¹ In the Older Avestan corpus scholars include the Ahuna Vairiia (Y 27.13) and the Airiiaman Išiia (Y 54.1), and some but not all also include the Aəm Vohū (Y 27.14) and the Yehē Hataam (Y 27.15). See H. Humbach, “Gathas I,” Encyclopædia Iranica 10: 321–27; and J. Kellens, “Avesta,” Encyclopædia Iranica, 3: 35–44. See also Hintze, A., “On the Literary Structure of the Older Avesta,” Bulletin of the School of Oriental and African Studies 65, no. 1 (2002): 31–51.CrossRefGoogle Scholar

² See Kellens, J., Chaire de langues et religions indo-iraniennes: Leçon inaugurale (Paris, 1994).CrossRefGoogle Scholar Kellens has of course modified his views somewhat, but the Gāthās are still for him liturgical compositions (not sermons), addressed to deities, etc. within the religious horizon of the Vedic hymns. Unlike the Veda, however, each Gāthā is a “compact text corresponding to a specific ritual” (Kellens, J., “Philology and the History of Religions in the Study of Mazdaism,” History of Religions 48, no. 4 (2009): 268CrossRefGoogle Scholar). See in particular Kellens, J., “Controverses actuelles sur la composition des Gâthâs,” Journal Asiatique 295, no. 2 (2007): 434–36.CrossRefGoogle Scholar

³ Henning, Compare B., Zoroaster, Politician or Witch-Doctor? (London, 1951), 14Google Scholar: “Inevitably, there is a large number of words in the Avesta whose meanings are unknown, and a further large number whose meanings are imperfectly known; and such unknown or imperfectly known words are particularly numerous in the Gâthâs. Then there are the words whose meaning is not in doubt; but even they, as all words, have a certain range of meaning, and from that range one can select an eccentric meaning. Now if one attributes an entirely arbitrary set of meanings to the unknown words, in such a way that this set of meanings is consistent within itself and conforms to a preconceived notion of the contents of the Gâthâs, and if one proceeds to select suitable extreme meanings for the known words, one can translate the Gâthâs (or for that matter any ancient text that carries a sufficient number of unknown words) in any way one likes.”

⁴ See, for example, Gershevitch, I., The Avestan Hymn to Mithra (Cambridge, 1959), 46Google Scholar; idem “Zoroaster's Own Contribution,” Journal of the Near Eastern Studies 23, no. 1 (1964): 13Google Scholar; idem “Approaches to Zoroaster's Gathas,” Iran 33 (1995): 6Google Scholar; Pettazzoni, R., Essays on the History of Religions (Leiden, 1945), 7Google Scholar; Gnoli, G., Zoroaster's Time and Homeland (Naples, 1980), 182–85, 210Google Scholar; idem, “Un monothéisme pré-zoroastrien?,” in Zarathushtra entre l'Inde et l'Iran (Wiesbaden, 2009), 100.Google Scholar

⁵ Gershevitch believes that Y 32.3 proves his monotheistic interpretation of the Gāthās—followed by other adherents of the thesis. Gershevitch's translation is more or less representative of this view: “But you gods all are a manifestation of evil thinking, and he who so-much worships you (is a manifestation) of falsehood and dissent” (Gershevitch, I., “Die Sonne Das Beste,” in Mithraic Studies (Manchester, 1975), 1: 79Google Scholar). Insler, S. (The Gāthās of Zarathustra (Leiden, 1975), 45)Google Scholar has: “But ye gods—as well as the one who worships you—all of you are the offspring stemming from evil thinking, deceit and disrespect.” Lommel, H. (Die Gathas des Zarathustra (Basel, 1971), 60)Google Scholar translates: “Aber ihr Götter alle seid Same (Abkömmlinge) aus schlechtem Denken, und wer euch hoch verehrt aus Lüge und Hochmut.” Gnoli, Compare G., Zoroaster in History (New York, 2000), 31–33Google Scholar: “An article by Gershevitch has been and still is an enlightening point of reference for me: since it was published, in 1975, I have gone back to it on several occasions … As far as Zoroaster was concerned, the meaning of daēva- remained the same: he condemned all of them … for the very reason they were the gods of polytheism … As Gershevitch rightly points out, they existed as ‘thoughts, conceived by erroneously thinking men.’ … Therefore the semantic development of Iranian daivas from ‘gods’ to ‘demons’ still holds good. It gets its origin from Zoroaster's condemnation of all the gods, manifestation of evil thinking (Y. 32, 3) or ‘Hirngespinste,’ as Gershevitch writes.” I argue in another article, “Old Persian duvītāparanam and the Gāthic daibitā(nā),” Studia Iranica (forthcoming), that this translation is in part based on the ad hoc assumption that in the stanza (and only here) the Gāthic adjective ciθra- means something like “manifestation” or “seed,” and on making the ablative akā manaŋhō serve a genitive function: “manifestation of evil thinking,” etc. Incidentally, my translation of Y 32.3 is: “But you all, the daēvas and the leader who worships you, clearly are (i.e., proceed) from bad thought, (and are involved) together in actions inspired by druj and negligence, for which you are notorious (even) in the seventh clime.”

