Indian Antiquary Vol 62

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

Page 61

MARCH, 1933)
THE INDIAN ANTIQUARY
According to sutras 424 and 425 of Eluttadigîrum of Tolkappiyam, the ancient Tamil grammar, gyoo andu, on ihdu and 2.009 uhd "retain " the dydam only if they are fol. lowed by words with initial vowels, e.g., go on ahdadai, whereas when the next word begins with a consonant, the dydam " is dropped," 0.g., adu pal.
Further, yogis eroployed with the iyian ia expressions like 802.ande (indeed! all right!) carrying with them a certain amount of accent.
We have to remember that adu, ahdu, idu, ihdu, etc., are derived from demonstrative particles a, i, etc. These demonstrative particles in Tamil appear in certain contexts combined with -~-; but the original particles were undoubtedly devoid of -v-. When these original particles in their short condition) conibine with a word having a voiceless consonant initially, the dydam is generated immediately before the voiceless consonant, as in a + kadiya > ahkadiya.
These facts directly show that the production of the dydam was connected with the distribution of the accent. When the accent is thrown straight upon the syllable containing the short demonstrative and the immediately following plosive, the dydam is generated. All Buch instances are associated with sandhi where the meaning leads necessarily to the association of accent with the syllable mentioned above. In aht(der (that is the village), etc., the acoent falls on the syllable containing original a and t, consequent upon the intimate merging of atd). and ar, whereas when this merging is impossible, in cases like adu koridu (that is hard), the higher accent fails to be associated with ad. or original at-, and hence no agram appears. In add indeed! all right!) the higher accent is obvious from the meaning. In aḥkadiya, the merging is complete because of the absence of t, and therefore the higher accent falls on a-k, and the aydam is generated. It is therefore possible for us to infer that the demonstrative base al-, derived from an ancient demonstrative particle a and an original -t, gave rise to the accented form che in certain positions, while it was retained as adu (with the voicing of t. to .d.) in un&ocented positions.
[B] Other instances of what are commoniy described as Wu, i.e., dydam that is organic, occur in the following Tamil words
ahgu- (to be shortened, to pass away, to become closed or compressed as a flower); ahgam (food-grain); ehg. (to sift or scrutinise, to be unloosened, to lift, to olimb); eng-am (weapon, sharpness, etc.);
vehg- (to desire ardently). Julien Vinson observes in connection with these instances : On a suggéré que, dans ces mots, le finale ne doit être qu'une dérivative, et que le oc est une mutation euphonique d'un l ou I radical. This would mean that the above instances were originally of the combinative type, and that the dydam was produced in connection with an original l or combining with k. It may be interesting to find out how far this suggestion is true of the above instances, though no definitiveness may be possible in our analysis of these instances.
ahgu (to be shortened, etc.) has been compared by the Tamil Lexicon to alku or algu with the meaning to be shortened.' In view of the fact that the deiotic particle could, as usual in Dravidian, combine with various affix-morphemes of Dravidian and produce different deiotic meanings, it is not clear whether there was at all any relationship in structure between algu- and shgu.. The Kannada cognate akkudisu with the same meaning furnishes no clue to this problem.
angam (grain) has been comparest by the Tarail Lexicon to Skt. argha ; but we have in Dravidian itself a base ar. (to cut) from which Kannada akki (through arki) and possibly Tamil arisi (rice) have arisen. What may have been the relationship of ark. to ahgam, is not clear.