SOURCE: “Zoroaster's Own Contribution,” Journal of Near Eastern Studies, Vol. XXIII, No. 1, January, 1964, pp. 12-38.

[In the essay below, Gershevitch investigates the nature of the discrepancies between the doctrines Zoroaster puts forth in the Gathas and those beliefs attributed to him in the Later Avesta.]

I. THE THREE VERSIONS OF ZOROASTER'S DOCTRINE

Zoroaster's own verses, the Gathas, which may be dated to the first half of the sixth century b.c., form only a small part of the scripture that goes under the name of the Avesta. The difference between the doctrine which Zoroaster states or implies in the Gathas and the doctrine which is attributed to him in the remainder of the scripture has long attracted attention. To avoid confusion it is convenient to refer to the religion of the Gathas as “Zarathuštrianism” and to the doctrine of the Younger Avestan texts as “Zarathuštricism.” The term “Zoroastrianism” may then be reserved for the form which the doctrine takes in the much later, Sasanian, period.¹

In the Gathas Zoroaster reveals himself as a monotheist in that he worships one god only, Ahura Mazdāh. He is, however, also a dualist, because he assumes the existence of two aboriginal principles, Truth and Falsehood. The common denominator of these two apparently irreconcilable viewpoints is in Zoroaster's system Truth. For Truth is one of the organs, aspects, or emanations of Ahura Mazdāh through which the god acts and becomes accessible. Two religions, therefore, appear to have been syncretized by the prophet: a monotheism centered in a god of whom Truth is an emanation, and a dualism in which Truth is primordial.²

The prophet himself refers to God's aspects or emanations by expressions which we may translate as “Entities” (cf. below, p. 17), but in the post-Gathic parts of the Avesta they are called Amsša Spsntas, that is, “Holy Immortals.”³ One of these Immortals, in fact the chief one, since he is God's creative organ, is Spsnta Mainyu, the “Holy Spirit” (cf. AHM, pp. 11 f., 165 f.). This Spirit sometimes appears in the Gathas in the company of Angra Mainyu, the later Ahriman, or “Fiendish Spirit.” A profound disagreement separates the two, which arises from the different choice they have made. The Fiendish Spirit, says Zarathuštra, chose to do evil, thus siding with Falsehood, while the Holy Spirit chose Truth (Yasna 30.5). This choice is the prototype of the choice which faces each man as he decides between following the path of Truth or that of Falsehood.

While in one Gathic passage (Yasna 47.3) Zarathuštra refers to God as the “father” of the Holy Spirit, in another (Yasna 30.3) he describes the Holy and the Fiendish Spirits as “twins.”⁴ The conclusion that the Fiendish Spirit, too, was an emanation of Ahura Mazdāh's is unavoidable.⁵ But we need not go so far as to assume that Zarathuštra imagined the Devil as having directly issued from God. Rather, since free will, too, is a basic tenet of Zarathuštrianism, we may think of the “childbirth” implied in the idea of twinship as having consisted in the emanation by God of undifferentiated “spirit,” which only at the emergence of free will split into two “twin” Spirits of opposite allegiance. Truth alone being, in addition to a primordial principle, also an aspect of Ahura Mazdāh, the fact that the Fiendish Spirit had chosen Falsehood would all but obliterate his original connection with God. It would follow that the names “Holy” and “Fiendish” of the two Spirits did not in the prophet's view pre-exist the choice which they made, but accrued to either as a result of it.

The different election of the two Spirits contained the seed of a clash. But, as we gather from the much later, Zoroastrian, texts, conflict could not take place in a purely spiritual state. This is presumably why the Gathic myth reports no struggle, or even bitterness, between the two Spirits, but only hints at a struggle to come. To bring about conflict—a desirable aim, since only through conflict could either Spirit hope to prevail over the other—life, death, and after-life had to be created, mainly, we may take it, in regard to Man, who would serve as champion, battleground, and prize. This thought Zarathuštra expressed in the following words: “In order that they might meet (in battle), the two Spirits first created⁶ life and not-life (i.e. the Holy Spirit created life, the Fiendish not-life), and established⁶ how ultimately existence will be: (that) of the owners of Falsehood (will be) very bad, but the owner of Truth (will) have the best dwelling” (Yasna 30.4).⁷

Although in conformity with their common origin both Spirits are seen in Yasna 30.4 to have been “creators,”⁸ only the Holy Spirit acts as God's creative organ. In this capacity he is frequently mentioned in the Gathas, while the Fiendish Spirit, having through his election ceased to be an organ of God's, receives no further attention from the poet.

In one Gathic passage (Yasna 44.7) we are clearly told that it is through the Holy Spirit that God created everything. Nevertheless it is God, not the Holy Spirit, who dominates the Gathic scene. God did not have to choose between Truth and Falsehood, like the two Spirits, because Truth, inasmuch as it is not only an aboriginal principle but also one of God's aspects, is part of God and belongs to his definition. God thus stands outside the struggle which is waged between the forces of Truth and Falsehood. His only intervention consists in revealing to Zarathuštra the responsibility which rests on mankind: the support which each man lends to the side he has chosen will add permanent strength to it; in the long run, therefore, the acts of man will weigh the scales in favor of the one side or the other.⁹

It will thus be seen that Zarathuštra's dualistic attitude hinges exclusively on his postulate of the two principles, Truth and Falsehood. The two Spirits have no independent dualist status. They merely react to a pre-existing dualism.

So much for the Gathas. The composition of the bulk of the Younger Avestan texts appears to have begun more than a century after Zarathuštra's death, towards the last quarter of the fifth century b.c. (cf. below, p. 28). In the Younger Avestan texts the monotheism of the Gathas is strangely contaminated with what has been called a “pagan” polytheism. Beside Ahura Mazdāh numerous other gods are worshiped, Mithra, Anāhitā, Vsrsthraghna, Tištrya, Vayu, etc. There are passages in the hymns to these gods in which Ahura Mazdāh himself is represented as worshiping them. This polytheistic attitude is of course a travesty of Zarathuštra's intentions, even though the pious authors expressly ascribe it to him. In the Younger Avesta even statements which flatly contradict the prophet's own doctrine are blandly introduced with the prefatory remark: “Thus said Ahura Mazdāh to Zarathuštra,” as if the authors meant to forestall any objections to what they were about to state.

Where the dualist doctrine is concerned, however, the authors of the Younger Avesta conform fairly closely to Zarathuštra's teachings. They still distinguish between the followers of Truth and the followers of Falsehood and range the Fiendish Spirit as an Archdemon with the latter. The Holy Spirit, on the other hand, is not very prominent in the Younger Avesta. He now bears the epithet “creator” (Yašt 10.143), like Ahura Mazdāh himself, so that one might almost expect him to become absorbed and replaced by Ahura Mazdāh. Of such an absorption, however, there is no indication in the scripture, except perhaps in the last chapter of the Vendidad, a late Younger Avestan book which was composed in post-Achaemenian times. This chapter relates that the Fiendish Spirit created against Ahura Mazdāh 99,999 diseases, to rid himself of which the god engaged the services of a number of genii. With this one exception, whenever the opposite number of the Fiendish Spirit is mentioned in the Younger Avesta, he is, as in the Gathas, the Holy Spirit and not God himself.¹⁰

The situation is very different in the late Zoroastrian books of the ninth century a.d., which are written no longer in the Avestan language but in Middle Persian. These books reflect in the main the views of the Zoroastrian Church at the time when Zoroastrianism occupied the position of official state religion under the Sasanian kings of Persia (a.d. 226-642). At this stage the doctrine appears in a simplified form. The pagan polytheism of the Younger Avestan, Zarathuštric scripture, has been virtually abandoned: the once powerful rivals of Ahura Mazdāh, such as Mithra, Vsrsthraghna, Vayu, etc., have been reduced to little more than genii or angels surviving as Avestan reminiscences. In addition, the complications arising in the Gathas from a simultaneous belief in a monotheism centered in Ahura Mazdāh, on the one hand, and a dualism of two opposing principles, Truth and Falsehood, on the other, have been eliminated. In the place of Falsehood now stands the Fiendish Spirit, in the place of Truth, God himself. Zoroaster's religion has become an uncompromising dualism, in which two aboriginal deities, Ohrmazd and Ahriman, God and the Devil, face each other and contend for ultimate victory.

This simple formulation, so familiar to all, presents an interesting problem. Who was responsible for it, since neither the Zarathuštrian Gathas nor the Zarathuštric Younger Avesta can be regarded as its direct source?¹¹ The circumstances in which this un-Avestan doctrine came to be imputed to Zoroaster deserve to be investigated.

At this point we must remember that it is not among the Sasanian Zoroastrian priests that the formula Ohrmazd versus Ahriman makes its first appearance. From a fragment of the περi pιλοsοpίαs of Aristotle we learn that already in the fourth century b.c., while part of the Younger Avestan scripture was still being composed, the Magi preached the existence of two principles … referred to as Spirits …, one of whom, the good one, was called ’Ωρομάsδηs, the other, evil, ’Aρειμάνιοs.¹²

Aristotle and his circle were greatly interested in the teachings of the Magi. They would not have been slow to appreciate the significance of a dualism based on the opposition of Truth and Falsehood, as Zoroaster had preached it. If nevertheless the fourth-century philosophers thought that the essence of the Magian doctrine consisted in the opposition of Oromasdes and Areimanios, it is clear that the Magi professed a dualist doctrine which considerably differed from that of the Avesta.

Now, it is true that the περi pιλοsοpίαs attributes the formula Oromasdes versus Areimanios to the Magi, not to Zoroaster. Several considerations, however, suggest that the Magi were teaching this formula in the name of Zoroaster. That this was so at a later stage is known from Plutarch's attribution of the formula to “Zoroaster the Magus,”¹³ and from the fact that with the Sasanian Zoroastrians the formula passed as the essence of the prophet's doctrine. But even in the beginning of the fourth century b.c. or at the end of the fifth, the Magi could hardly have given out the doctrine implied in the formula except by referring it to Zoroaster. For this doctrine is quite unrelated to the information which Herodotus had collected on the Persian and Magian religion.¹⁴ It is unbelievable that, had the Magi identified themselves with this arresting doctrine by the middle of the fifth century, it would have failed to come to the notice of Herodotus. The absence of the formula from Herodotus' detailed account, and the presence of it alone in the περi pιλοsοpίαs fragment as a succinct description of Magian thought, necessitate the assumption that the Magi changed their outlook at some time within the second half of the fifth century. The most likely reason for such a change would be the adoption on the part of the Magi of Zoroaster as their prophet. This is only another way of saying that up to the middle, at least, of the fifth century the Magi were not Zoroastrians, but in the first half of the fourth century they were.

II. DARIUS AND XERXES

We are thus using a well-known fact, namely the appearance in the fourth century of the formula Oromasdes versus Areimanios, as a new argument in support of a conclusion which can also be reached on the strength of two other considerations. One is that Herodotus' description of the Persian and Magian religion lacks all reference not only to the formula but even to the Zarathuštrian premises out of which the formula appears to have been developed.¹⁵ One cannot therefore argue that the Magi were orthodox Zarathuštrians before they adopted the heterodox formula.