⁶ Gershevitch, The Avestan Hymn to Mithra, 46.

⁷ Gershevitch, “Zoroaster's Own Contribution,” 13.

⁸ “‘Willing, if it is not to be a sort of wishing, must be the action itself. It cannot be allowed to stop anywhere short of the action.’ If it is the action, then it is so in the ordinary sense of the word; so it is speaking, writing, walking, lifting a thing, imagining something” (Wittgenstein, L., Philosophical Investigations (London, 2001), 135Google Scholar). One cannot simultaneously speak and not speak, write and not write, etc.

⁹ Gershevitch, “Approaches to Zoroaster's Gathas,” 6.

¹⁰ See Descartes, R., The Philosophical Writings of Descartes (Cambridge, 1985), 2: 93–120.CrossRefGoogle Scholar

¹¹ See Kant, I., Critique of Pure Reason (Indianapolis, 1996), 582–86Google Scholar, 626–27.

¹² In Augustinian theodicy “free will” is meant to exculpate the monotheistic god. The principle of free will is important for Augustine because it allows him to redirect the responsibility for evil away from God. Compare Nietzsche, F., The Anti-Christ, Ecce Homo, Twilight of the Idols, and Other Writings (New York, 2005), 181Google Scholar: “The notion of will was essentially designed with punishment in mind, which is to say the desire to assign guilt.” Free will is nothing if it is not self-grounding, if it is determined by external causes. The extent to which this principle is only a theodicy motif can be seen in, e.g., Augustine's argument in De libero arbitrio (Turnhout, 1970)Google Scholar that humans are not able to have a good volition in the absence of grace (De lib. arb. 3.18.51). Nonetheless, they are responsible for their sins, where they use the “good” created by God for perverse ends. The original accusation raised against God is thus reduced to the seemingly manageable question: why does God create a creature with free will? One would suppose the answer is “because it is good.” It is true that this is a tautological answer, since everything God creates is good. Still, it redirects the original accusation, be it at the expense of the perspicuity of the explanation: one exchanges a culpable God for a mysterious God, which expresses nothing more than the limits of human intellect. But Augustine feels the weight of the question: why should a free creature turn away from its Creator, whence the perversion? “But perhaps you are going to ask: since the will is moved when it turns away from an immutable good to a mutable good, from whence does this movement arise? It [the movement] is actually evil, even though a free will is to be counted among the good things, since without it no one can live rightly. For if that movement, that is, the will's turning away from the Lord God, is without doubt a sin, how can we say that God is the author of the sin? Thus the movement will not be from God. From whence then will it come? If I respond thus to your querying—that I do not know—perhaps you will be disappointed—but nevertheless I would respond truly. For that which is nothing cannot be known” (De lib. arb. 2.20.54). Augustine's casuistry aside, his disowning of the final question has its logic: the unmotivated turn away from God has to be either completely arbitrary or grounded in human nature. Augustine chooses the first, since the second would again point the finger of accusation at God. It is meant to justify the creator God. See Blumenberg, H., The Legitimacy of the Modern Age (Cambridge, MA, 1983), 125–36.Google Scholar The concerns and motives of such a theodicy, whether purely dualistic or “ethical,” or indeed a mixture of the two, are in no way native to the Gāthās. They are pressing only for a mind steeped in Judeo-Christian monotheism and exercised by its theological problems. The western scholar imports into the Gāthās the theodicy motif along with his assumption of Gāthic monotheism: an accused God must be as intolerable for the Iranian “prophet” as it is for the Christian theologian. Then, he will have to deal with the problem of reconciling his monotheism with his “ethical” dualism, which is intended to exculpate the one God in the face of evil.

