The five Gāϑās and the Yasna Haptaŋhāiti (“Yasna in seven sections”) constitute the core of the Old Avestan (OAv.) sections of the Yasna (Y), a Zoroastrian ritual text commonly divided into 72 hāiti (“section, chapter”) and at the centre of many Zoroastrian rituals.Footnote ¹ As Cantera recounts, Scholars have long debated the structure of the Old Avestan texts, examining the age of the divisions of the Gāϑās into hāiti.Footnote ² In that same article, Cantera questions whether the Yasna Haptaŋhāiti (YH) originally consisted of seven hāiti, arguing that despite its name the Yasna Haptaŋhāiti “was not originally divided into seven chapters”.Footnote ³ Cantera's perceptive observations are based on the use of the YH in the ritual as reflected in one manuscript, but are not easily reconciled with the manuscript tradition that preserves a structure suggestive of a division into seven hāiti. Here, I will examine the extent of the concluding hāiti (Y 41) of the Yasna Haptaŋhāiti in light of the manuscript evidence and the hāiti's reception in Middle Persian (MP) texts, irrespective of the age and authenticity of the structural divisions. I am delighted to offer this paper to François de Blois in honour of his life's work.

In his edition of the Avestan texts, Geldner divides Yasna 41 into eight stanzas, while noting that the manuscript “Pt4 reckons 6 Strophes”.Footnote ⁴ Elsewhere, he assumes that the YH ends with Y 41.6.Footnote ⁵ NartenFootnote ⁶ and HumbachFootnote ⁷ posit six stanzas, Kellens and PirartFootnote ⁸ five, while HintzeFootnote ⁹ follows Geldner positing eight stanzas. By contrast, the exegetical manuscripts of the Pahlavi Yasna (PY) unambiguously assign six stanzas to Y 41, ruling out any additional strophe at the end of Y 41: šaš wīčast sē gāh.Footnote ¹⁰ Y 41.6 and the concluding instructions of the hāiti appear as follows in manuscript 0400 (Pt4):

The brief description of Y 41 in the Supplementary texts to the Šāyest nē Šāyest (Suppl.ŠnŠ) agrees with this count of the stanzas:

The priestly tradition, however, ostensibly contradicts itself when defining the extent or boundaries of the YH in an earlier paragraph:

Based on this passage, KotwalFootnote ¹⁶ views humatanąm (Y 35.1) as the first and final stanza of the YH, thus positing seven stanzas for Y 41.Footnote ¹⁷ This view is thus in disagreement with Suppl.ŠnŠ 13.22 and the number six given in the extent descriptions in the manuscripts.

While the first clause of Suppl.ŠnŠ 13.16 undoubtedly discusses the first section or chapter of the YH (yasn bun kardag), the referent of the following clause (u-š bun …) is grammatically ambiguous. It could refer to the YH as a whole or to Y 35. Both interpretations are grammatically permissible. Since the following paragraphs of Suppl.ŠnŠ 13 describe the individual kardag of the YH, one may be inclined to take the second clause as referring to Y 35 and not the whole of the YH. However, with only nine strophes assigned to Y 35 in Suppl.ŠnŠ 13.16 and at the end of the kardag in the exegetical manuscripts of the Yasna, it would be difficult to justify another stanza after Y 35.9. In fact, such a strophe, i. e. another humatanām (Y 35.1), is not attested in any of the examined manuscripts as a stanza or as a repetition.Footnote ¹⁸ Moreover, according to Suppl.ŠnŠ 13.51 the YH comprises 40 stanzas: ud yasn čehel wačast ud harw wačast-ē sē gāh “And the Yasna has forty stanzas and each stanza three lines”. The number forty, however, can only be accounted for if the YH starts at 35.1 (humatanām) and ends at 41.6. Therefore, if Suppl.ŠnŠ 13.16 does not constitute an otherwise unattested tradition, which would repeat Y 35.1 at the conclusion of the first kardag, we would have to concur with Kotwal that Suppl.ŠnŠ 13.16 refers to the end of the YH. Indeed, as mentioned above and in Table 2, the repetition instructions at the end of Y 41.6 indicate that Y 35.1 is to be recited twice at the conclusion of the YH. These repetitions, however, are not constituent stanzas of a hāiti and are not counted in the extent descriptions. Thus, the question emerges as to how we can explain the contradiction within the same text and exegetical tradition.

