汉语词素位置编码及时空整合的认知研究

发布时间：2017-12-26 20:34

本文关键词：汉语词素位置编码及时空整合的认知研究　出处：《电子科技大学》2016年博士论文　论文类型：学位论文

【摘要】：视觉单词识别是阅读的基本过程。在阅读中,读者需要对一定视野范围内字词的属性及其位置进行编码。本文以汉语合成词词素位置编码为研究课题,重点研究了词素位置编码的时间进程和空间加工特征,主要研究内容分为四部分。第一部分主要从时间维度上考察汉语词素位置编码对视觉注意瞬脱效应的影响。注意瞬脱是指在快速序列视觉呈现中,被试对第一个目标刺激的正确辨认影响了在时间上与该目标刺激相近的(通常500毫秒以内)第二个目标刺激的识别现象。本研究通过对汉语合成词的两个词素分离及其位置颠倒来研究其对注意瞬脱的影响。实验发现:在假词情况下,出现注意瞬脱现象;而在词语、词素换位非词以及逆序词情况下,均没有出现注意瞬脱现象,这说明汉语词素位置编码会导致注意瞬脱现象的消失。在词素换位非词情况下,对两个目标汉字顺序识别错误率最高,尤其是在短时间间隔情况下。该结果进一步说明,在汉语词素分离和颠倒词语的识别过程中,相比于认知资源理论,注意瞬脱更倾向于是一种认知加工策略。同时,该实验结果也说明在中文阅读中,合成词是作为整体加工的,词素位置编码不严格,读者能容易的识别具有词素位置分离以及颠倒的词语。第二部分主要考察了词素位置编码的时间进程。已有研究显示,词语词素位置编码发生在视觉单词识别的早期阶段,但并没有确定具体的时间进程。因此,本研究通过双任务快速序列视觉呈现范式并使用汉语合成词,词素换位非词,逆序词以及假词作为实验刺激材料来探讨词素位置编码及其时间进程。汉字刺激流中的每一个项目的呈现时间为30毫秒。实验结果显示,在词语,词素换位非词和逆序词中,第一个识别的目标汉字在整个时间流程中都促进了对另一个目标汉字的识别;而在假词中,第一个识别的目标汉字抑制了对另一个目标汉字的识别。对词素换位非词的两个目标汉字的识别正确率要低于词语和逆序词的两个目标汉字的识别正确率。此外,两个目标汉字顺序识别的错误率随着时间间隔的增加逐渐降低;在词语和逆序词中,目标汉字顺序识别错误主要发生在30毫秒至180毫秒,而在词素换位非词中,目标汉字顺序识别错误主要发生在30毫秒至240毫秒。总的来说,这些实验结果表明,读者可以轻松的阅读词素换位非词,但是相比于词语,词素换位非词还是存在一定的加工位置信息的认知负荷。在视觉单词识别的早期阶段,基词的整体表征已经激活,但是对词素位置的加工并不严格。第三部分主要通过双任务快速序列视觉呈现范式并使用词语,词素换位非词和假词作为实验目标刺激来探讨汉语合成词词素分离和颠倒在左右视野中的加工特征。在实验过程中,第一个目标汉字总是出现在0°离心率的位置,第二个目标汉字随机出现在左右视野0°,2°,4°和6°离心率的位置。同时,第一个目标汉字和第二个目标汉字之间随机出现100毫秒,300毫秒,500毫秒和700毫秒的时间间隔。实验结果显示,右视野优势效应在词语情况下出现,而该效应在词素换位非词情况下消失。读者能从中央窝视野和副中央窝视野提取相同质量的语义信息,不论汉语合成词的词素位置颠倒与否,但是在右视野边缘区域对语义信息的提取有一定的困难。此外,两个目标汉字的顺序识别错误率在中央位置和最短时间间隔的时候最高,并随着离心率和时间间隔的增加而逐渐降低。因此,我们认为,汉语合成词词素位置颠倒主要在右视野边缘区域敏感。词素位置错误主要发生在中央窝(0°离心率)和100毫秒的时间间隔,这表明,在视觉单词识别的早期阶段,对呈现在中央窝区域的词语词素位置更可能被颠倒。第四部分主要通过双任务快速序列视觉呈现范式并使用词语,词素换位非词和假词作为实验目标刺激来考察词素分离及其颠倒在上下视野中的加工特征。在实验过程中,第一个目标汉字总是出现在0°离心率的位置,第二个目标汉字随机出现在上下视野0°,2°,4°和6°离心率的位置。同时,第一个目标汉字和第二个目标汉字之间随机出现100毫秒,300毫秒,500毫秒和700毫秒的时间间隔。研究结果显示,当加工词语的时候,出现下视野优势效应;当加工词素换位非词和假词的时候,没有上下视野优势效应。此外,对目标汉字的顺序识别错误率在中心位置和最短时间间隔时候最高。上述结果表明,在上下视野中,汉语词素位置编码在下视野相对敏感,并且词素顺序错误编码主要发生在0°离心率和100毫秒时间间隔。综上所述,汉语词素位置编码使得注意瞬脱消失。词素位置错误编码主要发生在30毫秒至180毫秒这个词语加工时间段。此外,词素位置编码在右视野和下视野相对比较敏感,且词素位置错误编码主要发生在视野0°离心率和100毫秒时间间隔。以上结果表明,词素位置编码主要发生在视觉单词识别的早期阶段,并且对词素位置加工不是很严格。
[Abstract]:Visual word recognition is the basic process of reading. In reading, the reader needs to encode the attributes and positions of words in a certain field of vision. In this paper, we focus on the location and coding of morphemes in Chinese compound words. We focus on the temporal process and spatial processing characteristics of morpheme location encoding. The main research contents are divided into four parts. The first part mainly examines the effect of Chinese morpheme position coding on visual attention blink from the time dimension. Attention blink refers to the correct recognition of the first target stimulus in fast sequence visual presentation, which affects the recognition phenomenon of second target stimuli which are similar to the target stimulus in time (usually less than 500 milliseconds). In this study, the effect of two morphemes on the attention transients was studied by the separation of the morphemes of Chinese synthetic words and the reversal of their positions. It is found that there is attention blink phenomenon in the case of false words, but there is no attention blink phenomenon when words, morphemes transposition non words and reverse words, which indicates that the location encoding of Chinese morphemes can cause the attentional blink to disappear. In the case of morpheme transposition non words, the error rate of the two target Chinese character sequence recognition is the highest, especially in the short time interval. The results further indicate that in the process of identifying Chinese morphemes and reversing word recognition, compared with the theory of cognitive resources, attention blink tends to a cognitive processing strategy. At the same time, the experimental results also show that in Chinese reading, synthetic words are processed as a whole, and morpheme location coding is not strict. Readers can easily identify words with morpheme position separation and inversion. The second part mainly examines the time process of the morpheme position coding. Previous studies have shown that word morpheme location coding takes place at the early stage of visual word recognition, but does not determine the specific time process. Therefore, this study explored the morpheme location encoding and its time course through two tasks fast sequence visual presentation paradigm and Chinese compound words, morpheme transposition non words, reverse order words and fake words as experimental stimuli. The presentation time of each item in the Chinese character stimulation flow is 30 milliseconds. Experimental results show that, in the words, non words and morphemes transposition inverted words, the