¹³ The passage cited is my translation of their text in Kellens, J. and Pirart, É., Les textes vieil-avestiques, 3 vols. (Wiesbaden, 1988–91), 3: 43Google Scholar: “Les deux états d'esprit, initiaux parce qu'ils sont le fondement des trois niveaux de la conduite rituelle (pensée, parole, acte), sont réputés être des songes jumeaux au stade de la pensée … ils sont jumeaux parce qu'il est difficile de les distinguer l'un de l'autre. Pourtant, si insaisissables et indistincts qu'ils soient, au moment de l'acte rituel, l'un inspire le bon acte, l'autre le mauvais, et certaines divinités sont capables de les distinguer, d'autres non.”

¹⁴ Kellens and Pirart, Les textes vieil-avestiques, 3: 78.

¹⁵ Ibid., 3: 43.

¹⁶ The translations of the stanza to which I refer in the following are given in full in the Appendix. This article has benefited from the careful reading and suggestions of an anonymous reader.

¹⁷ Kellens and Pirart, Les textes vieil-avestiques, 1: 110.

¹⁸ Insler, The Gāthās of Zarathustra, 33.

¹⁹ Gnoli, Zoroaster's Time and Homeland, 207.

²⁰ Humbach, H., The Gāthās of Zarathushtra and the Other Old Avestan Texts (Heidelberg, 1991), 1: 123.Google Scholar

²¹ Since, generally speaking, for Humbach the gods are the interlocutor of the poet, are the “twins” presented to them?

²² Gershevitch (“Zoroaster's Own Contribution,” 32–33) thought the “locative” paouruiiē from the adjective paouruuia- cannot mean “in the beginning” but, used adverbially, has to be “at first, firstly.” Accordingly, he translated the first verse: “Firstly the twin Spirits [lit. the two Spirits who (are) twins] were revealed (to me), each-endowed-with-own-wish (= free will).” Does the adverb bear on the revelation: “the spirits were initially revealed”? What could this mean? Or is it an adverb of ordering topics of the discourse: “first of all, (let me say that) the twin spirits were revealed to me”? The same form (paouruiiē) of the adjective occurs in Y 45.2aa' a frauuaxšiiā ahuš mainiiū paouruiiē. Gershevitch (ibid.), again, makes it an adverb: “I shall mention firstly of the world the two Spirits.” But in neither case is a topic marked in whatever fashion as second or third in the discourse. The sole possible instance where such a particular sense (i.e. indicating the order of speech) is attested for an adverb from paouruiia- is Y 43.8. Generally speaking, adjectives are used adverbially in the accusative form. Moreover, ahu- does not seem to have a concrete sense (the “world”) in the Gāthās, but always an abstract meaning: “existence” or “state.” Finally, the object of the “proclamations” (a frauuaxšiiā) of the first six stanzas is each time marked by an adjective as something ultimate: 45.1 īm (ratūm) … ciθr; 45.2 ahuš mainiiū paouruiiē; 45.3 ahuš ahiiā paouruuīm … mθrəm; 45.4 ahuš ahiiā vahištəm; 45.5 vacā … vahištəm; 45.6 vīspanm mazištəm. There can hardly be any doubt that paouruiiē belongs with mainiiū. The adverbial interpretation was in any case abandoned by Gershevitch himself in a later article (“Approaches to Zoroaster's Gathas,” Iran, 33 (1995): 17), where he reads mainiiū paouruiiē as nom. acc. dual: “the two primordial thoughts.”

²³ Kellens and Pirart, Les textes vieil-avestiques, 1: 110.

²⁴ Kellens and Pirart, “La strophe des jumeaux,” 37–38.

²⁵ Ibid., 43–50.

