Speech rate effects on the realisation of multiple acoustic cues to the Japanese stop voicing contrast.

Abstract

The production of speech at different tempos has consequences for the articulation and perception of linguistic contrasts. For example, in fast speech, segments are often temporally constricted and subject to articulatory undershoot. Although listeners can compensate for rate differences in perceiving phonological contrasts, less is known about how the structure of multiple cues to a contrast is conditioned by changes in speech rate. This study explores how speech rate modulates seven temporal and nontemporal spectral cues to the Japanese stop voicing contrast in spontaneous speech. It is observed that individual cues are subject to variation as a function of speech rate, where all cues undergo reduction or neutralisation in fast speech and the relative importance of each cue changes as a function of speech rate. Ratios between the duration of the stop and surrounding vowels are most informative at slow rates, and the degree of closure voicing is most informative at faster rates. These findings illustrate how the realisation and informativity of multiple cues to a linguistic contrast are conditioned by the articulatory constraints present at different speech rates.