Quantitative investigations in Hungarian phonotactics and syllable structure

This dissertation investigates statistical properties of segment collocation and syllable geometry of the Hungarian language. A corpus and dictionary based approach to studying language phonologies is outlined. In order to conduct research on Hungarian, a phonological lexicon was created by compiling existing dictionaries and corpora and using a system of regular expression rewrite rules (based on letter-to-sound rules) in order to derive pronunciations for each word in the lexicon. The resulting pronunciation dictionary contains not only pronunciations in several transcription systems but also syllable counts and syllable boundaries, corpus frequencies, and vowel and consonant projections for each entry.;The highlight of the dissertation is an investigation into whether the rhyme or body can be posited as an intermediate node in the structure of the Hungarian syllable. While both consonant-vowel and vowel-consonant sequences in Hungarian exhibit both attracting and repelling connections, on average the language behaves neither as a rhymetype language (such as English) or a body-type language (such as Korean). It is suggested that a more-nuanced description of the Hungarian syllable is required, and an alternative representation is proposed.;The dissertation makes several contributions to research in Hungarian phonology. Results include insights into the statistics of phone-based n-grams for Hungarian, previously unknown phonotactic restrictions, and data on syllable structure with consequences for modeling the structure of the Hungarian syllable. In the cross-linguistic context, the dissertation inspires related quantitative phonotactic research on disparate languages while simultaneously suggesting several caveats that should be taken into account when performing similar studies. In particular, the difficulties of quantitative studies of the syllable structure of multisyllabic words are addressed.