § 2.3.1.2. The distinction into a textual kernel (which has a stable structure) and later textual additions (in Avestan and non-Avestan) will provide a certain insight into the historical variability of the Xorde Awesta. Therefore, the edition of the Xorde Awesta and the Yašts will consist of three groups:

i.i. Xorde Awesta stricto sensu

i.ii Additional texts to the Xorde Awesta

ii. Yašts (that are not part of the Xorde Awesta)

The performative information included in the edition will be limited to the ritual instructions contained in the manuscripts. Given that the edition is based on the Iranian manuscripts produced between the end of the 16th century and the beginning of the 18th, only the Middle Persian instructions will be edited. The New Persian ritual instructions that appear in later Iranian manuscripts and the Gujarati of the Indian ones will, however, be included in the apparatus of variants.

The linguistic shape of the constituted text (cf. § 2.3.1.3)

As for the choice of the right reading among the almost infinite phonetic and orthographical variants, here too our projected edition will stand apart from all its predecessors. Geldner intended to capture the Avestan text at a stage of phonetic development that was reached with a second, fundamental text editing in post-Sasanian times. In fact, however, just like Westergaard, he reproduced a version close to the one by Mihrābān Kayxōsrō (14th century India, but hailing from Iran). Hoffmann, on the other hand, sought to reconstruct the Avestan text as it appeared in a first recording dating from Sasanian times – the so-called Sasanian archetype, but his textual decisions were still very close to the Indian manuscripts.

Against this, I have set out to show that it is methodologically impossible to reconstruct the Sasanian archetype, and that it is also quite improbable that the archetype should have existed as the one and only source of later manuscript copies. I should instead proceed from a more dynamic transmission model. Despite repeated attempts at harmonisation, the Long Liturgy text was recited in different phonetic variants, reflected by the varying orthographical rules and tendencies in the manuscripts. For this reason, as in the case of the Long Liturgy, the Outer Ceremonies and the Xorde Awesta, we will edit the text according to the orthographical uses we find in the Iranian manuscripts of the Safavid time. Nevertheless, in the case of the additional texts to the Xorde Awesta and most of the Yašts, the edition must be based on Indian manuscripts from this period because of the lack of Iranian manuscripts.

2.3.2.5. Task 5: Gathering of the ritual information in external sources (cf. § 2.3.1.2)

In order to represent the geographical and historical dynamism of Zoroastrian ritual life, we will compare not only the ritual information from the manuscripts, but also from the other available sources. Accordingly, we propose two PhD theses: one consisting in the gathering and evaluation of the ritual information in the Middle Persian meta-ritual literature (especially, the Nērangestān, but also other unpublished smaller texts), and in Middle Persian Zoroastrian literature in general; the other will focus on the information contained in the New Persian Rivāyats, many of which have yet to be edited or even translated. The information gathered will be published separately as independent monographs, and also included in the edition as apparatuses. We therefore propose this should be done in the first years of the project.

2.3.2.6. Task 6. Development of the Virtual Manuscript Room

We are already working on the customization of VMR CRE for its specific use for the edition of Avestan texts, with funding from the ERC-Advanced grant project MUYA and the Institute for Iranian Studies at FU Berlin. The main collaborators are the developer of the software, Troy Griffiths, and Martin Sievers (Kompetenzzentrum, Universität Trier), who is responsible for developing the Online Transcriptions Editor. The customisation will be ready over the coming months.

This project’s contribution to the development will be as follows:

  1. A better adapted and more powerful module for the display of the digital edition
  2. A closer integration of TEI or XML encoded texts into the output layer of the VMR CRE

We want the new module to be highly customisable by the end user and recognise the needs of displaying ritual texts. Accordingly, core features of the module for digital edition will be 1. user customisation; 2. display of the ritual variation; and 3. display of additional ritual information from other sources. The proposed new module for the digital edition will incorporate highly customisable apparatuses. They will enable the user to create alternative editions on the basis of the transliterations made available. Thus, alternative editions or collations, which derive from the data developed by the project, will be available to a larger scholarly and lay community, encouraging further research into the primary sources for the study of Zoroastrian rituals. They could further decide upon the range of linguistic variations to be displayed in the apparatus, instead limiting it to real variants or showing all the myriad orthographical and phonetic variations that appear in the manuscripts.

The platform will also develop several tools for the display of the different kinds of ritual variety (specifying, contextual-determined and typological) that are to be displayed in different ways. The specifying variation (day, time, place, priest, etc.) will be displayed through links that show the complete range of possible choices for each passage. Furthermore, for each passage where there is a typological or contextual-determined variation depending on the variant of the ritual performed, the user can choose which variant of the ceremony to see, or see them all synoptically.

