This thesis examines the issue of morphological variation in the Slovene language, by examining the ways in which variation can be investigated and defined using large electronic text collections – corpora. Morphological variation, understood as multiple morphological possibilities of the same grammatical form, is inconsistently and heterogeneously recorded in the freely available online lexicon Sloleks as well as in existing language manuals and dictionaries, and the data are sometimes outdated and need to be reevaluated. Since large computer-readable text collections are available for Slovene, which allow empirical research, we address the problem with a corpus-based approach. This means that the morphological variations that need to be addressed are checked in a reference corpus for the Slovene language, and the results obtained are also compared with the current description of morphological variations in existing language manuals. The key questions, which also form the main hypotheses, are: a) whether and how the corpus-based solutions differ from those currently prevailing in the reference language manuals, and b) whether a given morphological variation can be attributed to a group of words in a systematic way. The main results of the master thesis are a detailed analysis of the description of morphological variation in language manuals, a review of corpus analysis of variation, and a consistent and systematic description of the variants with the information on the frequency of occurrence of each variant.
|
