The study examined second language acquisition of Japanese discourse structure, focusing in particular on frequent sentence patterns applied by learners of Japanese, compared with the native speaker's speech. I analyzed the conversations of two different types of learners: a learner who has lived in Japan, and a learner who has not. The main research topics are following: 1) Do the types of learning environments (school instruction vs. natural input) affect their use of discourse organization? 2) What kind of structure patterns are applied by learners, compared with native speakers of Japanese? The results showed that the experience that a learner has received, in the amount of input from natural Japanese discourse, plays an important role in an examination of the acquisition of the discourse structure. The typical sentence structures seen in the data were (1) full sentences, (2) copula/predicate omission, (3) postposing, and (4) collaborative completion.
For example, the learner A who has not lived in Japan used full sentences more frequently than the learner B who has lived in Japan for a year and the native speakers of Japanese. The learner A used them 52 times out of 278 utterances (18.7%), compared with the learner B's 16 times out of 345 utterances (4.6%). In contrast, the learner B and the native speaker often omitted copulas and even predicates in their speech. The research is certainly relevant to the issue of what target-like discourse in Japanese is; thus, it will lead to some explanation for the second language acquisition of discourse competence in Japanese. The observation of the effect of natural input would further lead to the question of what should be included in the foreign language teaching.