Slide Slam D16
Neural responses to explicit processing and implicit representation of lexical tones in Mandarin speakers with Amusia
Nikolay Novitskiy1, Caicai Zhang2, Fang Liu3, Patrick Wong1, Alice Chan4; 1Department of Linguistics & Modern Languages and Brain and Mind Institute, The Chinese University of Hong Kong, Hong Kong SAR, China, 2Research Centre for Language, Cognition and Neuroscience, Department of Chinese and Bilingual Studies, The Hong Kong Polytechnic University, Hong Kong SAR, China, 3School of Psychology and Clinical Language Sciences, University of Reading, UK, 4Linguistics and Multilingual Studies, School of Humanities, Nanyang Technological University, Singapore
Introduction The present study investigates whether pitch deficits in congenital amusia extends to lexically contrastively pitch patterns and whether such deficits affect both explicit processing and implicit representation of lexical tones. Mandarin-speaking adults with and without amusia engaged in two sets of experiments that target neural processing and representation of Mandarin tones and musical pitch. Methods Thirty-six native Mandarin Chinese speakers participated in the study with half identified as amusics by the Montreal Battery of Evaluation of Amusia (MBEA). The subjects participated in two fMRI sessions (processing and representation). The processing session consisted of explicit music and lexical tone discrimination tasks in separate blocks. The representation session comprised a loudness judgement task for a sequence of lexical tone pairs while in half of the blocks the lexical tone repeated across the syllables. We compared subjects with and without amusia for both the tone and music processing tasks and between repeat and non-repeat trials in the representation task. GLM analysis was conducted using SPM12. Regions showing significant task and group differences were used as ROIs in the subsequent correlational and SVM-classification analysies. Results Behaviorally, subjects with amusia were less accurate than those without amusia (control subjects) in both music and tone discrimination tasks. The brain data suggested that, for the processing task, the control group produced stronger activation in the right superior temporal gyrus (STG) in two non-overlapping areas, separately, for the tone and music stimuli, in the ventral and dorsal parts of the gyrus correspondingly. The behavioral accuracy in the music task across all subjects correlated with the activity in the right dorsal STG area that showed the group effect for music. fMRI data classified the participants into amusic and control groups as well as behavioral data in the music task and better than behavioral data in the tone processing task. The representation task revealed one left dorsal STG area with a marginally significant group effect, and the group classification based on activation in this area was at chance level. Discussion We found significant group effects in right STG in the tone processing task, but not in the implicit representation task. Consistent with previous fMRI studies, speech was associated with more ventral part of STG than music (Norman-Haignere et al, 2015; Rogalsky et al., 2011). The performance in music but not tone processing task was correlated with fMRI activity in brain areas showing group effects. Traditional GLM analysis was confirmed by SVM classification analysis. Altogether, our data suggest that while different cortical systems support language and music pitch processing in individuals with and without amusia, such distinction is not shown in the implicit representation task. Lexically relevant pitch deficits in amusia may be restricted to processing only.