The use of prosodic cues in sentence processing by prelingually deaf users of cochlear implants

Objectives: The purpose of this study is to assess the use of prosodic and contextual cues to focus by prelingually deaf adolescent users of cochlear implants (CIs) when identifying target phonemes. We predict that CI users will have slower reaction times to target phonemes compared with a group of normally-hearing (NH) peers. We also predict that reaction times will be faster when both prosodic and contextual (semantic) cues are provided. Design: Eight prelingually deaf adolescent users of CIs and 8 adolescents with NH completed 2 phoneme-monitoring experiments. Participants were aged between 13 and 18 years. The mean age at implantation for the CI group was 1.8 years (SD: 1.0). In the prosodic condition, reaction times to a target phoneme in a linguistically focused (i.e., stressed) word were compared between the two groups. The semantic condition compared reaction time with target phonemes when contextual cues to focus were provided in addition to prosodic cues. Results: Reaction times of the CI group were slower than those of the NH group in both the prosodic and semantic conditions. A linear mixed model was used to compare reaction times using Group as a fixed factor and Phoneme and Subject as random factors. When only prosodic cues (prosodic condition) to focus location were provided, the mean reaction time of the CI group was 512 msec compared with 317 msec for the NH group, and this difference was significant (p < 0.001). The provision of contextual cues speeded reaction times for both groups (semantic condition), indicating that top-down processing aided both groups in their search for a focused item. However, even with both prosodic and contextual cues, the CI users' processing times remained slower, compared with the NH group, with mean reaction times of 385 msec for the CI users but 232 msec for the NH listeners (p < 0.001). Conclusions: Prelingually deaf CI users' processing of prosodic cues is less efficient than that of their NH peers, as evidenced by slower reaction times to targets in phoneme monitoring. The provision of contextual cues speeded reaction times for both NH and CI groups, although the CI users were slower in responding than the NH group. These findings contribute to our understanding of how CI users employ/integrate prosodic and semantic cues in speech processing.