Leksiskās daudzveidības kvantitatīvā analīze latviešu prozas izpētē

Abstract

The notions of richness of language and richness of the vocabulary are characteristics that are often mentioned in the context of practical literary criticism, however, from the theoretical standpoint, the precise meaning of these terms is not clear. Considering that this area of research is not well known in Latvia, the paper first outlines the field of study of vocabulary richness and lexical diversity and maps out the challenges that are related to it, including issues pertaining to selecting the focus and methodology of the study, selecting data, and evaluating the quality of data. Furthermore, the interpretation of the results of the analysis of lexical diversity is discussed, arguing that the vocabulary richness of a literary work does not necessarily indicate only the level of language competencies and artistic skill, it can also be interpreted as a trait of a literary genre or a subject of the artistic choice. Two methods of accounting for lexical diversity are analysed more closely – mean segmental type-token ratio (MSTTR) and moving window type-token ratio distribution (MWTTRD); the main idea and application, as well the place of these approaches among other methods are explained. To conduct the test study, the authors are using a small corpus that entails six Latvian prose writers – Alberts Bels, Augusts Deglavs, Regīna Ezera, Jānis Jaunsudrabiņš, Ivande Kaija, and Zigmunds Skujiņš. In the course of the study it is established that higher values of lexical diversity are achieved by the authors who represent more recent generation – Skujiņš, Ezera, Bels. To suggest further possibilities of employment of these methods in the studies of the stylistics of literary prose, the authors of the paper turn attention towards the diversity of adjectives in the works of the corpus. The MSTTR and MWTTDR of adjectives are measured; the percentage of adjectives that are shared by all authors and the percentage of unique adjectives for each author are being calculated. In conclusion, further developments of quantitative and qualitative digital analysis of prose texts are being suggested.