AN effort definitely to establish the date of the reform of Iranian religion by Zoroaster has been made by J. Hertel,1 who seeks to trace the activity of the prophet from 559-522 B.C., a result of much importance in view of the probability that no very great interval of time can intervene between the Rigveda and the Gathas of the Avesta.
Stress is naturally laid by Hertel on the fact that the religion of the Persians as described by Herodotos contains nothing specifically Zoroastrian, while it accepts much that Zoroaster rejected, especially the direct veneration of nature gods, and the animal sacrifice; moreover Herodotos records the drunkenness of the Persians, which may be connected with the Haoma offering, the fear of defiling fire or water, and the practice of the Magoi of permitting the bodies of the dead to be torn by birds or beasts, matters which are not recognized in the Gathās. Hence it may be concluded that Zoroaster’s activity fell at no great date before the reign of Xerxes I. Nor can we form any reasoned judgement to the contrary on the strength of the view of Xanthos of Lydia, a contemporary of Artaxerxes I (465-424 B.C.), who places Zoroaster 6,000 years 2 before the expedition of Xerxes, or the slightly variant account of Aristotle which makes him 6,000 years before the death of Plato.3
The Persian inscriptions show Darcios I as a believer in Auramazdā, who is clearly the god of Zoroaster, not the sky god of the nature worship of the Persians, whose name Herodotos heard under a form akin to the Indian Dyaus. The devotion of Darcios to this god can be accounted for only by the view that he was the deity of his branch of the family of the Achaimenidai. Dareios had to struggle against the hostility of the Magoi, representing the traditional faith, and it was his earnestness under the influence of the prophet’s teaching, which secured the sudden predominance of the Zoroastrian view, despite its abstract character and lack of popular clements. But this means that Zoroaster had lived just before Darcios or was still alive, for if he had lived some centuries before, and his religion had still at the time of Darcios been of so small account, it could never have suddenly expanded to important dimensions. This view, in Hertel’s opinion, is strengthened by the terms used by Dareios himself, when he asserts that the carlier kings had failed to accomplish what in one year he had wrought by the aid of Auramazdā, and when he advises his successors to punish the adherents of the lie. His successors omit this warning, which was only in place when a new god was being first proclaimed. Under Artaxerxes II (404-359 B. c.) we find Auramazda worshipped along with Mithra and Anahita, showing the contamination of Zoroastrianism with the older nature worship under the influence of the Magoi, who had accepted the duty of preaching the new faith, but naturally permeated it with ideas of the older belief.
1 Die Zeit Zoroasters (1924).
There is a variant 600; Moulton, Early
Zoroastrianism, p. 412 (a work unknown to Hertel). It is preferred by
Rapp, ZDMG. xix. 25; Maspero, Passing of the Empires, p. 572, n. 8. Pliny, N.H. xxx. 2. 1; sec Jackson,
Zoroaster, pp. 152 ff.
The Age of the Avesta and the Rigveda
615
Ingenious as are the arguments of Hertel, it is clearly impossible to accept them as adequate to prove his thesis that Zoroaster was an earlier contemporary of Dareios I. The attitude of the king is certainly that of the adherent of a faith who believes strongly in his god; it does not prove that the faith was a new one; it might well have existed for some centuries in comparative obscurity, to be brought into prominence by the moral earnestness of a king who shared the spirit of the founder of the reformed faith and of his successors. The evidence a priori is simply negative in direction, and the conclusion drawn by Hertel is wholly uncertain, unless it can be supported by other evidence.
Such evidence Hertel finds in the Parsi tradition which gives something like 595-594 B. C. as the opening of Zoroaster’s ministry and makes him live about 660-583 B.C. The date is earlier than his own, but in view of the deplorable nature of the Parsi tradition generally, he deems this of no importance, and refers to the fact that Anquetil du Perron1 arrives at the date 559 B. c. for the beginning of Zoroaster’s work, on the strength of the fact that a religious sect, which immigrated into China about A.D. 600 and which is evidently of Zoroastrian origin, had an era dating approximately from that time, which Anquetil interpreted as fixed by the prophet’s leaving his home and entering on his mission. For the accuracy of the tradition the best argument to be adduced is that, if Zoroaster had not lived at a comparatively later epoch, the Parsis would not have brought his date down, since no sect willingly depreciates the age of its founder. Here, however, we are left to mere guess-work, and we have a very formidable argument to confront in the fact that tradition does not make the patron of Zoroaster, Viśtāspa, identical with the father of Dareios until as late as Ammianus Marcellinus (5th cent. A.D.), whose evidence in this regard, Hertel very frankly admits, is worthless.
