Alocação de esforço cognitivo em uma tarefa de retradução: estudo sobre desempenho no par linguístico francês-português

Abstract

This thesis is affiliated to the discipline of Translation Studies, more particularly to processorientedstudies of translation. It draws on Malta (2015) in order to further characterize thetranslation process in (re)translation tasks, wherein more than one text serve as input for theproduction of a target text: the source text and one or more existing translation(s) of thissource text in the target language. An experiment simulating this kind of task was conductedand data were collected using eye-tracking and key-logging in order to investigate theperformance of foreign language students in the language pair French/Portuguese. During theexperiment, participants had their gaze path monitored and their keyboard and mousemovements recorded in real time. The (re)translation task was preceded by a copy task,conducted in order to introduce participants to the (re)translation task layout. In theexperiments third and final stage, participants were asked to engage in verbal protocolsrecalling their experience, these being first free protocols recorded while visualizing thekeyboard and mouse operations they performed during the (re)translation task; andsubsequently guided protocols with their response to a set of questions covering issuesconsidered relevant to this research. In the data analysis, four areas and four micro areas ofinterest covering the four texts involved were defined, namely, the source text, the target textand two translations of the source text already published in the target language. The microareas were established on the basis of a translation problem mappped through a text analysisbased on Systemic-Functional theory (HALLIDAY, MATTHIESSEN, 2014; FIGUEREDO,2007, 2011). Allocation of cognitive effort to the areas and micro-areas of interest wasinvestigated by analyzing the eye-tracking variables considered indicative of cognitive effort:visits (number and duration), visual transitions and fixations (number and duration) (ALVES,2002 ; ALVES, PAGANO, SILVA, 2011; HVELPLUND 2011, 2014; MALTA, 2015;O'Brien, 2006; SJØRUP, 2013). Recall protocols were recorded and transcribed, allowingtriangulation of eye-tracking data with the verbal protocols, according to Alves (2001,2003). The results partially corroborated Malta (2015), indicating greater cognitive effortallocation to the target text area, which serves as a link for visual transitions. The allocation ofvisual attention on input texts (previous translations) is lower than on source and target textsreinforcing Maltas conclusions regarding the auxiliary role played by the input texts ifcompared with the prevailing role of the relationship between source text and target text. Thecoordination of cognitive resources in the areas regarding the source and target texts wasfound to be impacted during a problem solving, although source text processing didnt exertmore cognitive effort than target text processing, as detected in Malta (2015). Verbalprotocols elucidated the eye-tracking results and reinforced Malta's (2015) finding thatparticipants seek an authorial translation during (re)translation tasks, i.e., the production ofa target text different from the input translated texts available in the task design.