17th Workshop on Building and Using Comparable Corpora, BUCC 2024, Torino, İtalya, 20 Mayıs 2024, ss.125-134
This paper focuses on multiple discourse relations, which refer to more than one sense relation between a pair of discourse segments. It shows how they are realized in English texts and their translations into Lithuanian, Turkish, and Portuguese in TED Multilingual Bank, an annotated corpus of English TED transcripts and translations into multiple languages. The paper overviews the annotation procedure and shows the change and variation of multiple discourse relations in the translations, such as omitting the and-component of multiple relations. The cross-linguistically framed analysis reveals that while both senses of a multiple relation can be explicitly conveyed, the salient sense is generally rendered through an overt connective. Even when it is not overtly expressed, it remains inferable and annotated during the annotation stage. By describing the different discourse structures arising from multiple relations and highlighting the implicitated components in translation, the research contributes to the understanding of discourse, aims to raise the awareness of translators and translation educators, and bridge the gap between discourse analysis and translation.