The idea that ritual reproduces the cosmic order (re-enacts the original creative act and thus preserves the established order) has found wide acceptance and application among students of Zoroastrianism. The context for the adoption seems to be the realization of the central role of ritual in Zoroastrianism. This awareness underlies all the major approaches to the Avesta today, despite their differences in the interpretation of various topics. In a recent publication Jean Kellens divides current interpretations of the Gāthās into three types: “mythological,” “prophetic-pious” and “liturgical.” These are supposed to be fundamental conceptions of the type of discourse the Gāthās are.Footnote ¹ The awareness of the centrality of ritual in Zoroastrianism is of course historically justified. Religious practice in the ancient world comprised two areas of engagement: performance of ceremonies and rites, more or less elaborate, more or less solemn, by variously defined groups, and quotidian (private) observances, which can also be described as ritual inasmuch as they adhered to defined rules. Furthermore, Kellens's basic types converge in their conceptions of (Zoroastrian) ritual: it reproduces or maintains or helps maintain the cosmic order (in imitation of the creative act).Footnote ² Once ritual is given this role, one has to explain how it actually achieves the ascribed purpose.

We may, for example, restrict the meaning of ritual to “sacrifice,” understood as communication or exchange with the gods in words and material offerings.Footnote ³ It is possible to interpret, as Almut Hintze seems to do, the YH sacrifice in this way.Footnote ⁴ The worshippers adopt and “perform” the creative god's intention, words and actions and thereby refresh the divine order: Y 35.3 “O Wise Lord, because of beauteous truth [or ‘order'Footnote ⁵] we have certainly chosen this: that we may think, speak and perform those existing actions which may be best for both existences.”Footnote ⁶ The instrumental phrase Y 35.3 aā srīrā is thus understood as stating the reason for the actions undertaken. What are these actions? They are said to be replications of the ones accomplished by the god:

The efficacy of the sacrifice in reinvigorating the divine order (and not just the god) is not stated as such, but it may be implied: Y 35.3 aā srīrā could be understood in the sense of “because of the order established by the god” and, further, interpreted as “in order to maintain the cosmic order.” In any case, if there is a place for such an interpretation, the restorative efficacy of the worshippers’ words and actions must be grounded in their re-enactment of the god's actions. The consecration (yazamaidē) of mental and corporeal beings presumably reproduces the original creative act.Footnote ⁸ Therein lies its power. Although there is no mention of deleterious forces in the YH, perhaps even here sacrifice is inscribed in the antagonistic frame: dark forces are in the background. The phrase “because of the divinely established order” would imply upholding that order against the forces of destruction and deception. We could thus explain the sense in which the consecration of corporeal and mental creations is understood to help maintain “order,” whether directly or by way of consolidating the supreme god's power.

The idea that sacrifice has the role of preserving the cosmic order is very rarely attested in the ancient world. Aside from the Vedic Puruṣa (RV 10.90), I am aware of a few instances (in ancient Egypt) where a ritual seems to have had such a purpose.Footnote ⁹ As J. Z. Smith convincingly argues, the Babylonian New Year festival should not be understood in this sense. In the Akkadian Epic of Creation, the ascending god Marduk destroys the aboriginal water goddess Tiamat in battle and creates the world from her body.Footnote ¹⁰ This is obviously not a creation through sacrifice. The Brāhmaṇic sacrifice is the “path” of immortalization, “discovered” by the (immortal) gods and “revealed” to humans thanks to Yama and the “inspired poets.”Footnote ¹¹ All in all, the idea is absent in ancient Greece and Rome,Footnote ¹² Mesopotamia and Western Semitic religions.Footnote ¹³ This consideration alone obliges students of Zoroastrianism to be wary of ascribing such a function to sacrifice. However, the idea has been put forward time and again by Avesta scholars, without really explaining in what sense it can be understood. One can appreciate the attraction of the idea. How can we account for the significance of ritual in Zoroastrianism? The application of Eliade's general theory of sacrifice as the renewal of existence through the “dialectic of the sacred” provides a ready and seemingly suitable answer, especially in view of the pivotal role of the (confusing) notion of “order” (aa) in the Avesta.Footnote ¹⁴ The notion seems to connect ritual and cosmic planes. In Skjærvø’s words, “the ritual Order matches the divine Order.”Footnote ¹⁵ The order of sacrifice reproduces the cosmic order. It seems to provide the evidence for the interpretive scheme, so clear that requires no further elaboration or scrutiny. Let me give a few examples.

