Abstract :
Chronological ordering of Sanskrit texts engaged numerous researchers over centuries. Early subjective efforts, based on historical tradition, astronomical references, and comparative linguistics, graduated to statistical techniques; however, intimate knowledge of the language was still required. In this backdrop, we propose an information-theoretic measure, namely the entropy rate, as a chronological indicator, and buttress our thesis with experimental evidence. The proposed measure, requiring minimal linguistic knowledge, could be attractive for its simplicity, and its basis in rigorous theoretical foundations. In particular, we establish prediction by the partial matching (PPM) algorithm as an efficient entropy estimation technique for Sanskrit texts, and showed that the resultant estimate indicates the chronology of landmark texts starting from the Rigveda, to Kalidasa´s works, to Rupa Goswami´s Padyavali.