Four more paragraphs follow the description of Y 41 (Suppl.ŠnŠ 13.22), before Suppl.ŠnŠ 13 moves on to Y 43, the next OAv. hāiti within the Yasna:

Kotwal makes a number of observations on Suppl.ŠnŠ 13.22–26. Firstly, that the numerological interpretation of Y 41 is missing in Suppl.ŠnŠ 13.22 and that the paragraphs 13.22–26 are descriptions of those YH stanzas that are repeated more than once.Footnote ²⁰ Secondly, that 13.25, a reference to Y 41.5, has possibly been dislocated, as it occurs after the paragraph on Y 41.6's yeŋ́hē hātąm in 13.24;Footnote ²¹ finally, that mizd refers to the drōn ceremony.Footnote ²² However, the expectation of a numerological reading of Y 41 in 13.22 is not compelling as the first hāiti of the YH (Suppl.ŠnŠ 13.16) also lacks a numerological speculation. The assumption that the text is describing stanzas of the YH with multiple repetitions is contradicted by the fact that two stanzas with two repetitions, namely Y 39.4 & 41.3, are not mentioned in Suppl.ŠnŠ 13. Therefore, the text does not merely list stanzas of the YH with multiple repetitions, unless we presume that the latter were mistakenly omitted. Furthermore, the manuscripts do not support Kotwal's assumed disturbances in the text. His critical apparatus, for instance, does not indicate an omission in 13.22, and no text appears to be missing in F35 either. In my view, the question is whether Suppl.ŠnŠ 13 describes the OAv. corpus as a text or whether the exegesis follows the course of a ritual?

In the manuscript F35, Suppl.ŠnŠ 13 starts with the slightly indented words čim ī gāhān ēn “The meaning of the Gāϑās is this”, with the rest of the text following immediately.Footnote ²³ Accordingly, this chapter is concerned with the interpretation of the Gāϑās, at times describing individual stanzas, which, contrary to the subject matter, are not exclusively OAv. The description of the individual stanzas resembles the repetition instructions found in the manuscripts of the PY with the addition of a numerological interpretation. That Suppl.ŠnŠ 13 describes a ritual emerges from the frequent references made to passages from the Wisperad (Wr). In fact, Suppl.ŠnŠ 13.15 explicitly mentions the Wisperad ritual and interprets the ritual action connected with the Ahunauuaitī Gāϑā.Footnote ²⁴ Likewise, Suppl.ŠnŠ 13.9 refers to ritual actions. The fact that the text does not mention passages from the Wīdēwdād, suggests that Suppl.ŠnŠ 13 describes the Wisperad rather than the Wīdēwdād ritual.Footnote ²⁵ As Table 1 shows, with the exception of Wr 14, the intercalation of the Wisperad passages described in Suppl.ŠnŠ 13 is identical to the scheme laid out by Geldner.Footnote ²⁶

Table 1: Suppl.ŠnŠ 13 & Wr ritual

Thus, Suppl.ŠnŠ 13.26Footnote ²⁷ largely agrees with the intercalation of the Wisperad ritual, where Wr 16 is inserted between Y 42 and Y 43. The text ignores Y 42 because it is not part of the OAv. corpus and links Wr 16 (ātrəmca) to the YH by associating it with the fire Wāzišt. With Suppl.ŠnŠ 13.27 the text moves on to Y 43 and consequently mentions Wr 18–21 & 23, which are inserted at the end of each of the succeeding Gāϑās that the Suppl.ŠnŠ analyses. It appears, therefore, that the content of Suppl.ŠnŠ 13 is not confined to a textual discussion of the OAv. corpus, but follows the course of the Wisperad ritual. If this is indeed the case, then Suppl.ŠnŠ 13.25 is not dislocated, as Kotwal suggests, but refers to Y 41.5 as one of the stanzas recited at the end of the YH. Moreover, rather than representing the drōn ceremony, mizd in that same paragraph refers to Y 41.5c where the reward (Av. mīždəm) is explicitly mentioned: ē mizd ō manīgān frāz dahē “May you give my people the reward”. The paragraph only seems dislocated in comparison with the corpus, where Y 41.5 precedes the yeŋ́hē hātąm.

