Mandarin-Learning Preschoolers With Cochlear Implants Prioritize Prosody Over Semantics When Recognizing Emotion in...

Accurate identification of emotion in speech requires integrating semantic (what is said) and prosodic (how it is said) cues, which can conflict in contexts such as irony, humor, insincerity, etc. Like adults, Mandarin-learning preschoolers with typical hearing (TH) preferentially use prosody when cues conflict, indicating early, adult-like cue-weighting strategies. However, because cochlear implants (CIs) only provide pitch information with reduced spectrotemporal resolution, preschoolers with CIs have limited access to speech prosody and might rely more on semantics, potentially showing a different strategy from TH peers. Therefore, this study examined cue weighting in Mandarin-learning preschoolers with CIs when identifying emotions with conflicting cues. We hypothesized that children with CIs would either show a prosody-semantic balanced strategy or a semantic-dominant strategy.

Thirty-one Mandarin-learning preschoolers with CIs (4 to 7 yr) and 31 age-matched TH controls completed a 2-alternative forced-choice task (i.e., happy versus sad). Stimuli were semantically positive or negative sentences spoken with happy or sad prosody, creating prosody-semantics incongruent expressions. Children judged the speaker's emotion on each trial. The proportion of prosody-based responses (judgements matching the emotion expressed by prosody) was calculated and compared between groups (note that 8 children with CIs who did not perform above chance in an unambiguous control condition were excluded from the main analyses, leaving 23 CI participants for comparison with 31 TH controls).

Both TH and CI groups provided prosody-based responses significantly above chance (50%), indicating a preference to follow prosody rather than semantic content when cues conflicted. Furthermore, there was no significant group difference in the proportion of prosody-based responses, suggesting that preschoolers with CIs prioritized prosody to a similar extent as their TH peers, at least in the current closed-set task involving two extreme emotions (happy and sad).

Despite receiving auditory input with reduced spectrotemporal resolution, Mandarin-learning preschoolers with CIs who were able to use prosody exhibited TH-like, prosody-dominant cue-weighting strategies when interpreting conflicting cues of two basic emotions (i.e., happiness versus sadness). These findings might suggest early adaptive use of preserved prosodic information (e.g., duration, intensity) and support incorporating training on nonpitch prosodic features into rehabilitation to enhance emotion understanding and social communication in preschoolers with CIs.