DISTRIBUTIVE MODELS IN THE PROJECTION OF THEORETICAL APPROACHES OF LEXICAL SEMANTICS (ACCORDING TO THE RNC)
Abstract:
The article analyses the data of the semantic vector model of the main corpus of RNС. Groups of noun associates whose contextual proximity exceeds the threshold of 0.5 are considered. Theoretical methods of analysing lexical-semantic relations are applied to the corpus data of the associate groups. More abstractly, the main question of the article can be formulated as follows: to what extent do existing theoretical ideas about lexical-semantic relations between lexemes in a language, such as hyperonymy (genericspecies relations), meronymy (part-to-part relations), synonymy, correlate with corpus data on the proximity of the contexts of use of these lexemes? The article shows that the attraction of data on the proximity of semantic vectors makes it possible to see a more complexly organised system of lexical-semantic relations. This system allows us, on the one hand, to compare pairs of lexemes within a type of relation on the closer- farther scale and thus to identify the analogue of semantic prototypes on the basis of corpus data. On the other hand, we show the dependence of lexical-semantic relations on extra-linguistic parameters, such as their distribution in the sphere of textual use, the association of lexemes’ use with certain textual genres, as well as socio-cultural factors infl uencing the dominance of certain contexts.