Two-kanji compound words in the Japanese mental lexicon

Terry Joyce, Ph.D.
Postdoctoral Researcher, Large-Scale Knowledge Resources - 21st Century COE Program
Tokyo Institute of Technology, Tokyo, Japan 　

Abstract:
Given the productivity of compounding as a word-formation process, the Japanese language is especially suitable for investigating the extent of morphological involvement in the organization of the mental lexicon. This presentation introduces some of the constituent-morpheme priming experiments that I have been conducting to examine the lexical representation and retrieval of two-kanji compound word in the Japanese mental lexicon from a morphological perspective (Joyce, 1999; 2002; 2003a; 2003b). Essentially, these experiments are comparing the patterns of facilitation on lexical decision responses to compound words due to the prior presentation of a constituent kanji (relative to a baseline condition), across a number of word-formation conditions such as modifier + modified, verb + complement, complement + verb, and synonymous pairs. The results from Joyce (1999; 2002) were that reaction times in both constituent conditions were significantly faster than the baseline condition and, in the majority of cases, at similar levels across the word-formation conditions. However, the finding of a significant difference between the constituents only in the verb + complement condition, where the first constituent condition was faster, would seem to suggest an effect of verbal semantics. Further support for this notion comes from experiments (Joyce, 2003a; 2003b) that manipulated the positional frequency of the verbal constituent in the verb + complement and the complement + verb compound words, and observed a reversed pattern of priming across the high positional frequency conditions. These results are discussed in terms of the Japanese lemma-unit model, as a connectionist model of the Japanese mental lexicon (Joyce, 1999; 2002; 2004).