A phonotactic analysis of the content of syllables on word boundaries in spoken and written Czech texts
Abstract
This study aims to provide and analyze a representative list of Czech initial syllable onsets and final codas along with their frequencies of occurrence in running text (token frequencies) and in the vocabulary of unique word forms extracted from it (type frequencies). The frequency data are important because many experiments have demonstrated that phonotactics is not categorical, but rather gradient in nature.
Importantly, the study analyzes and compares both spoken and written texts, using the Czech National Corpus, and the two modalities are hypothesized to yield different outcomes. All words in the sample were transcribed phonemically and analyzed.
A general preference was found for phonotactic structures that are simple in the context of the attested inventory, and the two corpora differed most in the repertoire of complex onsets/codas (some sequences being unique to one modality) as well as in their respective frequencies. The results are discussed in relation to previous studies of Czech phonotactics, and evaluated with respect to implications for phonological theory, focusing on spoken/written and type/token comparisons.