Abstract:
Despite the greater average duration of spontaneous sentences relative to matched fluently read sentences, it is difficult to identify the physical basis for determining whether an utterance is spontaneous or read. The present experiment compared judgments of spontaneity with judgments of duration for the same speech samples. The test material comprised 25 sentence pairs in which one sentence was spoken spontaneously and one sentence was read, both produced by the same speaker. On each trial, subjects identified the sentence of the pair that was spoken spontaneously, and the sentence that was longer in duration; tests were blocked by judgment type. The greater the absolute duration difference was between the two sentences, the better subjects performed on the duration judgment. The physical differences in duration did not predict the spontaneity judgment, nor did the duration judgment correlate with the spontaneity judgment. The results suggest that physical duration alone is not an effective attribute underlying the perceptual differentiation of spontaneous and read utterances. [Work supported by NIDCD.]