| Abstract | Three experiments examined the conditions under which repeated words undergo durational shortening in speech. Exp 1 found no shortening when words were produced in lists. In Experiment 2, reductions were observed for the same words produced in meaningful prose.
Words preceded by homophones did not undergo shortening. Shortenings may reflect talkers' exploitation of a word's redundancy in the context of a discourse. Experiment 3 found more shortening of content words produced in a communicative context than in the same discourse,
transcribed and read into a microphone. The tendency to shorten may be due to the presence of a listener, or it may reflect the faster speech rate characteristic of spontaneous as compared with read speech. |