________________
268
THE INDIAN ANTIQUARY.
[SEPTEMBER, 1892.
preceding remarks make evident with what particular reservations we should here surround the application of this principle. In themselves, the Sanskrit forms are certainly more arch ic; they are historically older than the Präkțit forms of the time of Piyadasi. Yet that does not prevent Sanskrit, as a whole language, in the form in which we know it now, having only succeeded in conquering for itself an existence long after the rise of his Prakrit. So it is with the different Prakrits. The general phonetic appearance of Pali is certainly more archaic than that of Maharashtri. Have we any right to conclude that therefore it actually existed, in its definitive form and orthography, before Maharashtri? In no way. In short, we must carefully distingush between the constituent elements of the dialect, considered directly, and their atilization in the shape of a particular literary dialect, adapted to a certain order of production. We cannot apply to literary idioms, in part artificial and learned, the same measure as that which we apply to purely popular languages. They, the former, can, in a sense, go up the regular stream of their linguistic development. This is the very fact which we have proved for Mixed Sanskrit. When I speak of inquiring into the age of the Literary Prakrits I mean, not to determine the epoch to which the elements, morphological and phonetic, of which they are composed, can be traced up, but to fix the moment when they were arrested, crystallized, in & definite form for literary use. For this purpose the forms which are the most altered are those which are most instructive. They can be made to prove that such & dialect cannot be earlier than suoh & given epoch. The better preserved forms prove nothing. They may have been either subsequently reconstructed in the light of the learned language, or preserved for a greater or less period by tradition before receiving their place and their consecration in the special dialect of which they finally formed an integral part.
The criterion, therefore, founded exclusively on the general phonetic appearance of the dialects must be resolutely put to one side, if we wish to avoid misconceptions regarding the most certain, the most characteristic features of the history which we are endeavouring to build up.
This being settled, a two-fold object of inquiry presents itself. On the one side, the relation existing between the Prakrits of the monuments and that of the books, and of the other, the relation existing between the literary Prakrits and Sanskrit.
To set to deliberately, to convert, by systematic work, popular dialects into literary dialects with forms fixed for ever, is not so simple an idea that it would suggest itself of itself, and that it should not require any explanation. Such an undertaking must evidently be regulated on a prototype, on some pre-established model. India possesses a type of this description, Sanskrit. Indeed, if we pay heed to the names, prákorita and samskrita are correlative terms. The actual bond which connects together the two series of facts is certainly no less close than the formal relationship of the names which designate them. Historically, the earlier term is Sanskřit. On that point there can be no possible doubt. It is the very elaboration and diffusion of Sanskpit which has served as the basis and model for the elaboration of the Prakrits. They have been regalarised in imitation of it. The recollection of this origin is perpetuated in the teaching of the grammarians. They take care to establish that Prakrit has Sanskțit for its basis and for its source (Hémach. I. 1, and Dr. Pischel's notes). It would be a mistake to attribute to the Hindus, on the strength of such a remark, the idea of a linguistic genealogy founded on comparative analysis. When Vararachi and others (cf. Lassen, Instit. Ling. Prakrit, p. 7) declare that the prakriti of Saarasêni is Sanskrit, and that of Maharashtri and of Paisachỉ the Saurasêni, it is quite clear that we must not take the proposition in an historical sense. It is nothing but a manner of stating that Saurasêni, in various characteristics, approaches Sanskrit orthography more nearly than the other dialects, - that it is in a fashion midway between the learned language, and the dialects with a more altered orthography. It is not a genealogical classification, but an entirely practical one. It is something like a direct recognition of the method according to which these languages have received their grammatical fixation. This working has taken for its basis the grammar of the learned language, and for its principle the gradation of each of them on a determined level below the stage of Sanskrit.