Different talkers have different habitual speaking rates, and a given talker will produce a large variation in the rate at which they speak. Meaningful contrasts in speech are cued by duration, and changes in speaking rate alter the distribution of these cues.

Despite this constant restructuring of the relationship between acoustic cues and phonetic categories, listeners experience perceptual constancy, compensating for differences in speaking rate both between and within talkers. We use the identification of sounds at different rates to investigate the extent to which listeners' phonetic category definitions are governed by generalizations, exploring the relative impact on perception of two cues defining the intrinsic linguistic qualities of phonetic categories (place of articulation and voicing), in the context of an independent paralinguistic cue (speaking rate).

Results support the conclusion that listeners do not weight all cues equally, but preferentially weight the voicing cue when resolving ambiguous stimuli.