Abstract: Chinese and Korean are paronym, and there are three kinds of type of Chinese-Korean homograph: the full synonym homographs,the partial synonym homographs and the homograph homographs. In order to investigate how the homograph of the high proficient Chinese-Korean bilinguals second language accesses its lexical representation, the masked translation primed lexical decision task was used. The experiment takes the translation direction and the word type as the independent variable. The reaction time and the correct rate are the dependent variables. Subjects are 30 senior students in a Normal University, who major in Korean and have passed Korean TOPIC6 Exam and are all high proficient Chinese-Korean bilinguals. The materials in the questionnaire are chosen from the studies of Lianzheng Kim(2009) and Zhixuan Wu(2011).180 homographs, 60 to each type, were appraised its degree of familiarity by other 30 high proficient Chinese-Korean bilinguals. Those words whose degree of familiarity are less than 4 were deleted. Finally, 99 homographs were chosen as experimental materials.Other 99 translation independent words and 99 fillers were also chosen. Filler words formed by the change of one word of the true double character words to a paralog, and for balancing the experiment, half of the fillers were changed the first word and the else were the second. Subjects should make the decision as soon and accurate as possible,half of the subjects do the direction of Chinese-Korean firstly and the other do the opposite direction firstly. The program runs by eprime 1.0 on IBM computer, and the data processing uses SPSS 19.0.Screen resolution is 1360×768. The results shows a significant masked translation priming effect for both directions, which is affected by the type of the homograph. The main effect of word type was significant ,F(3,87)=15.66, p＜.0001.The translation priming effect from Chinese to Korean of the full synonym homographs and partial synonym homographs was significant,which means both of these two type of homographs sharing lexical representation;The significant effect of full homograph homographs was only in Korean to Chinese translation direction,which means separate lexical representation for this homographs. Contrasting the full homograph and the translation independent words,there is only one difference—the full homograph homographs have similar pronunciation while the translation independent words have not,so in a way the priming effect may be attributed to the reliance of phonology in Korean reading. Concept and phonology both facilitated the translation prime, with a greater priming effect of the former. It appeared that the bilingual interactive activation model best accommodates the data on bilingual word recognition in Chinese-Korean homographs. The analysis of data shows that both of the direction, from Chinese to Korean and the reverse, has significantly translation priming effect of the full synonym homographs and the partial synonym homographs，while the effect is asymmetric,the direction of Chinese to Korean is significantly stronger than the other direction, for t(29)=6.21,p＜.01 to the full synonym homograph and t(29)=6.55,p＜.01 to the partial synonym homograph,which means that connection exists between both direction only with the difference of strength. The results supports the the hierarchical model by Korll(1992). The experiment researches the lexical representation of high proficient Chinese-Korean bilinguals, using masked translation primed task.Chinese and Korean are two special languages in the world, as only two countries use them,and have the same development history.In addition,Chinese and Korean have many same and different points,such as pronunciation and morphology,but until now there is no related research for these two languages,in China,researches of bilinguals memory organization model are mostly about Chinese and western languages or Chinese and minority languages, so the experiment for Chinese and Korean is the first time and can fill in some blanks in this field.

Key words: high proficient Chinese-Korean bilinguals, masked translation effect, homographs, lexical representation

摘要： 采用隐蔽翻译启动范式的词汇决定任务，研究了三类中韩同形词的词汇表征，被试为30名熟练中韩双语者，以翻译方向和词汇类型为自变量。发现中韩双语间存在不对称的隐蔽翻译启动效应，这种不对称性与同形词类型有关；语义的启动效应大于语音的启动效应。说明完全和部分同义同形词的词汇表征是共享的，异义同形词的词汇表征是分离的；中韩双语同形词的词汇表征为双重心理词典模型。研究结果支持了修正层级模型。

关键词: 熟练中韩双语者, 隐蔽翻译启动, 同形词, 词汇表征