________________
THE INDIAN ANTIQUARY.
[SEPTEMBER, 1892.
be surprised at sporadic instances of orthography such as purisadatáva (Nâs. 24), bhayáva vélidatava and uyaraka (C. T. I., p. 17, No. 23), beside the usual ovaraka, and the terminations in áya. So, also, we find in the inscriptions of the north, side by side, samvatsarayé, athasatatimaé, tachhasilayé, puyaé (Taxila), &c. We find kaliana (Kanh. 13, 24, &c.) as well as kaliyana and pulumai, pulumáyi, and pulumávi (Nâs. 12, 13, 15); dhutua, mútua (Kaph. 27) beside dhutuya, mátuya, &c.; ya and ja are used indifferently the one for the other, when it is necessary to represent an etymologicalj: on the Wardak vase we read puyaé, beside raja, at Taxila, raya beside puyaé, and, to confine ourselves to the cave inscriptions, Kaph. 18 reads puyatha[m], C. T. I., p. 16, No. 20, vániyiyasa, Amr. 26 B, vúniyasa; while on the other hand, beside the usual bhayá (equivalent to bháryd), we have bhajayá, Kanh. 19, bharijayé, Nâs. 11 B.
260
Inversely, a hard consonant is sometimes substituted for a soft one. For example, nékama, beside néguma (C. T. I., p. 60, No. 2), nákaṇaka (Kanh. 2), nákaniká (Amr. 121), nákachanda (Amr. 56), in the frequently occurring ma[m]tapa, beside mandapa and mamḍava; Kanh. 16 reads bháka for bhaga; Amr. 222, légáticha equivalent to lékáditya, and bhagapatô for bhagavató.
Although the palatal nasal n is not unknown, its use is very irregular. Kârli 20 has and equivalent to anyaḥ; Kanh. 5, anani, Kanh. 27, punam equivalent to punyam and nati equivalent to jñáti. The same spelling núti occurs again at Amravati, e. g. in Nos. 232, 249, while, on the other hand, I have noted in two inscriptions (C. T. I., p. 53, Nos. 28 and 30) kalianaka.
Similarly, other modes of orthography sometimes bring us nearer to, and sometimes take us further from, the learned standard. I may mention amasa[n]taka, Nâs. 11 B; bammaniya beside bammhana, C. T. I., p. 14, No. 15; these methods of writing are the more worthy of note because, long before, at Kapur di Giri, we regularly find the spelling bramana. C. T. I., p. 46, No. 14, writes shanuvisa equivalent to shaḍvimsati, an absolutely sporadic instance of the use of sha in this Prâkrit: a similar inscription, no less Prâkṛit, writes putrasa beside putasa, (C. T. I., p. 40, Nos. 3, 5, 6, 7).
These inconsistencies of orthography are all sporadic. That they certainly do not depend on differences of time, can easily be proved by reference to the monuments from which the examples have been drawn.
These monuments are dispersed over a very wide area. Now, between the inscriptions of Gujarat or of the caves of the Western coast, and those of Amravati at the mouth of the Krishna, those of Khandagiri in Orissa, of Sâñchi in Mâlava or of Bharhut in Bihar, we find no trace of differences of dialect. They extend over at least four centuries, from the second century B. C. to the 3rd century A. D., without disclosing, between the most ancient and the most modern, any appreciable variation. In an area so extensive, the vulgar tongue certainly could not have failed to divide up into numerous dialects. This is a phænomenon escaped by no language. Literature bears witness to it for the following period, and no one can be tempted to imagine that the fact was then a new one. On the other hand, it is clear that a language cannot pass through four or five centuries in the mouth of the common people without decay and transformation. The earliest literary specimens which we possess of the Prakrits, the stanzas of Hâla, and the Prakrits of the most ancient dramas, although in origin but a short distance from the end of the period to which we refer, reveal a phonetic alteration which was much further advanced. Let us, therefore, bring ourselves face to face with the orthographical facts which have just been pointed out.
The parallel employment of forms unequally altered, belonging to different strata of the language shew that this dialect of the monuments, however near we may suppose it to the living popular language, is neither its direct expression nor its faithful imitation. It conceals under a level in part conventional, a more advanced degeneration of the current language - a degeneration of which the distortions are reflected in those more corrupted spellings which accidentally escaped the engravers.