It is noteworthy that certain strophes positioned after Y 41.6 in Suppl.ŠnŠ 13.22–26 are not mentioned in the manuscripts in the repetition instructions at the end of the chapter. As Hintze points out, the sequence and repetition mentioned in the final section of the YH agree with those in Nērangestān 47.41: humatanąm (Y 35.1) to be recited two times, yaϑā ahū vairiiō four times and the aṣ̌ǝm vōhū three times.Footnote ²⁸ This sequence, however, conflicts with Suppl.ŠnŠ 13.22–26, which in addition mention Y 35.4 (“huxšatrōtemāi”), Y 35.7 (“aṣ̌ahiiā āat̰ sairī”) and Y 41.5 (“ϑβoi statarascā”). A comparison between Y 41.6's repetition instructions found in the various manuscripts of the PY and the information provided in the Suppl.ŠnŠ 13 leaves us with two different schemes (see Table 2).Footnote ²⁹

Table 2: Extent descriptions in Suppl.ŠnŠ & Y

As Cantera has observed, concluding instructions with a similar pattern occur at the end of each hāiti of the Gāϑās: the first strophe of the respective Gāϑā (2x); yaϑā ahū vairiiō (4x); aṣ̌əm vohū (3x); title of the hāiti + hāitīm yazamaide and the concluding yeŋ́hē hātąm (1x).Footnote ³⁰ Although Cantera states that a similar pattern can be observed at the end of the YH, in the manuscripts we find the concluding block untypically preceded by two repetitions of yeŋ́hē hātąm, while Suppl.ŠnŠ 13 offers a completely different view (see Table 2 for both).Footnote ³¹ As the manuscripts largely agree on these instructions at the end of the YH, we can perhaps propose that the pattern for the concluding block of prayers differs between the Gāϑās and the YH.

Be that as it may, it is difficult to establish whether Y 35.4, 7 & 41.5 in Suppl.ŠnŠ 13 are intended to replace the respective concluding prayers or if they are to be recited after them. If we take Y 35.1 and the second yeŋ́hē hātąm as a frame of agreement between the two sources, then it is likely that the stanzas mentioned in Suppl.ŠnŠ were substitutions for the prayers mentioned in the repetition instructions of the canonical manuscripts, thus representing a variant of the instructions perhaps for a variant ritual.

Recent scholarship has brought into focus the complex structure of Zoroastrian rituals.Footnote ³² Karanjia, for instance, has shown the complexity of the Bāj-Dharnā ritual, otherwise also known as the Drōn Yašt.Footnote ³³ Recently, Cantera has advanced the idea that the transmitted Avestan texts do not represent a rigid ritual structure.Footnote ³⁴ To the contrary, he argues, the intended ritual could trigger certain variations and combinations of the texts which are not always transmitted in the extant manuscripts, so for instance the lists of the textual ratu which could differ according to the type of ceremony.Footnote ³⁵ The manuscripts, however, do not record these variations. Elsewhere, Cantera notes the Wisperad Gāhānbār of Y 60.12 as one example, where the yaϑā ahū vairiiō is repeated ten times rather than the prescribed four.Footnote ³⁶ More importantly, Cantera regards the Wisperad ritual as “the basis for the celebration of other variants of the [long] liturgy”.Footnote ³⁷ Suppl.ŠnŠ 13 claims to explain the meaning of the Gāϑās. It does this by including the YH as part of the Gāϑās and intercalating them with material from the Wisperad. In doing so, the text sets the OAv. corpus within the context of a ritual in which it is then interpreted. In this light, the disagreements between the repetition instructions and N 47.41 on the one hand and the Suppl.ŠnŠ 13 on the other, do not constitute inconsistencies, but rather Suppl.ŠnŠ 13 could refer to a different ritual with an alternative sequence of prayers at the end of the YH. That the YH might have played a significant role in this hitherto unknown ritual emerges from the fact that the major differences noted here relate to the sequence and repetitions of this text's stanzas at the end of Y 41. The exact nature and textual structure of this ritual, however, remain elusive and the question arises as to whether the omission of Wr 14 after Y 34 is an error or a deliberate part of this alternative ritual. We may also ask as to why Supp.ŠnŠ 13 does not mention Y 27 and Wr 12?

Admittedly, the exegesis of the Gāϑās in Suppl.ŠnŠ 13 is enigmatic. While the text approaches a ritual text from an exegetical and numerological vantage, its reading of the text does not appear systematic at first. As a result, Kotwal makes frequent references to misplaced passages within Suppl.ŠnŠ 13.Footnote ³⁸ Although dislocation and loss of content must remain a distinct possibility, I would like to propose that Suppl.ŠnŠ 13 follows the course of a hitherto unknown ritual or a manuscript with a set of variations in its repetition instructions. At this point, it might be instructive to visualise the manner by which Suppl.ŠnŠ 13 approaches the gāhān and their hāiti (see Table 3).

