Alberto Cantera How Many Chapters Does the Yasna of the Seven Chapters Have?

Transcription

1 Iranian Studies, volume 45, number 2, March 2012 Alberto Cantera How Many Chapters Does the Yasna of the Seven Chapters Have? This article argues that the Avestan text Yasna Haptaŋhāiti ( Yasna of the seven chapters ) was not originally divided into seven chapters. It was divided into three parts that were not considered hāitis (chapters) in the same way as the gathic hāitis. The division of the Yasna Haptaŋhāiti into seven parts and its corresponding title is based on the arrangement of the Old Avestan texts. These texts try to reproduce the structure of the prayer Ahuna Vairiia; three verses with two hemistichs each correspond to three pairs of compositions with 7, 4 four and 1 hāiti each of the Old Avestan texts. The Old Avestan texts in isosyllabic meters appear in the long ceremony divided into five units, Gāθā (song). In this liturgy each Gāθā in its turn is divided into smaller units called hāiti (chapter). There has been a long discussion about which was the primary textual unit of these compositions. The main alternatives are: 1. the textual unit is the hāiti, which roughly corresponds to the Vedic hymns, and the Gāθās are collections of different hāitis arranged according to their meters, 1 but not representing the original ordering. 2. the textual unit is the Gāθā, and the hāitis are later divisions, introduced in the arrangement of the Gāθā for the long liturgy. 2 This would imply a clear difference with the Vedic hymns, at least concerning the length. That the hāitis are the actual textual units of the Old Avesta has recently been demonstrated in different ways. But it has been also recognized that Molé s arguments about the inner unity of the Gāθās are compelling. Each Gāθā is thus also a textual Alberto Cantera is a Professor at the University of Salamanca. His current research is focused on the Avesta in its Avestan and Pahlavi versions and on its transmission. He is the director of the Avestan Digital Archive ( 1 See James Darmesteter, Le Zend-Avesta, 3 vols. (Paris, ), 1: xcviii; Wolfgang Lentz, Yasna 28: Kommentierte Übersetzung und Kompositions-Analyse (Mainz, 1955); Hans-Peter Schmidt et al., Form and Meaning of Yasna 33 (New Haven, CT, 1985). 2 See Marijan Molé, Culte, mythe et cosmologie dans l Iran ancien (Paris, 1963). Note that the expression long liturgy, introduced by Jean Kellens in Considérations sur l histoire de l Avesta, Journal asiatique, 286 (1998): , refers to the different variants of the Yasna ceremony celebrated in the Zoroastrian temples by priests. It corresponds to the term inner ceremonies used by other authors. ISSN print/issn online/12/ The International Society for Iranian Studies

2 218 Cantera unit of its own, not just an accidental collection of compositions organized according to their meters. So in the latest discussions, 3 the real existence of the hāiti as a textual unit in its own right is unanimously admitted, but without excluding further larger units. All through the arrangement of each Gāθā, as well as in the arrangement of the complete core of the Staota Yesniia (the texts comprised between the Ahuna Vairiia and the Ariiaman Išiia), there are identifiable structural and sequential patterns. Both compositional levels are mutually compatible, and even more levels may have been at work. A bigger textual unit is, for example, the complete Yasna collection. In this paper I shall deal with the hypothesis that the hāitis of the Yasna Haptaŋhāiti (YH) were, as has been supposed for the Gāθās, only secondary divisions of a unitary text. However, my arguments for supporting this hypothesis are totally different from the starting points of Molé or Kellens concerning the Gāθās. There is cogent evidence in favor of the hāitis being true textual units for the Gāθās, but I do not think that the same is true for the YH, or at least not in the same sense. The sections of the YH are not used in the long liturgy as proper hāitis. Notwithstanding, this is the only composition of the whole Avesta bearing a title which mentions the number of hāitis it consists of. This paradox is the starting point of my investigation. In my view these contradictory data can in fact be explained as a result of the long process of constitution of the corpus of the Staota Yesniia in its actual form. While scholars still discuss whether the Gāθās are unitary texts or collections, nobody seems to doubt that the YH is but one composition consisting of seven chapters. The title given to this composition within the long liturgy, Yasna Haptaŋhāiti (Yasna of the Seven Chapters), has prevented any debate. As a matter of fact, the unitary character of the YH seems very likely, and the structure proposed most recently by Hintze is convincing. Its division into 7 hāitis, however, seems to me much more suspect. Nevertheless, Narten accepts the transmitted division, since the seven chapters are clearly divided according to their contents. 