The later Avesta in fact and the Parsi tradition ascribe the patron of Zoroaster to a Kavi dynasty, founded by a certain Kavata; moreover it is silent regarding the great Persian kings who favoured the faith. It also, like the Greek tradition, gives Zoroaster the position of one of the Magoi. In Hertel’s view the dynasty of Kavi is a pure invention, Kavi Viśtāspa in the Avesta really denoting mercly the king 2 Viśtāspa; the omission to mention the Achaimenidai is due to the opposition of Dareios to the Magoi, and the connexion of Zoroaster with that body is mythical. On the contrary, in vv. 8 and 9 of the Gatha (Yasna, liii.) Hertel finds a direct encitement to Viśtāspa to overthrow the usurper Gaumāta and the Magoi, the passage dating shortly before Dareios acted on the hint in lieu of his father, and brought about by treachery the destruction of the Magoi. This suggestion, however, scems wholly implausible. The verses form part of the utterance of the prophet at the narriage of his daughter Pourucista; what he says naturally applies generally to evil-doers and to their punishment by the Lord, and it is illegitimate to read into them an admonition, carefully disguised from motives of prudence under a double entendre in the later Indian style.
4
Hertel naturally rejects the argument in favour of an earlier date of Zoroaster derived from the occurrence of Mazdaku as a royal name in Media
1 Jackson, op. cit., p. 165.
2 But Viśtāspa does not appear as a king
in the inscriptions; Herodotos (iii. 70) makes him only mapxos in Persia.
3 So, for very different reasons, Moulton,
op. cit., pp. 116, 197.
• Meyer, KZ. xlii. 16. See Moulton, op.
cit., pp. 30 f., 422 ff.
Hertel ignores
C. Clemen’s important work, Die griechischen und lateinischen Nachrichten ūber die persische Religion (Giessen, 1920), where the date, not later than 1000 B. c., is contended for, and where (pp. 54-77) a fair case is made out for the view that Cyrus was a Zoroastrian; cf. Keith, JHS. xl. 282.
616
Appendix A
about 715 B. C., on the ground, also adopted by Jackson, that the name need not be derived from Mazdah as a divine name, but merely from the word in its signification of wisdom, parallel to Sanskrit medha. The occurrence of Assara Mazāś as a divine name in an inscription of the seventh century B. C. probably, in view of its form, carries back the existence of Ahura Mazdah to a period before the life of Zoroaster, so that, even if the royal name were theophoric, it would prove nothing for the date of Zoroaster.1
It remains, therefore, impossible to establish with any certainty Zoroaster’s date. The views of Hertel rest on a priori reasoning so far as they have any validity; the tradition which supports them as regards date contradicts them on the vital point of ancestry, and there are certain considerations of high importance to be set against his contentions. In the first place, we find Greek reports as early as the fifth century B. C. placing Zoroaster in the very remote past; in the second place, the younger Avesta already treats Zoroaster as a mythical figure. It is very difficult to reconcile these facts with the theory that Zoroaster was alive as late as 522 B.C., and it must be recognized that even the Parsi tradition declines to place him as late as that, for it is out of the question to take seriously the suggestion of Anquctil, on which Hertel relies for his upper date of 559 B. c. as the beginning of the work of the prophet.
It seems, therefore, that the date of the Gathas of the Avesta must remain unascertained; it remains, however, to consider the value of the evidence adduced by Hertel regarding the position of the Rigveda. He finds in that text evidence of hostility to the tenets of Zoroaster openly expressed, in the condemnation of the brahmadvis2 and the devanid. The doctrine of the Brahman was held by persons who insisted on negating the pre-existence and future existence of each individual; the believers in the Daena of Zoroaster, for which the younger Avesta has Fravaśi, held opposing views, and thus can be censured as brahmadvis. Similarly the Zoroastrians were the only persons known to us who degraded the gods of the Aryans to demons. The answer in either case seems irresistible; the devanid of the Rigveda is most naturally and normally applied to the non-Aryan focs of the Vedic Indians, and the brahmadvis is even more obviously not to be interpreted in the strange way suggested by Hertel, whose assertion that the believers in the doctrine of Brahman did not accept an individual pre-existence and future existence of the soul is a pure hypothesis, quite unsupported by anything which we know of any historical Brahman doctrine as opposed to that invented by Hertel,* in which, doubtless, the individual is to be regarded as the result of the streaming out of the cosmic fire, into which he is reabsorbed. Hertel does not state how he connects this view with the obvious belief in transmigration of the Upanisads. Even more remarkable is his assertion that in Rigveda vii. 104. 2 the term brahmadvis is applied to a PiŚãca, the l’iŚacas being a cannibalistic race found in the Indus valley and in the Hindu Kush. Apart from any dubiety as to the correctness of Grierson’s views on P’iŚacas," there is the uncontrovertible fact that PiŚãca does not appear in the passage cited at all, which clearly and obviously applies to a Rakṣas, the hymn being directed. against these demons. Even on Hertel’s own theory it is obvious that it is
"
1 Maspero (op. cit., pp. 455, 572) argues that the name l’hraortes denotes Fravartiś, confessor of the Zoroastrian faith, which, of course, would give an early date as Deiokes’ father bore the name (Herodotos, i. 96), but the etymology is dubious (Moulton, Early Zoroas-
trianism, pp. 269 ff.).
