Fred Cummins
Depts. of Linguistics and Cognitive Sci., Indiana Univ., Bloomington, IN 47405
An experiment was conducted to elicit stable forms of rhythm in spoken English. Stimuli were constructed, each of which consisted of a repeated series of the words take and cards. While the interval from one take to the next was a constant (1.5 s), the timing of cards relative to this interval was systematically varied. Subjects were instructed to repeat the phrase take a pack of cards, so that the relative timing of take and cards matched that of the stimulus. After some time, the stimulus was switched off, and subjects attempted to maintain the prescribed timing pattern on their own. Although there were eight target timing patterns in the stimuli, the obtained distribution of subjects' productions was trimodal rather than octomodal. The three modes are hypothesized to correspond to three stable, preferred forms of rhythmic production for this phrase. These data strongly suggest that there is an observable rhythmic basis to English speech production. [Supported by ONR.]