Soo-Ok Kweon

Acquisition of the constraints on wanna contraction by advanced second language learners: Universal grammar and imperfect knowledge

Contraction of want to to wanna is subject to constraints that have been related to the operation of Universal Grammar. Contraction appears to be blocked when the trace of an extracted wh-word intervenes. Evidence for knowledge of these constraints by young English-speaking children has been taken to show the operation of Universal Grammar in early child language acquisition. The present study investigates knowledge of these constraints in adults, both English native speakers and advanced Korean learners of English. The results of three experiments — using elicited production, oral repair, and grammaticality judgments — confirmed native speaker knowledge of the constraints. A second process of phonological elision may also operate to produce wanna. Learners also showed some differentiation of contexts, but much less clearly than native speakers. We speculate that non-natives may be using rules of complement selection, rather than the constraints of Universal Grammar (UG) to control contraction.

Korean engineering students’ perceptions of English-medium instruction (EMI) and L1 use in EMI classes

ABSTRACT Engineering schools have spearheaded the implementation of English-medium instruction (EMI) in Korean higher education. Three major engineering universities were chosen for this study with its goals to examine engineering students’ perceptions of EMI and L1 use in EMI classes and to make suggestions for the directions that Korean engineering schools should take in relation to EMI. A questionnaire survey was conducted among the undergraduate students of the three universities. Five hundred twenty-four students participated in the survey, and the results show that more students felt insufficient about their English ability than sufficient and that the majority of the students preferred Korean-medium instruction over EMI and did not find EMI beneficial for their English ability. Nevertheless, most of the survey participants believed that EMI should be maintained but with changes in the schools’ mandatory policies. In addition, the students without distinction as to school were supportive of L1 use in EMI classes: approximately 90% of the respondents from each school believed that L1 should be used to facilitate their learning.

Generating grammar questions using corpus data in L2 learning

This paper examines how grammar questions are automatically generated for L2 learning by applying a sequential labeling technique to learner corpora. We developed a model that helps detect possible error positions and select the most appropriate form among choices. Discriminant models such as conditional random field and maximum entropy are used to generate the error identification question. Questions generated by the proposed method corresponded highly to questions that experts made. Our data-driven approach lends itself to any language without costing expensive expertise.