A linguística de corpus no ensino de inglês: um experimento com atividades com "to" e "for"

• Ano de defesa: 2012
• Autor(a) principal: Marlei Rose Renzetti Tartoni
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: "to" e "for"; Linguística de corpus; Lingüística; Língua inglesa Métodos de ensino; Gêneros textuais; Lingua inglesa Anedotas, satiras, etc; Produção de textos; Língua inglesa Estudo e ensino
• Link de acesso: http://hdl.handle.net/1843/LETR-8Y2P5C

The research work described in this master thesis has been based on an exploratory empirical study put through with 9th-grade students of a public school in the state of Minas Gerais. We investigated the implications of developing activities prepared with Corpus Linguistics resources in order to improve those students accuracy. Starting from genre-based teaching, the decision was to focus on biographies and jokes. Therefore, corpora of small dimensions were compiled with 100 texts from biographies and jokes, amounting to 19.850 words. Starting with these corpora and the frequency of the words in them, the focus of the study was on the several uses of to and for. The primary objective of this research work came out to be how linguistic awareness provided by the examination of concordance lines generated from the corpora could collaborate to the appropriate use of the structures with to and for. The participants in the research, 36 students, were divided into the following groups: treatment (TG) and control (CG). Firstly, they performed text comprehension activities, together, and later, they produced one text of each genre, as a pre-test. From this moment on, the groups worked separately with exercises on to and for. The TG performed exercises which followed the principles of data-driven-learning (JOHNS, 1994), based on concordance lines and the CG performed traditional exercises based on dictionary definitions. By the end of this empirical study, the groups rewrote their productions, as a post-test. The productions of all the participants were typed and treated by WordSmith Tools 5.0 (SCOTT, 2008), in order to generate the concordance lines to be analyzed quantitatively and qualitatively. During the qualitative analysis we tried to assess the degree of accuracy shown by the participants. With the results of this analysis, it was possible to confirm the hypothesis that the participants in the TG, being exposed to a linguistic awareness based on corpora evidences, were able to produce more combinations within the patterns studied, when compared to the CG. It was concluded by this research work that second language learners can develop their linguistic accuracy when observing patterns in concordance lines, even when these items are complex and present different usages.