4 Hintze essentially repeats the argument: Each [hāiti] has its own theme and internal structure. However, in her analysis of the structure she does not make use of the divisions in hāitis, but keeps to the division of the YH into three parts. My point is that some details within the recitation of the Yasna Haptaŋhāiti in the different variants of the Yasna ceremony raise doubts about the original character of its articulation in seven sections. In the long liturgy each hāiti of the Gāθās ends with a closing sequence that always follows the same pattern: 1. two recitations of the first strophe of the Gāθā; 2. four recitations of Ahuna Vairiia (only in the Ahunauuaitī Gāθā); 3 See Almut Hintze, On the Literary Structure of the Older Avesta, Bulletin of the School of Oriental and African Studies, 65 (2002): 31 51; Stephanie W. Jamison, The Rig Veda between Two Worlds = Le Rgveda entre deux mondes (Paris, 2007); Jean Kellens, Controverses actuelles sur la composition des Gāthās, _ Journal asiatique, 295 (2007): See Johanna Narten, ed. and trans., Der Yasna Haptaŋhāiti (Wiesbaden, 1986).

3 How Many Chapters Does the Yasna of the Seven Chapters Have? three recitations of the aš əm vohū; 4. title of the hāiti + hāitīm yazamaide; 5. one Yeŋ hē Hātąm. Now, in the case of the YH this closing pattern appears only after Yasna (Y.) 41. The end of each of the seven sections of the YH is marked in the liturgy only by the introduction of one or two Yeŋ hē Hātąm. This indicates that for the arrangers of the long liturgy the YH does not consist of seven proper hāitis, at least not in the same sense as the hāitis of the Gāθās. Nonetheless, a division of the YH into seven parts was certainly known to them. The very fact that Y. 35, 39 and 41 are closed by two Yeŋ hē Hātąm instead of one suggests an alternative division of the YH into three sections: an introductory Y. 35, a core consisting of Y (the pairijasāmaidē [Y. 36] and yazamaidē sections [Y ], probably the text for the meat offering to the fire), and a concluding section (Y ). This traditional division better reflects the true structure of the YH than its division into seven parts. Some modern authors 5 also propose a threefold division. But unlike the arrangement in the long liturgy, they ascribe Y. 36 to the introductory section. The texts of the long liturgy clearly ascribe it to the central core, and this seems to me consistent with the ritual course of the ceremony. Y. 36 seems to mark the beginning of the ritual action accompanied by YH, the offering of meat to the fire, with the consecration of the fire on which the victim is going to be placed. Furthermore, the textual coherence of the third section (Y ) is unquestionable. The repetition of hiiat mīždəm mauuaiϑim fradadāϑā daēnābiiō mazdā ahurā in Y and Y opens and closes this section of the YH. But even these three main divisions of the YH were not treated as proper hāitis within the long liturgy. The liturgy as it appears in the manuscripts presents the YH as a unitary text divided into three main sections that could be subdivided into 1, 4 and 2 parts, respectively. None of these are considered to be hāitis in the same way as the gāθic hāitis. Let us compare the closing formulas of the complete YH with those of the Gāθās. The end of each Gāθā is marked by the same closing formula as each hāiti of the same Gāθā. However, in the final yazamaide formula, instead of the simple title of the hāiti, we find the following sequence: title of the hāiti in accusative + hāitīm yazamaide. title of the Gāθā in accusative + gāϑąm ašạonīm ašạhe ratūm yazamaide. title of the Gāθā in genitive + gāϑaiiā hanḍātā yazamaide 5 See Jean Kellens and Éric Pirart, eds. and trans., Les textes vieil-avestiques, 3 vols. (Paris, ), 3: 124; Mary Boyce, Zoroastrianism: Its Antiquity and Constant Vigour (Costa Mesa, CA, 1992), 89ff.; Almut Hintze, ed. and trans., A Zoroastrian Liturgy: The Worship in Seven Chapters (Yasna 35 41) (Wiesbaden, 2007), 6ff.

