Examined how a new topic was introduced, maintained, and changed in the Japanese-English interlanguage data of a 45- minute interview between a native and a non-native speaker of English. Findings revealed that although the topic marking system of the interlanguage shared some features with the first and second languages, it maintained features…

Discusses variation in interlanguage and suggests two hypotheses to explain such variation as multiple internal representations of a form and processing errors. Suggests that second language learners can initially represent new forms as prototype schemas, and that such non-discrete representations are a third source of variation in interlanguage.…