The second core feature that we would like to see in the VMR CRE is a closer integration of TEI or XML encoded texts into the output layer of the VMR CRE. The XML-based features of the texts encoded in the transcription stage are not currently accessible for further research. We would like to investigate the integration of the core functionalities of the open-source database eXistdb (www.exist-db.org), an XML database programmed in Java. An integration of an XML aware core such as eXistdb has two main advantages. On the one hand, eXistdb would enable teams to quickly build web applications based on XML data. As the project grows and more texts are encoded and stored, a tighter integration of encoded data will be essential for the project. Thus, texts and texts groups can easily be represented while taking advantage of encoded data. The representation of the various apparatuses discussed above could be outsourced to the web application functionalities offered by eXistdb. On the other hand, it is important to integrate XML-related standards into the process of textual research. The most relevant standard for us, beside representation-related standards such as XSLT, is XQuery. It would be desirable, for instance, to search stored data according to criteria such as personal names, geography or dates. These features may not be relevant at this stage of the project, but would ensure the longevity of this endeavour. For instance, a future integration of lexicographical and bibliographical information would hugely benefit from the availability of the XML features provided by eXistdb.

To research and implement the above features we require a programmer, who will be employed full-time for the first year. Within the next years, we foresee three months of maintenance work distributed throughout the year.

2.3.3. Time table

[Key: LL: Long Liturgy; OC: Other Ceremonies (Xorde Awesta –XA-, Yašts, Outer Ceremonies); Tr.: Transliteration (revision and tagging in VMR); S: Staff; C: Coordination; PI: Principal Investigator. Cantera, Alberto; PD1 Postdoc 1; PD2: Postdoc 2; GS: Graduate Student; TF: Task Forces; Tech: Computer Technician]

 

Term

LL edition

C: PI

S: IP, PD1

PD2 (last two years)

OC edition

C: PI

S: PI, PD2

Transliteration

C: PD1, PD2

S: TF

Field Work

S: PI, PD1, PD2

PhD

S: GS

Technical

S: Tech

1st Year

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Edition and publication of Y0-27

Analysis of the OC manuscripts: Prolegomena

 

LL: Y0-11

OC: XA texts

 

Two months in India

(Cama Oriental Institute)

 

One month in Iran

The Pahlavi Zoroastrian ritual literature

Adaptation of all VMR tools to the Project.

2ndYear

Edition and publication of XA: Alphabet, Nērang ī kustī bastān, Srōš Baj, Niyāyišn, Hōšbām

 

LL: Y12-21

OC: Yašt 1-10

 

Two months in India

(Cama Oriental Institute)

 

One Month in Iran

Technical support

( = Three months per year)

3rd Year

Edition and publication of XA: Sīrōza, Gāhs, Āfrīnagāns

 

LL: Y21-27

OC: Yašts 21-13

 

4th Year

Non-Avestan and additional Avestan texts (Nām Stāyišn; Patets; Nērangs, Āfrīns, Bāǧs, Namāz, Šnūmans)

 

LL: Y28-Y42

One month in India

 

One month in Iran

The Persian Zoroastrian ritual literature

5th Year

Edition and publication of Y28

-59

Edition and publication of Yašts

 

LL: Y43-VrS27

6th Year

LL: Y52-59

OC: Outer Ceremonies

7th Year

LL: VrS32-Y69

One month in India

 

One month in Iran

The Zoroastrian "Outer Ceremonies"

8th Year

Edition and publication of VrS32-

-Y72

Edition of Drōn

LL: Y70-72

9th Year

Edition of Farroxši

LL: V1-7

10th Year

Edition of Afrīnagān ceremony

V8-13

One month in India

 

One month in Iran

Edition and commentary to the Wīštāsp Yašt

11th Year

Edition and publication of

Videvdad

OC edition finished.
PD2 works in LL edition.

V14-22

 

12th Year

No more transliterations required for this period.

