| Research on the effects of bilingualism on cognition has generated many theoretical debates and implications for bilingual education. In particular, the Threshold Hypothesis (Cummins, 1976, 1978, 1981, 1984, 1987), that postulates language proficiency plays a key role in cognitive development, has greatly affected bilingual policies and teaching practices. The purpose of this study is two fold: testing of the Threshold Hypothesis and examining the relationship between bilingual proficiency and acquisition of metalinguistic ability.;The data were gathered from 121 fifth grade Korean-English speaking bilingual children who are socio-economically and socio-culturally homogenous. Their proficiency in both languages was measured by oral interviews and reading comprehension tests. According to the scores on oral production and the reading comprehension tests, bilingual proficiency types were determined as balanced bilinguals, Korean-dominant speakers, and English-dominant speakers using the Discriminant Function Analysis. Metalinguistic ability was tested by a grammaticality judgment task. The relationship between types of bilingual proficiency and metalinguistic component was examined via the Split Plot Factorial design (SPF−3. 24) (Kirk, 1995).;The results of this study indicated that the analysis metalinguistic component was acquired independent of proficiency of learners and influenced by the language of instruction. The cognitive control component was acquired only when the bilingual learner's dominant language matches the language of instruction. Also, the results of the study have indicated that the degree of exposure to a language might influence the acquisition of the cognitive control metalinguistic component. This research yielded implications for the Threshold Hypothesis in that bilingual development may occur along with many other factors including the language of instruction and the degree of exposure to language learning. The data in this study indicated that instruction in a second language was not effective in acquiring metalinguistic ability for the learners whose proficiency was below the threshold in the language of instruction; however, when a second language should be introduced to the speakers of the non-instructional language is open to further research. |