The other consideration is, that while in the Achaemenian inscriptions Darius I (522-486 b.c.) and Xerxes (486-465) name only Ahura Mazdāh as their god, Artaxerxes II (405-359) proclaims himself as worshiping Ahura Mazdāh, Mithra, and Anāhitā, and Artaxerxes III (359-338) invokes both Ahura Mazdāh and Mithra. It is generally thought that the religion of the Achaemenids from Artaxerxes II onwards must have been similar to the Zarathuštricism of the Younger Avesta, if not identical with it. At any rate, there can be little doubt that a relation existed between the change from the religious attitude of Darius and Xerxes to that of Artaxerxes II on the one hand, and the change from the monotheism of the Gathas to the polytheism of the Younger Avesta on the other. This relation, and the relevance it may have to the date of the conversion of the Magi, we must try to ascertain, but not without first attempting to define as closely as possible the religious attitude of Darius, Xerxes, and Artaxerxes I (465-425).

The inscriptions of Darius leave no doubt that his religion was a monotheism centered in Ahura Mazdāh. It is true that Darius occasionally refers to Ahura Mazdāh as “the greatest of the gods” and frequently invokes Ahura Mazdāh with the gods,” which formula is replaced in one inscription (Persepolis, d) by “Ahura Mazdāh with all the gods.” But such collective reference to “gods” is generally taken to reflect not a polytheistic attitude on the king's part but his tolerance of the numerous gods other than his own who were worshiped within his vast empire.

I believe we can be a little more precise. There is evidence which can be interpreted to the effect that Darius was tolerant enough towards the foreign, non-Iranian gods, but not altogether towards Ahura Mazdāh's Iranian rivals. The wording “Ahura Mazdāh with (all) the gods” does not yet appear in the Behistun inscription, the first which Darius dictated. In it we find, instead the wording “Ahura Mazdāh bore me aid, and the other gods who are.”¹⁶ In translating this expression the Elamite translator added to the name of Ahura Mazdāh the explanation “the god of the Iranians.”¹⁷ It seems reasonable to infer that in his opinion at least, the “other gods who are” were not Iranian. If this inference is correct, he must have had reasons to think that Darius held no brief for any “other” Iranian god that he, the translator, knew of,¹⁸ including, of course, Mithra,¹⁹ of whose existence the translator cannot have been unaware.

At all events, Darius's exclusive attribution of his success to Ahura Mazdāh, the fact that on the eve of his murder of Gaumāta he “prays”²⁰ to Ahura Mazdāh alone, and his almost Gathic statement that “Ahura Mazdāh is mine, I am Ahura Mazdāh's” (Susa, k), leave no room for doubt that he was a monotheist. Since in addition he speaks of “Falsehood” (Drauga) as if it were a personification of the principle whose opposition to “Truth” constitutes the basis of Zarathuštra's dualism,²¹ I agree with those scholars who consider Darius to have been a follower of Zarathuštra.

It has often been argued against this view that Darius mentions neither Zarathuštra nor the Amsša Spsntas.²² The latter, it will be remembered, play a prominent role in the Gathas as the seven Aspects of Ahura Mazdāh. They are called “Holy Immortals” in the Younger Avesta, but Zarathuštra, who mentions them collectively only on two occasions, has a different way of referring to them. In one passage (Yasna 51.22) he describes them by the phrase “those who have been and are,” in the other (the Yeehē Hātγm prayer) by the plural of the present participle of the verb “to be,” which requires the translation “those who are, the being ones,” cf. AHM, p. 164 f.²³ It is on the strength of this definition that modern scholars sometimes refer to the Holy Immortals as “Entities.”

The objection that Zarathuštra and the Amsša Spsntas are not mentioned by Darius has, of course, no probative value, since in inscriptions of a political character there was no necessity to refer to either. In fact, however, I believe that Darius does refer to the Amsša Spsntas, and we shall see later (p. 28) that the Greek rendering of Zarathuštra's name as “Zoroaster” makes it at least likely that Darius was familiar with the prophet's name.

To take the Amsša Spsntas first, we have noted in the Behistun inscription Darius's wording “Ahura Mazdāh bore me aid, and the other gods who are,” which strangely differs from the formula he used in his later inscriptions, “Ahura Mazdāh with (all) the gods.” Even assuming that by the time Darius used the second formula he really meant to refer to foreign gods, can one be sure that he employed the earlier formula with the same intention? As the puzzling relative clause “who are (or exist)” has not yet been explained,²⁴ I venture to suggest that it was prompted by Zarathuštra's definition of the Amsša Spsntas as “Entities.”²⁵ If this interpretation is acceptable, it will remove the last doubt that the definition of Ahura Mazdāh which Darius carried in his mind was the one which Zarathuštra had given.

The manifestation of Ahura Mazdāh in seven Aspects or “Shapes,” as the Younger Avesta calls them (Yašt 13.81), is no less enigmatic a notion than the Christian Trinity. If the expression “gods who are” refers to the Amsša Spsntas, it is a fair assumption that only initiates of genuine Zarathuštrianism would understand it correctly. It would therefore be only natural for the Elamite translator to assume, and indicate to his compatriots, that by “the other gods who are” the king meant the gods of Elam, Babylonia, Egypt, etc. Foreigners probably placed the same interpretation on Darius's usual formula “Ahura Mazdāh with (hadā) the gods.” If they were right, one would have to interpret this phrase as a curious abbreviation of “Ahura Mazdāh with all the gods,” which is in fact the wording Darius used in the Persepolis d inscription. The phrase could, however, stand for “Ahura Mazdāh with the gods-who-are (viz. the Amsša Spsntas),”²⁶ in which case the adjective “all” in Persepolis d would be pleonastic.²⁷

As far as Xerxes is concerned, he clearly followed or tried to follow in his father's steps in religious matters. He, too, only worships Ahura Mazdāh but frequently invokes for protection “Ahura Mazdāh with the gods,” as Darius had done. In addition, Xerxes claims to have destroyed a “temple of the daiva's,”²⁸ that is, of certain gods collectively so described, whom Zarathuštra had singled out for disapproval.

It has been argued, however, that Xerxes cannot have been a Zarathuštrian (or Zarathuštricist), because for him the word artāvan had a different meaning than for the prophet.²⁹ In considering this argument it will be of help to bear in mind that, from all accounts and on the evidence of the inscriptions, Xerxes lagged considerably behind Darius in intellectual power and curiosity. Provided both took an interest in Zarathuštrianism, one would expect Darius not only to have applied his mind more intensely to it because he was the first to promote it over other beliefs, but also to have penetrated more deeply into its meaning than his son would be able to do. In particular, Darius would have been able to cope more easily than Xerxes with a difficulty which any ancient Iranian confronted with Zarathuštra's terminology was bound to meet, namely the re-interpretation implied in the prophet's utterances of certain pan-Iranian religious terms. We have already credited Darius with understanding and approving of what Zarathuštra had meant by “those who are.” What Xerxes thought on this point we have no means of telling; but he certainly shows no knowledge of the new meaning which the prophet had given to the term artāvan, while Darius, I suspect, does.

The word Arta, “Truth”, of which artāvan is a derivative, represents a basic religious notion which the ancient Persians, like the Eastern-Iranian speakers of Avestan, had inherited from Indo-Iranian times. The seat of Truth is located according to the Rig Veda in the other world, and the Younger Avesta has preserved the pre-Zarathuštrian belief that “the souls of the dead dwell in the radiant quarters of Truth” (Yasna 16.7). Accordingly before Zoroaster's time the adjective artāvan, which literally means “one who has acquired Truth,” must have been used by all Iranian speakers in the sense of “one who has gained Paradise, one of the blessed dead.” The other-worldly connotation of artāvan is not only attested in the “Daiva” inscription of Xerxes, but is attributed by Greek lexicographers to the grecized form of the world, aρταιοι, and is presupposed by the meanings “demon” of Khwarazmian arδāw, and “spirit, genius” of Ossetic (i)dāwag³⁰ (cf. AHM, pp. 155, 321).

For Zarathuštra, however, Truth and the state of grace it held out were within reach of the faithful already in this life. Accordingly in the prophet's terminology artāvan (or rather, as he pronounced the word, ašāvan) denoted additionally a living person “who had gained possession of, and was consequently acting in accordance in the Behistun inscription (column V, 19 f., 34 f.) seem to require a restoration which would show that Darius was aware of Zarathuštra's extension of the meaning of artāvan. Xerxes, on the other hand, incurious about the refinements of Zarathuštra's doctrine, continued to use artāvan in the sense in which it had been generally accepted from times immemorial.³¹

The same tendency of Xerxes to continue in the rut of traditional Old Persian religious parlance shows itself in his association of Arta, “Truth,” with the age-old term brazman, “rite” (the etymological and factual equivalent of Vedic bráhman³²), a term which is conspicuously absent from the Avesta.

Xerxes thus appears to have been a Zarathuštrian in intention. But he was not sufficiently versed in Zarathuštrian terminology to be aware of all the instances where the prophet had discarded a pan-Iranian religious word or invested one with a new meaning.

III. ARTAXERXES I AND THE CALENDAR

While the inscriptions of Darius and at least the Daiva inscription of Xerxes bear the stamp of a personal style, those of Xerxes' successors have little individuality. Apart from a few innovations in the substance matter, such as the mention of Mithra and Anāhitā as protective deities by Artaxerxes II, they merely reproduce the wording of earlier texts. This lack of a personal stamp suggests that the later inscriptions were not dictated by the kings themselves but were drafted at their command, or even at the request of mere builders or goldsmiths, by scribes whose initiative rarely extended beyond copying existing inscriptions.

That the dependence of the later inscriptions on wording coined by Darius is due to their being copy-work, and not to a persistence of Darius's words in the memory of his royal successors, may also be inferred from the absence in them of sentences and phrases which Darius had used only in the inscriptions of Behistun and Naqš-i Rustam: these inscriptions were placed too high for would-be imitators to be able to read and copy them.

The reason why no Achaemenian king later than Xerxes insisted on having his own words inscribed was presumably that, as a result of the Aramaic language and writing having become the chief means of written communication, the scribes had lost familiarity with the letters and spelling conventions of the Old Persian cuneiform script and could not be expected to do more than to slightly adapt existing patterns to new needs. Even within these limits the scribes were apt to fall back on Aramaic spelling conventions, cf. Appendix IV, p. 33.

It is with these considerations in mind that we must evaluate the religious references which are found in the few inscriptions of Artaxerxes I (465-425 b.c.) and Darius II (424-405). These references consist of the following three formulas, which Darius I and Xerxes had abundantly used: “A great god is Ahura Mazdāh” (Artaxerxes I), “Ahura Mazdāh with the gods” (Darius II), and “by the will of Ahura Mazdāh” (Artaxerxes I and Darius II).

That such formulas need not in the case of Artaxerxes I and Darius II imply the same monotheistic attitude as is vouchsafed by additional statements in the inscriptions of Darius I and Xerxes, can be seen from the fact that under Artaxerxes II (405-359) the third formula is used, and under Artaxerxes III (359-338) the first, each in an inscription where a few lines later Mithra and Anāhitā, respectively Mithra alone, are invoked in addition to Ahura Mazdāh.