 

2.3.4. Team

The principal investigator and general coordinator of the project is the applicant, Alberto Cantera. He is to be assisted in the general assessment of the scientific work by a scientific board made up of two philologists and two experts in Zoroastrianism and Zoroastrian rituals:

-       Almut Hintze, Zartoshty Brothers Professor of Zoroastrianism, School of Oriental and African Studies, University of London

-       Antonio Panaino, Professor, Dipartimento di Beni Culturali, Filologia, Religioni e storia dell’Iran, University of Bologna

-       Ab de Jong, Professor of Comparative Religion and Religions of Antiquity, University of Leiden

-       Michael Stausberg, Professor, Study of Religion, Department of Archaeology, History, Cultural Studies and Religion, University of Bergen

Two Postdoctoral fellows will be responsible for the two main blocks of the work:

  1. Jaime Martínez Porro and the principal investigator will share the edition of all the texts related to the Long Liturgy
  2. Götz König will be responsible for editing the Outer Ceremonies, the Yašt, and the Xorde Avesta

Additionally, we will be counting on Arash Zeini., scientific assistant at Institute for Iranian Studies of the Freie universität Berlin, as a consultant in Digital Humanities, and very especially in TEI standards.

2.4       Data handling

Each digitised manuscript is to be transcribed in the TEI-aware "Online Text Editor" (OTE) integrated in the VMR CRE. In this step of the process, the text of each individual manuscript is transcribed in Unicode (UTF-8), and encoded according to the latest TEI Guidelines. The OTE provides a WYSIWYG interface for the TEI tags. This feature speeds up the encoding process, while minimising the error rate.

The encoding will be twofold. Individual manuscripts will be described according to the TEI Guidelines for the description of manuscripts (currently Ch. 10 of the guidelines). The intellectual content of the manuscripts will be encoded according to the guidelines for the representation of primary sources (Ch. 11), while transcribing the manuscripts in the OTE. All project-specific TEI or XML elements will be documented in the header and through the use of appropriate schema. To ensure compatibility and reusability, headers will be generated through the "Webformular für die Metadatenerfassung" of "Deutsches Textarchiv" (DTA). We are aware of the DTA's various guidelines for the "Basisformat", and take them fully on board. Concerning the TEI headers and the general practice of encoding, we will also use the versatile XML editor oXygen, which incorporates specific customisations for TEI.

The description of the manuscripts will encode information such as sigla, library where the manuscript is kept, provenance, affiliation, dates, scribes, patronage, etc.

The encoding of the intellectual content of the manuscripts, here referred to as text, will comprise all textual features typically represented in or required for the compilation of a critical edition. Examples are lacunae, gaps, substitutions, abbreviations and shift of hands in the text.

The transcribed and encoded texts represent the raw material for this project. They will, however, be open access and available to the public and the scholarly community. To ensure the availability and longevity of the raw material, we take a multifaceted and multilayered approach to the representation of the material. Firstly, all transcribed texts and their manuscripts will be made publicly available through the VMR CRE platform, which provides stable URLs for the material published on the platform. Secondly, the encoded texts will be stored and offered to the public through the open-source Java XML database eXistdb, which successfully combines a number of well-established XML-related standards such as XSLT and XQuery. The use of these standards ensures flexibility in the representation of the texts through XSLT, and partly CSS, but also the ability to take advantage of the XML and TEI encoded information through powerful and meaningful searches through XQuery. The server for both applications is hosted and maintained at the FU Berlin, with guaranteed funding for the next 20 years though the in-house resources of the Institute for Iranian Studies. Thirdly, and perhaps most importantly, the transcribed texts will also be available through FU Berlin’s DINI certified eDOCS document server. This service provides stable and unique long-term URLs for the uploaded documents, thus ensuring open and long-term access to the project's generated output. These files will be made publicly available as they are generated by the team, and placed under a creative commons license. This means the project’s digital output will be available to a wider scholarly community. Thus, other researchers can always replicate and test the project's apparatus or editorial views, or produce alternative editions by using the transcribed texts in the collation module of the VMR CRE.

The project’s print output will be published in a new and dedicated series of the Brill “Corpus Avesticum”, to appear as a sub-series of the Handbook of Oriental Studies. A. Hintze and me will be the sub-series’ joint editors. To guarantee the publisher's rights and interests, the constituted text published in this series will be subject to a 12-month embargo. At the end of the embargo, the constituted texts will also be uploaded to FU Berlin’s eDOC servers in PDF format. This approach has no bearing or negative side effects for the wider scholarly community, as the data and the tools for the generation of the apparatus and the constituted text will be publicly accessible. The very aim of this long-term project is not only the production of an edition, but also to encourage and enable other scholars to renew their editorial approach to the Avestan texts and rituals.

