O texto traduzido sob a perspectiva do avaliador: um estudo exploratório

Abstract

This dissertation seeks to contribute to the discipline of Translation Studies, specifically to studies on translation quality assessment, by reporting on an analysis of the translated text as assessed by different groups of raters. Eighteen raters (six non-translator researchers from thefield of engineering, six linguists / applied linguists and six professional translators) were asked to assess eight translations into English of a dissertation abstract originally written in Brazilian Portuguese. The subjects who translated the texts were four professional translatorsand four non-translator field specialists on radioactive waste. Concerning the raters, our first goal was to determine the correlation between the raters profile, his/her assessment method, and the aspects of the texts on which they focused more frequently; and the second objectivewas to explore whether there was agreement within and between rater groups concerning the grades and rankings assigned to the translations. Regarding the subjects who produced the translated texts, our first goal was to find out which group of subjects (professional translatorsor non-translator researchers) received the best grades from the raters; the second objective was to investigate whether any subject was more successful than the others from his/her group; and the third goal was to inquire the correlation between the data from questionnairesfilled out by the experiment subjects and their performance when these subjects had their translations assessed. The methodology from the studies on Expertise and Expert Performance was applied, and the think-aloud protocols (ERICSSON; SIMON, 1993; WOLFE, 1997;DEREMER, 1998; TAN; LEE, 2005; TRICKETT; TAFTON, 2007) were the main methodological tool used. In this research, in order to investigate translation quality assessment, TAPs were collected concurrently with the assessments, aiming at collecting data on the thoughts and strategies used by the raters while assessing the translations. The assessment method was characterized according to the time spent by the raters during the assessment session, the nature of their comments, and the movements (the alternation between topics approached throughout the assessment) made by the raters while assessing the texts. Inaddition, questionnaires filled out before translation data collection were used to obtain supplementary data on professional translators and on non-translator researchers education and reading and writing habits both in Portuguese and in foreign languages. The resultsshowed that there was agreement within groups, but there was no agreement across all three groups of raters. Domain knowledge (SCARDAMALIA; BEREITER, 1991) had little impacton assessment processes, and, regarding the subjects who produced the translations, one nontranslator researcher (R2) and one professional translator (T2) stood out among their peers, and to some extent R2s and T2s education profiles and answers to the questionnaires correlate to their superior performance.