As seemingly simple and straightforward constructions, demonstratives are taught to foreign language learners at a rather early stage in their language instruction. For native speakers of Japanese, English "this" and "that" seem fairly easy to acquire, just as the Japanese demonstratives ko, so, and a seem like unproblematic constructions for native speakers of English. However, language teachers often find that even fairly advanced learners of Japanese or English have trouble with many of the less transparent issues surrounding demonstrative usage.

The present paper focuses on the demonstratives "this," "that," ko, so, and a and the peculiar problems that they pose for L2 students. We will show that in accordance with Strauss (1993a, 1993b) and Kinsui and Takubo (1990, 1992), instruction of demonstratives based on the traditional analysis of plus/minus proximity is inadequate. Data from intermediate and advanced L2 learners as well as from native speakers of each language are examined according to recent models (i.e., Strauss' focus schema and Kinsui and Takubo's domain theory of the speaker's experience/perception), which prove to be promising alternatives in teaching demonstratives to L2 learners of Japanese and English.