Rhythmic pattern of American English: An articulatory and acoustic study

The rhythmic pattern of American English is studied within the framework of the Converter/Distributor Model (Fujimura, 1992; 2000a; 2000b). The data used in this study comprises the articulatory and acoustic signals recorded at the University of Wisconsin Microbeam facility for four speakers of American English. Subjects were asked to repeat the same correction of one digit in a three-digit sequence consisting of "five" or "nine" followed by "Pine Street." All target syllables share the same low vowel as their nucleus. By analyzing jaw movements, syllable magnitudes are evaluated, in order to infer a linear syllable pulse train for representing the rhythmic organization of the utterance. Articulatory syllable duration is derived from the syllable magnitude based on the assumption of the C/D model. The pattern of syllable magnitudes and boundary magnitudes are then studied to describe the rhythmic pattern of semi-spontaneous speech in reference utterances. The phrasal reorganization in the context of contrastive emphasis, when the utterance is repeatedly corrected is also discussed. Acoustically determined syllable and boundary durations are compared with the articulatorily derived syllable/boundary durations. The alignment of tone and stress is also studied. Findings on the acoustic correlate of jaw opening are also reported by analyzing F1 characteristics for unemphasized and emphasized syllables.