2 Die Zeit Zoroasters, p. 62.
IF. xli. 200. The term occurs only in RV.
i. 152. 2; vi. 61. 3.
• Ibid. 185 ff. Cf. above, Chap. 27, § 2. Cf. above, p. 73.
The Age of the Avesta and the Rigveda
617
unlucky that the Rigveda should apply the same term brahmadvis to the Zoroastrians and to a race of eaters of dead bodies, who are severely censured and denounced in the Avesta.1 Nor does he adduce any reason why, even on his own theory, the term should not always apply to PiŚãcas, if it actually once does so.
6
It appears impossible also to derive any results as to the age of the Rigveda or the Avesta from the discovery of Aryan elements in the speeches of Asia used in the Hittite Empire, Sumerian, Babylonian or Akkadian, Kanisian (Hittite), Luvian, Balaian, Charrian, Proto-Hittite, and that of the Manda." That Indo-European elements of centum type may be traced in Kanisian and Luvian is quite natural in view of the obvious activity of the Indo-European tribes. More immediate value attaches to the evidence of the Aryan speech of the Manda tribes on the borders near Mitanni, to whose speech Forrer ascribes the names of Mitanni gods, which as pronounced he holds to have been Midrassil, Ur(u)vanassil, Nasattiyana, and Indar. The first three names are preceded by the determinative An (-Meś) which marks them as denoting groups of gods, which may as regards Mitra and Varuna serve as an indication of their being closely united. The variant Arunassil is regarded as due to influence by the Kanisian word arunas, ocean, a view interesting if it may be taken as indicating a connexion between Varuna and the ocean in the minds of the Manda tribes. As Manda also he classes the terms of horse-breeding and training found in a text emanating from a man of Mitanni; they include in his view, aika-vartanna, one round’, teravartanna or terōrtanna, three rounds’, panzavartanna, sattavartanna, satvartanna, navartanni vasannasaya, in nine rounds of the stadium’, and auzomewa, ‘run’, in which he sees the -wa of an infinitive termination comparable to Vedic -mane and -vane, drawing the further conclusion that in a period prior to the period of Indo-European speech unity there existed a spirant which in Kanisian became a v sound, in Greek and Luvian an m sound. This is much too speculative; what is clear is that the numerals are by no means Vedic and that they can be set down as Aryan with equal plausibility. We are, therefore, still left without any definite evidence to aid us in dating the distinction of Aryan into Iranian and Indian, and we should probably revise our conception of this division. In an area of considerable extent over which Aryan was spoken we may assume dialectical differences sprang up, accelerated in development by contact with different racial elements, and the fragments of Mitanni speech akin to Aryan found of late represent developments of what may conveniently be called Aryan, not either Iranian or Indian. This natural hypothesis removes the need of imagining movements back from India to the west, while the possession of some gods in common well accords with the Aryan character of the speech.
3
Hertel accepts as proved Ipson’s effort 4 to establish the theory that the period of Indo-European unity lasted until the period 2000 B. c. The evidence, however, for this thesis is inadequate; it rests merely on the view that the prototype of the Indo-European variants of the word star 5 could only have been borrowed then. It is, however, obvious that, even if we assume that the word was borrowed, the evidence is quite insufficient to establish either the date or the unity of the Indo-Europeans at the time of the borrowing. What
1 Vend. vii. 23 f.; viii. 73 f.
- E. Forrer, ZDMG. lxxvi. 174-269; cf. J. Friedrich, ibid. 153-73; IF. xli. 369-76; Bloomfield, JAOS. xli. 195- 209; Prince, ibid. 210-24; Sayce, JRAS. 1922, pp. 563 ff.; cf. A. Ungnad,
Die āltesten Völkerwanderungen Vorderasiens (1923); ZDMG. lxxvii. 87 ff.; A. E. Cowley, The Hittites (1918).
- IIQF. ii. 7.
IF. xli. 174 ff.
Greek dorp, Latin stella, OE. steorra,
618
Appendix A
is fairly well established is only the existence in the latter part of the first half of the second millennium B. c. of dialects which may generically be styled Aryan, and which have developed beyond the hypothetical language, which may be regarded as Indo-European.
