Abstract
The present research is an attempt to explore Japanese EFL learners' speaking ability in explanatory structure. 63 undergraduate students participated in a series of performance tests, including two types of oral performance tests and a writing performance test. Spoken and written production was transcribed into an interlanguage corpus, which was gathered for the purpose of analyzing lexical and syntactic features. An explanatory multiple regression, with the scores on each OPT by three raters on an intuitive basis as the dependent variables and four types of ratio as independent variables, yielded some explanatory relations. Based on the results, a structure equation modeling was undertaken. It was found that there did exist two types of speaking at a linguistic level. Furthermore, it provided the pedagogical implication that some knowledge of syntactic rules on clause structure and information structure was acquired through different processes, which gives us some insights into enlightened and eclectic approaches, new ones for these two types of speaking ability.
