Padrões prototípicos de segmentação na descompactação e recompactação de unidades de tradução em tarefas de tradução direta e inversa

Abstract

This master thesis presents a study of prototypical segmentation patterns by means of an empirical-experimental approach to translation process research. It involves two languages,Portuguese and English, and aims at assessing the impact of directionality Portuguese into English and English into Portuguese on the processing of translation units while unpacking and repacking linguistic information during the drafting and revision phases of the translation process (JAKOBSEN, 2002, 2003). The study focuses on three major points: productivity, editions and processes of problem solving and decision making. Quantitative and qualitativedata were collected using Translog© (JAKOBSEN & SHOU, 1999), retrospective protocols (ERICSSON & SIMON, 1980, 1984/1993), the Tobii T60 eye tracker, and questionnaires previously filled in by the eight professional translators under investigation. Translog© .xmlfiles were analyzed using Litterae (VALE, 2010; ALVES & VALE, 2009), to identify different editing categories (P1, P2 and P3) (ALVES & VALE, 2011) and a textual production category (P0) observed during the translation process (ALVES & GONÇALVES, forthcoming). The use of triangulation as a methodological option (JAKOBSEN, 1999; ALVES, 2001, 2003) for data collection and analysis allowed for the exchange of data and agreater possibility of pointing to generalizations. As the data show, directionality did not have impact on the number of editing procedures (P1, P2 and P3) nor on the number of occurrences in the production category (P0). However, directionality appears to have an impact on theidentification of profiles and subprofiles and points to a different profile or subprofile in inverse translation. Directionality also influenced the size of segments, so that there were more instances of peak performance in direct translation. Furthermore, it does not seem to have impact on the number of occurrences of short segments, since those segments aredistributed similarly in direct and inverse translations. Directionality did not seem to have influenced pauses regarding their duration and nature, but it did influence the type of support used during these pauses, with greater use of external support in inverse translation than indirect translation. Despite the small impact of directionality on the dependent variables analyzed in this thesis, the study contributes to the mapping of expertise in translation and shows that there is a balanced cognitive rhythm among the professional translators whosetranslation process data were analyzed. It was also possible to identify alternation patterns between pauses and textual production during the drafting and revision phases of translation process before instances of peak performance and instances of low textual productivity, andrelate them to the translators metacognitive monitoring of the translation process.