The Effects of Training Variability and Pitch Aptitude on the Overnight Consolidation of Lexical Tones

Abstract

Purpose: Although variability of training materials has the potential to benefit the learning of lexical tones, the benefit is contingent on an individual's pitch aptitude. Previous studies did not segregate immediate learning and consolidation after an overnight interval, and little is known about how pitch aptitude differences affect consolidation. This study examined whether pitch aptitude predicts overnight consolidation of Cantonese level tones through high-variability (HV) and low-variability (LV) training. Method: Two groups of Mandarin-speaking participants were first assessed in terms of pitch threshold and tone discrimination, which tapped into different aspects of pitch aptitude. They then received Cantonese level tone identification training in either an HV or an LV condition. The participants were trained in the evening, were tested after training, and returned after 24 hr for overnight consolidation assessment. Results: The results indicate that pitch aptitude, measured through pitch threshold, may have predicted overnight consolidation and training progress of the HV group but not those of the LV group. In the HV group, compared with high-aptitude learners, low-aptitude learners benefited temporarily from training variability but did not consolidate the tonal knowledge as much as their high-aptitude counterparts did after 24 hr. Conclusions: The findings suggest that individual learners had difficulty learning nonnative tones by virtue of memory consolidation. Higher pitch aptitude ability (pitch threshold) may provide protection against the decay of learned tones and facilitate tone consolidation. The findings imply that the early emergence of tonal representation is a dynamic process among individuals of nonnative speakers who are exposed to training variability. Supplemental Material: https://doi.org/10.23641/asha.20539998

Phonology, homophony, and eyes-closed rest in Mandarin novel word learning: An eye-tracking study in adult native and non-native speakers

“…Eye movements in Test Phase II only were analyzed. For preprocessing, data were output as a sample report using the SR Research EyeLink Data Viewer and then preprocessed using the VWPre package (v1.2.4; Porretta et al, 2016). To index each unique recording sequence of eye movements, Event was created as a combination of Subject and Trial.…”

“…In terms of data visualization, averaged proportion looks to different interest areas were plotted using the VWPre package (v1.2.4; Porretta et al, 2016). Additionally, as the summary of GAMMs does not indicate the exact shape of regression lines and whether they differ from each other significantly, we used the itsadug package (v2.4;van Rij et al, 2020) to plot the nonlinear smooths and significant difference curves.…”