The idea of sacrifice as the medium of the renovation of cosmic order in general is already found in the great work of Marijan Molé, Culte, mythe et cosmologie dans l'Iran ancien (1963). Based on a number of Pahlavi texts, he applies Eliade's theory to the Gāthās.

Molé is the first scholar who explicitly formulates the idea in relation to the Gāthās and tries to demonstrate its applicability. Following Dumézil's interpretation of the “system” of the Gāthic “entities” as the Iranian adaptation of trifunctional system of divinities, Molé views their presence in the ritual as the reflection of the cosmic order and the intention to represent the latter. “Each of the Entities presents a divine aspect, a human aspect and a symbolic aspect. The sacrifice represents the world; it is clear that all the functions must be represented.”Footnote ¹⁷ I have tried to show elsewhere the problems with the application of the Dumézilian scheme to the Avestan pantheon.Footnote ¹⁸ Many students of ancient Iran, such as Kellens, Cantera, Herrenschmidt and Skjærvø, have taken up Molé’s conception.Footnote ¹⁹ It seems to give a fitting answer to the question of the significance of ritual in Zoroastrianism. The “result of the yasna ritual” for Skjærvø, who according to Kellens's classification advocates a “mythological” interpretation of the Gāthās, is “the re-ordering and rebirth of the cosmos.”Footnote ²⁰

There are two basic questions here. The first one is: in what sense does the ritual order “match” the divine order? We saw that Molé uses Dumézil's theory to argue that the ritual “represents” the divine order in its fundamental constituents. For Mary Boyce the basic cosmic elements (sky, earth, water, plant, animal, human and fire) are (symbolically) represented in the ritual.Footnote ²¹ But even if granted, how does their ritual manipulation and arrangement “reproduce” the god's act of creation and instituted order? The confusion between the two senses of the word “order,” namely the orderliness of ritual activity and a specific concrete arrangement, underlies Herrenschmidt's formulation of the idea. “In Mazdean Iran, ritual functions as the repetition of the god's cosmogonic activity and reproduces the exact order of divine creation. It is characterized by exact order in its language, gestures, and speech.”Footnote ²² Ritual “reproduces the exact order of divine creation”—how? Is this a claim made by the Avestan priest for the ritual he performs, or a theoretical assumption of the scholar? What (textual) evidence are we given in support of this theory?

The same semantic confusion is found in Kellens.

The “symbolic reproduction” of the god's creative act is achieved by the perfect execution of the ritual. The adjective “symbolic” releases the interpreter of the burden of describing the mechanism by which the precision of ritual procedures produces the effect credited to it. What is done in ritual must be (i.e. by definition) the reproduction of the cosmogonic act, and if it is executed perfectly, it brings to bear the power of the divine institution.

The second question is: on what does the power of the ritual bear? What is it that the ritual is thought to preserve or reproduce, presumed to be the keystone of the cosmic order? According to Skjærvø, it is the daily rise of the sun: sacrifice is the “means of making Order and the sun reappear.”Footnote ²⁴ He thus gives sacrifice a cosmological function. The Gāthās and the YH, Skjærvø maintains, “are ritual texts that must have accompanied sacrifices (rituals) whose purpose was to renew the existence of either the new day or the new year with their light, warmth, and fecundity after periods or darkness and infertility.”Footnote ²⁵ The fixed coincidence between the New Year and the vernal equinox in the official calendar is a post-Sasanian development.Footnote ²⁶ More importantly, if, as Skjærvø maintains, ritual intervention is required to combat the deathly forces of cold and darkness and ensure the return of warmth and light, would it not make more sense to place the intervention in the period when darkness and cold are rampant (i.e. winter solstice), as Cantera mentions,Footnote ²⁷ rather than the vernal equinox? As for the new day—that is, the rise of the sun—Skjærvø makes it the raison d’être of the daily yasna ceremony: it “serves the purpose of renewing the daily existence, that is, the removal of darkness and return of the sun.”Footnote ²⁸ He thus places the entire rite before the sunrise, and interprets the Yasna texts accordingly.Footnote ²⁹ One obvious problem with this interpretation is that the yasna service in fact starts after dawn.Footnote ³⁰ Has there been a displacement? How does Skjærvø explain it? Would it not mean the nullification of the raison d’être of the service? But Skjærvø goes beyond the ascription of this cosmological function to the yasna ceremony.