²⁶ Ibid., 50.

²⁷ Ibid., 60–61.

²⁸ Humbach, The Gāthās of Zarathushtra, 1: 123.

²⁹ Lommel, Die Gathas des Zarathustra, 41. The nominative attributes (“better and bad”: neuter) in Lommel's scheme do not agree with the subject (“spirit”: masculine) in gender.

³⁰ Gnoli, Zoroaster's Time and Homeland, 207.

³¹ Gershevitch, “Approaches to Zoroaster's Gathas,” 17.

³² Insler, The Gāthās of Zarathustra, 33.

³³ Compare Beekes, R., A Grammar of Gatha-Avestan (Leiden, 1988), 139.Google Scholar

³⁴ Kellens and Pirart, “La strophe des jumeaux,” 50–51.

³⁵ See de Vaan, M., The Avestan Vowels (New York, 2003), 422–24CrossRefGoogle Scholar for the edition to paouruiiē (Yasna Pahlavi) as opposed to Humbach's and Kellens' pauruiiē (Yasna Sade). The adjective *paouruiia- regularly means “first, primitive, former.” Bartholomae, Ch. (Altiranisches Wörterbuch (Berlin, 1961), 874)CrossRefGoogle Scholar gives: “‘der erste, primus,’ nach Zeit und Ordnung,” what, then, comes before or first in a sequence. For its Vedic counterpart pūrvyá-, Mayrhofer (Etymologisches Wörterbuch des Altindoarischen (Heidelberg, 1992–2001), 2: 157) has: “vormalig, früher dagewesen, frühest, vorzüglichst,” and for the adverb pūrvyám he has: “zuerst, früher, zuvor.” Thus the Vedic adjective means “what precedes,” comparatively or absolutely, in a sequence, and from the latter sense it naturally develops the meaning “excellent, prime.” The Gāthic adjective apaouruiia- is “without precedent.” In none of its Gāthic occurrences does the adjective (or the adverb) have the sense of the hierarchical fundament. Used as an adverb paouruuīm means “primordially, originally, first, in the beginning.” I think the point is important and has to be stressed. Thus the adjective, contra Humbach, Kellens and Pirart, cannot mean “fundamental.”

³⁶ Hintze, A., A Zoroastrian Liturgy. The Worship in Seven Chapters (Yasna 35–41) (Wiesbaden, 2007), 112–13.Google Scholar

³⁷ Humbach, The Gāthās of Zarathushtra, 2: 48.

³⁸ See Hintze, A Zoroastrian Liturgy, 112–13; and Narten, J., Der Yasna Haptaŋhāiti (Wiesbaden, 1986), 139.Google Scholar

³⁹ See de Vaan, The Avestan Vowels, 423; and Kellens and Pirart, “La strophe des jumeaux,” 61.

⁴⁰ Hoffmann, K. and Forssman, B., Avestische Laut- und Flexionslehre (Innsbruck, 2004), 85.Google Scholar

⁴¹ The translation given in the text agrees with one of the two “variants” they consider “les seuls choix raisonnables” (Kellens and Pirart, “La strophe des jumeaux,” 63). In the alternative variant the sense of x ^vafənā is completely artificial: “les mainiiu (jumeaux) originels/fondamentaux qui sont connus/ont été connus pour leur sommeil (jumeau) au moment de penser et de dire” (ibid.). What information are we given by “the two spirits that are known for their sleep during thought and speech” or even “in respect of thought and speech”? That at the level (“stade”) of thought or of speech during (the) ritual the two spirits are “sleep” or “dreaming,” but then become active at the level of gesture? None of this makes much sense.

⁴² Insler, The Gāthās of Zarathustra, 165.

⁴³ Kellens and Pirart, Les textes vieil-avestiques, 2: 41.

⁴⁴ See Narten, J., “Avestisch frauuai-,” Indo-Iranian Journal 28 (1985): 35–48.Google Scholar

⁴⁵ I should perhaps say that the image of “twin spirits as dreams” would not be a strange notion in ancient thought, where supernatural beings are sometimes taken to be dreams, rather than, say, send dreams. See, for example, Detienne, M., La notion de daimōn dans le pythagorisme ancien (Paris, 1963), 43–46.Google Scholar I think, however, that in this stanza as elsewhere in the Gāthās the poet first draws attention to his own powers and status before speaking about the ultimate things. As I will argue further on in the text, dream is in ancient religious thought a privileged mode of access to the invisible.

