________________
MAY, 1892.]
THE INSCRIPTIONS OF PIYADASI.
155
It is certain that the orthography of our inscriptions does not always exactly reflect the actual pronunciation. It is unequal to the task when it neglects to notice dooble consonants or long vowels, and it overshoots the mark when, at Girnar, it retains a long vowel, either before an usvára, or before a group of consonants. Besides this, it elsewhere gives evidence, as, for example, in the notation of the groups which contain an r, of a significant indifference in regard to phonio expressions which belong to diverse periods of the development of the language. It is, therefore, sure that this orthography, in a certain number of cases, obeyed (as we call them) learned historic influences. Like the modern languages, like the mixed Sanskțit of the Gathás, it is full of words or methods of writing, which constitate so many graphic tatsamas, and which consequently form an artificial and learned elemeni. There is no ground for citing against this proposition the ignorance of the engravers. They may be responsible for certain material errors, for certain inconsistencies, but not for a system of orthography. They applied that system, it is true, but, however imperfect it may have been, it must have been founded by persons who were educated, skilled men, Even at the present day, it is evidently the learned caste that takes these loans, which, entering the popnlar language, gradually oxtend themselves to the most ignorant. In its generality, therefore, the principle appears to me to be unassailable, and those facts, which are certain, justify by themselves important conclusions as regards the light in which we should consider the language of our inscriptions.
Other facts, such as those which concern the groups st, 81, tp at Girnar, allow more room for contradiction, and I only claim probability for my opinion regarding them. I have merely one more observation to add. It is specially at Girnar and at Kapur di Giri that we meet these semi-historic modes of writing. If my interpretation of them is accepted as correct, they will add seriously to the balance in favour of the conclusion to which the undisputed facts tend.
This conclusion has a corollary. It implies that the differences of dialect between the popular languages, which are reflected by the various versions of our inscription, are less decided than we should at first be induced to consider them, judging from the appearance of the orthographies. If they are really separated by some characteristics, they have, in general, arrived at nearly equivalent stages of phonetic corruption. The more prominent points of difference, which attract our attention at first sight, have their origin in tendencies, more or less accidental, of word-borrowing or of modes of writing, - in the greater or less use of tatsamas. This result is in itself à priori so probable, that it might almost be invoked in favoor of the conclusions which I have endeavoured to establish. It is, assuredly, scarcely probable that, by its mere natural movement, by its spontaneous development, the same language should, in the same time, have reached, in neighbouring provinces, stages of decay so unequal as a comparison between the orthography of Girnar and, for example, that of Khálsi would suggest. The views which I have put forward explain this anomaly. For inadmissible inequalities of phonetic development they substitute the very simple notion of different orthographic systeme in parallel use in different regions. If, as everything tends to shew, the epoch to which our inscriptions belonged was still, so far as regards the art of writing in India, a period of feeling the way and of uncertainty, if it is anterior to the regalarisation or at least to the general expansion of the Sanskrit orthography and to the codification of the literary Prakrits, the parallel existence of these divergent imperfectly established systems is easily explained. I shall shortly indicate what circumstances seem to have conduced to favour their geographical distribution in the manner to which witness is borne by the evidence of our monuments. These circumstances equally concern the distribution of the dialectic differences properly so called.
(To be continued.)