Transferência Linguística e Transferência de Treinamento na Interlíngua do Falante de Português -L1/Inglês-L2

Abstract

This dissertation aims to describe aspects of the interlanguage of Brazilian learners of English as a second/foreign language and to discuss the influence of mother language knowledge and teaching procedures upon interlanguage configuration (L1 transfer andtransfer of training, respectively). English causative-have construction, illustrated by (1) John had his hair cut, and Brazilian Portuguese (henceforth BP) agent-beneficiary alternation, illustrated by (2) O cabeleireiro cortou o cabelo de JoãoBEN and (3) JoãoBEN cortou o cabelo were chosen as the linguistic scenario for this investigation. Bothstrategies are a means of assigning prominence to the beneficiary argument of the sentence, which appears in the subject position. Bearing in mind that a sentence such as (4) John cut his hair does not violate the syntactic rules of the English grammar and that its literal-equivalent in BP corresponds to (3), it was asked whether Brazilian learners of English as an L2 would produce sentences such as (4) in which a beneficiary role isassigned to the subject of the sentence, as in sentence (3). Results have shown that the subjects produce both structures (1) and (4) in their interlanguage performance. This finding has led to the question of whether the production of sentences like (4) by Brazilian speakers was due to language transfer from BP into the interlanguage. Theavailable evidence did not allow us to conclude whether language transfer is present or not. Brazilian speakers have also shown to be more conservative, in that they tend to reject sentences such as (4) in <subject-beneficiary> contexts a lot more than native speakers of English do. Some of them have also shown a greater tendency to assign anagent reading to such sentences, not accepting the assignment of a beneficiary role to the subject. It has been hypothesized that these behaviors result from the presence of an interlanguage rule, inferred from pedagogical input. A total of fifty-nine subjects took part in this study, having been assigned to different experimental groups, according tothe specific goals of which phase of the study.