Indian Antiquary Vol 59

Author(s): Richard Carnac Temple, Charles E A W Oldham, S Krishnaswami Aiyangar, Devadatta Ramkrishna Bhandarka
Publisher: Swati Publications

Page 197

SEPTEMBER, 1930.]
TAMIL ARISI (RICE) AND GREEK ORUZON
181
If, as we have shown above, the Dravidian forms (with and without the initial labiodental fricative) are native, could we not connect the Greek word oruza directly with an old Dravidian form varisi or varizi which we can reasonably presume to have existed? The change of initial va- into Greek o and the representation of Greek u for Dravidian i are quite normal. Initial va- in Dravidian, we may observe here, even to-day in dialects like Tuļu and Kannada, has sometimes nearly the same value as (o).
Thus we come baok again to Caldwell's opinion, which, though summarily stated by him, appears to have justification on closer analysis. It is true that Caldwell himself did not work out this relationship and that he contented himself with pointing out the superficial resemblance; but, closer inspection, in the light of what we have stated above, & reveals that his view of the origin of Greek oruza cannot now be dismissed by us as lacking any justification whatsoever.
So far as Sanskrit orihi is concerned, the root suggested by Sanskrit grammarians is orí, to choose,' but the semantio derivation of the meaning 'rice' from the meaning of this radical appears difficult. Could it not be that the Sanskrit word was adapted in a slightly modified form from Dravidian with the meaning it had in Dravidian, and then a Sanskritic derivation was attributed to it by Sanskrit grammarians ? This is a view which deserves to be taken up by students of Sanskrit philology. The question of the relationship of Dravidian vari, etc., and Sanakrit uríki is a little complicated by the fact that certain forms cognate with Skt. prthi oocur in Iranian and Central Asian Aryan dialeote ; but there is nothing inherently impossible in the view that seeks to trace the Indo-Aryan forms to Dravidian. For one thing, the nature, chronology and extent of the influence of Dravidian on Indo-Aryan (in pre-Vedio as well as in post-Vedic times), when oarefully investigated, should offer valuable assistance in this direction. Again, if we consider that the Dravidian forms with u- are native,-28 we may have to, in viow of the above discussion,--and if, further, & relationship between the Dravidian and the Indo-Aryan forms is envisaged, then the possibility is all on the side that Dravidian may have been the lender and Indo-Aryan the borrower.
Another interesting fact in this connection is that, while the Indian Austric dialects (like Santali, Mundári, etc.) show only a few forms for 'rioe' or 'paddy,' which are allied to Dravi. dian, distant Austrio dialects like Malay and Javanese show vari, pari, broi, beras, padi, etc., with the meaningspaddy' and 'rice.' Is the resemblanoe aocidental ! Or was the close contact between Dravidian and Austrio in pre-historic times (a view which is gaining great popularity among scholars to-day) responsible for the presumable borrowing of the word by Austric from Dravidian ?
Beset as these questions are with oonsiderable difficulties, and much as we have still to investigate before definitive conclusions could be laid down in regard to these inter-relationships, we yet have to say that Caldwell's view about the origin of Greek oruza as having been borrowed from Dravidian cannot be dismissed, especially in view of what has emerged from the above discussion, viz., that a hypothetical variji or varizi could be postulated for Dravidian, from which the Greeks oould, with characteristio modifications, have borrowed their word.
8 The conditions under which a full bilabial has become incorporated in prothetic positions require to be sifted and classified in the different dialecta ; but, for the purpose of our discursion, it would be enough if we keep in viow what is a well-recognised phenomenon of Dravidian, vis., that appears before initial dorsal vowels as a glide fully evaluated in pronunciation in Kannada, Tulu and Gandi.