________________
110
THE INDIAN ANTIQUARY
(JUNE, 1920
According to this principle of lessening the quantity of a vowel, new sets of diphthongs have found their way into our vernaculars. But in these, the sounds were changed owing to the shifting of the accent or stress, sometimes on the first and sometimes on the second element, and also because of the peculiarity of pronunciation of the first element, i.e. a, in different vernaculars.
As regards the wide sound of è and o in Marwâri and Gujarati, I think, it is due to the accent, or stress, on the first part of a i and a*. And so it goes without saying that here I am at one with Mr. Divatia, who has clearly said (Indian Antiquary, Feb. 1918, p. 41) " that when in the vocalio group of or 3, 4 is accented the <3 getting subordinate, the ultimate result is the wide sound & off." But I am at one with Dr. Tessitori in his disagreement with Mr. Divatia when the latter says that "STK and 7 pass through an intermediate steps and TT (eventually and ) before assuming the wide sounds and 371." As regards the narrow sound of é and 6, it owes its existence to the accent on the second part of a-i (dissyllable) and a-u (dissyllable) and not of a: (monogyllable) and a monosyllable). Here the accented i and u become e and o respectively, and they are narrow, because there are no other elements whatever to widen them; and then the preceding a is assimilated with the following e and o according to the Prakritic rules for assimilation. I should note here what Mr. Divatia has stated (ibid.) in this connection: "When in these groups (and ) the 7-7 are accented, their prominence leads to the uniting of andy and wand 3 into the narrow and wit."
Examples are needed to illustrate what I have to say, but before producing them I must briefly discuss the question of accent above mentioned. At the very outset it should be borne in mind that accentuation of words has undergone a great many changes from the Vedic time downward. The accent which was the cause of original corruption of a word does not necessarily continue to be in its place so long as the word remains either in the same or other form ; for it may have a different acoent resulting, in some cases, in its assuming a further new form. I do not wish here to enter into details, but simply to say as a general proposition that accentuation in Prakrita has played a great part in forming the words of different vernaculars. Evidence has been given by Sir R. G. Bhandarkar in his Wilson Philological Lectures, of the accent on the penultimate or final syllable, from which a number of new words have found their place in our vernaculars. But he did not cite any Prakritic words in support of his view though such words are not wanting. Mark the following words : vdülla also vdüla (Skt. vyakula); daïvva (Skt, daiva); paravvasa (Skt. paravasa). Here the accent is evidently on the penultimate syllable. It can, however, also be shown that sometimes in Prakrita the first syllable of a word is accented. Let me here draw attention to the following words which will support this contention : nivvedayitavva (Skt. nivedayitavya), Samardichchakaha, Bibliotheca Indica, p. 134; sapparihasa (Skt. sa parihása), Shadbhasha, p. 47; jovvana or juvvana (Skt. yauvana ), Kumarapala, III, 18: sochchia (Skt. V buch+tvd or ya), Prakritarúpavatára, XII, 78; etc. These words invariably carry the accent on the first syllable which accounts for the doubling of those consonants which have to bear the brunt of the stress. The same thing happens in different vernaculars. Mark the following Bengali expressions : (i) sakkale jdne, all know '; (ü) kakkhano nd, never'; (ii) kichchhu dibo na, nothing I will give '; (iv) badda garam, 'too hot'; (v) chotto gachh, 'a very small tree '; etc. So in Hindi : Kausikka and Gautamma for Kausika and Gautama respectively (Hammira-ráso, Nagrįpracârimî, p. 9). The subject of accentua
3 Sometimes doubling is made to suit a metre. Instances are abundant in Ppthvirajaráso.
