(Des)compactação de significados e esforço cognitivo no processo tradutório: um estudo da metáfora gramatical na construção do texto traduzido

Abstract

This dissertation seeks to contribute to the discipline of translation studies, particularly to studies of the translation process. It examines aspects of text production in translation, focusing on problemsolving until completion of a translation task performed by two groups of subjects: non-translator field specialists, and professional translators. Drawing on the notion of grammatical metaphor, as developedby Systemic-Functional Linguistics (Halliday & Matthiessen, 2004), decisions made by the subjects are analyzed to gather insights as to how they deal with metaphorical meanings in the source text.Those meanings are assumed to be instantiations in the source text potentially demanding cognitive effort on the subjects part, constraining lexico-grammatical realizations in the target text and havingan impact on some of its rhetorical relations. Two hypotheses were formulated building on Steiners (2001a, 2001b) and Tirkkonen-Condits (2005) claims, namely that (1) renditions in the target language tend to be metaphorically analogous to those in the source text; and that (ii) instances of cognitive effort can be tapped in the translation process, particularly when more metaphorical variants are demetaphorized. To test these hypotheses, 16 Brazilian and 16 German subjects, proportionallydistributed as field specialists and professional translators, were asked to perform a task of translation of one of two versions of an English-language source text into their mother tongue. The task wasrecorded through key-logging (Translog©) program developed by Jakobsen & Schou (1999). Both versions construed analogous meanings, but they had either the most or the least metaphorical variantsof ten agnate realization pairs (five of each in each version). From a process-oriented perspective, the key-logging data was analyzed to determine the number of renditions per variant and their respectivemetaphoricity levels. From a product-oriented perspective, subjects renditions were analyzed using the framework of Systemic-Functional Linguistics (SFL) and Rhetorical Structure Theory (RST) toinvestigate the impact of their choices on the explicitness and implicitness of the target texts and on the rhetorical relations between text spans. Additionally, subjects recall protocols collected uponcompletion of the translation task and readers answers to an on-line questionnaire on the source and target texts were analyzed to identify a notion of translation that could have guided experimentparticipants decision makings and that could inform on issues underlying the product and process findings. The results confirm the hypothesis that the production of target text with lexico-grammaticalrealizations analogous to those in the source text is a default procedure and requires less cognitive effort. Nevertheless, more metaphorical variants in the source text do not necessarily require morecognitive effort than the congruent variants. The cognitive effort indices of only two subjects were significantly higher than the others as a result of their (de)metaphorization choices, which had in turn substantial impact on the rhetorical structures of their texts. Regarding a prototypical notion of translation, the analysis of both translation processes and on-line questionnaires revealed that both readers and experiment participants opt for lexico-grammatical realization that imply metaphoricityanalogy between source and target texts, but the former are relatively more flexible to variance than the latter. This suggests the importance of investigating gatekeeping until the release of a translatedtext and seems to point out that corpora research and process-based translation research have different objects of study, even though they are referred to as translation. From the perspective of translationpedagogy, the notion of translation that can be observed in the translation process and surveys of readers may inform translation courses as to to what extent, or under what circumstances, it is desirable to render target texts with metaphoricity levels higher, lower than or analogous to those inthe source text and how translators can monitor their process to cope with metaphoricity issues.