Quantifier scope in non-native Japanese : a comparative interlanguage study of Chinese-, English- and Korean-speaking learners

This thesis investigates native language (L 1) influence and innate linguistic knowledge (i.e., Universal Grammar) in non-native language (L2) acquisition by means of a comparative interlanguage study of quantifier scope interpretation in L2 Japanese, by adult native speakers of English, Chinese. and Korean. The phenomena investigated are: 1. the availability of object-wide scope in sentences with an existentiallyquantified subject and universally-quantified object (e.g., ,)'omeone read every hook.); 11. the availability of a pair-list reading in questions with evel),one as the subject and what as the object (e.g., What did everyone huy?). Picture-sentence match tasks are developed to investigate these two phenomena in native Japanese, English, Chinese, and Korean, as well as in English/Japanese Chinese/Japanese and Korean/Japanese interlanguage. The native experimental data confirm that, with respect to (i), the object-wide scope interpretation ("for each book, someone read if) is readily available in English but not in Japanese. Chinese, or Korean; and with respect to (ii), a pair-list answer (e.g., .\'lIm hOIlJ,!hl apples, Jane bought pears, Sue hought ... ) is readily available in English. Chinese and Korean, but not in Japanese. These cross-linguistic differences are exploited in the investigation of two main predictions based on Schwartz & Sprouse's (1994. 1996) Full Transfer/Full Access model of L2 acquisition: (1) the L2 learner groups will show divergent development with respect to Japanese scope interpretation due to the distinct scope interpretation possibilities in their respective LIs; (2) advanced L2 learners of Japanese will demonstrate native-like knowledge of quantification phenomena even under severe poverty of the stimulus, due to L2 acquisition being constrained by UG. The results support both predictions. On the basis of these findings. it is concluded that both the L 1 and UG are privileged sources ofknowJedge in the L2 acquisition of phenomena at the syntax-semantics interpretive interface.