The inscriptional evidence, therefore, although it requires us to suppose that Artaxerxes I and Darius II continued to worship Ahura Mazdāh, does not exclude that their religious attitude may have been closer to that of Artaxerxes II than to the monotheism of Xerxes. There is in fact reason to think that by the year 441 Artaxerxes I had come to support, though not necessarily to practice, a polytheism which closely resembled the one of the Younger Avestan scripture.

The year 441 has emerged from S. H. Taquizadeh's calculations (Old Iranian Calendars, pp. 13, 33, and passim) as the most likely date at which the so-called “Zoroastrian” calendar was introduced as civil calendar throughout the empire. Such a step could not have been taken unless the Great King had approved of the religious implications of the calendar.

The calendar is called “Zoroastrian” by modern scholars because it was used by the authors or some of the authors of the Younger Avesta. In the scripture each day of the month and each month of the year has a religious name. Divinities honored by the names include Ahura Mazdāh,³³ six “Holy Immortals,” Mithra, Anāhitā,³³ Tištrya, Fire, the Fravašis,³⁴ Sun, Moon, Sky, Earth, Wind, the Soul of the Cow, the Religion, Discipline, Reward, in short, most of the deities of the Zarathuštric religion, with the notable exception of Haoma (cf. below, p. 26, n. 43).

Although apart from the Avesta no source belonging to the Achaemenid period mentions this calendar, its appearance in later times both in Sogdiana and Chorasmia in the East, and in Armenia and Cappadocia in the West, leaves no doubt that it had been in general use as a civil calendar before the Achaemenian empire broke up. We may note, however, that while the Cappadocian month-names fully agree with the Avestan ones, whose Persian equivalents are still used today, some of the Armenian, Chorasmian, and Sogdian months have names which differ from their Avestan counterparts.

The chronological system of the “Zoroastrian” calendar as attested in all sources has been shown by Taquizadeh to have been the result of a reform of an earlier religious calendar which we shall henceforth call the “unreformed” one. The latter may have been instituted in about 510 b.c., and was used until 441 beside the Old Persian civil calendar which we find in the Behistun inscription of Darius and the Elamite tablets of Persepolis. The date of thereform, 441, must also have been the date of the official promotion of the religious calendar to civil use. Thereafter the Old Persian civil calendar was presumably abandoned.

While the Old Persian civil calendar as used by Darius followed a Babylonian pattern, the unreformed religious calendar must have been modeled on the Egyptian calendar with its vague year of twelve months of thirty days each and a yearly intercalation of five days. The reform of 441 consisted of two measures: the vague year of the unreformed calendar was changed into a fixed year by means of a new intercalation system; and the beginning of the year was set near the vernal equinox. One result of these measures was that the month which in Avestan is called “(the month of) the Fravašis,” having come to fall at the beginning of spring in 441, remained the first month of the year ever after.

At the time when the unreformed religious calendar was instituted, the first month of the year must have borne the name of the Iranian chief god. The apparently illogical fact that in the reformed, “Zoroastrian,” calendar the month dedicated to “the Creator” occupies the tenth place is easily explained within Taquizadeh's scheme by the shift of position which all months had to undergo when the month of the Fravašis became the first. But this convincing explanation necessarily implies that the month which after the reform became the tenth was dedicated to “the Creator” also before the reform.

Were the names of the other months of the “Zoroastrian” calendar, in particular the “pagan” ones, also taken over from the unreformed calendar? If they were, we could reasonably conclude that the Zarathuštric-looking polytheism which is attested for a later period in the inscriptions of Artaxerxes II, had received the official recognition of the Court not later than 441. If on the other hand the “pagan” month-names of the “Zoroastrian” calendar replaced names of the unreformed calendar which belonged exclusively to the Gathic-looking Mazdāh-worship of Darius,³⁵ we shall have to consider two alternatives: either the replacement was part of the reform, in which case we would again be driven to the conclusion stated in the preceding sentence; or it took place at some later date within the Achaemenid period. The odds are heavily in favor of the first alternative.

For, once the new civil calendar had been promulgated throughout the empire, its promulgators would scarcely revoke part of its terminology and thus cause much inconvenience and confusion in the imperial administration, unless they had very strong reasons for doing so.³⁶ Such reasons, in the given case, could only have been religious. But it is difficult to see why even Artaxerxes II or Artaxerxes III, who more than any other Achaemenid king display in their inscriptions a religious attitude which agrees with the one implied by the “Zoroastrian” calendar names, should have felt that a Gathic-looking calendar would be unrepresentative of their religion. Do not both kings, despite their polytheism accord pride of place to Ahura Mazdāh, as do the authors of the Younger Avestan scripture? And as the latter, despite their polytheism, call themselves Mazdayasnians, that is, “worshippers of Mazdāh,” would not the two kings consider themselves as faithful Mazdayasnians as Darius had been? A Gathic-looking set of month-names, we may surmise, would therefore have seemed to the two kings as venerably archaic and authentic, as the Gathas themselves must have seemed to the Younger Avestan authors, and as little in need of “modernization” as were the Gathas in the eyes of Zarathuštra's polytheistic epigones.

Thus we reach the conclusion that, whatever may have been the relation between the month-names of the reformed and the unreformed calendar, one of which at least, “the Creator,” was the same in both, the reformed calendar was characterized already in 441 b.c. by the month-names, and presumably also the day-names, which we know from the Younger Avesta.³⁷

IV. ARTAXERXES I AND ZARATHUšTRICISM

On first consideration one might think that the adoption of the “Zoroastrian” calendar under Artaxerxes I means that this king had by 441 b.c. embraced or at least approved of the Zarathuštric religion of the Younger Avesta, which the calendar names would in this case faithfully reflect.³⁸

And yet, a basic question remains to be asked. Artaxerxes I could not have been influenced by the Zarathuštric religion unless this religion was there to influence him. But this religion, in the form in which it appears in the Younger Avesta, is in fact a perversion of Zarathuštra's original message. For such a perversion to be not only tolerated but actually enjoined by the Eastern Iranian Church which Zarathuštra had founded, a strong incentive must have come into being. Can we discover such an incentive before the introduction by Artaxerxes I of a polytheistic civil calendar? If not—and the answer has so far been “no”—would we be justified in reversing the tables and assuming that, far from the king having been influenced by the Zarathuštric religion, it was he who by his acknowledgement of the pagan gods brought this religion into being?

While the carefully balanced structure of Zarathuštra's original doctrine is pleasing even to our sophisticated minds, one may well wonder as to who may ever have derived satisfaction from the bewildering contradictions of the Younger Avesta, in which monotheism, dualism, and polytheism are all preached simultaneously. The lack of appeal of this scripture is illustrated by the fact that, when Zoroastrianism re-emerges from the dark ages in Sasanian Iran, its priests regard the sacred texts of the Avesta with affectionate veneration, translate them into the Middle Persian language, and write commentaries on them, but in fact preach the very different Magian version of the doctrine.

Nevertheless, it will be said, the mixed scripture did exist; however contradictory its “system,” must there not have been at some time people who approved of it as it stands?

The answer, to my mind, is that the religious mixture of the Younger Avesta is precisely what it appears to be: not a single religion which anyone would defend as a system, but a sum, an agglomeration, of different creeds to which different communities were attached in the fifth century in the distant Eastern Iranian homeland of the Avesta.

Thus viewed the Avesta is not the religious nightmare which it is sometimes, and pardonably, treated by religious historians. As far as the religious data which it preserves are concerned, the Avesta may be regarded simply as a record of cults, an anthology, as it were, of Old Iranian religious lore. It transcends, however, the character of an anthology, in that it is pervaded and held together by the pious fiction that the whole of it was revealed by Ahura Mazdāh to Zoroaster.

It is in this fiction that we can recognize the wood through the maze of trees. If the purpose of the authors of the Younger Avesta had been to provide an attractive syncretism of existing beliefs, their failure to avoid glaring contradictions would be incomprehensible. But if their intention was to restate existing beliefs for the sole purpose of claiming that Zoroaster had recommended them all, the motley assortment presented by the scripture ceases to be surprising. As a “proof” of Zoroaster's sponsorship of all the Iranian cults that are mentioned in it, the Younger Avestan scripture makes perfect sense.

On the other hand, for a pious fiction of this kind to take shape and gain acceptance, there must have existed a strong motive, the time must have been ripe, the conditions favorable. So long as the Great Kings gave official recognition to Ahura Mazdāh alone it is very unlikely that the Eastern Iranian Zarathuštrian priests would claim that their prophet had approved of the pagan gods whose worship he had in fact prescribed. But a powerful motive and a favorable climate of opinion for advancing such a claim would come into being from the moment an Achaemenian ruler had publicly stated that the pagan gods were as worthy of worship as Ahura Mazdāh. The introduction in 441 of the polytheistic civil calendar amounts to a public statement to this effect, the earliest to our knowledge. At that date, therefore, the calendar names were more likely a premonition than an echo of the mixed, Younger Avestan, scripture.

V. THE MAGI

If the Zarathuštric religion of the Younger Avesta was not the source of the polytheistic calendar names or, more generally, the stimulus for the recognition and eventual adoption of polytheism by the royal House, what else may have been the source or stimulus? To answer this question we must try to reach a closer definition than Herodotus provides of the religious data which he quotes.

The Iranian gods whom the Persians worshiped according to Herodotus, are “Zeus,” Mithra, Anāhitā,³⁹ the Sun, the Moon, the Earth, Fire, Water, and the Winds. All these divinities are also objects of worship in the Younger Avesta and duly appear among the calendar names. The probability is that on the soil of Western Iran the same variety of cults existed as in the Eastern Iranian homeland of the Avesta. Yet in Western Iran only one priesthood is known to have officiated: the Magi.

To all the gods mentioned, the Persians are said by Herodotus to have offered sacrifice; but, Herodotus adds, “without a Magus it is not lawful for them to offer sacrifice” (i.132). This statement permits the hypothesis that the Magi were qualified and expected to minister in the service of any of a number of different gods worshiped by different people in Persis. Now Darius, we know, worshiped Ahura Mazdāh. We also know that at the time of his accession he had a grudge against the Magi. The manner in which he settled it does not suggest that the Magi were the priests of his religion: his first concern after obtaining the throne was to order a slaughter of Magi. Year after year the anniversary of this slaughter remained an occasion of great celebrations.⁴⁰ Nevertheless, from all accounts, the position of the Magi at the Achaemenian court, at any rate from Xerxes onwards, remained unchallenged, and this despite the fact that the Magi did not hold the religion of their rulers in such high regard that it should be brought to the notice of Herodotus.

