________________
Vol. XVI, No. 2
non from the Indo-European period we can say that the tendency is more r than I even in Indo-European.
Here a question may be raised whether the interchanges between rand I can be possible. In a sense I was not a cerebral sound in Indo-European' (=IE), because there is no cerebral sounds in Greek, Latin etc. and that is why the interchanges between r and I was possible, because the place of utterance of both was dental. And that is why IE *l becomes r in sanskrit. For example IE *leugos=OIA rocas etc. and because of that reason in sanskrit the dictum ralayorabhedaḥ has come into existence. In this connection it could be mentioned that Värttika-sútra r-!. varnayormithaḥ sävarnam vacyam, is made, by Kātyāyana, because in his time the place of utterance of r and I were not the same, Kātyāyana, atleast, understood this much that their place of utterence though not the same, they should be considered as homogeneous, so that in sandhi between r and I we could have either r or ! or both.
Apart from this, we can trace even from prakrit that r was not originally cerebral. This is evident from the treatment of r in prakrit. r in conjunct with t, that is r+t, has two-fold development. In one place r+1>it, and in the other r+t>tt. Hemacandra has a sútra which indicates this twofold development of this conjunct rtt. Hemacandra has classed them in a group of words where rtt become ty (rtasy-adhũrtādau 2.0), for example, kaivartaḥ> kevatto, vșttiḥ> vasti etc. where rtt becomes cerebral fi.e. because ris a cerebral sound, it has influenced the dental sound which is turned into a cerebral.
The other group where rtt does not become cerebral is refered to by Hemacandra as adhürtādi class. For example, dhůrtaḥ> dhutto, kirtihkitti etc. Though in most of the cases rtt becoming cerebral. the instances where it does not occur reminds us the historicity of the development of this sound r. Accidentally because I was a dental sound originally it did not develop into cerebral and because those words are small in number they belong to the dhūrtādi class and in most of the cases the rest of the combination rtt become cerebral. because by that time the pronunciation of r is changed from denta to cerebral, and the first Sanskrit grammarian who mentioned r as cerebral sound was Candragomin between 5th and 7th centuries A.D. It means that Candragomin first recorded this fact or the nature of r as cerebral. It seems probable, that the tendency started before the time of Pāņini as his rules for ņatvaşatvu, particularly after r, suggest the cerebral nature of r. But the records of Prātisākhya do not also
Jain Education International
For Private & Personal Use Only
www.jainelibrary.org
