| This study investigated Hispanic second language learners' meta-linguistic ability to recognize and correct errors in their second language, English. Three measures of error recognition and three measures of error correction were used to operationally define the dependent variable error judgment. Ability to judge the errors was predicted from the second language learners' ability in the mother tongue, their years of schooling in L(,1), their length of residence in the host country, and their years of schooling in L(,2). Burt and Kiparsky's (1975) global-local distinction of second language learners' errors defined the language construct, judgment of errors. A 43 item instrument representing global and local errors was constructed and validated in three pilot studies previous to its use as a research instrument in the study.; The results of the validation study defined global errors as those errors which seriously violate the deeper semantic structure of English, regardless of surface clausal complexity. It also defined local errors as those which are language specific and which violate only the arbitrary conventions of standard English.; The results of the predictive study revealed that the better the native language skills and the more years of schooling in L(,1) the better the second language learner could recognize global errors, but the more years of schooling in L(,2), in addition to native language skills and L(,1) schooling, the better the second language learner could recognize and correct local errors. The results appear to support Cummin's linguistic interdependence hypothesis and Krashen's Monitor Model of Language Learning. These results also bear significant implications for bilingual education and second language learning for adults. |