That the Magi might not initially have greatly cared for the Zarathuštrian Ahura Mazdāh is understandable. As far as we can tell, this modified version of the pan-Iranian god Ahura (cf. n. 1) was introduced in Western Iran by Darius after he had accepted the substance of Zarathuštra's message.⁴¹ Like most long-established priesthoods, the Magi were probably inclined to be conservative and would not lightly accept a new form of worship. If the new form were introduced by a king they hated and feared, this would commend it even less to their affection. But the Magi could hardly afford openly to flout the one god cherished by Darius and Xerxes and to refuse to serve him. Rather we may suppose that the reason why they were allowed to become again the priests of the Court was that, despite their resentment, they had agreed to officiate in the Zarathuštrian service of Ahura Mazdāh, along with their service on behalf of the polytheistic communities whose priests they had been for centuries.

How then would the mind of a Magian priest appear? Eclectic in the extreme, we may suppose, as far as the Iranian gods were concerned. The Magi would be, not the representatives of one particular religion, but technical experts of worship, professional priests who, equipped with barsman twigs and all the paraphernalia of a meticulous ritual, would conduct the service of any Iranian god to whom an employer willing to pay them should wish to render homage.

Such an interpretation would also account for the fact that Herodotus has not really anything to say about the Magian religion. It would appear that there was no Magian religion before the Magi became Zoroastrians. The peculiarities of Magian behavior could be observed only in the minutiae of ritual or purification, such as their mode of disposing of the dead or their zeal for destroying animals which they considered impure, both of which Herodotus duly noted.

If the Magi were a professional priesthood to whom Zarathuštrianism was merely one of the forms of religion in which they ministered against payment, much as a professional musician earns his living by performing the works of different composers, then we are in a better position to understand why Zoroaster and his doctrine remained unnoticed by the Greeks for such a long time. The Greeks relied for information on the Magi and for the Magi of the first six or seven decades of the fifth century Zarathuštrianism need not have meant more than an idiosyncratic treatment of the age-old Ahura, whom they continued to regard, within the pantheon of the non-Zarathuštrian gods whom they continued to serve on behalf of the majority of their Persian employers, as an approximate equivalent of the Greek Zeus. If the information we read in Herodotus was provided by a Magus, the latter may well have felt that by mentioning Ahura (=“Zeus”) first, he had taken sufficient account of the special regard in which the god was held by the Great Kings.

Although the Magi must have become thoroughly familiar with the terminology Zarathuštrianism, their eclectic upbringing would prevent them for a long time from penetrating to the core of the prophet's doctrine. Not being prepared to revere Ahura Mazdāh as sole god, they would be unable to appreciate the inner necessity of the complicated relationship, possibly even to understand the relationship itself, which Zarathuštra had been led logically to envisage, between Truth, Ahura Mazdāh, the Holy Immortals, and the two Spirits. What presumably the Magi understood and liked best in Zarathuštrianism was its condemnation of the Daivas, which agreed with their own exorcistic bent and which they may have associated with the Gathic account of the incompatibility of the two Spirits (cf. below, p. 30).

But if by the second or third decade of the long reign of Artaxerxes I the Magi had become perfectly reconciled with Zarathuštrianism, within the limits of their understanding of it, and without surrendering to it the cults in which they had officiated for centuries, the members of the royal house would also be due for a change of heart. The effect of allowing, over a period of several decades, the spiritual as well as to some extent the political affairs of the State to be handled by a priesthood which served with professional impartiality both the King's god and many others, could hardly have been but to loosen the monotheistic exclusiveness on which Darius had insisted. The resulting laxness can be seen at an advanced stage in the invocation by Artaxerxes II of Mithra and Anāhitā beside Ahura Mazdāh. The introduction in 441 of a polytheistic civil calendar may mark its beginning.

Here then lies the answer I would suggest to the question we posed at the beginning of this chapter: the stimulus for the adoption by the royal house of a polytheism which looks Avestic was provided by the consistently eclectic attitude of the Magi.

VI. THE SCRIPTURE

Let us now return to the “Zoroastrian” calendar. Shortly before 441 b.c., the year of its inception, the Great King and his advisers, among whom Magian opinion was no doubt represented, must have decided to introduce a reform of the civil calendar, as part of which the numbers until then assigned to the days of the month were to be replaced with names, and new names were to be given to the months. As the new calendar was intended to replace also the old, “unreformed,” religious calendar, it was decided that the names of days and months should all be religious. At least one name, “the Creator,” was taken from the “unreformed” calendar. Of the other names we cannot tell whether they had been used as calendar terms before. The result, at any rate, was a calendar whose names were representative not of one cult only but of all the main cults in which the Magi officiated as professional priests. Even though the lion's share of the names belongs to Zarathuštrianism, the religion of the royal house, it would be misleading, with reference to the year 441, to describe such a calendar as “Zoroastrian.” Being a selective index of the Magian repertory, a more appropriate term for it would be “Magian.”⁴²

In due course the new civil calendar would reach the Zarathuštrian communities of the homeland of the Avesta, in faraway Eastern Iran, and enter into daily use with them. Inevitably, however, its names would mean more to them than merely a new device for distinguishing days and months. The names, insofar as they belonged to a civil calendar, would constantly remind Zarathuštrians that important concessions had officially been made to cults other than their own. In the days of Darius and Xerxes the Eastern Iranian Zarathuštrians were no doubt aware that their sole god, Ahura Mazdāh, was also the sole god of the Great King. After the year 441, however, it would become increasingly clear to them that the religious climate of the Court had changed. The calendar provided, as it were, a summary of what the Court had officially come to regard as the national religious patrimony.

We now encounter the incentive I alluded to earlier …, for the Eastern Zarathuštrian Church to take the initiative and to produce a new scripture. The fact that the Great King himself was prepared to countenance non-Zarathuštrian beliefs would lend them respectability even in the eyes of Zarathuštrian priests. Zarathuštrianism per se, with its exclusive worship of Ahura Mazdāh, was obviously on the decline at Court, perhaps also in Eastern Iran. To cling to it alone, against the trend of the time, would have meant for the Zarathuštrian priesthood connivance at its own extinction. But the Zarathuštrian Church had two invaluable assets by which it could try to insure its survival. One was the prestige of its founder, whose fame was growing the more he receded into legendary antiquity. As that fame grew, the legends must have grown, and with them the possibility of attributing to Zoroaster views he had never held. By a masterstroke of priestly wisdom, if this is how things happened, the Zarathuštrian Church claimed that the national religious patrimony now sanctioned by the Court was not a mere co-existence of unconnected or loosely connected beliefs but had a hitherto unperceived unity, in that Zoroaster himself had preached it all.

There may be more than one way in which a determined Church might attempt to substantiate such a claim. But the most effective would undoubtedly be, to produce a scripture that proved it. Here the second asset of the Zarathuštrian priests would come into its own: the fact that they already had a scripture consisting of Zarathuštra's own words, as well as the skill, accruing to them from the literary tradition of which they were the heirs,⁴³ to engage in creative literary activity on a large scale. A readily acceptable excuse for adding to their existing scripture would be, that what was now being composed and given scriptural status was what Zarathuštra had taught but failed to embody in verses. Accordingly the late disciples of Zarathuštra could set to work with a good conscience and provide a scripture for the motley religious patrimony of which the Magi had been in charge. All they had to do was to inform themselves of the contents of the hymns and liturgies of the various cults concerned and compose new texts in which these contents were presented as part of Ahura Mazdāh's revelation to Zarathuštra.

Thus, by an intense literary effort, the Zarathuštrian priests, whom we may now call Zarathuštric, transformed the eclectic standpoint of the Magi into a system purporting to be Zarathuštra's. The logical weakness of the system would not concern them. This was not a case of thinking out what would persuade, but merely of reproducing, elaborating, and combining what everyone, from the Great King downward, was ready to accept. The success fully justified the effort. By producing a new, all-embracing scripture which exploited to the full the propagandistic value of Zarathuštra's authority, the priestly authors turned the narrowly denominational character of their Church into a truly pan-Iranian one and ensured the survival of Zoroastrianism on a national scale down to the conquest of Islam and, on a more limited scale, down to the present day.

Supposing this to have been the course of events, what would be the reaction of the Magi? Whereas in early Achaemenian times there does not seem to have existed a single, definable religion that could be called Magian, the scripture of which the Magi now came to hear presented the whole of the Magian repertory as a single, pan-Iranian religion, such as could properly only be called “Magian” but in fact claimed to be “Zarathuštrian.” Had the Magi fought this claim on the ground that they had had no share in the composition of the scripture and were scarcely mentioned in it,⁴⁴ they would merely have cut their own flesh. For if they did not even know that the cults in which they were officiating had all been founded by the greatest of all sages, what excuse would they have had for continuing in their ministry?

The only sensible reaction of the Magi would have been to accept the claim and to acknowledge Zoroaster as their prophet. We are on safe ground in assuming that this is precisely what they did. For the later sources show that the Magi appropriated Zoroaster altogether: the prevalent opinion expressed in Greek texts is that Zoroaster was a Persian or a Mede, and the Sasanian Zoroastrian writers not only regard Media as Zoroaster's homeland but locate in Media a number of Avestan toponyms which, according to the geographical horizon of the scripture, must in fact have belonged to Eastern Iran.⁴⁵ It is clear that this misleading information emanated from the Magi; and that the most suitable occasion for their inventing it would be at the time when they felt impelled to claim that they, and not the Eastern Iranian Zarathuštrian (or Zarathuštric) priests, were the true heirs and custodians of Zoroaster's doctrine.

If our reconstruction of the course of events is correct, the composition of the new Zarathuštric scripture will have begun soon after the year 441. The adoption by the Magi of Zoroaster as their prophet may have followed within the next two decades. From then onwards something which could be called a Magian doctrine took shape and became known to the Greeks before the middle of the fourth century, while the prophet's name, once the Magi began to take a personal interest in him, reached the Greeks already at the beginning of the fourth century.⁴⁶

It reached them at first, however—and this has so far been insufficiently appreciated—not in its Median form ^*Zaratuštra, which was to reappear, slightly altered, as Zardušt in Sasanian Persia, nor in its Avestan form Zaraθ-uštra, but in the characteristically Old Persian form ^*Zarauštra which the Greeks, because of the prophet's alleged preoccupation with astrology, transformed into Zωρο-άsτηρ by popular association with their own word for “star. …”⁴⁷

The unmistakably Persian origin of the name Zoroaster seems to me to confirm that the original sponsors of the prophet in Western Iran, and his sole supporters for a number of decades, were not the Median-speaking Magi, but an influential group of Persians. On whose initiative this group adhered to the prophet's doctrine is a question to which the religious purport of the Darius inscriptions supplies a reasonably clear answer.

By the time the Magi took control over Western Iranian Zoroastrian affairs, their own Median form of the prophet's name could no longer prevail over the one which had gained currency in Europe: the name Zoroaster is still with us as an indictment, as it were, of the slowness with which the Magi embraced the cause of the prophet.⁴⁸

Conversely, the fact that the Sasanian Zoroastrian priests used neither the Avestan nor the Persian form of the prophet's name but the Median one confirms that these priests were not called “Magi” for nothing.⁴⁹ Their using this title, their referring to the prophet as “Zardušt,” their preaching a doctrine based on the formula Ohrmazd versus Ahriman, are all clear indications that although these priests were by then and remained the sole custodians of the Avestan scripture, they were the direct heirs, not of the Avestan Zarathuštric tradition, but of the Magian tradition as it had crystallized in the fourth century b.c.

