________________
MARCH, 1933 ]
THE INDIAN ANTIQUARY
(2) The fact that dots are used in Tamil and in Sanskrit need not necessarily dis
prove the native origin of the sound whose secondary character was probably
fixed and recognised by Sanskrit-knowing Tamilians. (3) This point again raises, if at all, only the secondary character of the sound in
Tamil. (4) The argument about prosodic lengthening' would not apply to instances of
mut't'rûydam like thgu, which have no alternants. (5) The resemblance between the dydam and the Sanskrit spirants can lead to no
inference, in the absence of any direct evidence. (6) The ancient Tamil grammarians who could well distinguish Sanskrit sounds from
native ones, have nowhere referred to the dydam as a borrowing or as an 'invention.' Apart from these arguments, there are certain other facts also which I shall urge here in favour of the native origin of this sound in Tamil. That the sound was not a common one in Dravidian admits of no doubt; but a discussicr of the phonetic aspects of its growth with comparative reference to a similar development in the central Indian Dravidian dialect Gôndi, would tend to show that the aydam was a native though secondary sound in Tamil. It is possible that recognition was given to it by Sanskrit-knowing scholars.
[D] WAS THE ÂYDAM A MERE ORAL FRICATIVE, OR DID IT INVOLVE AN ELEMENT OF THE GENUTNE ASPIRATE, I.E., GLOTTAL FRICATIVE ALSO ?
(a) Nannûl describes the sound as being produced in the head' with an 'open mouth.' This description may apply to fricatives of the velar, uvular and glottal varieties alike. Whether the sound was originally a genuine glottal sound is not made clear by the description in Nannal. We learn, however, one fact from these references to the head' and the open mouth,' and this is that the sound so described could not possibly have been labial, dental or palatal. It is clear therefore that at the time of the composition of Nannůl, the sound should have been either an aspirate or a back fricative of the velar or uvular type.
(6) Caldwell Vinson (who oslle the sound an'aspiration gutturale') and Mr. Pillay regard the sound as a genuine aspirate. The modern value of the spirantic g when texts are read is (as Mr. Pillay has observed) probably only due to the characteristic modern tendency of giving the velar fricative value to intervocal aspirates, as shown for instance by the Tamilian pronunciation of Skt. muhúrtam as mugúrtam, the intervocal -- being evaluated as a velar fricative.
(c) Mr. Sastri would consider the sound to be a fricativo, whose value may be labial, dental, palatal or velar according to the character of the immediately following surd. He is of opinion that the velar value given to it today when texts are read is a mistake.' The description given in the Nanna! and the uniformly velar value given to it today would show that no such 'mistake' could have crept in after the time of Nannûl. In the absence of evidence to show that there was really a mistake,' we have to regard the sound as a 'back sound originally, whose exact value (i.e., whether it was only velar or whether it was glottal) has to be determined by a consideration of other factors.
The analogy pointed out to the jikvdmuliya and upadi.maniya sounds of Sanskrit leads to nothing conclusive. For one thing, we have no evidence to prove that the Tamil sound was copied from these. Secondly, these Sanskrit sounds, grammatical abstractions' themselves (as Whitney puts it), probably had an aspirate value also beside the fricative values depending upon the immediately rollowing surds ; vide &$ 69 and 170 (d), Whitney's Gr.
The velar fricative value given uniformly to the dydam today, whatever the value of the surd concerned may be, taken along with the description given by Nannûl would point to the value of the sound having shared a common aspirate element from a very early stage.