Manasara Series (I. A Dictionary of Hindu Architecture; II. Indian Architecture according to Manasara-Shilpashastra; III. Manasara on Architecture and Sculpture: Sanskrit Text; IV. Architecture of Manasara: Translation; V. Architecture of Manasara: Illustrations; VI. Hindu Architecture in India and Abroad; VII. An Encyclopaedia of Hindu Architecture) v2.c17.108

v2.c17.108

"This language has, therefore, no fixed characteristics at all. We have seen that in such words as *karma*, *jaga*, and *yaśa* above, the final consonant is dropped, and these as in the Pāli and the Prākrits made nouns in *a*. But *yaśas* the original Sanskrit form, is also used as in *kīrtiyaśaścha* and these are instances in which other final consonants are preserved. Along with such a Pāli form as *suṇishyāti* noticed above, such a Sanskrit one as *śṛiṇvanti*, is found. It therefore appears to me that this is not an independent language; but that the writers of the *gāthās* knew the spoken language or Pāli, and that they were imperfectly acquainted with Sanskrit, knowing enough of it to see that the assimilation of consonants was a vulgarity, but not acquainted with its grammar. They intended to write in the more polished or literary language, but not knowing it well often used unconsciously the grammatical forms and the peculiar words of the vernacular. (At the time when the *gāthās* were written, the claims of the Pāli to be considered a separate language were probably not recognized, and it constituted the speech of the uninstructed.) Those who in this condition of things wished to write could not think of doing so in that form of speech, and therefore wrote in what they considered the language of educated men, but they knew it imperfectly, and hence produced such a heterogeneous compound as we have seen.”