________________
68
S. N. Gh osal
Here we find Saurasení spoken by Vasantasenā, Radanikā, Madanikā etc, Avanti spoken by Viraka and Candanaka, Prācyā by Viduşaka, Magadh by the servants of Sakāra, Cărudatta and Vasantasenä, sakārı by Sakāra, Candāli by two Candālas and Dhakki by two gamblers. Although there is no scientific edition of the Mrcchakatika, which makes any statement about
the it hazardous, it is undoubtedly true that the speeches distributed to characters do not agree with the account, as provided by the Narvasästra or by the Sahityadarpaņa. The Natyagästra assigns Ardhamāgadhi to the slaves, Rājputs and gentlemen, which we do not actually find in the drama. Again it suggests that the rogues should speak Avanti, which is certainly not the case as they speak a different kind of dialect, which is called Dhakki. Keith by an incidental study of the features of the dialects reduced all the Präkrit specches of the drama into three dialects only-namely Sauraseni, Māgadhi and Dhakki. According to his estimate Avanti and Präcya are in no way different from Saurasenr; sakāri and Cāndali are in fact varieties of Māgadhi and Dhakki is of an indefinite character.
From all these facts particularly the one that the Prākrit grammarians ignore many dialects mentioned by the dramaturgists and the same remain unrepresented in most of the Sanskrit dramas we come to certain conclusions. First many of these speeches e.g, Prācyā, Ayantijā, Bāhliki etc. were minor dialects confined with narrow areas and as such they had extremely local application. Some again were sectarian speeches, which were utilised by the minor communities of people living likewise within restricted areas. They might be presumed as class-dialects as they were not in circulation in all substrata of the people. Ābhiri, Cāņdāli and some others might be considered as belonging to this class. As these minor dialects were sub-dialects they integrated theniselves to some bigger Prakrit speech, which was current in the arca as a popular literary speech. Secondly major dialects like Māhāraștri, Saurasenj and Māgadhi were perhaps generic speeches and were not unitary in character. Each contained a number of subdialects within itself. They were presumably contiguous and might have contributed certain common elements, which stood as the substratum of each of these generic speeches. In the case of Māhārāştri we can make such an assumption owing to the fact that in this dialect the declension and conjugation systems show multifarious forms, which can hardly be expected in a unitary speech. I refer particularly to the forms of the first and second persons. Such varieties of forms can never appear in one particular speech homogenous in character but they are expected to develop in a speech, which is split into a number of dialects or subdialects, each of which contributes elements for the standardization of the generic speech, under which all these subdialects occur.
In the case of Māgadlii the generic feature of the speech is all the more clear. I can show it with the help of a specific example. Vararuci