The insertion of the sacrifice within an antagonistic frame as a general proposition is plausible enough for Zoroastrianism. The struggle against the destructive forces is pervasive, to which ritual in particular must be directed. But the idea that Ahura Mazdā established the cosmos by way of a sacrifice is not found in the Avesta, nor is the idea that humans reproduce the god's primeval sacrifice in order to maintain the ordered cosmos. In any case, it seems that in Skjærvø’s view the actual substance of these ideas come down to what he says about the purpose of the yasna service, namely the daily return of the sun.

Kellens has recently attempted to read the Gāthās as a text describing and accompanying a ritual course.Footnote ³² In his 2011/12 Collège de France lectures Kellens suggests that each of the Gāthās was recited during one of the five ratus of the day. The recitation of the second Gāthā took place at night, for instance, while that of the first was just before the sunrise (3 February 2012).Footnote ³³ Gāthā Ahunauuaitī describes and accompanies a ritual course that ends in a pre-dawn sacrifice.

Kellens attempts to locate in the text two sacrificial events: the libation of haoma in Y 33, in the supposed analogy with the Hōm Stōm, and the immolation and offering of an animal victim in Y 34.

In fact, according to Kellens, in each of the multihāitic Gāthās, the “sacrificial moment” appears to be “an obligatory phase of the liturgical process of which the Gāthā is the recitative.”Footnote ³⁶

According to Kellens, the purpose (“la finalité”) of the supposed GA sacrifice is eschatological “The sacrificial course of GA expresses the aspiration to the permanence of a perfect union of two perfect states.”Footnote ³⁷ It is not clear, however, whether the aspiration is pursued on behalf of the individual or the world, as a “great future event.”Footnote ³⁸ In any case, Kellens affirms “the eschatological function of sacrifice” in the Gāthās.Footnote ³⁹ But he also gives a cosmological function to the sacrifice, without really questioning how the two functions may be related.Footnote ⁴⁰ The libation of the haoma enables the earth (which Kellens sees behind ārmaiti, e.g. in Y 33.13) “to give momentum to the dawn.”Footnote ⁴¹

Kellens proposes to locate in Gāthā Ahunauuaitī the key constituents of his thesis that the rite underway serves the cosmological purpose of ensuring the return of the sun and making the world “prosperous” (i.e. the haoma rite in Y 33.5–14), and the eschatological purpose of winning the mižda- of a perfect life (“the permanence of a perfect union of two perfect states”) by way of animal sacrifice (Y 34). As I mentioned, Kellens hesitates as to whether “the aspiration to perfection and permanence is fragmented as assurance for each individual soul or fixed in the expectation of a great future event.”Footnote ⁴² But it seems to me that the description he gives of the desired outcome (permanent and perfect union between mental and material states) commits him to historical eschatology. Let us now examine how Kellens argues the two aspects of his thesis.

Can we find a haoma rite in Y 33? Kellens points out that in the Yasna ceremony “the third pressing of haoma begins with the recitation of the last verse of Y 33.4.”Footnote ⁴³ He asserts that the composer of Y 27 “savait que 33.5–14 correspondait à la libation de haoma … Pour l'auteur du Y27, c'est clairement l'appel à entendre le bruit des pressoirs qui structure l'offrande haomique du Y33.” His attribution of the knowledge of the content of Y 33 (i.e. that Y 33.5–14 is about haoma preparation and libation, whose recitation accompanies the process in the “primitive” GA rite) to the author(s) of Y 27 is based on the citation of Y 33.11–14 in a context that invokes, among other ritual texts and acts, the pounding of haoma twigs. In Y 27.6 Sraoša (clearly allegorized) is present at the crushing. I am not sure how Kellens can conclude that “for the author of Y 27 it is clearly the invitation to hear the sound of the crushing that structures Y 33,” or even what this affirmation could mean.Footnote ⁴⁴ But let us not worry about this, and look at Kellens's analysis of Y 33 and his reasons for seeing in Y 33 a haoma rite.

