The timing of vowel and consonant gestures

Cross-language differences in the temporal organization of articulatory gestures for vowels and consonants may relate to the overall rhythmic structure of a language. Two models of these differences are compared. In one, successive vowels are timed in relation to each other; in the other, intervening consonants are also taken into account.;The two models of timing were compared using Articulatory Phonology (Browman and Goldstein 1986), in which articulatory gestures are the basis of phonological representation. Movements of the tongue and lips in Japanese and Italian utterances with single and geminate consonants were recorded using the NIH X-ray microbeam facility at the University of Wisconsin, Madison, and the timing of the articulatory gestures was investigated by relating the abstract gestures to measurements of these movements. Virtually every measured interval in the Japanese utterances was longer when the intervocalic consonant was a geminate, suggesting that the coordination between gestures is affected by the length of the consonant, as predicted by the combined vowel-and-consonant timing model. The Italian speakers showed a tendency to maintain a constant interval between the times at which the two vowels reached their target positions, supporting the hypothesis that a change in consonant length does not affect the relative timing of the vowels, as in the vowel-to-vowel timing model.;Temporal patterns of the two languages were modeled by specifying the coordination among gestures in terms of stiffness and phasing relations. These models, when used to predict the durations of the measured intervals, showed that the observed patterns of durational differences between utterances with single and geminate consonants could be produced using an organization based on combined vowel-and-consonant timing for Japanese, and an organization based on vowel-to-vowel timing for Italian.;Finally, some suggestions are made as to how the difference in the timing organization of Japanese and Italian relates to other prosodic characteristics of these languages, particularly the traditional descriptions of Italian as syllable-based and Japanese as mora-based.