Search Abstracts | Symposia | Slide Sessions | Poster Sessions
An investigation of the functional neuroanatomy of lexical tone encoding in overt Mandarin word production with functional near-infrared spectroscopy (fNIRS)
Poster Session A - Sandbox Series, Thursday, October 24, 10:00 - 11:30 am, Great Hall 3 and 4
This poster is part of the Sandbox Series.
Xiaocong Chen1, Tai Yuan1, Fumo Huang1, Caicai Zhang1; 1The Hong Kong Polytechnic University
Introduction: Current psycholinguistic models of Chinese word production assume that the encoding of lexical tone mainly consists of two stages: the retrieval of the tonal information (either as independent segments or as part of the metrical or word-shape frame), and the tone-to-syllable mapping/integration process (i.e., assigning the lexical tone to the syllabic or segmental contents) (Alderete et al., 2019; J. Chen, 1999; O’Seaghdha et al., 2010; Roelofs, 2015), which was supported by our recent EEG study (Chen & Zhang, submitted). But the functional neuroanatomy of these subprocesses prior to articulation is largely unknown. Thus, we aimed to investigate the neural bases of lexical tonal encoding during overt Mandarin word production using functional near-infrared spectroscopy (fNIRS), which has reasonable tolerance of head motion and spatial resolution. Based on previous neuroimaging studies on phonological encoding of speech production (e.g., Chang et al., 2014; Liu et al., 2006; Indefrey, 2011), we hypothesized that the tonal encoding may involve several brain regions, including superior temporal gyrus, inferior parietal lobe, inferior frontal gyrus and motor areas. Methods: Forty-nine native Mandarin speakers (25 females, aged 19-32) completed a phonologically primed picture naming task, where they produced a disyllabic name upon seeing an object picture, with their naming latencies and fNIRS signals recorded. Each picture was preceded by a monosyllabic visual prime, which was presented simultaneously with its spoken form auditorily. To disentangle the two tonal encoding sub-processes (tonal retrieval and tone-to-syllable integration), we orthogonally manipulated the tonal relatedness and syllabic relatedness of the primes to the first syllable of the target picture names (e.g., 鹦鹉, ying1-wu3, ‘parrot’). This created four conditions of primes: a homophonous prime with both syllabic and tonal overlap (e.g., 英, ying1), a syllabic prime with only syllabic overlap and no tonal overlap (e.g., 硬, ying4), a tonal prime with only tonal overlap and without segmental overlap (e.g., 滋, zi1), and an unrelated control prime without any tonal and syllabic overlap (e.g., 志, zhi4). Preliminary results: The behavioral data (naming latencies) revealed significant main effects of tonal relatedness and syllabic relatedness and a significant interaction effect. Specifically, the tonal relatedness effect was modulated by syllabic relatedness, as there was a tonal facilitation effect when the syllable was shared between the prime and the target word, but a tonal interference effect when the syllable was different. This showed that the lexical tonal overlap could not be planned in advance, but only when the lexical tone and syllable were both available, the production could be facilitated, which indicated there was a tone-to-syllable integration process. The fNIRS data analysis is currently in progress. If a significant main effect of tonal relatedness with an absence of a significant interaction between tonal and syllabic relatedness is found in a channel, this will indicate the tonal retrieval process. If a significant interaction effect between tonal and syllabic relatedness is found in a channel, especially if the tonal relatedness effect is modulated by the syllabic relatedness as in the behavioral data, this will indicate the tone-to-syllable integration process.
Topic Areas: Language Production, Phonology