Language-specific realizations of syllable structure and vowel-to-vowel coarticulation.

Abstract

This paper investigates the effects of syllable structure on vowel-to-vowel (V-to-V) coarticulation using Thai and English data. Languages differ in syllable complexity and their realizations of syllable structure. It was hypothesized that languages with complex syllable structure (English) would allow more V-to-V coarticulation than languages with simple syllable structure (Thai). Onset and coda consonants are different acoustically, articulatorily, typologically and perceptually. Onsets are generally 'stronger' and more stable than codas because they are longer, louder, and involve tighter articulatory constrictions. It was hypothesized that closed syllables (that end in a consonant C, i.e., VC#V) would allow more V-to-V coarticulation than open syllables (V#CV). /C(1)V(1)#C(2)V(2)/ and /C(1)V(1)C(2)#V(2)t/ sequences were recorded from six native speakers in Thai and six in English. First and second formant frequencies were measured. Results show that English allows more V-to-V coarticulation than Thai regardless of the intervocalic duration and vowel quality difference, but open and closed syllables only affect V-to-V coarticulation minimally. In addition to syllable structure, other possible factors contributing to the language difference in V-to-V coarticulation are discussed.