Each volume in the series will also be published in a printed version and as an e-book. While the printed edition will contain only a selection of readings of the apparatus, the Phonetica et Orthographica apparatus will contain all the readings from the manuscripts, and will be available only as an e-book, not in hard-copy. The following volumes are planned:

  1. Synoptic edition of the Long Liturgy (in 3 volumes: 1. Paragnā + Y0-21; 2. Y22-59; 3. Y60-72)
  2. The intercalated texts:

                                          i.    Vidēvdād (in 2 volumes: V1-8; V9-21))

                                         ii.    Vištāsp Yašt

  1. Xorde Avesta (2 volumes)
    1.                                           i.    Xorde Awesta stricto sensu
    2.                                          ii.    Additional texts to the Xorde Awesta
    3. The Yašts (1 volume)
    4. Outer Ceremonies: Drōn, Farroxši and others (1 volume)

 

3       Bibliography

Andreas, C. F. (1903) Die Entstehung des Awesta-Alphabets und sein ursprünlicher Lautwert, in Verhandlungen des XII. Internationalen Orientalisten-Kongresses in Hamburg 1902, p. 93-106. Leiden.

Andrés-Toledo, M. A. (2015) A revision of Geldner's Critical Edition, in A. Cantera (Ed.) The transmission of the Avesta, p. 433-438. Wiesbaden.

Andrés-Toledo, M. A. (2015) Ceremonies in the Xorde Avesta Manuscripts: the Drōn Frawardīn Yašt, in Estudios Iranios y Turanios 2, p. 29-43. Barcelona.

Anquetil-Duperron, A. H. (1771) Le Zend-Avesta. 3 vols. Paris.

Bailey, H. W. (1943) Zoroastrian Problems in the Ninth-Century Books. Oxford.

Brockhaus, H. (1850) Vendidad Sade: die heiligen Schriften Yaçna, Vispered und Vendidad. Leipzig.

Cantera, A. (2012a) Building Trees: Genealogical Relations between the Manuscripts of the Wīdēwdād, in A. Cantera (Ed.) The transmission of the Avesta, p. 279-346. Wiesbaden.

Cantera, A. (2012b) Why do we really need a new edition of the Zoroastrian long liturgy? in A. Cantera (Ed.) The transmission of the Avesta, p. 439-478. Wiesbaden.

Cantera, A. (2013) Talking with god: The Zoroastrian ham.paršti or intercalation ceremonies, in Journal Asiatique 301, p. 85-138. Paris.

Cantera, A. (2014) Vers une édition de la liturgie longue zoroastrienne: pensées et travaux préliminaires. Paris.

Cantera, A. (2015) The Problems of the Transmission of the Avestan Texts and the Tools for Avestan Text Criticism (TATEC), in T. Andrews & C. Macé (Eds.) Analysis of Ancient and Medieval Texts and Manuscripts: Digital Approaches, p. 95-118. Leuven.

Cantera, A. (2016a) Substantial Change in the Approach to the Zoroastrian Long Liturgy. About J. Kellens's Études Avestiques et Mazdéennes, in Indo-Iranian Journal, pp. 139-185. Leiden.

Cantera, A. (2016b) “On Avestan text criticism (2): the accusative singular of the ū̆- and ṷa- stems in the long liturgy”, in Pirart, É. (Ed.)Études de linguistique iranienne: in memoriam Xavier Tremblay. (Acta Iranica 57), p. 89-156. Leuven.

Darmesteter, J. (1892) Le Zend-Avesta. 3 vols. Paris.

Geldner, K. F. (1886-1996) Avesta, the sacred book of the Parsis. 3 vols. Stuttgart.

Henning, W. B. (1941) The disintegration of the Avestic studies, in Transactions of the Philological Society 41, p. 40-56.

Hintze, A. (2012) On editing the Avesta, in A. Cantera (Ed.) The transmission of the Avesta, p. 419-432. Wiesbaden.

Hoffmann, K. (1971) Zum Zeicheninventar der Avesta-Schrift, in Eilers, W. (Ed.) Festgabe deutscher Iranisten nur 2500 Jahrfeier Irans, p. 64-73. Stuttgart.

Hoffmann, K. (1986) Avestisch ṣ̌, in R. Schmitt & P. O. Skjaervø (Eds.) Studia grammatica Iranica, Fetschrift für Helmut Humbach, p. 163-183. Munich.

Hoffmann, K. & J. Narten (1989) Der Sasanidische Archetypus. Wiesbaden.

Humbach, H. (1959) Die Gathas des Zarathustra. 2 vols. Heidelberg.

Karanjia, R. P. (2010) The Bāj-dharnā (Drōn Yasht). Bombay.

Kellens, J. (1998) Considérations sur l'histoire de l'Avesta, in Journal Asiatique 286.2, pp. 451-519. Paris.

