CASTE: IDEOLOGY AND CONTEXT 5 view of caste organization may be a more or less accurate description of its morphology in modern times, but this does not mean that such a complex caste organization could not have its beginnings in the simpler varna organization. This is not to argue in favour of a linear development; nor to account for the proliferation of jatis within the varna structure in terms of fission of primary; categories through the mixed (varna samkara) marriages in the manner visualized by the Dharmasastra writers; but to assert that whereas varnas emerged in small-scale later Vedic societies as tribes disintegrated with the growth of economic disparities, the jatis; emerged in post-Vedic times when the varna-based society began to expand spacially and demographically. Both the processes of internal fragmentation and incorporation of tribal communities produced the complex structure of jatis within the varna framework. The pattern of the structuring of the wider society was provided by the principles of the varna ideology, which was concerned not only with the grading of statuses but with the perpetuation of hierarchy through insistence on the hereditary nature of occupations and rankings and devaluation of the status of women in-built in the notions of hypergamy and endogamy, the two being two sides of the same coin. ° However, interpreting varna as 'category' or 'estate' and jati as caste, Dumont distinguishes the two on the ground that although there is 'homology' and a certain degree of osmosis between the two systems, their principles are different, as in the former stress is laid on function rather than birth but heredity is more important in the caste system, l^ we may point out that the importance of the factor of heredity hi the varna scheme is obvious from its linkage with theory of gunas, the three basic qualities namely satya (goodness or purity) rajas (passion or glory?)^0 anc| tamas (darkness, ignorance). The Bhagavadgita clearly says that owing to their natural, inborn, qualities (gunas) the four varnas have been assigned different functions. * Dumont argues that the fact that ruling dynasties of different origin could acquire ksatriya status in later times shows that function rather than heredity was important for the varna categorization. But he does not take into account the fact that the ascription of ksatriya status to people of different origins was itself dependent on their achieving putative hereditary links with the ancient ksatriya lineages or heroes through invented genealogies and those rulers who did no seek such validation continued to be regarded as members of lower varnas, for example, Harsavardhana of Kanauj and Reddi kings of Andhra. The hereditary nature of varna identification is indicated by references to Ksatriya merchants (vaniks) in the records of the early centuries of the Christian era in the Andhra country . and in a Gurjara-Pratihara inscription^^ of the tenth century from the Doab region of U.P.