4 220 Cantera Now, at the end of the YH we find basically the same pattern (Y ), but the formula is just: yasnəm. sūrəm. haptaŋhāitīm. ašạuuanəm. ašạhe. ratūm. yazamaide The absence of the mention of hāiti at this point is consistent with the absence of any proper hāiti closings at the end of each section of the YH. And the formula with hanḍatā is also omitted. The arrangers of the long liturgy treated each Gāθā as a collection of different constitutive parts (hanḍatā). However, for them the YH does not consist of hanḍatā. The underlying tradition that does not consider the YH as divided into hāitis have left traces even in the written transmission. In the important Iranian manuscript Vīsperad-Sāde G18b, 6 dating from the seventeenth century, the divisions of YH into seven parts are consistently marked. However, whereas each hāiti of the Gāθās is introduced by the designation hāh, which corresponds to Av. hāiti, the seven divisions of the YH are not called hāh but karde, like the sections of the Yašts and other texts. This is no idiosyncrasy of G18b alone, since it is confirmed by the Nērangestān. The expression ēsn ī kardag mayān in Nērangestān clearly refers to the middle section of the YH (Y ), as the most recent editors of the Nēragestān have rightly recognized. 7 All this is consistent with the probable fact that YH was never ritually divided. In the ritual the YH always appears as an indissoluble unit, which is proved by the lack of intercalations within its recitation. In some ceremonies based on the Visperad, Young Avestan texts are sometimes intercalated within the core of the Staota Yesniia, but although in the Vīsperad, Vīdēvdād and Vīštāsp Yašt Young Avestan texts are intercalated in the midst of the Ahunauuaitī Gāθā, there is no intercalation at all within the YH. Moreover, while the exegetic Nasks (books) of the Great Avesta 8 (Sūdgar, Warštmānsar and Bayān Nask) introduce an exegetic intercalation after the four prayers (Ahuna Vairiia, Aš əm Vohū, Yeŋ hē Hātąm and Ariiman Išiia) and after each hāiti of the Gāθā, there is only one exegetical intercalation after the complete YH, and none after each of its hāitis. The origin of such intercalations can be a ritual similar to the actual Vīdēvdād ceremony, in which exegetical texts were intercalated within the recitation of the Staota Yesniia. Through these intercalations the core of the Staota Yesniia becomes divided into 22 sections. This number of fragards agrees 6 This manuscript is now preserved at Meherji Rana Library. Its electronic reproduction, prepared by myself, is available in ADA ( Frēdōn Marzbān Frēdōn Wahrom Rōstām Bundār Šāhmardān joined in Kermān in the seventeenth century, in a single manuscript to be sent to India, two older manuscripts containing the first one, the Vīštāsp Yašt, and the second one, the Yašt Vīsperad. He copied the Vīštāsp Yašt Sāde with Nērangs (G18a) from a manuscript copied by Husrōšāh, son of Anōšāgruwān Rustām, year Y. (= 1344), one of the oldest Avestan manuscripts known. The Vīsperad was copied from a manuscript written for him by Manušcihr Ardašīr Wahrom Sfandyād Ardašīr in the year 996 of Yazdegird 20 (= 1647). 7 See Firoze M. Kotwal and Philip G. Kreyenbroek, The Hērbedestān and Nērangestān, vol. 3, Nērangestān, Fragard 2 (Paris, 2003), The expression Great Avesta refers to the compilation of Zoroastrian writings ordered in three groups of 7 Nasks that it is described in the 8th book of the Dēnkard and in other Pahlavi works.

