Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Daytime nap consolidates learning of Cantonese tones in non-tonal language speakers: sleep-EEG evidence.
Poster Session D, Thursday, October 1, 4:30 - 6:30 pm, Wangari Maathai
Quentin Zhen Qin1, Jiayu Liang1, Susu Lai1, Caicai Zhang2; 1The Hong Kong University of Science and Technology, 2The Hong Kong Polytechnic University
Prior research shows that post-training sleep facilitates the consolidation of newly learned speech sounds, such as consonants. Lexical tones, which use pitch variations to signal word identity, are a useful case for examining sleep-mediated speech learning because they are challenging for adult second-language learners. A daytime nap study of (Mandarin-speaking) adults who speak tonal languages found neither a nap-specific behavioral benefit nor sleep-EEG support: both the nap and wake groups showed offline consolidation, and sleep-EEG parameters, spindle density, and slow oscillations (SOs) did not predict the nap group’s performance. Recent work examining (English-speaking) adults who don’t speak tonal languages, however, found that nap facilitated higher-level Mandarin tone categorization relative to wakeful rest, but this nap-specific benefit was not associated with expected sleep parameters, such as NREM sleep. While nap-mediated consolidation of Cantonese tones would likely occur in non-tonal language speakers, its association with sleep-EEG parameters needs further investigation. This study addresses this among the Hong Kong ethnic-minority population from South/Southeast Asia, who speak non-tonal languages and face persistent challenges acquiring spoken Cantonese. Seventy-eight participants with limited spoken Cantonese exposure were randomly assigned to either a nap group or a non-nap group (n = 39 for each). They were trained to identify Cantonese words contrasting in contour and level tones (240 trials per contrast, order counterbalanced). After training, participants completed tone identification test 1 with trained stimuli (ID1; 80 trials per contrast) as a pre-test. They then completed a 90-min manipulation: the nap group napped with an EEG recording, whereas the non-nap group remained awake for non-verbal activities. Following this manipulation, participants completed an identical identification test 2 (ID2) as a post-test, with changes in ID2–ID1 indexing consolidation effects. Spindles were identified across the ten scalp electrodes from nap participants’ NREM (i.e., N2 and N3) sleep, and SOs were identified in frontal electrodes from NREM-N3 sleep using Yet Another Spindle Algorithm (YASA). Mixed-effects models were fitted to the accuracy of ID1 and ID2 for the nap and non-nap groups. The model revealed a significant Group × Session interaction (z = -2.18, p = .029). Post-hoc analyses showed that the nap group improved from ID1 to ID2 (z = -2.80, p = .005), whereas the non-nap group did not. The analyses of the sleep-EEG data showed that nap participants’ spindle density (z = 3.73, p < .001) and SOs density (z = 3.01, p = .002) positively predicted consolidation, with better nappers showing greater improvement. The results provide behavioral and sleep-EEG evidence that a daytime nap supports Cantonese tone consolidation in non-tonal language speakers. Specifically, only the nap group showed an ID1-to-ID2 improvement, suggesting nap-mediated offline consolidation of newly learned tone categories. This effect was further supported by sleep-EEG evidence, as higher spindle and SOs density predicted greater ID1-to-ID2 gains in the nap group. Together, these findings suggest that sleep-related neurophysiology contributes to the consolidation of newly learned Cantonese tones, highlighting a key role of sleep-dependent memory consolidation in supporting lexical tone learning among non-tonal language speakers.
Topic Areas: Speech Perception, Prosody