The interaction between phonological information and pitch type in the processing of lexical tones at attentive stage
【摘要】：Previous studies have revealed two types of information, acoustic and phonological information that are important for tone perception. The acoustic information refers to the physical features of lexical tones as assessed by fundamental frequency(F0), such as pitch height and pitch contour, while the phonological information means the different word meanings expressed by different tonal categories. For example, Mandarin syllable/fu/means skin when it is pronounced with Tone 1(/fu1/), but means father when it is pronounced with Tone 4(/fu4/). Though several studies in the literature have explored how these two types of information in lexical tones were processed at pre-attentive and attentive stages, most of these studies confounded the effect of phonological information with the effect of specific acoustic information, that is pitch type(pitch height and pitch contour) on the processing of lexical tones. One exception is Yu et al.,(2017), which found the interaction between phonological information and pitch type on the extent and time course of lexical tone processing at pre-attentive stage. However, it still remains unclear whether the phonological information interacts with pitch type in the attentive processing of lexical tones. Thus the present study aims to dissociate these two types of effects and further investigate the processing of lexical tones at attentive stage by event-relate potential(ERP) technique. We used real and pseudo Cantonese syllables to differentiate the phonological information from the acoustic information in lexical tones, and chose Cantonese level tones and contour tones to distinguish pitch height from pitch contour(as level tones mainly differ in pitch height while contour tones mainly differ in pitch contour). An active oddball paradigm with four types of deviant stimuli(real syllable with level tone, real syllable with contour tone, pseudo syllable with level tone, pseudo syllable with contour tone) was adopted in the experiment. The results showed significant interactions between syllable type(real vs. pseudo) and pitch type(pitch height vs. pitch contour) in the peak latency of both ERP components, N2 b and P3 b. Specifically, for both N2 b and P3 b, the peak latency of real syllable with contour tones was significantly earlier than that of pseudo syllable with contour tones, while the peak latency of real syllable with level tones was similar to that of pseudo syllable with level tones. In addition, the peak latency difference between real and pseudo syllables with level tones was significantly smaller than that between real and pseudo syllables with contour tones. But no significant interactions were found in N2 b and P3 b amplitude. The results revealed that the phonological information interacted with pitch type in the time course of lexical tone processing at the attentive stage, which was different from the findings at pre-attentive stage(Yu et al., 2017). Participants' attention and decision-making at the attentive stage may result in the different findings between attentive and pre-attentive stages. The study further provided neurophysiological evidence to the attentive processing of acoustic and phonological information in lexical tones, and contributed to the understanding of the neural mechanism of lexical tone processing.