Comprehending speech at artificially enhanced rates

Abstract

This study is part of a larger study comparing the production, perception, and long-term comprehension of natural fast speech and artificially sped-up speech. In the 1950's, Fairbanks and colleagues made the interesting claim that artificially compressing speech to half its duration (twice the rate) and listening to it twice is comprehended better than listening to the original natural speed utterance once. Since this finding has been disputed [T. Sticht, J. Expr. Ed., 37, 60-62, 1969], we focus here on providing baseline results for perception of naturally produced speech at conversational and fast rate and of speech sped up to twice its original speed. Specifically, we compare how well the different stimuli are perceived when presented once, twice in rapid succession, and twice with delays of hours and days between presentations. We evaluate perception for both audio-only and audiovisual stimuli. © 2013 Acoustical Society of America.