VII. THE MAGI AND THE SCRIPTURE

At the beginning of this article we saw that the Holy Spirit is in the Gathas Ahura Mazdāh's creative Aspect and is referred to as “Creator” in the Younger Avesta (Yašt 10.143), where conversely Ahura Mazdāh, in his capacity of creator, is often addressed as “Holiest Spirit” (Yašt 8.10, 10.73, etc.). This terminology may seem almost to invite the inference that God and the Holy Spirit are identical, from which it would follow easily that the opponent of the Fiendish Spirit must have been Ahura Mazdāh himself, and only a little less easily that the Fiendish Spirit was as primordial as God. Yet, logical, or fairly logical, as it might have been to deduce the formula Oromasdes versus Areimanios from the position as it is stated in the Avesta, such a step was neither taken by the authors of the scripture in Achaemenian times nor explicitly approved in Avestan texts that were added to the scripture in post-Achaemenian times.

There is, it is true, the last chapter of the post-Achaemenian Vendidad, where Ahura Mazdāh himself faces the Fiendish Spirit (see above, p. 14). But it is a far cry from the 99,999 diseases with which the Fiendish Spirit infects the god in that chapter to the integrated design of a cosmic conflict resulting from the primordial co-existence of Ohrmazd and Ahriman, as it is outlined in the Middle Persian books and presupposed by the two aρχαί of the περi pιλοsοpίαs. Had the authors of the Vendidad approved of this design, it would be hard to understand why it is not prominently displayed in this “Book against the Demons”—which is what “Vendidad” means—for which it would have made an admirable framework.

We may then say that in Achaemenian times the Zarathuštric priests, who probably continued adding to the scripture until the collapse of the empire, either did not learn of the Magian formula or deliberately ignored it as a heresy. The authors of the Vendidad, on the other hand, who can hardly have been unaware of the formula, probably regarded it as a heresy but were nevertheless somewhat affected by it in their outlook.

That the formula Oromasdes versus Areimanios was in fact a heresy is obvious from its absence from the oldest Avestan texts. That it was an original and elegant heresy is also undeniable. Nevertheless originality and elegance are hardly qualities we have reason to associate with the Magi. Once the Magi had adopted the great sage as their prophet, it seems to me that, far from wishing to be different, they would be likely to try hard to conform to his message. But to conform to it they would have to understand it. And to understand it, at a time when Zoroaster's own Avestic epigones were crediting him with the very views he had condemned, they had only one guide to turn to: the Gathas.

Zarathuštra's Gathas today count among the most difficult products of world literature. This is not merely because we are so remote from them. The conciseness of the wording, the ambiguity of the inflectional endings, the abundance of technical terms that do not recur even in the Younger Avesta, the apparent attempt to depict by means of the word order obscure sequences of thought, all these are factors which must have made the Gathas extremely difficult to understand at all times. In addition, it must be realized that there was a considerable language-barrier between the ancient Persians and the speakers of the Avestan language. The difference between Median and Avestan must have been less great, but a difference undoubtedly existed.

I said earlier that Darius appears to have accepted the substance of Zarathuštrianism (above, p. 24). It is unlikely that much more than the main drift of this subtly balanced doctrine could have been apprehended by a foreigner, as Darius was in respect to Zarathuštra. But if our reading of the relevant hints in the Darius inscriptions is correct, Darius understood the principles of the faith far better than anyone else in antiquity of whom we have any record, apart, of course, from Zarathuštra himself. In any effort to inform himself of the message of the alloglot prophet, Darius would have been favored by an opportunity forever denied to those who after him endeavored to understand the Gathas: his chronological nearness to Zarathuštra. In Darius's youth and early manhood, at a mere two or three decades from the death of the prophet (v. Henning, Zoroaster, 41), the possibility still existed for Persians or Medes to have the Gathas explained to them by men who had known Zarathuštra, provided that one of those concerned was bilingual. By the middle of the fifth century such opportunities had vanished, not only in Western Iran but also in the homeland of the scripture. Even the authors of the Younger Avesta, who must have known the text of the Gathas by heart, had only an imperfect understanding of their contents.⁵⁰

How much less is it to be expected that the Magi who, if our theory is correct, never claimed to be Zoroastrians until the bold bid of the Eastern Zarathuštrian Church compelled them to do so, would belatedly extract from the Gathas the correct details of the prophet's theology. Their new approach to Zoroaster's doctrine would be conditioned by their old one. We need not assume that, because the Magi for a long time failed to identify themselves with Zarathuštrianism, they had not formed their own views of it quite early. And if it is true that they had missed the essence of the prophet's message to such an extent that they saw no objection to serving both his religion and other cults, then we need not be surprised if they misinterpreted his message in respect of the identity of the protagonists of his dualism. It may have been, at a wild guess, in the context of the Daiva-worship proscribed by Xerxes, that the Magi came to the conclusion that Zarathuštra's Fiendish Spirit was the aboriginal chief of all demons, the “daivic” counterpart of Ahura Mazdāh.⁵¹ There is nothing to show that Xerxes, if confronted with such a view, would have been in a position to refute it. What is more, we need not expect that such a misapprehension would have shocked Western Zarathuštrians nearly as much as it would shock us. What mattered most to them, witness Darius, was the exclusive worship of Ahura Mazdāh and the rejection of the principle “Falsehood” which the Fiendish Spirit had chosen. Zarathuštrians who conceived of the prophet's message in such simple, practical terms, would scarcely give much thought to the implications of a replacement of the name “Falsehood” by “Fiendish Spirit.”

By the time the Magi adopted Zoroaster as their prophet and the Gathas as their guide to his meaning, these difficult poems could no longer act as a corrective of the Magian dualist formula. Their obscure wording might just as well have been thought by the Magi to confirm it. More than that, even if the formula had not yet been evolved by the Magi, they could at this late stage deduce it from the Gathas themselves, without realizing that they were misrepresenting the prophet's views. For although Truth is in the Gathas an aboriginal principle distinct from God, and the two Spirits are emanations of God's, both Truth and the Holy Spirit are identical with Ahura Mazdāh insofar as not only the Holy Spirit, but also Truth, is an aspect of his (cf. above, n. 3). The Magi could therefore hold in perfect good faith that in Zoroaster's opinion “Truth” and “Holy Spirit” were no more than parts of the definition of the aboriginal δαίμων Ahura Mazdāh. With such a premise it would be difficult to avoid the fallacy that the Gathic “Falsehood” was, like “Truth,” a mere aspect of an aboriginal δαίμων, for whose role the Fiendish Spirit must have seemed to the Magi to have been ideally cast.

Considering the odds against them, the Magi did very well. The simple and arresting formula Oromasdes versus Areimanios, by which they replaced the complicated situation of the Gathas, may have been misleading and heretical. But it was destined to exert a far greater and more lasting appeal than Zoroaster's own rarefied doctrine. Moreover, the Magi deserve credit for having understood and transmitted to posterity Zoroaster's emphasis on the dignity of man as a free agent, on whose recognition of Truth and consequent support the very cause of God depends.⁵² This moral teaching is, after all, what would matter much more to ordinary men and women than the niceties of the relationship between God, the two Spirits, the Immortals, and Truth. It is no doubt the satisfying insistence on the intelligence and trustworthiness of man which more than anything else has enabled Zoroastrianism to remain a living religion to this day.

To the student of Iranian civilization, however, the merit of the Magi lies not so much in what they made of Zoroaster's doctrine, as in the fact that, by adopting Zoroastrianism, they ensured the preservation down to the present day of its Avestan scripture. Without this scripture the bright dawn of Iranian thought would be hidden from us. Much of it, indeed, is still unclear, after nearly two centuries of study. But the scripture is with us, and it is for us to make the best of this Magi-sent opportunity. In the present article I have really only attempted to show where Zoroaster's own contribution does not lie. A close study of the Gathas, however, permits, even at the present stage of imperfect understanding, an analysis of certain ingredients of his doctrine as belonging to earlier, Indo-Iranian thought, while other religious elements appear for the first time in the Gathas. It is thus possible, to some extent, to discern where Zarathuštra innovated, where he took over notions we can trace elsewhere, and even where he expounded ideas which cannot be traced anywhere else but of which nevertheless he is unlikely to have been the inventor.

Our purpose here has been the more modest one of trying to account historically for two drastic and more or less contemporaneous innovations on Zoroaster's doctrine, the one represented by the Younger Avesta, and the other inherent in the formula Oromasdes versus Areimanios; and to suggest reasons why either innovation was presented not as such, but as Zoroaster's own contribution.

APPENDICES

I

The introduction of the term “Zarathuštricism” is intended as an improvement on H. Lommel's and my own earlier terminology, according to which “Zarathuštrianism” denotes the prophet's own doctrine, and “Zoroastrianism” both its Younger Avestan and its Sasanian modifications. The objection to applying the term “Zoroastrianism” to the Younger Avesta (except when precision does not matter) is that its authors called the prophet “Zarathuštra,” not “Zoroaster,” while the ancient Greeks, who called him “Zoroaster” (see above, p. 28) never learnt to know the Younger Avestan form of the doctrine. What the Greeks believed to have been Zoroaster's teaching is substantially the version we find centuries later in the Sasanian books. It is therefore to this version alone that the term “Zoroastrianism” should strictly speaking be applied.

II

The details of the “twin” passage, Yasna 30.3, are controversial, but all translators except one agree that in it the dual ysmā, “twins,” refers to the dual mainyū, “the two Spirits.” The divergent view is W. Lentz's (A Locust's Leg, Studies in Honour of S. H. Taquizadeh, [1962], pp. 132 f.), whose translation, however, defeats itself: one has only to remove the explanatory parentheses to discover, even through the disguise of the English wording, that the two duals belong together. In my view the verse should be translated as follows:

at tā mainyū paouruyē
yā ysmā x(v)afnā asrvātsm
manahičā vačahičā šyaoθanōi
hī vahyō aksmčā
asčā hudaehō srsš
vīšyātā nōit duždaehō

“Firstly the two twin Spirits [lit. the two Spirits who (are) twins] were revealed (to me), each-endowed-with-own-wish (= free will). Their (hī) two ways-of-thinking, ways-of-speaking, and ways-of-acting are (respectively) the better and the bad. And between these two (ways-of-thinking, etc.) it is the wise, not the fools, who choose correctly.” This means taking hī as a general dual form of the enclitic pronoun, comparable in its multivalence to Vedic sīm, and x^va-fnā as a nominative dual, comparable in formation and meaning to Skt. sva-cchanda, “following one's own will or pleasure,” and displaying, in conditions of compound reduction (cf. Av. srsdva-fšna-), the base ^*fan- which is known from Khot. (pa-, us-)phan-, “to be pleased,” and Oss. fœndon, “wish” (see H. W. Bailey, Trans. Philol. Soc. [1956], p. 121, and M. J. Dresden, Trans. Amer. Philos. Soc., XLV [1955], 478, b). However, since x^vafna- occurs elsewhere in the sense of “sleep; dream,” there remains the possibility that the first line means: “At first the two twin Spirits were revealed (to me) through a vision.”

