Poster A12, Thursday, August 16, 10:15 am – 12:00 pm, Room 2000AB

The impact of alphabetic script acquisition on Cantonese phoneme judgement: a combined behavioral and ERP study

Yubin Zhang1, Jessica Ka Yui Leung1, Chotiga Pattamadilok2, Dustin Kai Yan Lau1, Mehdi Bakhtiar1, Caicai Zhang1,3;1The Hong Kong Polytechnic University, 2Laboratoire Parole et Langage, CNRS, 3Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences

Acquisition of an alphabetic script induces remarkable behavioral and cerebral changes in speech processing. For example, it can lead to enhanced phonological awareness — the ability to segment the spoken language into smaller sound units. However, only a few studies have examined the extent that these changes also occur in logographic readers, like Chinese readers, who have learned a Romanization system of their native language during adulthood. In this study, we focused on the impact of learning Jyutping, a Romanization system of spoken Cantonese, on explicit phoneme judgement and its underlying neural activities in native Cantonese speakers. A GROUP (Jyutping + Chinese vs. Chinese) × STIMULUS (speech vs. pure tone) × LIST (global vs. segmental) design was adopted. Fourteen native Cantonese speakers with both logographic Chinese and Jyutping proficiency (Jyutping + Chinese group) and 12 matched controls only literate in Chinese (Chinese group) were tested on a Cantonese initial phoneme judgement task (speech) and pure tone judgement task (pure tone). The two groups were matched on their English proficiency. On the initial phoneme judgement task, participants judged whether pairs of Cantonese CV(C) pseudowords had the same initial, while on the pure tone judgement task, they judged whether pairs of tone-triplets shared the same initial tone. The pure tone task served as a control condition, as the impact of Jyutping acquisition was not expected to generalize to other auditory discrimination tasks. The test materials for both tasks were further divided into two lists – global and segmental. The global list included pairs of pseudowords with the same rimes (e.g., paa2 and faa2) and pairs of tone-triplets with the same non-initial tones (e.g., 200Hz1600Hz800Hz – 200Hz1600Hz800Hz), whereas the segmental list contained materials with more irrelevant variations, namely different rhymes (e.g., paa2 and paai2) or different non-initial tones (e.g., 200Hz1600Hz800Hz – 200Hz1400Hz1800Hz). These variations could interfere with the judgement of the initial sound, pushing listeners to adopt a ‘segmentation’ strategy, especially in the speech-segmental condition. Both behavioral (RT and accuracy) and EEG data were collected for analysis. Accuracy results showed that both groups performed significantly worse in the speech-segmental condition than in other conditions. As for RT, a marginally significant three-way interaction indicated that the group difference in RT on correct trials was largest in the speech-segmental condition, where the Chinese group performed much slower than the Jyutping + Chinese group. Electrophysiologically, compared with the Chinese group, the Jyutping + Chinese group exhibited a prolonged negativity in the 200-300 time-window (N200) and the following 300-400 ms time-window (P300), while processing the second syllable in the speech condition. This group difference might be attributed to their different phonological processing abilities, namely, phonological awareness and working memory. Enhanced phonological processing skills were found in Cantonese native speakers who had learnt Jyutping, at both the behavioral and neural levels. This finding suggests that speech processing, especially explicit phoneme judgement, can be changed even for logographic readers who have acquired an alphabetic script late in adulthood. This underscores the plasticity of the human brain even in adulthood.

Topic Area: Phonology and Phonological Working Memory