Kellens compares the ritual components of the Hōm Stōm and Y 33. They have “three stages in common”: “the preliminary purge of social circles… the growth of the body of the sacrificer subsequent to the ingestion of the juice… and the offering of the uštana of this transcended body.”Footnote ⁴⁵ For the “stage” of “social purge” he invokes Y 33.4 and Y 9.28–29. The comparison, however, is not probative, since, just like the very similar Y 31.18 (and, more loosely, other passages where social groupings and customs are subjected to religiously coded dualistic antagonism), Y 33.4 is embedded in a discourse on the requirements of the partisanship of the religious view urged by the poet. In fact, Y 33.4 is the pivot that connects the social context of the partisanship with the poet-priest's conception of his role and activity. Expression of social antagonism in religious texts is common, although it is true that how the expression is formulated is likely to be significant. In Y 9.28–29 the god Haoma is said to be the master of the four social circles, thus articulating his universal domain, and in that capacity he is asked to protect the circles from those in their midst who want to harm them. It is interesting to note that the punitive measures the god is urged to take all have to do with the faculties of the body. Y 33.4 must be understood against the preceding stanza. The person who actively and thoroughly promotes the poet's cause (y aāunē vahištō, x^vaētū vā a vā vərəzniiō / airiiamnā vā … ) “will be in the pasture of aa- and good thinking” (Y 33.3). In my mind, the image is eschatological. In any case, against this demand for a thorough penetration, articulated in terms of social circles, the poet gives his own activity the same reach as a ritual officiant in Y 33.4. Thus, the articulation of social circles in Y 33.4 has nothing to do with a “structural stage” of the “haoma-offering” rite. Nor is it at all clear that the articulation in Y 9.28–29 is meant as a stage (“social purge”) in the ritual. Here, too, the passage must be read against Y 9.27 where Haoma's universal domain is spelled out in terms of the Avestan four social circles.

Kellens's second point of comparison, the “growth of the body of the sacrificer,” is also questionable. In Y 10.14 and 11.10, the sacrificer “give[s] over this body” (paiti … daδąmi imąm tanūm) to the god Haoma “who conducts to aa” (aauuāzō). The eschatological horizon of the operation is evident in Y 11.10. In Y 33.10cc’ the poet seems to say something like: I thrive in body (or in person) thanks to good thinking, power and aa: vohū uxšiiā manaŋhā, xšaθrā aācā uštā tanūm. Neither the (literal) meaning of “growing in body” (does it mean “thriving in person”?) nor the sense of the instrumentals is clear. Kellens's translation makes the eponymous gods the agents of the bodily growth: “(Me,) the good Thought, Power and Arrangement [i.e., Aa] make me grow with my whole body.”Footnote ⁴⁶ Perhaps. Two fundamental facts, however, undermine the assimilation of the two contexts (Y 33.10 and Y 11.10). First, the body in Y 33.10, whatever “growing” may signify, is not alienated, whereas in Y 11.10 it is, apparently experienced in an ecstatic state (maδa- “ecstasy”). When the underlying act and the resulting status of the body, in respect of which the comparison is made, are different in the two contexts, what is the basis of the assimilation? Second, in Y 33.10 Haoma is not found at all among the agent gods, whereas in Y 11.10 the god is the sole promoter of the ecstatic experience. There is nothing in Y 33.10 that suggests that “growing in body” is due to the consumption of the sacred drink. Kellens finds in the previous stanza, Y 33.9, an allusion to the clank of mortar and pestle, to which, according to him, the appearance of the morning light is attributed. The material world appears (in the emerging light) during the recitation of Y 33.9–11. Here is his translation of Y 33.9: “Let the alternate movement bring with good Thought, to you, O Mazdā, the passion of the two (pressing stones) which resound with increasing Arrangement (and), to me, wellbeing, (for) the accordance of these two (with the passion) to which (our) souls are attached has been set in motion.”Footnote ⁴⁷ The adjective sarəidiia—is used in the dual apposite to aaoxšaiiaṇt- “makes aa grow.” Kellens maintains that the adjectives refer to mortar and pestle. The probable Vedic cognate śárdhya—means something like defiant or challenging, from √śardh “show one’s strength, withstand, challenge.”Footnote ⁴⁸ If the Avestan term should be given the same semantic range, “which resound” would hardly be appropriate. If the context demands, one may interpret a ringing sound as bravado, but describing a challenging speech or posture as a ringing sound loses important semantic content. The obscure noun maēθā- if related to √miθ “alternate” could perhaps mean “alternation”—e.g. between two states—but hardly “alternate movement” as Kellens has it. Here, too, the clank of mortar and pestle pounding of the haoma twigs is already assumed.Footnote ⁴⁹ One has to admit that Kellens's translation of the stanza is bedeviled with expressions that are hard to understand: “carry the passion of mortar and pestle to the gods” (why translate mainiiu- as “passion” here and not “avis”?Footnote ⁵⁰) or “the setting in motion of the accord of the two” (one would rather expect verbs like √dā or √kar).