Since Av. paouruya-, “first,” is never used as a noun, its adverbial locative paouruyē should not be translated “in the beginning,” as is usually done, but “at first, firstly.” This would always have been understood, had not the first line of Yasna 45.2 obscured the issue: at fravaxšyā aehsuš mainyū paouruyē literally means: “I shall mention firstly of the world the two Spirits,” that is: “I shall mention the two Spirits before I mention anything else that belongs to the world.” Here the normal construction of the adj. paouruya- with the partitive genitive (cf. Yasna 57.2: “Sraoša … who (is) the first of Mazdāh's creation [= the first out of all the creatures of Mazdāh's]”) has been transferred to the adverb paouruyē.

III

To import the formula Oromasdes versus Areimanios into the Gathas one would have to assume that (1) the two Spirits were unrelated, and (2) the Holy Spirit was identical with Ahura Mazdāh.

The first assumption would require an evasion of the prime meaning of the word “twin,” to which indeed Lommel, Die Religion Zarathustras, pp. 27 f., took recourse, on the ground that unless one did so the prophet's doctrine would be marred by the “unthinkable” contradictions which he lists on pp. 26 ff. These contradictions, however, are all of Lommel's own making, as they only arise from the unwarranted supposition that if the two Spirits were real twins, their father would have been in Zoroaster's opinion not Ahura Mazdāh but Zurvan.

The second assumption is so firmly refuted by Yasna 47 (as even Lommel grudgingly admits on p. 20, lines 12-20), that not even a comparison, not made by Lommel, of Yasna 30.5, where Zoroaster describes the Holy Spirit as “being clad in the most steadfast skies,” with the Younger Avestan statement that “Ahura Mazdāh wears the sky as a dress” (Yašt 13.3), can reduce its improbability. Being clad in the sky, to judge from the similar attribute of having the sun as eye, which the Younger Avestan Ahura Mazdāh (cf. Yasna 1.11, 3.13, etc.) and the Vedic Varuna have in common, was probably a characteristic already of the Indo-Iranian god Asura (on whom cf. above, p. 12, n. 1). By transferring the heavenly garment to the chief son of God, Zoroaster would indirectly maintain its traditional association with Ahura and yet remain true to his conception of God as being devoid of naturalistic or anthropomorphic traits.

We may note in this connection that the dual deity which the Indo-Iranian Asura formed together with Mitra may also have been thought of as being clad in the sky, since its Younger Avestan representative, Mithra-Ahura, is invoked in association with the heavenly luminaries (Yašt 10.145), while its Vedic descendant, Mitrā-Varunā, has, like Āsura Varuna alone, the sun as its eye. See below, Appendix IX.

IV

It need hardly be stressed that even after the conquest by Darius of Hinduš, Iranian Mithraists worshiped the Iranian Mithra, and not the Old Indian Mitra. The occasional spelling Mitra in the inscriptions of Artaxerxes II is not due to borrowing from “Indic,” as R. G. Kent has it (Old Persian, p. 9, end of §9), but betrays the influence of the Aramaic spelling MTR of Miθra's name. The scribe who prepared the cuneiform draft for Artaxerxes II's stonemasons would normally be employed to write in Aramaic. When so engaged, he would have frequent occasion to write MTR as part of Iranian theophoric proper names such as the attested MTRDT and MTRWHŠT of the papyri. The awkward form paradayadā- (written pa-ra-da-ya-da-a-), … in an inscription of Artaxerxes II is similarly best explained as a mere transposition into cuneiform characters of the Aramaic spelling ^*PRDYD' of OPers. ^*paridaidā-.

V

It does not seem to have been adequately realized how closely Darius's personified Drauga agrees with the personified form which Zarathuštra's Drug (“Falsehood”) sometimes assumes in the Gathas (cf. E. Benveniste, The Persian Religion. p. 36). According to Darius it was Drauga who “made the countries rebellious inasmuch as these (rebel leaders) deceived the people” (Beh. Inscr. IV, 34), and Drauga “comes” to a country as do hostile armies and famine (Persepolis, d). According to Zarathuštra the wicked “destroy the creatures (or property) of Truth by the commands of Drug” (Yasna 31.1), Drug can be “removed from ourselves” (Yasna 44.13), a wicked man is “a son of Drug” (Yasna 51.10, cf. AHM, p. 169), and Drug should be “placed into the hands of Truth” (Yasna 30.8, 44.14). If Zarathuštra had been asked to name a supernatural instigator of mutiny among men he would certainly, on the evidence of the Gathas, have named Drug and not the Fiendish Spirit who in the Gathas entertains no relations with men. That Darius and Zarathuštra use slightly different though etymologically very closely related words for what each of them treats as a proper name meaning “falsehood” is of course no argument for denying the identity of the two personifications. Old Persian and Avestan also have slightly different words where the meanings “brick” (išti-: ištya-), “daughter” (^*duxθrī-: duγδar-, v. Benveniste, BSL 1951, 22), “eye” (čaša-: čašman-), and others, are concerned.

VI

There may be a connection, overlooked hitherto, between the usual Avestan designation of Ahura Mazdāh's Aspects as “Immortals,” and Zarathuštra's definition of them as “those who have been and are,” since this definition can be understood as an equivalent of “those who never cease to be.” One may consider two possibilities: either the term “Immortals” was introduced by Zarathuštra's disciples as a simplification of the definition given by the prophet, or, more likely (cf. AHM, p. 165), the prophet's definition represents his metaphysical elaboration of the term “Immortals,” the latter having been either coined or inherited by him. It should be borne in mind that in ordinary Iranian parlance the “Holy Immortals” are the “elements,” in which meaning the term may have been used before Zarathuštra's time (cf. AHM, p. 10).

VII

It would be highly irregular if, in the sentence “Ahura Mazdāh bore me aid, and the other gods who are,” the Old Persian relative pronoun were used in the sense of “whoever they,” or “as many as.” King and Thompson translated “and the other gods, (all) that there are.” This is certainly permissible, since the later formula “with the gods” has a variant “with all the gods”; but the assumption that “all” was implied still fails to account for the addition of the relative clause “who are.” That a metaphysical definition, such as the one of Yahwe in Exod. 3: 14 (“I am He who is” in E. Schild's convincing translation, Vetus Testamentum, IV [1954], 296 ff., to which Mr. J. V. Kinnier Wilson obligingly drew my attention), would have been offered by Darius of gods other than his own, can be safely excluded. Upon enquiry as to whether Darius's wording could be a calque of a Semitic phrase meaning “God X and all the other gods,” Mr. Kinnier Wilson very kindly sent me the following note: “Akkadian could say ilū mala bašû (as The Epic of Creation, V, 106) ‘the gods, as many as there are,’ or ilū gimiršun (ibid., III, 130), meaning the same thing. Akkadian certainly has no formula equivalent to ‘god X and the other gods who are,’ or even (and this may be important to you) to ‘god X and all the other gods,’ because in the latter instance the word ‘other’ (which is quite rare in Akkadian) would not be expressed. I cannot quote an example, but to express this idea it would be quite natural for Akkadian to write simply ‘god X and all the gods,’ using one of the phrases given above. In fact, in the line in question [see above, p. 17, n. 18], the word šanūtum ‘other,’ strikes me as being very probably ‘translation Akkadian,’ in which case the whole phrase need not have been a native expression at all.”

VIII (SEE P. 18, N. 29)

R. C. Zaehner's doubts as to the correctness of the usual translation of OPers. daiva-dāna- by “daiva-temple or sanctuary” (The Dawn and Twilight of Zoroastrianis´m [1961], pp. 159, 331) are unfounded. The word has its expected counterpart in ^*baga-dāna-, “god-temple, altar, sanctuary,” which survives in Armenian, Bactrian, and Sogdian, as well as, indirectly, in the name of the town of Baghdad; see W. B. Henning, BSOAS, XVIII (1956), 367, who refers inter alia to Sogdian [c]xwδ-βγδ'nyy, “the temple of the Jews = synagogue.” The difficulty raised by Zaehner, viz. that -dāna- normally means “receptacle” (from the base dhā-), which, as he says, “seems a very odd way of referring to a temple,” can be overcome by an analysis of the word in these two compounds as a derivative of the base dam-, “to build” (cf. Asia Major, New Series, II, 136), with the expected meaning, “building, house, structure”: dāna- (IE ^*dmHno-) would then stand to dmāna-, “house,” (IE ^*dmeHno-), as Av. gata- (IE ^*g^wmto-) stands to OPers. gmata- (IE ^*g^wmeto-).

IX (SEE P. 24, N. 40)

It has long been realized that in calling the Persian Aphrodite Mιτραν (i. 131), Herodotus confused Anāhitā with Mithra. If the fault was his and not of his informants, it would seem that not only had he failed to record the name of the Persian Aphrodite, but the notes which he brought back did not contain a clear statement on the function of Mithra, as an individual god, either. The latter deficiency would hardly have occurred, had his informants told him that they regarded Mithra as the Persian sun-god (cf. E. Benveniste, The Persian Religion, p. 27), as indeed Mithra scarcely was at such an early period (cf. AHM, pp. 38, 41 f.). What Herodotus might have been told is that the Persians worshiped a god of the sky who was sometimes referred to as Ahura (this being the un-Zarathuštrianized Iranian descendant of the Indo-Iranian Asura, see above, p. 12, n. 1) and sometimes as Mithra-Ahura. The latter, dual, deity, being the equivalent of the Vedic Mitrā-Varunā and the Younger Avestan Mithra-Ahura, can confidently be supposed to have been worshiped by the Persians from the earliest times, even though it is named in Western sources only from Plutarch onwards, as Mεsορομάsδηs and Mίθραs ’Ωρομάsδηs (cf. AHM, pp 4 f., 44 ff., 320, bottom). We have seen in the last paragraph of Appendix III that this dual deity, like Ahura alone, may have been thought of as being clad in the sky.

If then Herodotus says that the Persians call the whole circuit of the sky “Zeus” … and “Zeus” is his rendering of Ahura, we may surmise that his notes also contained a reference to ^*Mιτραν Δία as the dual deity of the sky. Being unfamiliar with this Indo-Iranian type of dual deity and consequently finding himself at a loss, back at home, to understand what the name Mίτραs was doing in front of what he had written down as “Zeus,” and noticing, moreover, the vacancy in his notes for the Persian name of Aphrodite, he filled the vacancy with the redundant Mίτραs, possibly under the misapprehension that the juxtaposition ^*Mιτραν Δίαν referred to “Zeus” and his (Persian) spouse. On this assumption, a confirmation that Mithra had indeed been described to Herodotus as Ahura's partner in the rule of the sky may be found in the specification, in the passage in question, of Aphrodite as “Urania”: “In addition, they learned from the Assyrians and Arabians to worship Urania. Aphrodite is called Mylitta by the Assyrians, Alilat by the Arabians, Mitra by the Persians.”

