汉语口语产生中音节和音段编码的认知机制

PSYCH OpenIR > 认知与发展心理学研究室

	汉语口语产生中音节和音段编码的认知机制
其他题名	The Cognitive Mechanism of the Encoding of Syllables and Segments in Mandarin Speech Production
	岳源
	2014-11
摘要	口语产生即为将思想转化为口头言语表达的过程，其中包含了概念准备、词条选择、词素音韵编码、语音编码和发声五个主要阶段，在口语产生的内部准备和外部发声过程中同时伴随了自我监控。其中，词素音韵编码阶段作为语音信息提取的关键阶段，受到了广泛的重视，针对这一阶段中编码单元的探索是词素语音编码阶段研究的热点之一。在汉语口语产生的研究领域中，这一问题存在着争论。一部分研究者认为汉语口语产生中音韵编码阶段的编码单元是音节，同时也有一部分研究者认为这一单元并非音节，而是音素或音段，两种观点均有其相关实验证据支持，在前人研究的基础上，为了进一步探索音段和音节在汉语口语产生过程中编码的认知机制，本研究有针对性地设计了五个实验，分为两个研究方面，研究一采用行为实验对音节和音段的编码机制进行探索，研究二采用事件相关电位技术，利用其高时间分辨率的优点，在行为实验的基础上进行进一步探究音节与音段、音素效应产生的时间进程。研究一中，与无关条件相比，在包含音韵编码、语音编码和发声阶段的即时命名任务（实验一）中，干扰词与目标图名称音节相关和音段的重合显著地缩短了图画命名时间，表现出音节和音段促进效应；在只包含发声阶段的延迟命名任务（实验二）中，音节相关和音段相关条件显著地延长了图画命名的时间，表现出音节和音段抑制效应；在包含语音编码和发声阶段的延迟命名和发音抑制结合的任务（实验三）中，音段相关条件显著地延长了图画命名时间，表现出音段抑制效应。结果表明，音节和音段的促进效应发生在汉语口语词汇产生中的音韵编码阶段，音节和音段的抑制效应可能发生在语音编码或者发音阶段。效果量（Cohen d）的分析表明音节的促进效应强，而音段的促进效应弱，音节是音韵编码过程的合适单元，为合适单元假设提供了支持证据。与音节相比，音段在语音编码和发音阶段的效应量较大，表明音段在运动执行过程中可能起了相对重要的作用，支持了口语产生中词汇表征准备阶段与运动阶段分离的观点。研究二中，实验四将即时命名的图词干扰范式与事件相关电位技术相结合，得到了与实验一一致的结果，且音节效应的发生在图画出现后 200ms，早于音段效应约150ms，为合适单元假说提供了证据。实验五结合了脑电技术与内隐启动范式，同样发现音节的效应发生早于音素条件。基于研究一和研究二的结果，得出结论： 1）音节在汉语口语产生过程中的激活早于音素和音段； 2）支持合适单元假说，在汉语中音韵编码阶段的合适单元是音节，但是音素和音段会在音节激活后被提取。
其他摘要	Speaking is a process to transform inner thoughts to outer expression. It involves stages of conceptual preparation, lemma selection, morphophonological encoding process, phonological encoding process and phonetic encoding. Along with the whole speech preparation, there’s a self monitoring which can detect errors and stop them in time. As a key stage of speech production, morphophonological encoding attracts widespread concern. And the encoding unit of this process is a research focus. In the field of Chinese speech production researching, this issue remains controversial. It’s been widely accept that the encoding unit of this process in Chinese is syllable. However, there’re still disagreements that claim this unit is phoneme or segments. Both positive and negative experimental evidence exist. Base on former studies, present research applied two parts of study to further investigate the function of syllables, phonemes and segments in the phonological encoding stage of Chinese speech production. Study 1 applied three behavioral experiments to identify that in which process syllables and phonemes work. Study 2 combined Event-related Potentials to traditional paradigm to investigate the time course of the activation of syllables, phonemes and segments. In Study 1, Syllable and segments facilitating effects were found in the instant naming task (exp.1) which contains phonological encoding, phonetic encoding and articulation; Rhyme and syllable inhibiting effects were found in the delayed naming task (exp.2) which contains only the process of articulation; the inhibiting effects of rhyme and the beginning segment in the combination of delayed naming and articulation suppression (exp.3) which contains the process of phonetic encoding and articulation. By comparing these effects among three tasks, we suggest that syllable and segments facilitation effects localized at the stage of phonological encoding, whereas syllable and segments inhibition effects localized at the stage of phonetic encoding and (or) articulation. The analysis of effect size (Cohen d) suggests that the effect size of the syllable facilitation is larger than that of segments facilitation; support the hypothesis of the proximate unit. Meanwhile, the effect size of syllable inhibition is smaller than that of segments inhibition. This phenomenon supports the assumption of the interdependence between pre-motor stage and motor stages. In study 2, the exp.4 combined PWI paradigm and ERPs and got the same effect with the exp.1 and the effect of syllable starts 150ms earlier than the effect of segments at 200ms. The analysis of the waveform indicates that the effect of the syllable and the beginning segment locate in the frontal and central region of the brain and the effect of the rhyme locates in the posterior region. This result further supports the proximate unit hypothesis and the assumption of the interdependence between pre-motor stage and motor stages. In the combination of implicit priming task and ERPS (exp.5), the waveform of syllable homogenous group is differing from that of the heterogeneous group. Phoneme homogenous group is differing from the heterogeneous group as well. And the effect of phoneme starts later than the effect of syllable. Base on above researches, we draw to the conclusion: 1) The activation of syllable is earlier than that of phoneme and segments; 2) The hypothesis of proximate unit is reasonable and this unit in Chinese speech production is the syllable. Phonemes and segments will be activated after the retrieval of syllables.
学科领域	基础心理学
关键词	言语产生音韵编码单元图词干扰范式言内隐启动范式脑电技术
学位类型	硕士
语种	中文
学位专业	基础心理学
学位授予单位	中国科学院研究生院
学位授予地点	北京
文献类型	学位论文
条目标识符	http://ir.psych.ac.cn/handle/311026/19755
专题	认知与发展心理学研究室
作者单位	中国科学院心理研究所
推荐引用方式 GB/T 7714	岳源. 汉语口语产生中音节和音段编码的认知机制[D]. 北京. 中国科学院研究生院,2014.

条目包含的文件
文件名称/大小	文献类型	版本类型	开放类型	使用许可
岳源-硕士学位论文.pdf（1679KB）	学位论文		限制开放	CC BY-NC-SA	请求全文