O processamento cognitivo em tarefas de (re)tradução: um estudo baseado em rastreamento ocular, registro de teclado e mouse e protocolos restrospectivos

Abstract

This research was conducted at the Laboratory for Experimentation in Translation, at Faculdade de Letras (FALE), Universidade Federal de Minas Gerais (UFMG). Affiliated to process-oriented studies within the discipline of Translation Studies, it aimed at describing the translation process in a (re)translation task (TAHIR-GÜRÇALAR, 1998) by analyzing data collected through key-logging, eye-tracking and verbal protocols carried out upon task conclusion (ALVES, 2005). The aim was to identify indicators of cognitive effort in the translation process such as recursive movements, pauses and fixation count and duration (OBRIEN, 2006; ALVES; PAGANO; DA SILVA, 2009; JAKOBSEN; JENSEN, 2008). Translation students and teachers of Spanish as a foreign language with experience in translation received an 80-word excerpt from a short story in Spanish and two published translations of that excerpt into Brazilian Portuguese, and were requested to produce a new version in Portuguese language. The translation process of those participants was recorded in real time through the key-logging software Translog-II and eye-tracking software Tobii Studio with an eye-tracker Tobii T60. Upon completion of the task, free and guided verbal protocols were recorded in audio and later transcribed. Confirming Hvelplund`s (2011, 2015a; 2015b) findings, this study showed greater allocation of cognitive effort on processing the target text and prevalence of visual attention first on the target text, then on the source text and occasionally on both translated texts received as input. Regarding the stages of translation process, participants spent more time on drafting, then on orientation, and finally on revision. Revision proved to be the most heterogeneous stage, sometimes being conducted with a focus only on the target text, sometimes on a comparison of the source and target texts and, less frequently, on further processing of the target text and both translations. Similarly to Carl et al. (2011), this investigation showed variation regarding the time devoted to revision by participants. Verbal protocols corroborated the eye-tracking results, which indicated more visual attention devoted first to the target text, and then to the source text. They also confirmed less visual attention and effort spent on the translated texts that participants received as input to carry out the task. Additionally, they revealed aspects related to participants perceptions on that task.