⁴⁶ Insler, The Gāthās of Zarathustra, 165.

⁴⁷ Compare the passage from the Pythagorean Memoires (according to Alexander Polyhistor) preserved in Diogenes Laertius VIII, 32 (Laertius, Diogenes, Lives of Eminent Philosophers (Cambridge, MA, 2005), 2: 347–48Google Scholar): “The whole air is full of souls which are called genii or heroes; it is they who send men dreams and signs and illnesses, and not only to men, but also to sheep and cattle” (translation slightly altered). The idea is extremely archaic. The societies of the pastoralists included their domestic animals.

⁴⁸ As for RV X 36.4ab grāvā vadann ápa rákṣāṃsi sedhatu duṣvápnyaṃ nírRtiṃ víśvam atríṇam (let the ringing pressing stone repel fiends, bad dream, destruction and every atrin) neither the associated negative terms nor the remedy against them (the “ringing pressing-stone”) that the poet appeals to makes Insler's “ill rivalry” more attractive than “nightmare.” See also Kellens and Pirart, “La strophe des jumeaux,” 58.

⁴⁹ Gershevitch, “Approaches to Zoroaster's Gathas,” 17–18.

⁵⁰ For Bartholomae, see Kellens and Pirart, “La strophe des jumeaux,” 38–39.

⁵¹ R.C. Zaehner, Hindu Scriptures (New York, 1992), 82.

⁵² Detienne, M., The Masters of Truth in Archaic Greece (New York, 1999), 63.Google Scholar

⁵³ Detienne, The Masters of Truth, 45. “Like Mnēmosynē, Alētheia is the gift of second sight: an omniscience, like memory, encompassing the past, present, and future. The nocturnal visions of dream, called Alēthosynē, cover ‘the past, the present, and all that must be for many mortals, during their dark slumber.’ And the Alētheia of the Old Man of the Sea is knowledge ‘of all divine things, the present and the future.’ As a power of prophecy, Alētheia sometimes replaces Mnēmosynē in certain experiences of incubatory prophecy, as in the story of Epimenides. Thus magus spoke with Alētheia, accompanied by Dikē, during his years of retreat, in the cave of Zeus Diktaios” (Detienne, The Masters of Truth, 65). The cited passages are from Euripides, Iphigeneia in Tauris and Helena, respectively. See further ibid., 45: “In Hesiod's poem we find the most ancient representation of a poetic and religious Alētheia. What is the Muses' function according to the theology of speech deployed in the Theogony? The Muses proudly claim the privilege of “speaking the truth” (alētheia gērusasthai). The meaning of this Alētheia is revealed by its relation to the Muses and to memory, for the Muses are those who “tell of what is, and what is to be, and what was before now”; they are the words of memory. The very context of the Theogony thus already indicates a close connection between Alētheia and memory and even suggests that one should understand these two religious powers as a single representation.” See Vernant, J.-P., Myth and Thought among the Greeks (New York, 2006), 139–53Google Scholar, for the myths around the link between memory and death in ancient Greece. See also Gernet, L., The Anthropology of Ancient Greece (Baltimore, 1981), 220–26Google Scholar, on the notion of decisive proof. Under certain conditions, visions and words could immediately reveal the “real” (e.g. beyond time) origins of existence, and this revelation was always vouchsafed to specially privileged figures such as inspired poets. The belief in the mantic power of dreams, and in the equivalence of sleep and death with respect to occult knowledge, is also found among the ancients Celts and Germans. See Eliade, M., Shamanism (Princeton, 1964), 382–84.Google Scholar

⁵⁴ Detienne, The Masters of Truth, 42–43.

⁵⁵ Cited in Detienne, La notion de daimōn, 77.

⁵⁶ Detienne, The Masters of Truth, 84. Detienne cites this interesting tradition about Epimenides, the “magus”: “During the day Epimenides lay down in the cave of Zeus Diktaios and he slumbered in a deep sleep for many years; he conversed in his dreams with the gods and spoke with Alētheia and Dikē” (ibid., 123).