5 How Many Chapters Does the Yasna of the Seven Chapters Have? 221 with the number of fragards of Vīdēvdād. Hence it seems likely that a Vīdēvdād ceremony with 22 intercalations instead of 10 was possible, and that the exegetical Nasks reproduce a similar scheme. The 22 sections are the 17 hāitis of the 5 Gāθā + the 4 prayers (Ahuna Vairiia, Aš əm Vohū, Yeŋ hē Hātam and Ariiaman Išiia) + the undivided YH. These 22 sections appearing in the exegetic Nask of the 9th book of the Dēnkard and in Vīdēvdād are the exact correspondence of the Ahuna Vairiia + its 21 words. In this symbolic arrangement the YH counts as just one unit and not as seven, whereas in the Gāθās each hāiti counts as one unit. There seems to be also an alternative counting of the YH as 7 hāitis, but this is occasional and certainly banal. As soon as the division into seven hāitis has been introduced, the YH can count as 7. But in the most relevant numerical pattern (the 21 sections of the core of the Staota Yesniia), where each hāiti of the Gāθās counts as 1 unit, the YH as a whole counts as 1 unit as well. Windfuhr recognizes a numerical value as 7 for the YH in the 24 words of the Ariiaman Išiia, which according to him were conceived as equivalent to the 17 hāitis of the Gāθās + the 7 of the YH. 9 However, the tradition does not attribute any exegetical value to the 24 words of the Ariiaman Išiia, and this number finds no reflex in the arrangement of the ritual texts, in contrast to the 21 words of the Ahuna Vairiia, which Windfuhr correlates with the 21 northern constellations and with the 21 Nasks. The only sure numbering in which the YH is indeed counted as seven hāitis is the arrangement of the Yasna into 72 hāitis. Although some manuscripts, such as G18b, do not consider the divisions of the YH as hāitis, it is true that they are counted as seven in the general arrangement of the Yasna. What follows after the YH is Y. 42, not Y. 36. But I don not believe that this is an argument supporting an original division of the YH into seven chapters. Rather, I see here an indirect confirmation that the sevenfold division of the YH is secondary. The total number of hāitis of the Yasna collection is remarkable: 72 does not represent any significant numerical value within the Avestan compositions, and it is also no multiple of any of the recurrent numbers. 10 The most recurrent numbers in the patterned structures of the Yasna and its variants are the number and its multiples (especially 22 and 33) on the one hand, and on the other 1, 3 and 7, which in turn appear as closely related, since 7 is usually analyzed as 1 + [3 + 3]. The number 22 links both series, since it is analyzed as 1 + [7 x 3], being 7 = 1 + [3 + 3], and as 11 x 2. Therefore, 22 is the most perfect number, in which both numerical series are combined. It is the number representing the Ahuna Variiia (the whole 9 See Gernot Windfuhr, Zoroastrian and Taoist Ritual: Cosmology and Sacred Numerology, Zoroastrian Rituals in Context, ed. Michael Stausberg (Leiden, 2004), Windfuhr ( Zoroastrian and Taoist Ritual, 217) explains the number 72 as the result of two cycles in the long liturgy: one based on the number 6 (72 = 6 x 12), with high points in every third group of 6 (in the hāitis 18, 54 and 72), and a second one based on the number 9 (72 = 9 x 8) with the same high points (since the cadence is 18 = 9 x 2 and 6 x 3). However, I do not think that these cycles have any textual support, fragard 18 being the weakest point. 11 See the hendecatropism of the Gāθās recently proposed by Kellens again (see Kellens, Controverses actuelles, 418).

