Keywords: semantic types, polysemy, lexicography, lexicology, perception adjectives, corpus linguistics, Estonian
The article aims to identify the patterns of systematic polysemy for Estonian tactile adjectives using semantic types and corpus data. It also discusses the presentation of semantic patterns in a dictionary. This will help to provide a theoretical basis for lexicographers and to align the presentation of adjective polysemy in dictionaries. The dataset for the study consists of 265 tactile adjectives.
The corpus manager and text analysis software Sketch Engine (Kilgarriff et al. 2004, 2014) is used for corpus context analysis. The program allows to observe the collocations of a word based on various statistics, to generate word sketches, and to study words used in similar contexts by means of the thesaurus function. To find the patterns, semantic types are assigned to meanings and similarities are searched between meaning shifts. For tagging the adjective meanings, the semantic types developed to describe Estonian (polysemous) adjectives (Tuulik 2014; Tuulik, Langemets 2016) are used, and for noun meanings, the semantic types of the Estonian noun developed by Margit Langemets (2010) are used. The research method used for analyzing polysemy in the article is of a practical nature and has grown out of real-life lexicographical work.
A total of 13 polysemy patterns emerged in the tactile adjective group. The most common shift was from tactile characteristics to psychological characteristics (in the sense structure of 43 words). Shifts could be observed from pleasant tactile sensations to positive mental characteristics (e.g pehme padi ‘soft pillow’, pehme loomuga ‘soft-natured’) and from unpleasant touch experience to mentally unpleasant experience (okkaline roos ‘thorny rose’, okkaline iseloom ‘thorny character’). The study of systematic polysemy points to a strong link between the domains of tactile touch and feelings.
The lexicographic presentation of polysemous patterns varied. The pattern was represented as: 1) two distinct primary meanings; 2) primary meaning and submeaning(s); 3) a common aggregate meaning; 4) systematic polysemy that appeared in the corpus but was not presented. The analysis showed that lexicographers may perceive and present one and the same pattern in different ways; therefore, familiarity with polysemy patterns associated with the respective semantic type could contribute to a more systematic presentation.
Maria Tuulik (b. 1985), PhD, Institute of the Estonian Language, Lexicographer (Roosikrantsi 6, 10119 Tallinn), email@example.com