X

There is no reference to the Magi in the extant books of the Avesta, with the possible exception of the hapax moγu.tbiš-. Even if this compound meant “hater of the Magi,” as Bartholomae and Schaeder (OLZ [1940], 376 f.) thought, it would hardly have been quoted by the Magi and should not be quoted by us, as a proof that the composition of the Younger Avestan scripture, far from having taken the Magi by surprise, was wholly or partly the work of Magian priests. E. Benveniste, Les Mages dans l’ancien Iran, pp. 18 ff., has attractively argued that the term magu- denoted a particular social class in the proto-Iranian language, and continued to do so in Avestan. One might then suppose that in Media this social class became the priestly caste, so that the Persians, if within their own class distinction the term magu- had fallen into disuse, would know the word in no other sense than that of “a member of the Median priestly caste.” At all events the absence of Media and Persis from the lists of countries quoted in the Younger Avesta (cf. Appendix XI) suffices to refute the theory that the Magi were its authors. As to R. C. Zaehner's hypothesis that “the sacerdotal caste of the Magi was distinct from the Median tribe of the same name” (Dawn and Twilight, pp. 163 f.), this rests on a supposed identity of meaning between Old Persian magu- and Gathic Avestan magavan-. Even if this identity were granted, the absence of the term magavan- from the Younger Avesta would still militate against the identification of its authors with the Magi.

XI

It is invariably assumed that the country which in the Younger Avesta is called “the Zarathuštrian Raγa” is identical with the country Raga which Darius defines as being in Media, the medieval Rai. This assumption ill agrees not only with the absence of any reference to Media in the Avesta but also with the place which Raγa occupies in the relevant part of the list of countries in the first chapter of the Vendidad. There the succession is: Harax^vaitī (Arachosia)—Haētumant (Hilmand)—Raγa—Čaxra (possibly the Ghazna region)—Varsna (Buner, see W. B. Henning, BSOAS, XII [1947], 52 f.)—Hapta Hsndu (“the Seven Rivers,” that is, the Sapta Sindhavas of the Rig Veda, which included the Punjab), cf. A. Christensen, Le premier chapitre du Vendidad, pp. 36-54. The use of identical toponyms in different parts of Iran was common at all times. Suffice it to refer to the eastern mountain Harā Bsrsz of the Avesta (cf. also Khotanese Haraysä, H. W. Bailey, Khotanese Texts IV, 12), “across which the sun comes forth” (Yašt 10.118), against the western Alburz, or to the Avestan country Nisāya, which is described as lying “between Margiana and Bactria,” against the country Nisāya “in Media” of the Behistun inscription. Such homonymity is natural with descriptive toponyms: Harā Bsrsz means “high watchpost” (Av. har- “to watch,” cf. Yašt 10.51), Nisāya, “region of settlements,” Raγa, “plain, hillside,” etc. (cf. W. B. Henning, BSOS, X, 95, sect. 5, and W. Eilers, Archiv Orientální, XXII [1954], 301). The difference in the inflection of Ragā in Old Persian (ablative Ragāyā) and Avestan (ablative Rajōit) should also not be lost sight of. As the Avestan Raγa had been Zarathuštra's see (cf. AHM, p. 265), the homonymity of the two regions was bound to encourage the belief among Western Iranians of later generations that the scene of the prophet's activity had been in Media. We, by contrast, shall be more inclined to find the eastern location of the prophet's see confirmed by the consideration that the Gathic variant of the clearly Eastern Iranian Avestan language may well have been the vernacular of the “Zarathuštrian Raγa.”

XII

I assume with Benveniste, The Persian Religion, p. 16, that “it is in Plato, in the First Alcibiades (I, 121), the authenticity of which has been wrongly contested (written about 390 b.c.) that appears the first definite mention of the name of Zoroaster in Greece,” cf. Bidez and Cumont, op. cit., vol. II, 21 f. The reference to Zoroaster which Diogenes Laertius attributes to Herodotus' contemporary, Xanthus of Lydia, is of doubtful authenticity, cf. Bidez and Cumont, op. cit., vol. II, 9, top, and J. Marquart, Philologus, Suppl. VI, 531, 608, n. 353, end. The emigrated Magi … of Lydia, who would have been the most likely informants of Xanthus, were in any case scarcely Zoroastrians, as they seem to have been especially connected with the cult of Anāhitā (ibid., p. 5). To them, and generally to the μαγουsαιοι of Asia Minor, Zoroaster need have been no more than a legendary sage of great eminence. The alleged Xanthus quotation therefore makes sense at least insofar as the μαγουsαιοι were quite likely the source of the phantastic chronology (however cleverly linked with World-year speculation) according to which Zoroaster was placed at 6000 years before either Plato (thus Pliny, who probably wrongly attributed this information to Eudoxus of Cnidus, see ibid., vol. I, 12, n. 1, II, 11, n. 5, and above, p. 15, n. 13), or the expedition of Xerxes (thus the alleged Xanthus quotation, cf. ibid., vol. II, 8, n. 4). The Zoroastrian Magi, by contrast, were well acquainted with the date of Zoroaster, to judge from a tradition preserved by the Sasanian Zoroastrian priests, according to which 258 years had been counted from that date until the beginning of Alexander's rule over Iran in 330 b.c. (see W. B. Henning, Zoroaster, p. 40).

XIII

The discovery in the Twenties of the Middle Parthian form of the prophet's name, Zar(a)hušt, which goes back to the Avestan form Zaraθuštra, seemed to confirm Bartholomae's view (Gdr. d. Ir. Phil., I, 1, p. 39) that Zωροάsτηρ was a rendering of an Iranian pronunciation ^*Zarahuštra, cf. W. B. Henning, Deutsche Literaturzeitung (1932), 830. However, while the change of intervocalic θ to h is the rule in Middle Parthian, its taking place in this name at the Old Iranian period would be unparalleled. Already Hübschmann, Zeitschr. f. vgl. Sprachf., XXVI (1883), 604, concluded from the Old Persian form of Darius's name, Dāraya-va(h)uš, which goes back to an earlier ^*Dārayat-vahuš, that the proper Old Persian equivalent of Avestan Zaraθ-uštra (itself a secondary development from ^*Zarat-uštra) would have been ^*Zara-uštra, which form he regarded as the source of Greek Zωροάsτηρ. Since ^*Zarah-uštra is a phonologically improbable form in any Old Iranian language, while ^*Zara-uštra is predictable in Old Persian, we must reutrn to Hübschmann's explanation.

That ^*Zara-uštra was not also the Median form of the prophet's name appears from the Middle and Modern Persian form Zardušt (in Manichean Middle Persian Zardrušt < ^*Zarduršt [whence by dissimilation Zoroastrian Middle Persian Zarduxšt] < ^*Zarduštr), which can only have had as Old Iranian antecedent ^*Zarat-uštra. As the last form disagrees both with OPers. ^*Zara-uštra and Av. Zaraθ-uštra, we must conclude that it was Median. We thus incidentally recover a detail hitherto unnoticed of Median phonology: the preservation without change of a proto-Iranian postvocalic t at the end of first compound terms, and therefore, presumably, of any final postvocalic t.

The original meaning of the name may have been “camel-driver,” see H. W. Bailey, Trans. Philol. Soc. (1953), 40 f. For the contamination with Greek asτeρ cf. Bidez and Cumont, op. cit., vol. I, 6, and Henning, loc. cit.

Notes

1. See Appendix I, p. 32.

2. On the dual role of Truth in Zoroaster's system cf. AHM, p. 45 f. One may conjecture that the two components of the name of Zoroaster's god reflect the two religious systems which the prophet had syncretized, one in which a chief god, Ahura, guarded primordial Truth against primordial Falsehood, the other in which a sole god, Mazdah, had Truth as one of his aspects or emanations.

3. On the name “Immortal” see Appendix VI, p. 34. I retain here the usual translation “holy” of spsnta-, although the meaning of the adjective would be better rendered by a term conveying the idea of increase, such as “incremental.”

4. The Gathic “twin” passage is discussed in Appendix II, p. 32.

5. See Appendix III, p. 33.

6. For clarity's sake I translate dazdē twice, since its meaning is both “created” and “established.”

7. I agree with Humbach's interpretation of the third line and of the verb for “to meet” as implying hostility (ZDMG [1957], p. 369). His rejection of the traditional understanding of dazdē as 3rd dual perf. is, however, unjustified and makes nonsense of the first two lines. Av. dazdē stands to Ved. dadhāte as Av. daiδītsm to Ved. adadhātām, the reduced grade of ā being in the first case zero, in the second ī. The meaning is precisely the one which Humbach thinks the form should have, but denies it: “sie schufen für sich” (i.e. the two Spirits created respectively, each to further his own aim, life and not-life, paradise and hell). The 3rd dual middle jasaētsm may be injunctive or optative.

8. Cf. the Younger Avestan statement that “the two Spirits, the Holy and the Fiendish, created the creation” (Yasna 57.17, Yašt 13.76).

9. Cf. W. B. Henning, Zoroaster, p. 46. To the Gathic passages that illustrate the prophet's missionary aim, Yasna 28.1 may be added in the following revised translation:

   ahyā yāsā nsmaehā
   ustānazastō rafsδrahyā
   manysuš mazdā pourvīm
   spsntahyā ašā vīspsng
                       šyaoãnā
   vaehsuš xratūm manaehō
   yā xšnsvīšā gsuščā
                       urvānsm

   “I appeal firstly, O Mazdāh, with hands outstretched in prayer, to all (men) of (= who side with) the Holy Spirit with (= and) Truth, for actions of support of him (ahyā, viz. the Holy Spirit) with whom I am anxious to satisfy the will of (the Holy Immortal called) Good Mind, and the Soul of the Cow.”

   Here the meaning of xratu- approaches that of “injunction,” which is attested for a much later period in Armenian xrat, Middle Pers. xrd (see W. B. Henning, Trans. Philol. Soc. [1944], p. 114), and Khotanese grra, grata (see H. W. Bailey, BSOAS, X [1942], 901).

10. Between the Gathas and the Younger Avesta proper lies a group of seven brief texts known as the “Yasna of the Seven Chapters,” which may be as early as the sixth century. They contain no dualistic statements but are rich in polytheistic ones. Their authors worship not only Ahura Mazdāh and his Aspects but also fire, the waters, wind, earth and sky, the Haoma-plant, the souls of animals and pious people, a mythical donkey, cornfields, cross-roads, and much else. The great gods of the Younger Avesta, however, such as Anāhitā, Tištrya, Vsrsthraghna, Vayu, do not appear in these texts; neither does Mithra, unless he is to be identified with the unnamed “Protector” of Yasna 42.2 (cf. AHM, pp. 54 ff.). What therefore makes the Seven Chapters less un-Zarathuštrian than the texts we have called Zarathuštric, is that apart from a possible half-hearted concession to Mithra, their authors did not admit as objects of worship any personal gods acting in rivalry with Ahura Mazdāh and detracting from his omnipotence.

11. On the incompatibility of this formulation with statements found in the Gathas cf. Appendix III, p. 33.

12. See J. Bidez and F. Cumont, Les Mages Hellénisés, II, 9, 67²⁶. Aristotle (384-322 b.c.) wrote the περi pιλοsοpίαs after Plato's death (347 b.c.), at the same time as part of the first book of the Metaphysics, see W. Jaeger (transl. R. Robinson), Aristotle, p. 128. As the above information on the Magian doctrine is credited by Diogenes Laertius both to Aristotle and to Plato's friend Eudoxus of Cnidus (408-355 b.c.), who was an expert in Oriental lore, it is likely that Aristotle learned it from Eudoxus.