6 222 Cantera Ahuna Vairiia + its 21 words), the whole Avesta (the Avesta or its equivalent the Ahuna Vairiia + the 21 Nasks), the Staota Yesniia in at least one version, 12 and it is also the number of fragards of Vīdēdād and of the exegetical Nasks described in the 9th book of the Dēnkard. The number 72 cannot be analyzed according to these numeric patterns. If the YH did not count originally as 7 hāitis, but as 1 hāiti, then the total amount of hāitis would turn out to be the non-accidental number 66, a multiple of 11 and the double of 33. This agrees with the fact that the first hāiti of Yasna determines the number of the textual ratu parišhāuuani as 33, as well as with the further fact that a late tradition also fixes the Staota Yesniia as Although an analysis of the whole Yasna as an addition of ratu- parišhāuuani + Staota Yesniia is not likely, it seems that there was a trend to analyze the whole Yasna as consisting of collections of 33 hāitis. Even the analysis of the Staota Yesniia as 22 chapters could fit in well with this number. The Yasna would be a triadic structure of three times 22 sections, its middle section being the Staota Yesniia between the Ahuna Vairiia and the Ariiaman Išiia. Although in the actual arrangement of the Yasna, as well as in the actual arrangement of the central core of the Staota Yesniia, successive adaptations to numeric patterns have taken place, the final number of 72 hāitis could be the result of splitting the YH from its value as one section to its value as 7 sections, so that the original number of 66 hāitis became 72. There is hence an obvious contradiction between the title given to this composition in the long liturgy, yasna haptaŋhāiti or Yasna of the Seven Hāitis, and its usage at the same liturgy, where its sections are not treated as hāitis. There are two possible explanations for this situation: 1. the original divisions were removed from the long liturgy or 2. a unitary text was secondarily divided into seven parts The sevenfold division of the text is certainly possible, but it is in no way demanded by the text itself. Therefore, the most likely explanation would be precisely the solution that can be motivated best. So let us ask whether there is any plausible motivation for not considering the divisions of the YH as hāitis, and for counting the YH as 1 unit instead of 7, given that the rest of the hāitis each count as 1? Or are there, after all, plausible reasons for dividing a unitary text later on into 7 parts? To begin with, there is at least one easy explanation for why the hāitis of the YH did not receive the same treatment as the gāθic hāitis. It is highly plausible that only the hāiti in metric verse-lines with a fixed number of syllables were considered worthy of receiving the complex division markers reserved for the gāθic hāitis. Also the hāitisof the Young Avestan parts lack such boundaries. A further possibility which could account for the first alternative would be that the gāθic hāitis were the texts for specific 12 I proposed this recently in a communication at the colloquium Questions zoroastriennes II. La sort des Gāθās, held in Liège on April 22 23, See Darmesteter, Zend-Avesta, 1: lxxxvii.

7 How Many Chapters Does the Yasna of the Seven Chapters Have? 223 rites that required special boundaries. Nevertheless, at first glance the lack of any ritual action from the YH to the Ariiaman Išiia in modern ceremonies seems to weaken this explanation. Actually, on the one hand the YH, despite its title, counts nearly always as one unit and not as seven, that is, the problem is not that the seven divisions are not considered as hāitis, but they are not considered at all as true divisions. On the other hand, the divisions of the YH as they appear in the long liturgy are after all not exactly seven, but three (for which the structural arguments are obvious), further subdivided into 1, 4 and 2, and thus yielding the total sum of 7. Therefore, let us concern ourselves