The third point of comparison also collapses under scrutiny: the offering of “the uštana of this transcended body (Y 33.14, Y 11.18) directly modeled on the precedent” The idea of a “transcended body” is not found in Y 33.14, where the poet seems to be making his most valuable possessions a gift to the gods: tanuuascī x^vaii uštanəm “the life (or mobility) of his very own body.” I cannot see how this could imply a “transcended body” in the relevant sense—that is to say, left behind in ecstasy, be it feigned, which the ingestion of haoma is supposed to have induced. Moreover, the syntax of the stanza is not as straightforward as it may first appear. Kellens's translation is: “Here Zarathuštra, as a prime (or first) gift, gives to Mazdā and Arrangement the mobility of his own body and of his good Thought, the sonority and power of act and speech.”Footnote ⁵¹ rātąm … pauruuatātəm cannot mean “first gift,” but, if read together, something like “bestowed excellence” or at a pinch “bestowed best (part).” On what does the genitive manaŋhascā vaŋhuš depend, and with what is it coordinated? It is possible that uštana does not have the specific meaning it has in the Young Avestan texts, but means “mobility” or “vigor” in general, or something similar. If so, “Zarathuštra gives to Mazdā and aa as bestowed excellence the vigor of his very own body and of (his) good thinking, and the sraoša and the power of (his) action and (his) speech.” At issue are the poet's most valued faculties and abilities being made a gift to Ahura Mazdā and aa, and not the expression of an ecstatic experience. In effect, the poet does not alienate, for example, the vigor of his body, but places it in the service of the gods. This gift-making is at the same time a commitment of oneself to the gods’ protection, seemingly understood in an eschatological horizon: in Y 33.13cc’ the poet asks Ārmaiti to guide the vision-souls (daēnā) because of aa (frō spəṇtā ārmaitē, aā daēn fradaxšaiiā), the instrumental probably expressing the reason for the requested guidance.Footnote ⁵²

In Y 32.14 the poet attacks the haoma rite.Footnote ⁵³ Thus, an interpretation of Y 33 as the accompanying text of such a rite would need to explain the resultant discrepancy. This is how Kellens does it.

The problem with this explanation is that it leaves one wondering: why name the rite only when it is being condemned? How to understand the strange taciturnity implied in Kellens's interpretation? There is not a single term or image in Y 33 (and the Gāthās) that so much as suggests that the rite was positively valued and practiced. It is true, I think, that the daēva cult (and the embedded hoama rite) was nocturnal. But if the haoma rite survived the virulent condemnation of the cult in the Gāthās,—then how to understand the ban on (the positive mention of) the name of a rite that is supposed by Kellens to have been so central a figure?

