O processamento lexical de homógrafos interlinguísticos por bilíngues em tarefas de reconhecimento de tradução

Abstract

This research is developed under the area of Translation Studies from the perspective of Psycholinguistics. Understanding the cognitive processes behind language can help translators to develop their skills and to understand their translation decisions. The main objective of this study is to investigate the cognitive processes involved in the translation recognition process of interlinguistic homographs of the Brazilian Portuguese-English pair in two contexts of presentation, isolated words and sentences. More specifically, it is expected to answer the following research questions: (1) Is there a difference in the cost of processing for interlinguistic homograph words, Brazilian Portuguese-English, in a translation recognition task?; (2) What is the effect of the presentation context of words, isolated or in sentences, in a translation recognition task? The studies carried out by Schwartz and Kroll (2006), De Groot et al. (2000), Eddington (2011), Van Hell and De Groot (2008), Zhou et al. (2019), De Groot and Comijs, (2001), and also the bilingual lexicon organization models provided theoretical and methodological support to the present research. The experimental session consisted of two translation recognition tasks. First, a single word translation recognition task in which the 30 bilingual participants whose first language was Brazilian Portuguese and the second language English had to decide whether in a pair of words the second word was a possible translation of the first, 40 pairs of words were used as stimuli, 20 pairs with homographs and 20 pairs with control words. After that, a translation recognition task of words preceded by sentences. This second task followed the same procedures as the previous one, however, participants read a sentence before deciding whether or not the two words were possible translations, 40 sentences paired with 40 words were used, 20 homographs and 20 control words. In both tasks, reaction time and accuracy data were collected using the PsyToolkit software (STOET, 2010, 2017). A higher cost of processing for interlinguistic homographs was expected in this type of task when compared to control words, and also a higher cost for words presented isolated than for those presented in sentences. The results of the first task showed that the processing cost for the homographs was slightly less costly and with greater accuracy than for control words. This result is in disagreement with previous studies. Regarding the second task, the processing of homographs was slightly more costly and with less accuracy than for the control words. This result is in line with previous studies and with the hypotheses raised for this research. When comparing the two tasks, the results showed that, in general, the translation recognition of words preceded by phrases was more costly, but with greater accuracy, than for words preceded by other isolated words. These results contribute to the discussion about the use of translation as a tool for investigating lexical access, and also contribute with perceptions about the processing and the lexical organization of bilingual speakers, especially those of the Portuguese-English pair, in processes that involve translation recognition. Added to this, the present research contributes to studies with types of words processed differently by bilinguals such as interlinguistic homographs.