briefly with the opposite alternative: could there be any motivations for a secondary, artificial division of the YH into seven hāitis? In my opinion there is at least one clear one: the ceremonial introduction of the YH amidst the five Gāθās made it necessary in any case to adapt it formally to the structure of the five Gāθās. In recent times the central position of the YH in the ritual sequence has been explained repeatedly 14 because of its being the nucleus of a series of concentric circles. Furthermore, the YH participates in the structure of the Staota Yesniia as an extension of the Ahuna Vairiia. According to its position therein, it corresponds to the second hemistich of the first verse of the Ahuna Vairiia. The core of the Staota Yesniia is organized in its actual arrangement clearly as an expanded version of Ahuna Vairiia, exactly like the whole Avesta (as described in the Dēnkard) is also considered an expansion of the Ahuna Vairiia. 15 The Ahuna Vairiia represents the center of the revelation that the Avesta consists of as a whole. The rest of the Avesta and the Staota Yesniia are presented as expansions of the Ahuna Vairiia and tend to be arranged according to its constitutive elements: 3 verses divided into two hemistiches; words (7 words x 3 verses). In the exegetical literature, the Great Avesta is presented as sharing the same organization: it consists of 21 Nasks organized in 3 groups of 7 Nasks. Each group can be divided into two subgroups matching the hemistiches of the verses of the Ahuna Vairiia. There has been a conscious attempt to get the Staota Yesniia to fit inwiththis pattern. However, to obtain the correspondence of the 21 words, some violence was needed. The Ahuna Vairiia and the Ariiaman Išiia were segregated from their corresponding Gāθā, andtheaš əm Vohū and the Yeŋ hē Hātam received 14 See Gernot Windfuhr, The Word in Zoroastrianism, Indo-European Studies, 12, nos. 1 2 (1984): ; Hintze, On the Literary Structure. 15 For a recent account of the presentation of the Great Avesta as an expansion of the Ahuna Vairiia see Yuhab Sohrab-Dinshaw Vevaina, Enumerating the Dēn : Textual Taxonomies, Cosmological Deixis, and Numerological Speculations in Zoroastrianism, History of Religions, 50, no. 2 (2010): In fact, it has the numerical value 7, counting the whole prayer as 1 and its constitutive elements as 6 (3 x 2 or 3 + 3).

8 224 Cantera consideration as Staota Yesniia. The organization of three verse-lines with two hemistiches was matched by grouping the Gāθās in pairs according to theirnumberofhāitis. Accordingly, the correspondence to the six hemistiches in three verse lines were the five Gāθās+the YH. EachtwoGāθās withthe same number of hāitis built one unit, corresponding to one verse of the Ahuna Vairiia. Now, the number of hāitis within each pair of Gāθās is also not accidental. It follows throughout a well-known pattern in the Young Avesta: The Gāθās are thus arranged as a series with decreasing number of hāitis according to the following numerical pattern: 7 (= ), 4 (= 1 + 3) and 1. The structure of the Staota Yesniia as parallel to the Ahuna Vairiia is the following: Ahuna Vairiia yaϑā ahū vairiiō. // aϑā ratuš aš āt cīt hacā. vaŋhəūš dazdā manaŋhō // šiiaoϑananąm aŋhəūš mazdāi. xšaϑrəmcā ahurāi.ā // yim drigubiiō dadat vāstārəm. Gāθās +YH Ahunauuaitī Gāθā (7 hāitis) // Yasna Haptaŋhāiti (?) Uštauuaitī Gāϑā (4 hāitis) // Vohuxšaϑrā Gāϑā (1 hāitis) // Spə _ ntā.mainiiū Gāϑā (4 hāitis) Vahištōišti Gāθā (1 hāitis) Consequently, if the YH had to be introduced after the Ahunauuaitī Gāϑā, it had to consist of 7 hāitis, merely in order to fulfill the formal requirements attached to being the correlate of the second hemistich of the Ahuna Vairiia. So my conjecture is that the YH was divided into seven hāitis in order to make it able to fit into the structure of the Staota Yesniia as an expansion of the three verses of the Ahuna Vairiia. This explanation also provides us with a good reason for