Measured and perceived speech tempo: Comparing canonical and surface articulation rates

Studies that quantify speech tempo tend to use one of various available rate measures. The relationship between these measures and perceived tempo as elicited through listening experiments remains poorly understood. This study furthers our understanding of the relationship between measured articulation rates and perceived speech tempo, and the impact of syllable and phone deletions on speech tempo perception. We follow previous work in using stimuli from a corpus of unscripted speech, and in sampling stimuli in distinct ‘global tempo’ ranges. Within our stimulus sets, the differences between canonical and surface rate measurements are directly due to syllable or phone deletions. Our results for syllable rates suggest that listeners use both canonical and surface rates to estimate speech tempo: that is, deletions do not have a consistent effect on perceived tempo. Our results for phone rates suggest that surface phone rate also influences judgements, but canonical phone rate does not. Our results also confirm previously-reported effects of f0 and intensity on speech tempo perception, plus an effect of stimulus duration, but no effect of listeners’ own tempo production tendencies.