42 SOCIAL SCIENTIST
world-view and the Indo-European pantheon was a projection of the tripartite class system, which later hardened as varna in India and pistra in Iran. A critique of this theory was provided by John Brough4 who showed that a similar threefold ordering of the social order into priests, warriors and cattle-herders may be seen in Semitic societies portrayed in the Old Testament. More recently, Bruce Lincoln5 has shown that the Nilotic tribes of East Africa too have a similar three-fold division with priests enjoying a hierarchical superiority over the warriors. His basic argument that the roots of the separation of the priest and warrior elements lie in the ecology of cattle-keepers is quite plausible, although it is evident that this separation is only functional at this stage and is not lineage-based or genealogically determined. Even Emile Benveniste,6 who agrees with Dumezil in tracing the tripartite social division to fedo-Europeans clearly states that these were functional divisions not 'political* or 'genealogical';
these were not kin-based. In his view the Indo-European social units were family, clan, tribe and country in the manner of concentric circles, but there was no uniformity in this regard; each group of Indo" Europeans developed these institutions independently. Nevertheless, Benveniste like Dumezil traces the dichotomy of priest and warrior and the notion of hierarchy of social orders to the Indo-European phase.
But the idea that those who deal with the divine are superior to those who have temporal tasks is nothing unusual or typical of any one culture. Varna ideology is much more than a recognition of three sppal categories to which a fourth one, that of the sudras, was added due to the historical circumstance of the Aryans confronting and subjugating the non-Aryans in vedic and post-vedic times. The basic question is:
what led to the origin of an endogamous caste structure which derived its legitimacy from the vedic notion of a hierarchical grouping of occupational groups, when there is no trace of endogamous lineage-clusters among the Indo-Europeans or in the Rigveda. This is the question which puzzled D.D. Kosambi. He pointed out7 that the Yajurvedic four varnas were quite different from the four classes mentioned in the ancient Iranian sources, namely, the priest, the charioteer, the tiller and the artisan. Endogamy is nowhere mentioned in the Yasna and in ancient Iran all the four classes were equally honoured, but this was not the case with the ancient Indian varna or caste organization, which had both endogamy as well as hierarchy. According to Kosambi an internal fourfold caste system among the Aryan tribes in India developed due to the assimilation of the survivors of the Harappa culture with the conquering Aryans. The subjugation of the Harappan agrarian population, identified as the Dasas of the Ri^veda by Kosambi, formed the nucleus of the Dasa/sudra varna, but the adoption of the 'ritually superior' priesthood of the Harappa culture by the Aryan tribes proved