用听觉事件相关电位研究声调语言词的长时程记忆痕迹

发布时间：2018-06-08 13:01

本文选题：言语感知 + 声调　；参考：《中国科学技术大学》2012年博士论文

【摘要】：语言是人类交流的重要工具,也是人类文明保存和传承的重要载体。人类大脑加工语言的神经机制,是近现代脑研究技术发明以后兴起的重要研究领域。语言的认知神经科学研究可以分为听觉言语感知(speech perception)研究和言语输出(speech production)研究,视觉语言文字的感知和输出研究。在言语认知研究领域,近二十年来最大的进展是发现了言语声音在大脑中的长时程记忆痕迹(long-term memory traces)。一系列的实验结果发现了听觉言语单元,如元音(vowel)、音节(syllable)、词(word)和动词短语(phrasal verb)等在大脑中的长时程记忆痕迹。这些长时程记忆痕迹在言语信号进入大脑后被迅速激活,是大脑进一步加工理解复杂言语信号的基础。词在大脑中的长时程记忆痕迹以芬兰语为实验材料被首先发现。后来以英语和德语等为实验材料进一步验证。这些实验都是以被动oddball范式呈现听觉言语刺激,当真词作为偏差刺激时诱发的失匹配负波(mismatch negativity, MMN)比伪词作为偏差刺激时诱发的失匹配负波幅度更大,反映了真词在大脑中长时程记忆痕迹的激活。到目前为止,词的长时程记忆痕迹研究都是以非声调语言(non-tonal language)中的词为实验材料。这些词都是由音段音位(segmental phoneme)构成,而没有超音段音位(superasegmental phoneme)的参与。然而,世界上已发现的六千多种语言中绝大部分(60-70%)属于声调语言(tonal language)。在声调语言中,音节(syllable)的基频(fundamental frequency, f0)轮廓可以决定词义(word meaning),这样的基频轮廓称为声调(tone)。声调信息和音段(segment)信息(包括辅音和元音)是并行进入听觉系统的,并且一个声调可以跨越多个音段,因此声调是典型的超音段音位。本研究所要回答的问题是声调作为超音段音位是否也参与了词的长时程记忆痕迹的形成和激活。在本研究中,18名以汉语普通话为母语的受试者参与了研究。两个汉语单音节([huo4]和[kuo4])作为标准刺激(每一个的出现概率是41.67%),四个汉语双音节([huo4da2],[kuo4da4],[huo4da4]和[kuo4da2])作为偏差刺激(每一个的出现概率是4.17%),以被动的oddball范式通过耳机呈现给受试者。四个作为偏差刺激的双音节中,第二个音节所携带的声调决定了该双音节是真词([huo4da2],[kuo4da4])还是伪词([huo4da4],[kuo4da2])。实验结果发现,真词诱发的听觉事件相关电位(auditory event-related potential, AERP)比伪词诱发的听觉事件相关电位在真词的词识别点(word recognition point)之后的164毫秒附近出现显著的负向偏转。这个负向偏转反映了真词在大脑中的长时程记忆痕迹的激活。根据实验结果可以推测,在语言习得过程中,声调信息和音段信息一同参与了大脑中词的长时程记忆痕迹的形成。同样,在言语感知过程中,声调信息和音段信息一同参与了词的长时程记忆痕迹的激活。并且,声调语言词的长时程记忆痕迹激活的峰值发生在词识别点之后的164毫秒,同非声调语言词的长时程记忆痕迹激活的峰值时间一致(词识别点之后的约150毫秒),说明声调语言和非声调语言词的认知神经加工机制可能是相同的,声调与音段并行进入听觉系统并没有影响大脑感知词的时间进程。
[Abstract]:Language is an important tool for human communication, and it is also an important carrier for the preservation and inheritance of human civilization. The neural mechanism of human brain processing language is an important research field after the invention of modern brain research technology. The cognitive neuroscience research of language can be divided into speech perception research and speech output (sp Eech production) research, the perceptual and output research of visual language. In the field of speech cognition, the biggest progress in the past twenty years is the discovery of the long term memory trace of speech sounds in the brain (long-term memory traces). A series of experimental results have found the auditory speech unit, such as the vowel (vowel), the syllable (syllable), and the words. (word) and the verb phrase (phrasal verb) and other traces of long term memory in the brain. These long history memory marks are quickly activated after speech signals enter the brain and are the basis for further processing of the brain's understanding of complex speech signals.
The long term memory marks in the brain were first found in Finnish as experimental materials. Later, the experimental materials were further verified in English and German. These experiments were used to present auditory stimuli with passive oddball paradigm, and the mismatch negativity, MMN, as a result of the bias stimulus, was more than a pseudoword. The mismatch negativity induced by bias stimulation is more significant, reflecting the activation of long term memory trace in the brain.
So far, the long term memory traces of the word have been used as experimental materials in non-tonal language. These words are made up of segmental phoneme rather than superasegmental phoneme. However, most of the six thousand languages found in the world (60-7) 0%) belongs to the tone language (tonal language). In tone language, the syllable (fundamental frequency, F0) outlines can determine the meaning of word (word meaning), which is called the tone (tone). Tone information and segment (segment) messages (including consonants and vowels) are parallel into the auditory system, and a tone can be used. In order to cross a number of segments, the tone is a typical supra phoneme. The question to be answered in this study is whether the tone as a supra phoneme also participates in the formation and activation of the long term memory traces of the word.
In this study, 18 Chinese speakers of Mandarin were studied. Two Chinese Monosyllabic syllables ([huo4] and [kuo4]) were used as standard stimuli (the probability of occurrence of each one was 41.67%), and four Chinese double syllables ([huo4da2], [kuo4da4], [huo4da4] and [kuo4da2]) were used as deviations (the probability of occurrence of each one was 4.17%), with a passive OD The dball paradigm is presented to the subjects through the headphones. In four disyllabic syllables as a deviation, the tones carried by second syllables determine whether the double syllable is a true word ([huo4da2], [kuo4da4]) or a pseudo word ([huo4da4], [kuo4da2]). The experimental results show that the auditory evoked potential related potential (auditory event-related potential, AERP) is induced by the true word. The auditory event related potential induced by the pseudo word is marked by a significant negative deflection near the word recognition point (word recognition point) of the true word. This negative deflection reflects the activation of the long term memory trace of the true word in the brain.
According to the experimental results, it is conjectured that in the process of language acquisition, tone information and sound segment information are involved in the formation of long term memory traces in the brain. In the process of speech perception, tone information and segment information participate in the activation of long term memory traces together with the long history memory marks of the tone language words. The peak of trace activation occurs at 164 milliseconds after the word recognition point, which is the same as the peak time of the long time memory trace activation of the non tone language words (about 150 milliseconds after the word recognition point). It shows that the cognitive processing mechanism of the tone language and the non tone language may be the same, the tone and the sound segment enter the auditory system in parallel. The time process that affects the brain's perceptive words.
【学位授予单位】：中国科学技术大学
【学位级别】：博士
【学位授予年份】：2012
【分类号】：R767.92

【引证文献】