Table 3: The gāhān in Suppl.ŠnŠ 13

The text shows a thorough understanding of the structure of the gāhān, their ritual as well as numerological significance. The discussion of the gāhān is not strictly limited to the Gāϑās, but includes the YH, airiiaman išiia and Y 58. The final two paragraphs (50 & 51) of the text offer statistics on total numbers of stanzas, lines, words and syllables in the Gāϑās. They also define the first and final stanzas of the Gāϑās. The extent of the individual Gāϑās are clearly defined, each by a reference to the concluding twofold repetition of its first stanza. The Ahunauuaitī Gāϑā (Y 28–34) is the only exception, as two more paragraphs (13.13 & 14) interpret its content and one (13.15) discusses ritual actions for all the Gāϑās after the concluding Y 28.1. Unexpectedly, Wr 14 does not follow Y 34 in Suppl.ŠnŠ 13. Likewise, despite mentioning Wr 20–23, Suppl.ŠnŠ 13 is silent on the second YH, which is typically recited after Wr 20. As already pointed out, the numerological interpretation of some passages is missing in Suppl.ŠnŠ 13. For instance, the manuscripts mention the repetition of Y 33.14 but not Suppl.ŠnŠ 13. If we eliminate the possibility of unsystematic exegesis and missing or misplaced passages, then these perceived inconsistencies appear as features of an unknown variant of a Wisperad ritual. Although it is very unlikely, these could also have been characteristics of a manuscript that must have formed the basis for the exegesis in Suppl.ŠnŠ 13, a chapter that is preceded (Suppl.ŠnŠ 12) and followed (Suppl.ŠnŠ 14) by discussions of ritual matters. Perhaps the references to the nōzūd Footnote ³⁹ and nāwar ceremonies in Suppl.ŠnŠ 13.2 are not the results of a misplacement, as Kotwal suggests,Footnote ⁴⁰ but somehow the context of the ritual described and interpreted in Suppl.ŠnŠ 13.Footnote ⁴¹

If we thus accept that Suppl.ŠnŠ 13 describes an alternative ritual, possibly an alternative intercalation of the Yasna with the Wisperad, then we can also reevaluate Suppl.ŠnŠ 13.16: yasn bun kardag nō wačast u-š bun humatanām u-š sar humatanām. I subscribe to Kotwal's interpretation of the passage, that humatanām refers to the beginning and end of the YH.Footnote ⁴² However, in view of the fact that Suppl.ŠnŠ 13.22 differs from the repetition instructions found in the extant PY manuscripts, I would like to leave open even the possibility that Suppl.ŠnŠ 13.16 represents a ritual tradition that repeated Y 35.1 at the end of the text's first chapter.

Moreover, the discrepancies between the manuscripts of the Yasna and Suppl.ŠnŠ 13 reveal how the exegetes approached the OAv. corpus. As we have seen, the extant Yasna preserves a ritual text that was subject to a certain degree of variation, even if the extant manuscripts do not attest these clearly and widely. In the ritual, it was possible to add prayers in various positions, and stanzas or parts thereof reoccur in other passages. Indeed, the Yasna is the proverbial example of this practice. Y 58.8, for instance, quotes Y 36.6 verbatim, while Y 68.23 only quotes Y 36.6b. Similarly, Y 5 anticipates Y 37, causing its abbreviation in many of the manuscripts at its original place between Y 36 and 38. In addition, the prayers occur throughout the Yasna. In a tradition that permitted such modifications for ritual purposes, it was essential to keep track of the “original” text. This is not surprising as already in the YAv. Yasna and in the Wisperad, the OAv. corpus and particularly the YH were viewed as sacrosanct texts. Defining the extent of the OAv. corpus, allowed the priests to rearrange the text for the ritual without risking disturbances in the transmission of the texts. Therefore, descriptions such as šaš wīčast sē gāh, were crucial in maintaining the boundaries of the text. In this way, the exegetes could accommodate their desire to conclude the YH with its initial stanza (Y 35.1), an attempt at constructing a compositional cycle, while the borders and extent of what was regarded by the priestly tradition as the original composition were still maintained by counting the stanzas. The manuscripts thus transmit the ritual text along with variations and repetitions while such descriptions of the extent of the text preserve the necessary boundaries. In this way, the Zoroastrian tradition maintained a distinction between ritual performance and ritual text.