On the issue of completeness in second language acquisition

Abstract

The issue of completeness in adult second language acquisition is critical in the development of a theory of second language acquisition. Assuming the Chomskyan definition of core grammar as being those aspects of the language determined by the interaction of the innately specified Universal Grammar and the input to which the learner is exposed, we need to ask if it is possible for an adult learner of a second language to attain native-speaker competence in the core aspects of the grammar of the second language. This paper examines evidence for presence or absence of one principle of UG, Subjacency, in the grammars of groups of proficient nonnative speakers of English. There are three groups whose native languages - Korean, Chinese, Indonesian - differ from English with regard to Subjacency, Korean showing no evidence of it, Chinese and Indonesian showing partial evidence of it. There is one group whose native language, Dutch, shows the full range of Subjacency effects that English does. If all groups show the same Subjacency effects in English that native speakers do, then it must be the case UG is still available for adult second language learning and completeness in second language grammars is possible; if not, then completeness cannot be included as a possible characteristic of adult second language acquisition. Proficient nonnative university students with the above native languages were given grammaticality judgement tests on a set of sentences containing a variety of structures (islands) and Subjacency violations involving those structures. Analysis showed that though all groups were able to correctly judge grammatical sentences (containing islands) as grammatical, only the Dutch group was able to correctly judge ungrammatical sentences (containing Subjacency violations) as ungrammatical; the Korean subjects performed randomly on this task. This native language effect was shown not to be due to attribute variables, such as age of first exposure to English, number of months in an English-speaking country, number of years of English study, etc. The results support the conclusion that completeness is not a possible property of adult-acquired grammars since adults no longer have access to UG for the second language learning process.