⁵⁷ See Malamoud, C., Cooking the World (Delhi, 1996), 195–206.Google Scholar Thompson's essay on the notion of satyakriyā “truth-act” and its relation to the figure of the inspired poet is very important. See Thompson, G., “On Truth-Acts in Vedic,” Indo-Iranian Journal 41 (1998): 125–53.CrossRefGoogle Scholar The utterance of efficacious speech is immediately its realization. Compare The Masters of Truth, 73: “The speech of the diviner and of oracular powers, like a poetic pronouncement, defines a particular level of reality: when Apollo prophecies, he ‘realizes’ (krainei). Oracular speech does not reflect an event that has already occurred; it is part of its realization … while the visions of dreams in which words were not realized (akraanta) were opposed to dreams that did ‘accomplish the truth’ or ‘realize reality’ (etuma krainousin).”

⁵⁸ Both Pahlavi and Sanskrit translations of the verse treat it as an independent statement. The Pahlavi translation with the gloss (Kellens and Pirart, “La strophe des jumeaux,” 34) reads: “menišn ud gōwišn ud kunišn ān ī har dō az weh ud az-iz wattar (ēk ān ī weh menīd ud guft ud kard ud ēk ān [ī] wattar)”: “thinking and speaking and doing, they that were two, one [picked?] from the good and the other from the bad (one was the one who thought and spoke and did what is good, and one was the one who [thought and spoke and did] what is bad).” Incidentally, the “good” and the “bad” in the Pahlavi translation seem to describe the members of the triad and not the spirits (mēnōg)—according to the gloss in any case. Neryosang's Sanskrit translation (ibid., 35) reads: “manasi ca vacasi ca karmaṇi ca tad dvitayam uttama ca nikṣṭaṁ ca”: “in thinking and speaking and doing, those two [spirits] are the best and the corrupted.”

⁵⁹ Cited in ibid., 38–39.

⁶⁰ Insler, The Gāthās of Zarathustra, 33.

⁶¹ Monna, M.C., The Gathas of Zarathustra (Amsterdam, 1978), 17.Google Scholar

⁶² Kellens and Pirart, “La strophe des jumeaux,” 47–48.

⁶³ Kellens and Pirart, Les textes vieil-avestiques, 1: 9.

⁶⁴ Klein, J., Toward a Discourse Grammar of the Rigveda. Vol. I: Coordinate Conjunction. Part 1: Introduction, ca, utá (Heidelberg, 1985).Google Scholar The page numbers are given in brackets in the text.

⁶⁵ See Klein, Toward a Discourse Grammar of the Rigveda, 151–53.

⁶⁶ Kellens and Pirart, “La strophe des jumeaux,” 48 n. 26.

⁶⁷ Compare ibid., 49: “Dans une construction sujet + attribut, le locatif ne peut être que strictement circonstanciel. Il faut comprendre que hī est bon ou mauvais au moment de penser, au moment de dire, au moment de faire.” I have already pointed out the difficulties this “rule” gives rise to for their interpretation of Y 30.3a–b'.

⁶⁸ Lommel, Die Gathas des Zarathustra, 41.

⁶⁹ Bartholomae, Altiranisches Wörterbuch, 441.

⁷⁰ See the Appendix.

⁷¹ Kellens and Pirart, Les textes vieil-avestiques, 3: 204.

⁷² See Ahmadi, A., “Religious Regulation of Hospitality in the Gāthās (Y 46.5 & 6),” Studia Iranica 41 (2012): 7–24Google Scholar for further discussion.

⁷³ Cited in Kellens and Pirart, “La strophe des jumeaux,” 38–39.

⁷⁴ Gnoli, Zoroaster's Time and Homeland, 207.

⁷⁵ Humbach, The Gāthās of Zarathushtra, 1: 123.

⁷⁶ Insler, The Gāthās of Zarathustra, 33.

⁷⁷ Kellens and Pirart, Les textes vieil-avestiques, 1: 110.

⁷⁸ Lommel, Die Gathas des Zarathustra, 41.