Kellens interprets Y 33, as we saw, as describing a dawn haoma rite coinciding with “or” promoting the sunrise. Y 33.12 and 13, “grafted on the hemistich ādāi kahiiācī paitī of haoma libation,” “almost explicitly evoke the propagation of the daylight.”Footnote ⁵⁵ There is no necessity or reason to interpret Y 33.11 ādā- “offering” as the libation of haoma. But are the two stanzas about the sunrise nonetheless? “The sun is present from the first word of Y 34.12, uz + ar, which is precisely the verb of its rise, to the second word of 13, vouru.cašāni-… and the catalog of the mobilized forces, occupying the interval, seems to express the irresistible increase in its power.” The verb uz + √ar “arise or rise” is used of the sun in the Vidēvdād, of Tištar in his Yašt, of the dead in FrW, of the poet's vision or soul in Y 43.Footnote ⁵⁶ It is only in the Vidēvdād that the verb is used of the sun; describing it as the “very verb of the rise” of the sun is somewhat misleading. Kellens translates Y 33.12a us mōi uzārəšuuā ahurā as “Rise before me, O Ahura!” and maintains that the Ahura here is the sun. This is very unlikely. Firstly, the dative personal pronoun when used with a verb of motion does not mean “to me” or “before me,” but “for me”—i.e. for my sake or on my behalf. When the former is meant, the accusative (of relation) is used, e.g. with √gam (cf. 28.3c’ and 29.3c’).Footnote ⁵⁷ Y 33.12a us mōi uzārəšuuā ahurā means “arise for me, O Ahura!” To break the two vocatives in Y 33.12, as Kellens does, in order to have Y 33.12a ahurā refer to the sun and to have the listed “forces” contribute to its “irresistible rise,” makes Y 33.12b mazdā completely adventitious. To put it in another way: what is addressed to the sun, and what to Mazdā in the stanza? Kellens's apparent distribution is rather arbitrary: “Rise before me, O Ahura, from Proper-thought receive robustness, from the good presentation-of-offering (receive) vivacity, from Arrangement (receive) the aggressive dominant-force, from the good Thought (receive) fsəratū, (receive all that), O Mazdā, from the very prosperous passion.” As for the epithet Y 33.13a vouru.cašāni- “with vast vision,” it, too, has to refer to Mazdā. The demand for “help” (rafəδra- or rafənah-) is always addressed to Mazdā in the Gāthās, except in Y 54.1 where it seems to be addressed to the “tribe.”

Kellens's identification of the daēnā with the dawn and of ārmaiti with the earth in Y 33.13 is without any Gāthic basis—and the first even without any Avestan basis. Moreover, the passage does not allow it. “Under the appellation that refers to her immaterial part, it is to be understood that the earth is going to project onto the sky the auroral lights which burst from her in the east.”Footnote ⁵⁸ The idea that it is the earth that projects the light of the rising sun onto the sky is quite implausible and, in any case, unsubstantiated. The whole amalgamation is based on a thesis that nowhere finds support in the text. Rather, in Y 33.13cc’ we find the daēnā and ārmaiti in their typical psychopompic roles: “O, vitalizing Ārmaiti, guide the vision-souls, because of aa!”Footnote ⁵⁹ In the final stanza, according to Kellens, Zarathuštra proceeds to “the symbolic self-immolation of his body, which the dawn has made appear and which Haoma has transcended.”Footnote ⁶⁰ Now, all these three statements are problematic. The first one, I have already shown.Footnote ⁶¹ Second, one is hard pressed to know what is being conveyed in the idea that one's body appears thanks to the daybreak, beyond its being an instance of the general thesis that in the final three stanzas of Y 33 we have the “appearance and abandonment of the material world.”Footnote ⁶² Kellens fails to provide any support for his thesis. Third, the assimilation of the situation in Y 33.14 to the ecstatic abandonment of the body in the Hōm Stōm is rather questionable, as we have seen.

Kellens claims that being spəṇta “prosperous” (“faste”) is an effect of the libation of haoma. No sooner does the material world appear and is made “prosperous” by the rite than it is abandoned.

