Language and Linguistic Statistics in Evaluation Entropy and Information Energy

Professor habil. Gheorghe Savoiu, Ph.D

BS Math and MCS & M. Math. Matei Sandra


This paper is a first-order interdisciplinary research, in an interstice defined by linguistic statistics and it appeared by extending the statistical way of thinking in the universe of language. The research structure includes an introductory section that states the specific paradigm or the common statistical and linguistic concepts of the investigation, followed by a brief methodological section, simultaneously of a descriptive and applied type, with reference to laws, methods and statistical models possible in the initially quantitative universe and finally qualitative universe of language, to eventually reach the final section dedicated to an attempt to assess entropy and information energy, based on the results of research in its whole. From an evolutionary point of view, after the philologist initially investigates in an exemplary way and with philological accuracy, and the statistician approaches selectively and inferentially, but also with specific statistical rigor, the interdisciplinary point of view of linguistic statistics allows a rapid confrontation and a multiple evaluation of entropy through the information energy method, but also through other equally interesting methods. The justified hope of interdisciplinary thinking of linguistic statistics to optimally capitalize on simultaneously philological and methodological reasoning to validate or invalidate preliminary hypotheses and to allow the expression of conclusions, flollow-up analysis, outlines if not an answer, certainly at least several useful questions, which confirms practically and theoretically both the necessity and the veracity of the research in this article, according to the logic of a successful research in a Veblenian way.

Keywords: interdisciplinarity, language statistics, probabilities, random sample, (pseudo) stratified sampling, sample’s volume, limit error, statistical inference, correlation matrix, entropy, informational energy, Zipf law

[Full Text]

Romanian Statistical Review 4/2021