first recognition of the target Chinese characters have promoted the identification of another target in the whole process Chinese characters; and in the false words, the first recognition of the target recognition of Chinese characters suppressed another goal Chinese characters. The recognition accuracy of the two target Chinese characters of the morpheme transposition non words is lower than the recognition accuracy of the two target characters of the words and the reverse word. In addition, two target sequence recognition error rate Chinese characters with increasing time interval gradually decreased; in the words and morphemes words, target sequence Chinese characters recognition errors occurred within 30 ms to 180 ms, while in the non word morpheme transposition, target sequence recognition error Chinese characters mainly occurred in the 30 ms to 240 Ms. Generally speaking, these experimental results show that readers can read morpheme transposition non words easily, but compared with words, morpheme transposition non words still have certain cognitive load of location information. In the early stage of visual word recognition, the overall representation of the base word has been activated, but the processing of the morpheme position is not strict. The third part mainly discusses the processing characteristics of Chinese compound word morphemes separation and inversion in the left and right field of vision by using two tasks fast sequence visual presentation paradigm and using words, morphemes, transposition non words and pseudo words as experimental target stimuli. In the process of experiment, the first target Chinese character always appears at the position of 0 degree centrifugal rate, and the second target Chinese characters appear randomly in the position of 0 degrees, 2 degrees, 4 degrees and 6 degrees of centrifugation. At the same time, a random 100 millisecond, 300 millisecond, 500 millisecond and 700 millisecond interval between the first target Chinese character and the second target Chinese characters. The experimental results show that the advantage effect of right visual field appears in the case of words, and the effect disappears in the non word morpheme transposition. Readers can extract the same quality semantic information from the view of the central fossa and the visual field of the secondary central fossa, regardless of the location of the morphemes of the Chinese compound words. However, it is difficult to extract the semantic information from the edge area of the right vision. In addition, the recognition error rate of two target Chinese characters is the highest at the central location and the shortest time interval, and decreases with the increase of centrifugal rate and time interval. Therefore, we believe that the inversion of the morpheme position of the Chinese synthetic word is mainly sensitive to the right visual area. The location errors of morphemes mainly occurred at the time interval between the central fossa (0 degree centrifugal rate) and the 100 millisecond interval. This indicates that in the early stage of visual word recognition, the location of words and morphemes in the central fossa area is more likely to be reversed. The fourth part mainly investigates the processing characteristics of morpheme separation and its upside down in the upper and lower field by using two tasks fast sequence visual presentation paradigm and using words, morphemes, transposition, non words and pseudo words as experimental target stimuli. In the process of experiment, the first target Chinese character always appears at the position of 0 degree centrifugal rate, and the second target Chinese characters appear randomly in the position of 0 degrees, 2 degrees, 4 degrees and 6 degrees of centrifugation. At the same time, a random 100 millisecond, 300 millisecond, 500 millisecond and 700 millisecond interval between the first target Chinese character and the second target Chinese characters. The results show that when processing words, the effect of inferior visual field appears. When processing morphemes, there are no superior and inferior visual field effects when processing morphemes. In addition, the sequence recognition error rate for the target Chinese characters is the highest at the center position and the shortest time interval. The above results show that in the upper and lower field of vision, Chinese morpheme location coding is relatively sensitive to the lower field of vision, and the morpheme sequence error coding mainly occurs at 0 degree centrifugal rate and 100 millisecond time.
【学位授予单位】：电子科技大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：B842

【相似文献】