the title given to this yasna, that is, for the very fact that the number of hāitis is constitutive of the title only in the case of the YH, which was the only one not being originally constituted by hāitis. The title is thus programmatic, since it signals the formal condition which allows the insertion of the YH at this place without distorting the structure of the Staota Yesniia. Different scenarios can be imagined in which the need to adapt the YH to this schema was felt: The YH could have been recited together with the five Gāθās, but the Gāθās had been rearranged in order to adapt their number of hāitis so that the Old Avestan

9 How Many Chapters Does the Yasna of the Seven Chapters Have? 225 texts could appear formally as an extension of the three verse-lines of the Ahuna Vairiia, and the same had to be done with the YH. The five Gāθās already had their actual number of hāitis, and the YH was subsequently intercalated between them. Accordingly, it had to be adapted in order to complete the desired schema. Unfortunately, although our observation allows us to discover that there has been a readaptation of the YH in order to fit the schema of the Ahuna Vairiia, it does not allow us to reconstruct the historical facts behind this adaptation. However, even if it implies a certain degree of speculation, we can try to reconstruct a possible scenario for this adaptation. To begin with, several facts point to the direction that the YH could be a secondary addition. Metrics and style detach the YH from the Gāθās. And as a matter of fact, the five Gāθās withouttheyasna are mentioned as the group of texts to be recited in a certain ritual. Sraoša is said to have been the first who recited the five Gāθās (without the YH) in the context of a yasna ceremony (Y. 57.8):yō paoiriiō gāϑā frasrāuuaiiat yā pa _ nca spitāmahe ašạonō zaraϑuštrahe (who first recited the Gāθās, the five of Spitama Zaraθuštra, the supporter of Truth). Still in other texts the five Gāθās appear together with the mention of the YH, so that obviously the expression five Gāθās does not include the YH (e.g. Y. 71.6). 17 Furthermore, we have some evidence that the YH could be intercalated amidst the Gāθās for certain exegetical or liturgical reasons, since in the Visperad ceremonies it appears not only after the Ahunauuaitī Gāθā, but also after the Vohuxšaθrā Gāθā. The arrangement of the verbal utterances accompanying ritual actions in the frame of a complex ritual depends greatly on the structure of the ritual itself and on the distribution of the actions they accompany. Nevertheless, within Avestan studies the ritual words themselves have always been the main focus, both because we know little about the ancient ritual practices of the ceremonies and consequently the approach to the Avestan texts had to be mostly philological. This is why the ritual information provided by the Nērangestān, by the ritual instructions contained in the manuscripts and by the related Vedic ritual has not received due attention until recent times. My position is that the actual position of the YH could be explained in ritual terms. A close connection between the recitation of the YH and the ritual meat offerings to the fire has already been proposed long ago. 18 Although this point requires a more detailed discussion which I plan to offer elsewhere, the possibility that the YH consists of the utterances to be recited during the meat offerings to the fire (after the slaughtering of the victim prior to the recitation of the YH) seems very likely. In fact, the 17 See Almut Hintze, On the Ritual Significance of the Yasna Haptaŋhāiti, in Stausberg, Zoroastrian Rituals in Context, 306ff. 18 See Mary Boyce, Haoma, Priest of the Sacrifice, in W.B. Henning Memorial Volume (London, 1970), 68; Mary Boyce, A History of Zoroastrianism, vol. 1, The Early Period (Leiden, 1975), 117. Objections in Narten, Der Yasna Haptaŋhāiti, 249ff. and a short discussion by Hintze in A Zoroastrian Liturgy, 258ff.

