Sõnaliik leksikograafi töölaual: sõnaliikide roll tänapäeva leksikograafias

Abstract

Sonaliik on traditsiooniliselt keele kirjelduse osa, mis holmab sona morfoloogilist ja suntaktilist kaitumist, aga ka tahendust ning pragmaatilist funktsiooni. Vaatamata sellele, et sonaliigid on tinglike piiridega ning keeleteaduslikult kaugel uhtsest voi selgest nahtusest, on see kategooriana leksikograafias vajalik kui kompaktne viis andmaks ulevaadet sona kaitumisest kasutuses. Ka e-leksikograafias valtimatud automaatsed protseduurid nagu morfoloogiline ja suntaktiline margendamine ning tahenduste uhestamine eeldavad kas konkreetselt voi vaikimisi leksikaalsete uksuste eristamist. Kaesolevas artiklis anname sissevaate sonaliikide rolli leksikograafi toos. Vahendame tulemusi aastal 2019 tehtud uuringust, mille eesmark oli valja selgitada eesti keelt kirjeldavate leksikograafide probleeme ning vajadusi seoses sonaliikidega, aga ka ootusi keeletehnoloogilistele lahendustele sonaliigi maaramise holbustamiseks. ***  On word classes in contemporary lexicography: The lexicographers’ view This study examines the role of word classes in the contemporary lexicographic work. The analysis is based on a metalexicographic semi-structured interview carried out among 23 Estonian lexicographers. We focus on three main research questions: i) How necessary are word classes in lexicopraphers’ work? ii) What kinds of problems they do experience related to the word classes? iii) How do lexicographers solve these problems and how could digital tools facilitate the word class judgement processes? The results indicate that in lexicography, word classes provide an (implicit or explicit) frame for organisation of data, even if word-class tagging does not directly belong among the tasks of the lexicographer. The word class determination process in unclear cases is a complicated one. In addition to the usage of a word and the expected needs of the user, the lexicographer must consider the structural requirements of the particular dictionary (s)he is working with. Along with the development of digital resources, the demands of the data model behind the electronic database have to be considered; not to mention the theoretical linguistic insights. Linguistically, the most problematic issues appear in the grammaticalisation and lexicalisation processes related to inflectional suffixes (mainly adverbs and adpositions) or conversional transposition (adjectives). The typical word-class altering devices in Estonian such as suffixal derivation (see, e.g. Vare 2006: 199) do not emerge as particularly problematic. The interviewed lexicographers would welcome a digital application that would facilitate and quickened the word class determination process. The application would have an improved morphological tagger compared to the present tools and it would be more efficient in extracting statistical information from corpora. Lexicographers wish to acquire information about the prevailing word-class function together with the data this assessment is based on. Yet, the lexicographer makes the final decision.