CLASS SYSTEM i. In the Avesta



i. In the Avesta

The evidence for the existence of a highly developed class structure in the community in which the Avestan texts were composed is very slight, and the available information must be culled from sources chronologically as far apart as the Avesta itself (ca. 1200-600 B.C.E.?) and the Pahlavi texts (composed in the 8th-9th centuries C.E., though incorporating material from pre-Sasanian and early Sasanian times). In addition, both the absolute and relative dating of individual Avestan texts is extremely problematic, for they contain no reference to historical events.

The prevailing theory about early Iranian social divi­sions was proposed in the 1930s and has been developed by such scholars as Émile Benveniste and G. E. R. Dumézil, who have interpreted parallels in Greek, Latin, Indian, Old Iranian, and other early Indo-European languages as evidence that Indo-European society and sub­sequently the societies of individual Indo-European peoples were originally divided into three strata (see, e.g., Benveniste, 1969, I, pp. 279-92; Dumézil; Duchesne-­Guillemin, 1962, pp. 170-72). In the Avesta these strata are identified as the estates of the priests (Av. āθrauuan-; see āθravan-), the warriors (Av. raθaēštar-, lit. “standing in a chariot”; see artēštār), and the cattle breeders (Av. vāstriia- fšuiiant-, lit. “cattle-breeding pastoralists”; Benveniste, 1932; Gershevitch, p. 170). For example, in Farvardīn yašt (Yt. 13.88-89) it is said that Zoroaster himself was the chief priest, warrior, and cattle breeder. In one of the inscriptions of Darius I (521-486 B.C.E.; DPd 15-20; Kent, Old Persian, pp. 135-36) the king prays that his country be protected from the (foreign) army (hainā), famine (dušiyāra), and the lie (drauga), which have been interpreted as threatening the respective estates (Benveniste, 1938; idem, 1969, I, pp. 288-89; Kellens). A Scythian legend recounted by Herodotus (Dumézil, pp. 9­10) tells of gold objects falling from heaven: a cup, an ax, and a plough and yoke, possibly referring to the three social divisions (see Duchesne-Guillemin, 1962, pp. 170­-72; idem, 1964, pp. 128-30). A late survival of the threefold division is to be found in the Middle Persian Bundahišn, where the estates are connected with the Prophet Zoroaster’s three sons (Bundahišn, TD 2, p. 235; tr. Anklesaria, chap. 35.56, pp. 300-01; Benveniste, 1932, pp. 118-19; Boyce, Zoroastrianism I, p. 281).

In another Avestan passage (Y. 19.17), however, a fourth estate, consisting of artisans (Av. huitiš; quoted in Škand­gumānīg wizār 1.17, ed. Menasce, pp. 24-25), is also mentioned. This text, which is part of the Bag nask and contains a commentary on the Ahunwar prayer, is written in correct Young Avestan and must therefore be relatively old. It is this fourfold division of society that is normally mentioned in the Pahlavi texts (see iii, below); It also has parallels in India from as early as the 2nd millennium B.C.E. In Rigveda 10.90.11-12, for example, the origins of the four Indian castes are described: When the primal giant man (Puruṣa) was divided, his mouth became the brāhman (priests), his two arms the warriors (rājanyà), his two thighs the mercantile and agricultural population (vaiśya), and his two feet the lowest caste (śūdra). Similarly, in the Pahlavi Dēnkard (ed. Madan, I, p. 429, cf. Škand-gumānīg wizār 1.20-21; ed. Menasce, pp. 24-25) it is said that the body of man is in the likeness of the four estates, with priesthood at the head, warriorship in the hands, husbandry in the belly, and artisanship at the foot. Nevertheless, it remains to be established that the Indian and Pahlavi texts reflect inherited common beliefs, rather than independent developments.

The Gathas, which are among the oldest Avestan texts and were presumably composed by the Prophet himself, contain no clear references to established social divisions. The incomplete and ambiguous sources have thus been interpreted in various ways; most recently Mary Boyce (1982) has argued for a twofold division of Gathic society between warrior-herdsmen and priests.




E. Benveniste, “Les classes sociales dans la tradition avestique,” JA 221, 1932, pp. 117-34.

Idem, “Traditions indo-iraniennes sur les classes sociales,” JA 230, 1938, pp. 529-49.

Idem, Le vocabulaire des institutions indo-européennes, 2 vols.. Paris, 1969.

M. Boyce, “The Bipartite Society of the Ancient Iranians,” in M. A. Dandamayev et al., eds., Societies and Languages of the Ancient Near East, Warminster, Eng., 1982, pp. 33-37.

J. Duchesne-­Guillemin, La religion de l’Iran ancien, Paris, 1962.

Idem, Symbols and Values in Zoroastrianism, New York, 1966.

G. E. R. Dumézil, L’ideologie des Indo-Européens, Brussels, 1958.

I. Gershevitch, The Avestan Hymn to Mithra, Cambridge, 1959.

J. Kellens, “Trois réflexions sur la religion des Achémenides,” Studien zur Indologie und Iranistik 2, 1976, pp. 115-17.

Škand-gumānīg wizār, ed. J. de Menasce as Škand-gumānīk wičār. La solution décisive des doutes, Fribourg, 1945.

(Prods Oktor Skjærvø)

Originally Published: December 15, 1992

Last Updated: October 21, 2011

This article is available in print.
Vol. V, Fasc. 6, pp. 650-651