LEXICAL ACCESS TO CHINESE COORDINATIVE WORDS
Wenbin Wang
2001-01-01
Acta Psychologica Sinica
Abstract:Many studies suggest that affixed words, or at least suffixed words, are stored in their base form in the lexicon (Kintsch, 1972; Gibson & Guinet, 1971). Taft & Foster (1975) found that in a lexical judge ment task, prefixed words were analyzed into their constituent morphemes before lexical access took place. Taft & Foster (1976) did their research a step furthur by running five experiments on lexical storage and retrieval of polymorphemic and polysyllabic words. The conclusions they reached were as follows: (1) Polysyllabic polymorphemic words and polysyllabic monomorphemic words were accessed in the same way; (2) The search target in a compound was the first constituent and the second one was irrelevant.
In order to test these conclusions cross-linguistically, Zhang & Peng (1992) conducted an experiment on Chinese compound words. Their results suggested that representations of Chinese compound words were stored in morphologically decomposed form. However, although the lexical decision performance for coordinative words was affected by the morphological decomposition, the first and the second elements were similar in importance.
There are three different subtypes of coordinative words in the Chinese language. Our experiment investigated these three subtypes of coordinative words to see in lexical access if the first constituent is always decisive as proposed by Taft & Foster (1976) and whether the two constituents are equal in importance as advanced by Zhang & Peng (1992). The hypothesis in this experiment was that the process of word recognition of these coordinative words depended upon the idiosyncratic property of word compositionality. 30 native Chinese speakers participated in the experiment. The test was composed of 64 Chinese coordinative words, 64 compound fillers and 100 compound nonwords. Of the 64coordinative words, there were 24 words belonging to Subtype 1, such as "dao-lu" (road-road, meaning 'road') and "sheng-yin" (sound-sound, meaning 'sound'); 24 belonging to Subtype 2, such as "gu-rou"(bone-flesh, meaning 'kindred') and "ling-Mu" (collar-sleeve, meaning 'leader'); 16 belonging to Subtype 3, such as "guo-jia" (country-family, meaning 'country') and "shi-guang" (time-light, meaning 'time'). By using Psyscope 1.1 mn on Power Macintosh, the experiment revealed that the first constituent was not always decisive in the lexical decision. Besides, the two constituents in coordinative words were not always equally important in lexical recognition. It concluded that although morphological decomposition did take place in lexical judgement of Chinese coordinative words, semantic transparencies played a crucial role in lexical access. Moreover, it found that the lexical access to Chinese coordinative words called for a Confirmation Principle in mental lexicon.