Particular Aspects of the Quantitative Studies of theUkrainian Language

Abstract

This article describes the various calculations of the number of linguistic phenomena; the construction of their stochastic models based on these quantitative data; the testing of hypotheses about these various phenomena using statistical methods; and implementation of the obtained results in various fields related to the use and learning of natural languages. Each linguistic unit is researched as a component of a language system. Studies of earlier and newer collections of texts have made it possible to confirm both previously defined and new statistical parameters and patterns of textual structuring as well as the structural features of various languages. The article’s particular focus is on text attribution (authorship). Provided are the results of statistical studies of literary works by Vasyl Stefanyk, Mykhailo Iatskiv, Lina Kostenko, and Petro Karman’skyi carried out in the Department of Applied Linguistics of Lviv Polytechnical National University. Calculated were the frequency of alphabetic symbols and their euphony, and the required minimum sample size vis-à-vis the entire volume of an author’s texts is determined. It is posited that such samples must be formed by means of the random selection of textual fragments from all of an author’s works. Our tests obtained a negative result regarding the hypothesis that the frequency of such symbols may indicate authorship. The number of paragraphs in Stefanyk’s and Iatskiv’s works, of the number of sentences in each of their paragraphs, of the number of word forms in the sentences, and the word forms’ length were statistically established.