Estudo baseado em corpus: colocações em livros adaptados

Abstract

The present study aimed to analyze the most relevant collocations of the verbs make and take in graded readers. To achieve this objective, the study relied on Corpus Linguistics, drawing on authors such as Jonhs (1991), Halliday (1966), Sinclair (1991), Sardinha (2000), Biber (2015), and Brezina (2015), among others. The study worked with a specialized corpus composed of literary works with language adapted according to the Common European Framework of Reference for Languages (CEFR). The corpus was divided into subcorpora, classified as A1, A2, B1, B2, C1, and C2, according to their linguistic labeling, referring to the CEFR. LancsBox X software was used for corpus processing and analysis. Firstly, the GraphColl tool was used to identify the most relevant collocations based on the results of the LogDice statistical test. Secondly, the KWIC tool was used for the analysis of the structures and meanings of the selected collocations. Finally, using the Data-driven Learning (DDL) approach, a didactic sequence was developed based on the analysis results. The research results revealed that subcorpora A1 and A2 present collocations with direct and simple sentences, with few grammatical classes between the node and the collocate, usually having only one meaning; subcorpora B1 and B2 present collocations with sentences showing diversity in verb conjugation, with more grammatical classes between the node and the collocate, generally presenting more than one meaning; subcorpora C1 and C2 present collocations with more complex sentences, such as conditionals with 'IF' and sentences in the present perfect tense. In summary, this study contributed to understanding the most relevant collocations of make and take in a specialized corpus and researches involving the DDL approach by proposing an activity that integrates data into language teaching through the results generated in the analysis of collocations.