Department of Psychology, Renmin University of China, Beijing, 100872, PR China.
Department of Psychology, Renmin University of China, Beijing, 100872, PR China.
Neuropsychologia. 2020 Mar 16;140:107382. doi: 10.1016/j.neuropsychologia.2020.107382. Epub 2020 Feb 14.
Languages may differ in their proximate units of phonological encoding in spoken word production. The roles of syllables and phonemes when Mandarin Chinese speakers prepare spoken words remain controversial. In the present study, native Chinese speakers named disyllabic pictures that were preceded by briefly presented and masked prime words while electroencephalogram signals were recorded. Prime words were syllabically or phonemically related to the first syllable of targets or were unrelated. Variables of prime type (syllable overlap or phoneme overlap) and relatedness (related or unrelated) were manipulated in the experiment. Behavioral results demonstrated a critical interaction between prime type and relatedness, with a facilitation effect relative to unrelated primes appearing only in the syllable overlap condition. Spatio-temporal segmentation analysis indicated that syllable relatedness shortened the duration of Map 6 (around 285-410 ms), whose time window is associated with the phonological encoding stage. Comparatively, phoneme relatedness shortened the duration of Map 7 (around 410-480 ms) in a later time window. Onset latency analysis suggested an earlier onset for the syllable effect (284 ms) and a chronologically later onset for the phoneme effect (396 ms). Traditional waveform analysis revealed that the syllable overlap condition exerted significant effects on event-related potentials in the time window of 280-440 ms, whereas the phoneme overlap condition produced a weak effect in the time window of 400-440 ms in a few electrodes. Taken together, these findings substantiate the primacy of abstract syllables in producing Mandarin spoken words and provide initial topographic evidence that Mandarin speakers mentally represent subordinate phonemes at a later stage of word-form encoding.
语言在口语产生的语音编码的近体单位上可能存在差异。在汉语普通话使用者准备口语单词时,音节和音位的作用仍存在争议。在本研究中,母语为汉语的被试在观看短暂呈现且被掩蔽的启动词后命名双音节图片,同时记录脑电信号。启动词与目标词的首音节在音节或音位上相关或不相关。实验中操纵了启动词类型(音节重叠或音位重叠)和相关度(相关或不相关)这两个变量。行为结果表明,启动词类型和相关度之间存在关键的交互作用,只有在音节重叠条件下,与不相关启动词相比,相关启动词才会产生促进作用。时-空分割分析表明,音节相关度缩短了 Map6 的持续时间(约 285-410ms),Map6 的时间窗口与语音编码阶段相关。相比之下,音位相关度缩短了 Map7 的持续时间(约 410-480ms),时间窗口较晚。起始潜伏期分析表明,音节效应的起始时间更早(284ms),音位效应的起始时间较晚(396ms)。传统的波形分析表明,在 280-440ms 的时间窗口中,音节重叠条件对事件相关电位产生了显著影响,而在少数电极中,音位重叠条件在 400-440ms 的时间窗口中产生了较弱的影响。综上所述,这些发现证实了在产生汉语口语单词时,抽象音节的首要地位,并提供了初步的地形证据,表明汉语使用者在单词形式编码的后期阶段以心理形式表示次要的音位。