As stated previously, a climatic divide between winter and summer rainfall marked the passage of the Fertile Crescent economy from the Indus into the Ganges Valley A similar divide marked the passage from the Indus southwards into Peninsular India, since this was also a monsoonal region. The Neolithic populations of this region, beyond Maharashtra, essentially cultivated indigenous millets and grams, with occasional occurrences of West Asian wheat and barley, and later rice. In southern India, potteryusing cattle pastoralists with millets and grams established themselves in Karnataka by 3000-2500 Bc, evidently from a northwesterly origin in the general region of Rajasthan, Gujarat and Maharashtra. The migration of the bearers of this complex may well have taken place down a central zone of open grassland through the Deccan Peninsula, and agriculture appears only to have reached densely forested Sri Lanka around 1000 bc.65 The descendants of these Neolithic migrants speak Dravidian languages today
From a linguistic perspective, the immediate Dravidian homeland was located in the northwestern part of the present distribution of the family, also in Gujarat or Rajasthan, according to the presence there and perhaps also in Sind of a Dravidian place-name substratum. These regions today are occupied by Indic speakers. The most recent statement of Dravidian linguistic origins, by Frank Southworth and David McAlpin, derives the family ultimately from even further west on the Iranian Plateau, where it shared a very remote common ancestry with the ancient Elamite language of Iran. Proto-Dravidian speakers later spread, after 3000 bc, with an ancestral agricultural and pastoral vocabulary that included terms for cattle, sheep and goats, pottery, grain processing, and possibly some form of plow, but no specified cereals.66 The spread thus occurred with dominant pastoralism and small-scale (presumably millet and gram) cultivation, at roughly the same time as the ancestral Indic speakers were migrating along a more northerly route from the Indus into the Ganges Basin (Figure 7.3).
In contradistinction to the interior grassland route suggested above, an important linguistic detail about the early Dravidian movement through Maharashtra into Karnataka, revealed clearly by Southworth's (2005) discussion of Dravidian place names, is that it was essentially focused on the Konkan coastal plain of western India, rather than on the Deccan Plateau itself. The interior uplands, if we take the place name evidence seriously were settled by Indic speakers who might indeed have moved among pockets of Dravidian speakers, but who more likely were intruding upon an essentially Mesolithic cultural landscape of unknown linguistic affiliation. The Ganges Basin itself has never supported extensive Dravidian-speaking populations, and at 2000 bc the only other well-attested populations here, apart from Indic speakers, would have been ancestral Munda (Austroasiatic) speakers (Chapter 8).
The Dravidian language family, therefore, appears to have begun life in Iran as part of an independent linguistic macrofamily (Elamo-Dravidian) that also included Elamite. Its main distribution was created by a Neolithic migration of its ancestral speakers from the eastern edge of the Fertile Crescent into southern Pakistan and northwestern India. Its speakers became transformed economically during their movement into the dry monsoonal climate of Peninsular India, towards pastoralism on the one hand, and towards cultivation of a range of indigenous drought-resistant millets and grams on the other. The genetic profile of the Dravidian-speakers today is predominantly indigenous to South Asia, and presumably also to pre-Indo-European southern Iran.
World History
