The role of cue weighting and contextual information in speech perception by younger and older adults

Abstract

Normal-hearing older listeners are as accurate as younger listeners when perceiving native English words in quiet despite challenges in temporal processing. Older listeners may compensate for the declined use of fine-grained temporal cues by reducing the weight of temporal cues (VOT) and increase the reliance on other acoustic correlates (F0) of the sound contrast. In Experiment 1, younger (age 18–25) and older (age 55–65) normal-hearing listeners participate in an online 2AFC identification task with /d/-/t/ contrast varying in both VOT and F0. We predict that, while both younger and older listeners rely more on VOT than on F0, older listeners, because of their reduced temporal processing abilities, rely on F0 to a larger degree than younger listeners. Temporal processing not only involves local durational cues of the target segments, but also global contextual cues such as speaking rate. In Experiment 2, the same listeners complete another online 2AFC identification task with /dɑ/-/tɑ/ syllables that vary in VOT and vowel duration (short versus long). We predict that older listeners exhibit a smaller shift in the /d/-/t/ category boundary between the long and short vowel durations than younger listeners since older adults are less sensitive to contextual temporal information.