dissertation

III. Dissertation (in preparation)

• Proposal defense (Aug.2007)
• Title: Variability of onset consonants in relation to prosody: intra-/inter- language variation
• Abstract:

Previous research has established that prosody affects the phonetic realization of segments: that is, segments in strong positions are greater in the articulatory magnitude and stronger in the relevant acoustic dimensions (Klatt, 1976; Turk, 1992; Krakow, 1993; among others). In addition, prosodically conditioned variation has been reported to show language specific patterning (Lavoie, 2001; Keating et al, 2003), and to be a factor in sound change: e.g., relocation of aspiration in Dakhini Urdu (Moinuddin and Rahul, 2000) and consonant gemination in Proto-Lezgian (Topuria, 1974), both of which are conditioned by prosody.The presence of language-specificity in particular leads us to question how such variation is influenced by other linguistic factors.
This dissertation is concerned with patterns within segment variation under stress, pursuing the hypothesis that stress-induced variability is constrained by two linguistic factors: languages’ sound inventories and syntagmatic relations within the syllable. In other words, certain differences in consonant or vowel inventory will affect how consonants and vowels behave in relation to stress. The first factor can be verified from the studies by Manuel (1990) and Lavoie (2002), which respectively documented vowel variation and consonant variation referring to the inventory differences. The second factor is hypothesized assuming that syllable is the minimum unit for stress realization, and that onset and rhyme behave differently in manifesting stress. Perhaps more subtle and less frequent, there are stress effects on onsets: e.g., occurrence of aspiration, voicing change, and flapping in English. In other words, all syllable constituents are supposed to contribute stress manifestation based on their syntagmatic relations within the domain.
In order to test the hypothesis, empirical data will be collected from two groups of languages; 1) two dialects Spanish (Argentinean and Mexican Spanish) for intra-language variation and 2) three Germanic languages (English, German, and Dutch) for inter language variation. The data will be analyzed acoustically and described the phonetic details of variability among the groups. One of the expected contributions to scholarship is to provide the cross-linguistic speech data, based on which we may better understand allophonic variation and roles of linguistic factors in manifestating stress in languages.