I argued above that a Gāthic haoma rite (in Y 33) is an inarguable thesis, or, at best, an unproven conjecture. But even if one were to grant the thesis for a moment, the account of its supposed effect still counters further textual and logical problems. The statistics of the occurrence of the term spəṇta (four prior to the “sacrificial phase,” Y 33.5-34.10, as opposed to five during, according to KellensFootnote ⁶⁴) cannot, on its own, be an argument for the “spəṇta effect” of the haoma libation. Why does not Ārmaiti carry the epithet in Y 33.11 but she does in Y 33.13? Ārmaiti is by definition spəṇta. There can be no doubt about this (cf. Y 32.2). Why does mainiiu- not carry the epithet in Y 33.9 (a perfect place to signal the “spəṇta effect” according to Kellens's account of the stanza), but does in Y 33.12? This sort of mechanical interpretation of the presence or absence of a term is, in my mind, misleading. Having become spəṇta, men's “religious passion” invigorates the “gods of the diurnal world”—but why the “gods”? Even if one accepts Kellens's analysis and interpretation of Y 33.12, which is rather tendentious, to say the least, still it is only the sun that is invigorated there. Moreover, this idea (that the mainiiu-become-spəṇta is the means of invigoration of the gods) is in contradiction with the regular Gāthic conception that it is the power of good thinking that invigorates Mazdā or the gods (cf. 27.13).

According to Kellens's account, the man performing the GA rite symbolically immolates himself (in Y 33.14) and becomes spəṇta (in Y 34.2), thanks to the libation (or ingestion?) of haoma, and qualified to immolate the sacrificial animal (in Y 34.3). What does this achieve? What is the purpose of the animal sacrifice? Invigorating the gods? The “religious passion” of men having become “prosperous” already does this. Is it symbolic of the sacrificial death of the priest? But the “symbolic self-immolation” is already done. Is it supposed to accompany the vision-soul of the sacrificer?Footnote ⁶⁵ But Kellens has deprived himself of this basis by turning ārmaiti- into the earth and the daēnā into the dawn (in Y 33.13).Footnote ⁶⁶

The eschatological horizon of Y 34 is unmistakable, as Kellens recognizes.Footnote ⁶⁷ The poet's power of insight into the invisible is brought to bear on the fate of the soul after death (Y 34.9–10). The path along which the extraordinary vision is deployed is the path that the soul will eventually take (Y 34.12–13). It traverses the space between the earth and heaven: e.g., Y 43.16d–e’ x^vṇg.darəsōi, xšaθrōi iiā ārmaitiš / aīm iiaoθanāiš, vohū daidī manaŋhā “May Ārmaiti be in the sun-exposed domain (of Mazdā)! May she grant by way of actions inspired by good thinking the arrival (at Mazdā’s domain)!Footnote ⁶⁸ The “path of aa” (Y 51.13 aahiiā paθ-) is the way to the divine sphere, the “well-made road of good thinking along which the daēnās of the saošiiaṇts travel” (Y 34.13). The god is asked to “instruct” the mortals about the “easily traversed paths of good thinking because of aa” (Y 34.12). What place does animal sacrifice have in this economy? This is the question. Does the “soul of the cow” play the eschatological role of the “double” (or “ambassador”) for the sacrificer? This is quite plausible as a general proposition.Footnote ⁶⁹ But in the absence of textual evidence it cannot serve as a basis for interpreting, for example, Gāthā Ahunauuaitī as a “ritual course” that must comprise a “phase” of animal sacrifice. As we saw, Kellens attempts to locate such a sacrifice in the opening stanzas of Y 34: “in the primitive rite of GA, the stanzas Y 34.1-5 corresponded to the phase of meat offering.”Footnote ⁷⁰ Having become “prosperous” (spəṇta) and thus qualified for the task, thanks to the preceding haoma rite, the priest proceeds to sacrifice the animal. Y 34.3 is the pivot of his thesis, where animal offering (miiazda-) is explicitly mentioned.

Kellens translates the phrase related to miiazda-: “We place with veneration as the solid-offering [comme offrande-solide]… all the living beings.”Footnote ⁷¹ What is being offered to the gods? Living creatures. In what sense are they being given to the gods? Sacrificial offering. In other words, “living creatures” are being consecrated. What does this sacrifice achieve? The creatures are thereby committed to the care of the gods (placed in their power), so that they prosper.Footnote ⁷² The syntax forces us to understand miiazda- as expressing the sense in which the poet intends or understands the relation between the creatures and the gods. The poet may or may not have had a specific ritual in mind when he composed the stanza. But it seems impossible to give miiazda- a syntactic assignment other than the one stated. The living creatures are being consecrated. Thus, from the phrase one cannot conclude that the ceremony (“the primitive rite of GA”) involved the immolation and offering of an animal.

