composition and Sanskrit rhetoricians are emphatic in their condemnation of omission of sandhi.
Mammata says,”S (of these, wilful omission of sandhi, even rarely, is a fault)
132. There is little relevancy in going to Sanskrit grammarians to settle the usages of a Telugu literary dialect which is found in the works of the early poets. An examination of their language clearly shows that omission of sandhi has no place there except at the end of a sentence.
133. Mr. J. Ramayya Pantulu’s treatment of the question of sandhi differs materially from Mr. K.V. Lakshmana Row’s. In the opening sentence the former concedes that the question of sandhi admits of difference of opinion, and then disposes of the analogy of Sanskrit. “In Sanskrit as well as in the cultivated Dravidian languages these phonetic changes are carefully studied and embodied in rules. It is a mistake to suppose, as is sometimes done that the Dravidian languages have borrowed the rules of sandhi from Sanskrit. It will thus be seen that there is little in common between Telugu and Sanskrit in the matter of sandhi except in the nomenclature and method of treatment”. About the usage of Telugu poets he says, “In verse, it is regarded as compulsory with one exception, viz., at the end of a sentence. The word sentence is sometimes literally interpreted so as to include a verse line. But relaxation of sandhi even in these cases is rare in practice. Sandhi is the rule.”
134. Mr. J. Ramayya Pantulu advocates the omission of sandhi in some cases — but not uniformly — not certainly to the enormous extent to which it is practised by Vijnanachendrika, and by Mr. K. Veeresalingam Pantulu in his latest work. Mr. J. Ramayya Pantulu limits omission to cases where “making sandhi is likely to obscure the meaning or convey an incorrect idea of the form of a word —
