TEXT COMPLEXITY: STUDY PHASES IN RUSSIAN LINGUISTICS

The practice of Mathematics and Statistics application to determine text complexity in national Russian linguistics is subdivided into three stages: 1) quantitative characteristics period, 2) qualitative and quantitative characteristics period, 3) quantitative characteristics period. During the first period in the end of the 20th century the study of text complexity characteristics is primarily focused on quantitative parameters which are sentence length and word length (Lesskis (1964), Matskovskiy (1976), Mikk (1970, 1981), Tuldava (1975, 1979)). At about the same time the range of parameters was extended to the number and types of syntactic constructions and their frequency (Piotrovskiy (1973)), lexical characteristics: word abstractness (Mikk (1970, 1981), Rozenberg (1978), Sokhor (1975)) and polysemy (Tuldava (1975, 1979)). The second period at the turn of the 20th-21st centuries is the time of the first computer software for text analysis: Software KONUT (Mal'kovskiy (1997)), Russian Context Optimizer for DBMS (Ermakov (2002)) etc. At present, that is during the third period, scholars are aimed at upgrading the existing mathematical models of text complexity analysis. Qualitative parameters (sentence length and word length) (Krioni (2008), Oborneva (2006), Shpakovskiy (2007)) as well as the number and type of syntactic constructions (Ermakov (2002), Krioni (2008)) are still in the focus of a number of research works. The word length (number of letters or syllables) analysis is supplemented by the word derivation analysis (Nevdakh (2008), Pushkina (2004), Shpakovskiy (2007)). The number of homonyms (Ermakov (2002) and abstract elements of the text (words, terms, formulae, tables and diagrams) (Grechikhin (2007), Krioni (2008), Pushkina (2004), Shpakovskiy (2007)) are under the most careful consideration at the moment. The introduction of terms "informativeness" and "information" influenced the involvement of new characteristics of text complexity: co-reference (Tolpegin (2008)); co-ordinating cohesion (Krioni (2008)); reiteration, synonymy, hyponymy and hyperonymy, ellipsis, pronominal reference (Abramov (2011)). Text complexity is an up-to-date topic of numerous studies abroad. Coh-Metrix as a powerful tool for text complexity analysis based on five characteristics: narrativity, syntactic simplicity, word concreteness, referential cohesion and deep cohesion, is widely used for English texts.