Indian Antiquary Vol 58

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

Page 43

FEBRUARY, 1929 )
PLOSIVES IN DRAVIDIAN
37
the pronunciation and by the phonetic spelling, retain their purity. Indeed the number of such native forms containing pure inter-vocal surds is so large in these languages as to exclude altogether the possibility of any general law of convertibility, as in Tamil.
Telugu-kota (fortress); pašu (song); tafirichu (to flutter); ndţu (to plant); paka (hut); petuku (unpleasant), etc. Numerous inflexional endings and verb-forms in Telugu and Kanarese show inter-vooal surds.
Kanarese-katukolliri (sit down).
With regard to those forms in Telugu and Kanarese which show by their spelling and pronunciation the sonant variety as distinct from the corresponding Tamil forms, which give the surd-gymbols but the sonant sounds, the explanation is easy. There is a law applicable (in varying degrees) to all groups of languages that when a surd occurs intervocally, the sonority of the vowels is partially or wholly shared by the surd itself which consequently tends to become sonant. The history of many languages illustrates this, e.g., in many Old English words and also in the development of the Romanoe languages from late Latin.
We see therefore that the help of a distinctive Dravidian law need not be invoked for explaining the change of surds into sonants in Telugu and Kanarese. That such a Dravidian law does not exist is further shown by the presence in Telugu, Kanarese and the ruder dialects of numerous forms with pure inter-vocal surds.
Then we come to Malayalam. Gundert, Kittel and Caldwell have all maintained the existence of the law of convertibility in Malayalam exactly as in Tamil.
Two circumstances seem to have coloured their judgment in this matter (i) the close relationship that exists between Tamil and Malayalam has led them to ignore the differences involved in the actual pronunciation of Malayalam at the present day as distinct from that of Tamil. (ii) Dr. Gundert, who was undoubtedly a great Malayalam scholar, made his observations of Malayalam sounds from amongst the natives of the extreme north of Malabar who do not talk pure Malayalam.
Let us consider the actual facts regarding the pronunciation of Malayalam inter-vocal plosives in the Cochin State, the central portion of the Malayalam-speaking area, where the pronunciation has been admittedly recognized as the purest, i.e., free from the Kanarese influence of the north and Tamil influence of the south. (i) In the Malayalam literary dialect the inter-vocal surds are pronounced with entire purity. When books are read or discourses delivered, the surds have the value either of classes II or III or IV described above. (ii) In the colloquial among the cultured classes, the inter-vocal surds do not become pure media but assume the values of voiceless mediæ, i.e., class IV described above. (iii) In the colloquial of the lowest classes where purity of pronunciation is hardly cared for, and where elisions, contractions, and holophrastic changes are extremely common, not only inter-vocal surds but initial surds also become sonants.
From the above, it is clear that the "language-consciousness” of the Malayalamspeaking native does not admit of any such law of convertibility as in Tamil, Spelling only confirms this view. Unlike the Tamil alphabet, and like the Telugu and Kanarese alphabets, the Malayalam alphabet has symbols for both burds and sonants, and intervocal surds are all represented by the symbols for gurds only. If the values of the inter-vocal plosives had been only sonant at the time when the Sanskrit alphabet was taken over by Malayalam, it is at least strange why the Malayalîs, who are noted for their linguistic precision and exactness, did not represent the sonant sounds with sonant symbols. The only explanation possible is that the surds had not the value of sonants at the time of the adoption of the Sanskrit alphabet, i.e., some time about or before A.D. 1000.
Historical grounds also strongly support such a view. In a paper contributed by Prof. Bloch to the Indian Antiquary8 he has demonstrated, on the strength of a passage from the writings of a seventh century Sanskrit author Kumarila Bhatta, who quotes a number
. Vol. XLVIII, 1919, pp. 191-195.