________________
MAY, 1892.)
THE INSCRIPTIONS OF PIYADASI.
147
It is a universal rule in the Praksit, both in the dialects of the inscriptions and in the literary languages, that before anusvára a long vowel becomes short. In four or five instances, however, the long vowel of Sanskrit is retained : yátán (VIII, 1), susrusatán (X, 2), anuvidhiyatám (ibid.), samachérán (XIII, 7) at Girnar. It is plain that we have here purely and simply an orthography influenced by the learned language.
These last instances are mere accidents, but they enable us to judge better regarding those in which variations of orthography more nearly balance each other. In a certain number of consonantal groups composed of a mute and an t, instead of the disappearance of the r, compensated for by the doubling of the mute, we find at Girnar the etymological spelling, pra, tra, sra, rva, instead of pa (ppa), ta (tta), sa (sea), va (vva). This spelling is by no means fixed, -as may be seen from a reference to the text of any single edict, and it would be of little interest to quote here all the instances, one by one. It will be sufficient to state that we have the spelling pra about 45 times, as against the spelling pa 25 times : for tra, 30 times ta, 20 times tra: for rua, rva and va each about an equal number of times : for bra, once bra, against 6 or 7 times ba: once sra (for rsa, ráa), against once sa. Is it possible to contend that such an indifference represents the real spontaneous condition of the popular idiom, and that pronunciations corresponding to such different stages of phonetic decay, and that side by side in the same words, belonged actually to the same period of the normal development of the language ? If it were possible to have any doubts on the point, it would be sufficient to refer to later facts in the linguistic history. When we read, in Hindi, priya beside piya, putra beside púta, bráhmana beside bámhana, we have no hesitation. We know that the first of each of these pairs is an instance of learned orthography: that it is only a tatsama, that is to say, a word borrowed direct from Sanskrit, and restored to the current of the language. When in an inscription of the 24th year of Vásithîputa Pulamâyi (Karli No. 22, A. S.), we meet side by side the spellings puttasya, sõvasakasya, vathavasya, and budharakhitasa, upásakasa, prajá, parigahe, we are confident that these genitives in asya, this spelling of prajá, cannot, at such a period, have represented the true pronunciation of the people; that there also they are tatsamas. How can we avoid drawing the same conclusion from facts which, although more ancient, are none the less strictly analogous ?
It is therefore certain that these sanskritized forms do not represent the actual stage of the contemporary phonetic decay. One point, however, appears to be open to some doubt. The tatsamas of the modern languages actually enter into circulation, and that with either the ancient pronunciation, or with an approach to it. They are words of special origin, but at the same time real words of the current speech. The tatsamas of Mixed Sanskrit are, on the other hand, purely orthographical, for they belong to a purely literary language. That is to say while, in the modern tongues, the loans from the ancient language only deal with bases, and consequently have no effect on the grammar, in the Sanskrit of the Gáthás, the imitations extend even to the inflexions, i. e. to elements which would escape any arbitrary action of the learned in a really living language.
In which of these two categories are we to class the tatsamas of Piyadasi? We must, I think, consider them in the same light as those of the dialect of the Gáthás, and recognise them as 'orthographic' tatsamas, The examples given above show that little heed was paid to accurately representing the pronunciation and that the etymological form was readily adopted in cases in which the vulgar pronunciation must have been markedly different That is in itself a strong reason, but we shall see, besides, that the classical language had not yet been so developed into practical application at this epoch as to allow us to assume that it could have penetrated into the stream of popnlar ase. Moreover, in the different versions of the texts, the proportion of these tatsamas is very unequal. If it were a case of forms readopted into current spoech, such an inequality would be surprising; it is more easily explained by a
I shall refer to the dialect of the Gathús or Mixed Sanskrit in the following chapter.