Statistical Studies in Some Indian Languages with Applications to Communication Engineering

Abstract

The current interest in the problems of language is briefly surveyed and the role of language in communication is examined in the light of the information theory. The view is taken that as different languages having the same semantic content may be regarded as different codes for obtaining the same responses, it should be possible to compare the relative communication efficiencies of different languages in terms of the number of bits of information needed to transmit the same semantic content. Languages using different scripts may also be compared with regard to their efficiency of transcription from the verbal to the orthographic form.

A method developed for collecting data of the relative frequencies of speech sounds and speech sound digrams for Indian languages based on phonetic analysis is then explained. The relative frequencies of the different speech sounds in five of the major Indian languages, viz. Malayalam, Telugu, Marathi, Hindi and Tamil, thus obtained are then given. Based on these data a common telegraph code is constructed which does not fall short of the optimum codes for any of the individual languages by more than 7 per cent.