Kellens, J. (2006-2011) Études avestiques et mazdéennes. 4 vols. Paris.

Kellens, J. (2012) Contre l'idée platonicienne de l'Avesta ou les Considérations revisitées", in A. Cantera (Ed.) The transmission of the Avesta, p. 49-58. Wiesbaden.

Kellens, J. & É. Pirart (1988-1991) Les Texts Vieil-Avestiques. 3 vols. Wiesbaden.

König, G. (2012) Das Nask Bayān und das Xorde Awesta, in A. Cantera (Ed.) The transmission of the Avesta, p. 355-394. Wiesbaden.

König, G. (2016a) Yašt 3. Der avestische Text und seine mittel- und neupersische Übersetzungen. Einleitung, Estudios Iranios y Turanios. Supplementa 1. Girona, Sociedad de Estudios Iranios y Turanios (SEIT)

König, G. (2016b), “Bayān Yasn. State of the Art“, in: Iran and the Caucasus (2016), Leiden

Kreyenbroek, P. (2004) “Ritual and Rituals in the Nērangestān“, in: M. Stausberg (ed.), Zoroastrian Ritual in Context. Studies in the History of Religions. Leiden, S. 317-331.

Kreyenbroek, P. (2008) The Term Bāgān Yasn and the Function of the Yašts in the Zoroastrian Ritual, in M. Jaafari-Dehaghi (Ed.) One for the Earth: Prof. Dr. Mahyar Nawabi, Memorial Volume, p. 81.94. Tehran.

Martínez Porro, J. (2015) The Indo-Iranian group *sr/a_a in the Avestan manuscripts, in Estudios Iranios y Turanios 2, p. 151-162.

Mazdapour, K. (2008-2009) Čand dastnewis-e nouyāfte-ye awestā, in Nāme-ye Irān bāstān 8, p. 3-19. Tehran.

Mink, G. (2004) Problems of a Highly Contaminated Tradition: The New Testament. Stemmata of variants as a source of a genealogy for witnesses, in P. van Reenen, A. den Hollander & M. van Mulken (Eds.) Studies in Stemmatology II, p. 13-85. Amsterdam.

Modi, J. J. (1922) The religious ceremonies and customs of the Parsees. Bombay.

Morgenstierne, G. (1942) Orthography and Sound-System of the Avesta, in Norsk Tidsskrift for Sprogvid 12, p. 31-79. Oslo.

Redard, C. & J. Kellens (2013) Études avestiques et mazdéennes. Vol. 5. Paris.

Schmidt, D. und Colomb, R., 2009. Eine Datenstruktur zur Darstellung von Multiversionstexten online. International Journal of Human-Computer Studies, 67.6, 497-514. Internationale Zeitschrift für Human-Computer-Studien, 67.6, 497-514

Skjaervø, P. O. (2007a) The Videvdad: Its Ritual-Mythical Significance, in V. S. Curtis & S. Stewart (Eds.) The Age of the Parthians, pp. 105-141. London/New York.

Skjaervø, P. O. (2007b) The Avestan Yasna: Ritual and Myth, in C. V. Pedersen & F. Vahman (Eds.) Religious Texts in Iranian Languages. Symposium held in Copenhagen May 2002, Det Kongelige Danske Videnskabernes Selskab, pp. 57-84. Copenhagen.

Skjaervø, P. O. (2016) Review of the book Vers une édition de la liturgie longue zoroastrienne: pensées et travaux préliminaires by Cantera, A., in Bulletin of the Asia Institute 26, p. 163-183. Bloomfield.

Sommer, F. (2016) Review of the book Vers une édition de la liturgie longue zoroastrienne: pensées et travaux préliminaires by Cantera, A., in Bulletin of the School of Oriental and African Studies 79, p. 428-430. London.

Spiegel, F. (1853, 1858) Zend-Avesta. 2 vols. Vienna.

Stausberg, M. (2004) Zoroastrian rituals in context. Leiden.

Swennen, P. (2016) Xavier Tremblay et la liturgie longue proto indo-iranienne. Présentation, in É. Pirart (Ed.) Études de linguistique iranienne: in memoriam Xavier Tremblay, p. 1-18. Leuven.

Wachtel , K. (2008) Towards a Redefinition of External Criteria: The Role of Coherence in Assessing the Origin of Variants, in H. A. G. Houghton & D. C. Parker (Eds.) Textual Variation: Theological and Social Tendencies? Piscataway.

Westergaard, N. L. (1852-1854) Zendavesta or the religious books of the Zoroastrians. Copenhagen.