13. See Bidez and Cumont, op. cit., Vol. II, 71.

14. See C. Clemen, Fontes historiae religionis persicae, pp. 3-16, especially 5-7.

15. See E. Benveniste, The Persian Religion, pp. 25 ff. The religion described by Herodotus can be defined grosso modo as the residue obtainable by subtracting the Zarathuštrian elements from the Zarathuštricism of the Younger Avesta. Even after the subtraction, however, an important difference remains, in that Herodotus not only does not mention Haoma, but expressly states (i. 132) that the Persians “pour no libations” (although, according to Book VII. 43, the Magi did, at least to the Trojan artāvan's). Cf. also below, p. 24 and, on Haoma, n. 43.

16. Old Persian version, col. IV, lines 60 f., 62 f.

17. Elamite version, col. III, lines 77 f., 78 f. In the Akkadian version (lines 103 f.) there is a gap after “Ahura Mazdāh bore me aid, and the other gods […,” cf. Appendix VII, p. 34.

18. The gloss does not, of course, exclude the possibility that the expression “the other gods who are” may allude to Iranian gods that had not come to the notice of the Elamite translator. This is in fact what I believe to be the case. See below.

19. Cf. Appendix IV, p. 33.

20. Auramazdām patiyāvahyaiy, Beh. Inscr. i, 55, “I prayed to Ahura Mazdāh.” Kent translated “I besought help of A. M.,” but cf. W. B. Henning, ZII, IX, 174, line 8, and the Akkadian translation (line 22) us-sal-la (see Benedict and von Voigtlander, JCS, X [1956], 4).

21. See Appendix V, p. 34.

22. To a further argument, viz. that Darius fails to mention the Fiendish Spirit, Appendix V may serve as answer.

23. See Appendix VI, p. 34.

24. See Appendix VII, p. 34.

25. For Darius's use of the word “other” in this connection, note the Younger Avestan expression “Ahura Mazdāh and the other Amsša Spsntas,” cf. AHM, p. 12, n. The Younger Avestan word for “god,” yazata-, is applied to the Amsša Spsntas in Visprat 8.1 and 9.4.

26. Cf. Zarathuštra's wording in Yasna 50.4: “I worship you (plur. ad sensum), Ahura Mazdāh, with (hadā) [the three Amsša Spsntas called] Truth, Best Mind, and Power.”

27. Cf. the phrase “all the Amsša Spsntas” in the Seven Chapters (Yasna 42.6), where “all” is clearly pleonastic.

28. See Appendix VIII, p. 35.

29. Cf. F. B. J. Kuiper, Indo-Iranian Journal, IV (1960), 185 f., who rightly points out that the sense in which Xerxes uses artāvan must go back to Indo-Iranian times (as I had failed to realize in AHM, pp. 153-55), but draws in my opinion the wrong conclusion.

30. Provided, of course, that Oss. (i)dāwag does go back to ^*artāwa-ka-, see BSOAS, XVII (1955), 483-84. The alternative explanations of (i)dāwag, as proposed by G. Dumézil, V. I. Abayev, H. W. Bailey (cf. Hommages à Georges Dumézil (1960), pp. 7, 11), and E. Benveniste (Études sur la langue ossète, p. 133), fail to satisfy me. I remain convinced that this is a religious term of substance, like the izad (from yazata-, “god”) with which it is constantly associated in Ossetic literature, and that, as in the case of Sogdian xatu (see AHM, pp. 240 f., 329) a phonetically acceptable paper etymology will not by itself suffice to account for it. A derivation from ^*vitāwa(ka)-, as Abayev proposes, meets in any case a serious obstacle in the unsuitable meaning of Sogdian witāw-, “to endure, persevere,” as was pointed out already in BSOAS, XVII, 484, n. l. The realization that the restriction of artāvan to the dead is a pre-Zarathuštrian notion makes it unnecessary to see in the Oss. (i)dāwag's any Zoroastrian influence, and thus disposes at least of the conceptual objection which has been raised against my interpretation of these spirits.

31. Xerxes' statement runs: “The man who … worships Ahura Mazdāh … will become both šiyāta (= happy) while living, and artāvan (= one who has acquired Truth) when dead.” The translation of the Darius passage as restored in AHM, p. 250 f., reads: “Whoso worships Ahura Mazdāh [, Truth]h will fo[rever] be [his,] both (while he is) alive and (after he is) dead.” Here everything hinges on the restored word [arta]m, “[Trut]h,” whose m is merely an inflectional ending that is applicable to a large number of other nouns as well. The attested Old Persian abstract noun corresponding to the adjective šiyāta in the Xerxes passage, is šiyātiš (nom. sg.), “happiness,” which suits neither the gap nor the legible m. If one insists on the restoration “[happines]s,” an Old Persian synonym of šiyātiš of suitable length and ending will have to be invented, in which case the use which one would be forced to attribute to Darius of a different word for “happiness” from the one Xerxes had in mind, would be a valid objection to any restoration that would produce such a meaning.

32. See W. B. Henning, Trans. Philol. Soc. (1944), pp. 108 f., 117.

33. Ahura Mazdāh and Anāhitā are not referred to in the calendar by these names, but respectively as Ḑͣθušō, lit. “the Creator,” and Apam, lit. “the Waters” (of which Anāhitā was the goddess).

34. These twelve divinities account for the names of the months, but they also occur as day-names.

35. Names, that is, to choose from the day-names of the Younger Avestan “Zoroastrian” calendar, such as “Discipline” (Av. Sraošahe), “Reward” (Av. Ašōiš), or “(Soul of the) Cow” (Av. Gsuš (urvanō)).

36. In some provinces with strong traditions of their own, such as Sogdiana, the terminology introduced by the central government need in any case not always have prevailed in every detail over the existing local calendar names.

37. The above discussion and conclusion are offered as an attempt to close the loophole which I had felt obliged to leave in AHM, p. 18, n.

38. Cf. J. Duchesne-Guillemin, The Western Response to Zoroaster, p. 54 f., who infers from the calendar names that Artaxerxes I was a “Zoroastrian,” without, however, drawing a distinction, vital in the present context, between Zarathuštrianism and Zarathuštricism.

39. See Appendix IX, p. 35.

40. See W. B. Henning, JRAS (1944), 133 ff.

41. The circumstances in which Darius may have embraced Zarathuštrianism have been considered in AHM, p. 15 f. The picture would be somewhat different if the first Achaemenian convert to Zarathuštrianism were Cyrus, as has been suggested by W. Hinz, Zarathustra, pp. 146-52.

42. The definition of the calendar as “Magian” agrees with the fact that Haoma's name appears neither among the calendar names nor in Herodotus (cf. above, p. 16, n. 16), whose description, moreover, of the part played by worshiper and Magus at the sacrifice (i. 132) shows that no Haoma ritual was involved. Zarathuštric priests, by contrast, if the selection of the Younger Avestan calendar names had been left to them, would scarcely have failed to include the patron-god of their own class, whom they regarded as the divine priest of Ahura Mazdāh (Yašt 10.89 f.) to whom Ahura Mazdāh himself had entrusted the Mazdayasnian Religion (Yasna 9.26).

43. This tradition, of which the surviving “Yasna of the Seven Chapters” (see above, n. 11) affords us a glimpse, must have arisen in response to the need for simpler scriptural texts than the Gathas, on a level with the hymns and liturgies which the priests of the pagan Eastern Iranian communities continued to recite. These priests, like the Magi in the West, cannot readily be credited with a traditional literary skill in the second half of the fifth century. The texts which they recited had in all probability been composed centuries earlier, and the only effort required by the priests consisted in memorizing them. It should be understood, incidentally, that the terms “scripture,” “texts,” “literary activity,” etc., are a mere pis aller in the present context, as the Avesta had scarcely as yet been committed to writing in the fifth century.

44. See Appendix X, p. 36.

45. See Appendix XI, p. 37.

46. See Appendix XII, p. 37.

47. See Appendix XIII, p. 38.

48. It was scarcely from a Magian source that Hecataios of Abdera elicited towards the end of the fourth century b.c. the Avestan name of the prophet … and the true country of his activity, Aryana Vaējahh … the homeland of the Avesta, see Bidez and Cumont, op. cit., vol. II, 30 f. This information was ignored by the Greeks, no doubt because they believed the Magian version to be more authoritative.

49. Even 780 years after Darius had referred to Gaumāta as “a Magian man” (martiya maguš, Behistun I, 36), Šāpūr I in his Great Inscription at Naqš-i Rustam still called his priests mgw GBR'. …

50. Cf. Yašt 5.94, 9.25 f., 16.6, and the crude and deficient account of Zarathuštra's achievement in Yašt 13.88 ff., where one looks in vain for the essential tenets of the Gathas, let alone for an echo of the prophet's lofty vision of God's designs and man's destiny.

51. That the Magi may have been led to such a conclusion by acquaintance with the dualistic religion of the “Harrānians” is a possibility which, in view of Hildegard Lewy's most stimulating article on the subject (The Locust's Leg, Studies in honour of S. H. Taqizadeh (1962), pp. 139 ff.), deserves to be borne in mind.

52. Although the Iranian Manichees of the third century a.d. no longer had the technical term arta for Zarathuštra's Truth, they were still aware of what had mattered more than anything else to the prophet. In the words of a Parthian Manichean hymn-writer, “Zarhušt descended to the kingdom of Pārs, and showed truth (rāštēft)” ('wsxt zrhwšt 'w p'rs šhrd'ryft 'wš nm'd r'štyft, Andreas-Henning, Mitteliranische Manichaica, III, SPAW [1934], 879, lines 19 ff.).

Revised text of a lecture delivered at the Oriental Institute of the University of Chicago on 18 May, 1961. Several paragraphs have been replaced by references to my book, The Avestan Hymn to Mithra, here abbreviated as AHM. I have, however, abandoned the reconstruction proposed in that book, of an Avestan divine name ^* (Ahura) Vouruna. Whatever I wrote about the god ^*Vouruna should be understood as applying to a god called simply Ahura. This means a return to A. Hillebrandt's derivation of both the Iranian Ahura (who partly survived in Zoroaster's Ahura Mazdāh) and the Vedic Asura (who was generally referred to by his epithet Varuna) from an Indo-Iranian god whose name was Asura without further qualification: see WZKM, XIII (1899), 320, ZII, IV, 212, and his Vedische Mythologie (2d ed., 1929), II, 9; cf. also P. Thieme, JAOS (1960), 308 f. Hillebrandt's theory accounts better than any other for the co-existence of a close resemblance on the one hand, and a basic difference on the other, between the Avestan Ahura Mazdāh and the Vedic Āsura Varuna, cf. AHM, pp. 44 f., 321. A further divergence from AHM, which arises from an observation of F. B. J. Kuipers (see n. 30), will be found below, pp. 18 ff.