10 226 Cantera Nērangestān (fragard 47 [65]) places the animal sacrifice clearly in the sphere of the YH. According to this description it seems that the slaughtering of the sacrificial victim and the cutting of the offerings to be given to the fire had to be done during the recitation of three Ahuna Vairiia. This had to happen outside the ritual place and before reciting Y. 36.2, since at the beginning of this stanza fat from the victim is offered to the fire. According to the Nērangestān, during the recitation Y the priest presents a piece of fat to the fire, and while beginning to recite Y he offers it to the fire. The recitation of the middle section (Y ) takes place while the fire is consuming a small piece of meat hanging from a bone (Nērangestān 47.40) that the priest holds aloft over the fire. This passage of the Nērangestān offers little doubt that the YH consists of the utterances for the old genuine yasna, which consisted in bringing the meat offerings to the fire. In the actual ceremony the offerings of meat have been substituted by offerings of sandalwood and frankincense by the rāspī at the beginning of Y. 36. When the practice of the animal sacrifice was abandoned, all ritual actions corresponding to it were also dropped, and this is why there is no more ritual action in modern ceremonies until Y This interpretation also fits in perfectly with Kellens identification of the theogony mentioned by Herodotus with the YH. According to Herodotus, after cutting the pieces of meat a theogony was recited, and at the end the priest took the pieces of meat away. If this theogony is the YH, as seems likely, then this information corresponds exactly to the description in the Nērangestān: during the recitation of three Ahuna Vairiia before the Y the victim was cut in pieces, and afterwards the YH was recited and a small part of the victim offered to the fire. The witness of Herodotus confirms that the YH were the words of the rite of the animal sacrifice. They could also be recited independently of the Gāθās, just like the five Gāθās themselves could be used in rituals without including the YH. The introduction of YH amidst the Gāθās is likely to respond to a certain interpretative schema of the five Gāθās, according to which the animal sacrifice had to take place after the recitation of the Ahunauuaitī Gāθā and before the Spə _ ntā.mainiiū Gāθā. The animal sacrifice, performed at this moment, had to be accompanied by the most renowned utterances for this ritual action: the YH. In order to introduce the YH into the extremely patterned structure of the five Gāθās, and to build with it a structure parallel to the structure of the Ahuna Vairiia, some changes proved unavoidable. As we have already seen, to be included in the Staota Yesniia it had to count as 7 chapters. Although it did not, it was a plausible candidate for the correspondence of the second hemistich of the Ahuna Vairiia: the YH are the utterances of the ratu par excellence, 20 the ratu of the most important ritual action of the old yasna: the meat offerings to the fire, which go along with the identification of the soul of the 19 See Firoze M. Kotwal and James W. Boyd, A Persian Offering: The Yasna; A Zoroastrian High Liturgy (Paris, 1992), 112ff. 20 It is likely that the expression aš ahe ratu echoes the ratuš aš ātcit hacā of the first verse-line of the Ahuna Vairiia.

11 How Many Chapters Does the Yasna of the Seven Chapters Have? 227 victim with the soul of the sacrificer. The correct time for the central action of the yasna is represented textually by the YH. Accordingly, a secondary organization in seven chapters was introduced into it, although these new divisions never received the same status as the old divisions of the Gāθās. The YH is a ritual text with direct references to the ritual action it served. Since this ritual action disappeared as a consequence of abandoning animal sacrifice, such references are no longer easy to understand completely, but sometimes they can still be recognized. In the Gāθās, direct ritual references are not as obvious as in the YH. However, their arrangement depends on the ritual exactly like the arrangement of the YH. The position of the different Gāθās in the long liturgy depends on a certain exegesis of the text that interprets the Gāθās following the YH as speculative utterances about the union of the soul (uruuan) of the Cow/sacrificer with the vision-soul (daēnā) of the sacrificer. The ritual represents a journey. It starts with the slaughtering of the cow, after the Ahunauuaitī Gāθā and before the YH, and with the identification of the victim s offered meat with the body of the sacrificer (Y. 37.3), and it ends with the union of the sacrificer s soul with his vision-soul in the nuptial hymn of Y. 53. So the position of the YH in the long liturgy does not depend on its number of hāitis, but on its ritual function. Its number of hāitis is most likely secondary and the result of an adaptation to the number of hāitis of the Ahunauuaitī Gāθā, so that the structure of the core of the Staota Yesniia could match the structure of the Ahuna Vairiia. The whole composition of the Staota Yesniia is no accidental collection of older ritual texts inserted in a long ceremony, of course. But this does not mean at all that these texts are the result of a singular compositional decision by a single poet-priest. It is more likely to be the result of a longer historical process of ritual refinement, along which textual adaptations were undertaken as a result of ritual changes or of exegetical movements. The linguistic similarity between the different Gāθās and between the Gāθās and the YH in a very conservative milieu does not exclude slight chronological differences we are not able to evaluate exactly. Unfortunately, we lack any information about such a historical process. Only a detailed analysis of the treatment of the Staota Yesniia in the different long liturgies and of the underlying exegetical movements could provide us with a glimpse into this process.