CAJ | 학술논문

以汉语中的颠倒词为材料，探讨汉语重复知盲的发生水平。采用RSVP方式呈现包含颠倒词对（C1和C2）的句子列表或词语列表，让被试对列表中的词进行全部报告或报告是否存在重复词，记录报告C2的正确率。三个实验采用3（词语的重复性：颠倒重复、完全重复、不重复）×2（颠倒词对的意义相似性：意义不同、意义相似）两因素被试内设计，结果在全部报告和报告是否存在重复词的实验中都出现重复性和意义相似性的交互效应，但不同实验任务，数据模式不同。该结果表明，在RSVP任务中，汉语颠倒词加工存在RB效应，该效应产生的水平受制于不同的实验任务，在部分报告任务中，颠倒词的RB效应发生在知觉水平，在全部报告任务中，颠倒词的RB效应发生在语义水平。
이한어중적전도사위재료，탐토한어중복지맹적발생수평。채용RSVP방식정현포함전도사대（C1화C2）적구자렬표혹사어렬표，양피시대렬표중적사진행전부보고혹보고시부존재중복사，기록보고C2적정학솔。삼개실험채용3（사어적중복성：전도중복、완전중복、불중복）×2（전도사대적의의상사성：의의불동、의의상사）량인소피시내설계，결과재전부보고화보고시부존재중복사적실험중도출현중복성화의의상사성적교호효응，단불동실험임무，수거모식불동。해결과표명，재RSVP임무중，한어전도사가공존재RB효응，해효응산생적수평수제우불동적실험임무，재부분보고임무중，전도사적RB효응발생재지각수평，재전부보고임무중，전도사적RB효응발생재어의수평。
Repetition blindness （RB） refers to the reduced performance in reporting a repeated item in a rapid serial visual presentation （RSVP）（Kanwisher, 1987; Kanwisher ＆ Potter, 1989; Luo ＆ Caramazza, 1995）. The processing level at which the detecting deficit for the repeated item （C2） occurs is still under argument. For example, Token Individuation Theory （Kanwisher, 1987, 1991; Kanwisher ＆ Potter, 1989, 1990） proposed that the repeated items were recognized as types but not individuated as tokens, suggesting that repetition blindness occurred at the perception level. While the Construction and Attribution Theory （Masson, 2004; Whittlesea ＆ Masson, 2005） argued that construction and attribution processes led to repetition blindness when the repeated item was attributed to the wrong source, suggesting that repetition blindness occurred at the semantic level. Thus the present study examined the repetition blindness effect for Chinese reversed words to illustrate the processing level for the repetition blindness. We used a within-subject design with two variables, Repetition of words with three levels （completely repeated, reversedly repeated, and non-repeated）, and Semantic similarity for reversed words with two levels （different and similar）. The dependent variable was the accuracy rate for reporting C2. No matter the processing level at which the repetition blindness occurs, the accuracy rate in the reversedly repeated condition should be different from that in the non-repeated condition, showing a repetition blindness effect for reversed words. Further, if the accuracy rate for C2 in the similar meaning condition does not differ from that in the different meaning condition, it will suggest that the repetition blindness occurs at the perception level. Otherwise, it will suggest the repetition blindness occurs at the semantic level. Experiment 1 presented all words in sentences and asked participants to judge whether there were repeated words in a sentence, while Experiment 2 used the same stimuli, but asked participants to report the whole sentences. Results from a two-way repeated-measures analysis of variance （ANOVA） showed significant RB in both experiments. In Experiment 1, the accuracy rate for C2 in the similar meaning condition did not differ from that in the different meaning condition, but in Experiment 2, they did differ. Thus results in Experiment 1 supported that RB occurs at the perception level while results in Experiment 2 supported that RB occurs at the semantic level. To investigate whether the contradictive results were caused by different tasks, Experiment 3 presented all experimental material in lists of 6 words and asked participants to report all words in a list, similar to the full reporting task in Experiment 2. Results in Experiment 3 showed different accuracy rates for C2 in the two meaning conditions, thus again supporting that RB occurs at the semantic level. The results indicated：（1） In a RSVP paradigm, the RB effect occurred on Chinese reversed words. （2） The level at which the RB of Chinese words appeared depended on the experimental tasks. In a partial reporting task, the RB of Chinese words appeared at the perception level, and in a full reporting task, it appeared at the semantic level.