________________
32 SOUND SYNTHESIS IN IE, IIR & SANSKRIT
This shortening of a is an innovation in OIA and this may be partly due to prakritism (cp OIA ai>MIA e) and partly due to an intention of maintaining a contrast of the sandhi of ă+ildlr with the sandhi of ă te aijo aular. The later combination also results in ai, au and är and the former also historically should have become ai, au and ar, and the contrast of ått etc with ă+elai etc would have been lost. This might have motivated the change of a-ti to e etc (after the sandhi of a ti to e etc) instead of the historical ai. It should be remembered here that atf>e, a+å>o, a tr>ar is quite historical and a tie etc. show innovations.
On the contrary the reverse is also sometimes found i.e, a-t-ilăr is sometimes found as ai, au, ar instead of the historical e, o & ar respectively. e.g. paişayuh ( pra
+işayuḥ) (in. RV I. 120.5), upārchati (upa +scchati), There are also historically justifiable forms aindra (<at indra) (SV I. 2.1.45), ārti (<a trti) (VS). This clearly proves that there was a conscious effort side by side to retain the historical development of the attor in the form ai ( =IIr āi ), au ( IIr āu ) and är instead of e, 0, ar. This was more onthusiastially extended to forms. where there was an original el ca-ti etc), o atu) or ar (<a+r). In course of time when the e type of form was established for both ati and ati a few forms, which were established in the language with ai etc also remained as sporadic cases. 39. In external sandhi, OIA shows loss of a after e, o. In internal sandhi of course elota > aya apa, which is quite historical being from IIr ai au ta>ayalava etc.
The loss of a is comparatively late. In RV often this a is to be read as needed by metre. This loss of a after e, o in external sandhi is compulsory in CI Skt. This
Jain Education International
For Private & Personal Use Only
www.jainelibrary.org