The last verse line of the stanza seems to state the reason (zī) for the setting in motion of the “vitalization of the beneficent.” We may interpret the genitive hudŋhō … sauuō as subjective. The last verse line of the stanza then means something like: the vitalization of the beneficent one is set in motion in the divine sphere because of living creatures. The term hudāh- “beneficent” seems to designate the sponsor of the approved ceremony (at least here and in Y 34.13). In the Young Avesta it is regularly used of the “cow” or the “soul of the cow,” attached to it perhaps by metonymy. Since sauuah- seems to have an active sense in the Gāthās, it must bear on “living creatures”: the “vitalization” that the sponsor of the approved sacrifice sets in motion in the gods’ domain is because of or due to the consecrated living creatures. But even if such a reading is acceptable, the statement about (one of) the reason(s) of sacrifice does not necessarily imply that an animal sacrifice is underway. To take this step, one must have first decided that the text describes a ritual course. Further, if the reading suggested is plausible—that is to say, the reason for the sacrifice is the “vitalization” of the “living creatures”—should one understand this vitalization in an eschatological sense or in relation to their earthly living conditions? If the former, it would have to be a universal eschatology. The problem with the analysis and interpretation of Y 34.3cc’ given above is that, while the relation of animal sacrifice with individual eschatology is more or less describable and understandable, at least in general terms, its relation with universal eschatology is not at all clear.

It is of course possible to understand the genitive hudā˚ŋhō … sauuō in the objective sense, as Kellens seems to do: “It is indeed by them all, O Mazdā, that the swelling [i.e., invigoration] of the generous has been set in motion among you”Footnote ⁷³ The consecrated living creatures would then be the means or medium of vitalizing the beneficent one in the divine sphere. One may interpret this statement within the frame of individual eschatology (cf. Y 30.11). In Y 34.13, the hudāh- can look forward to the mīžda- “reward” dispensed by the supreme god (cf. Y 51.15 where the same verb √ciš “allocate” is used). The “reward” or “prize of victory” is eschatological, obtained at the successful completion of the “path” pursued by the daēnā, guided by Ārmaiti (cf. Y 33.13).Footnote ⁷⁴ In Y 34.14, the gods are envisaged to grant the “desirable prize” to the mortal who will have earned it thanks to his action inspired by good thinking.

The phrase astuuat- uštāna- “bone-possessed vitality” must be understood as a collective noun, the recipient of the mīžda-. Similar semantics may be assumed for Y 34.15 ahu- “existence”: xšmāka xšaθrā ahurā, fərašə¯m vasnā haiθiiə¯m dā˚ ahūm “O Ahura, by your power make existence true and splendid, (endowed) with (heavenly) light!”Footnote ⁷⁶ If interpreted in the light of preceding stanzas, this statement should be placed within the frame of individual eschatology, and not that of universal eschatology. In the Gāthās, aa- or Mazdā’s domain, the ultimate object of longing for mortals, is described as bathed in the sunlight. It thus makes sense to interpret the image of existence in such a place as referring to a desirable postmortem condition.

For Kellens the idea that the Gāthās are composed in order to accompany (and partly describe) a definite ritual course is programmatic. “What is not negotiable for us is that the Gāthās are texts composed for the purpose of serving as recitative in a liturgical ceremony and not doctrinal texts which ended up serving as liturgical recitative.”Footnote ⁷⁷ This may well be true. But why proceed from the assumption that we already know what type of rites they served, or indeed that these rites were homogenous and performed daily ? The idea that Gāthic sacrifice had a cosmological function (i.e. the reproduction of cosmic order) is at best a theoretical conjecture, which, despite its popularity, remains unsubstantiated. Kellens's attempt at sketching out the structure and constituents of the sacrificial “phase” of Gāthā Ahunauuaitī is debatable. A number of his key analyses of the text must be questioned. The way Kellens defines the purpose of animal sacrifice in Y 34 (i.e. permanent union of material and mental states) commits him to the thesis of a universal eschatology in Gāthā Ahunauuaitī, joined with a thesis about the cosmological purpose of the haoma rite of Y 33. How are these two related? Kellens's conception of the threefold “aurore mazdéene,” even if accepted, cannot combine the two postulated rites into a coherent “ritual course,” for the concept of daēnā in the Gāthās belongs to the frame of individual eschatology.