Peake, who argues for the Indo-Europeans, or Wiros, as of Nordic race, with heads rather narrow than broad, transparent skin, light hair, and grey eyes, settled in the steppes east of Dnieper about 3000 B. C., holds that drought about 2200 B. c. sent some of them to the Iranian plateau where they appear as the Kassites,2 while the Aryas proper, speaking Indo-Iranian dialects, were still living together in Russian Turkestan in 2000 B.C. He places about 1760 B. C., contemporaneously with the Kassite occupation of Mesopotamia, the separation of the Indians and the Iranians. The Mitanni barons he regards as a group of Aryas who spoke a language with Iranian affinities and separated themselves from the main body somewhat after 2000 B. c., rejecting Sayce’s suggestion that they were Phrygians, who entered their territory from Thracian homes in the west. Yet other Wiros established their power in the Hittite Empire," which they reached via Gallipoli. This reconstruction is interesting, but it lacks decisive proof of any kind. The Hungarian home of the Wiros claimed by Dr. Giles he rejects, save in so far as he admits that many of the Wiros may have been descendants of the late Aurignacian and Solutrean horse-hunters, and that they may have developed the rudiments of their language in some post-Solutrean time within the Carpathian ring. The Germanic origin of the Wiros is also wholly rejected, nor despite the arguments of Kossinna 5 and others does it seem tenable.6
Some would fix the Vedic period by first establishing the probable date of the Bharatan war, and then by reckoning of generations on the basis of epic or Purana genealogies dating certain kings who figure in the Rigveda. Thus the war is placed by various modern critics, contrary to tradition, about 950 B. C. or 850 B.C., whence it is deduced that some at least of the personages of the Rigveda flourished no more than a couple of centuries earlier. There is nothing impossible in such a result, but it must be confessed that it is difficult to put any faith on the epic or Purāṇa genealogies, and that the probative value of such reasonings must be admitted to be minimal. Least of all is it 1 The Bronze Age and the Celtic World, chaps.
xii. and XIV. CT. CAH. i. 82-5, where attention is called to their invention of wheeled transport and domestication of the horse.
This view, followed in CAH. i. 553, depends on the equations Buriaś Borcas; Surias Surya;
Maruttaś
6
=Marut; bugas-Slav bogu, Phrygian bagaios, god. But this is all very dubious; Keith, IHQ. i. 15 f. Op. cit., p. 158, where erroneously Wiro gods are ascribed as adopted by the Hittites, among whom Wiros were merely military adventurers (Sayce, JHS. xliii. 48 f.).
CHI. i. 67 ff.; CAH. ii. 28 f. Die Herkunft der Germanen (1911). Cf. C. Schuchhardt, Alteuropa (Berlin, 1919). An effort is made by C. Antran (Phéniciens, 1920) to assign Minoan culture the Caro-Lelegians of Caucasian origin.
• That Indo-European is a composite trade-
to
route language of Baltic-Mesopotamian origin is a mere guess, of an improbable character. That it was an easy language is equally implausible; all extant Indo-European speeches are vastly more simple than the original (see, c. g., Jespersen, Progress in Language).
Cf. Hemchandra Raychaudhuri, Calcutta Review, xiii. (1924), 67-77, who places Parikṣit in the ninth century B.C. When legendary accounts can be checked, their extraordinary inaccuracy is seen;
pseudo-Kallisthenes cf., e. g., the version of the history of Alexander the Great; the Charlemagne cycle with its colossal exaggeration of the reverse at Roncesvalles in the Roman de Roland (a case which should give pause to all arguments of a great Bharatan war); the utter confusion of the legend of Dietrich of Bern (Theodoric of Verona) –see Chadwick, The Heroic Age, pp. 23 f.; the amazing development of
The Age of the Avesta and the Rigveda
619
possible to find any support for the proposed date of the war in Jain tradition, which places Ariṣṭanemi, the second predecessor of Mahavira, contemporary with Kṛṣṇa. This can be made to support the date only by assuming a period of two hundred years between each Tirthamkara. For this idle assumption we have not the slightest evidence of any kind. The Jain tradition gives PārŚvanatha a semblance of reality by placing him 250 years before Mahavira, Ariṣṭanemi it banishes to the remotest antiquity, and we have not the slightest evidence of any kind that any Tirthamkara of that or any other name existed c. 950 B. C. Even the historical existence of ParŚvanatha is totally unproved. But at any rate, since the Jains themselves did not believe that Aristanemi was a figure of comparatively recent history, it is really otiose for us to rewrite their scriptures.