Some remarks on transcript translation in discourse analysis

Transcripts of second language interaction and within publications in languages differing from the data discussed often must be translated to make the results of discourse analysis accessible. Transcription follows strict rules to achieve a scientific standard recipients can rely upon. Its main task is to picture a piece of authentic interaction by means of writing as detailed as possible and necessary, to make readers able to follow an analysis and judge it by themselves. But what is true for transcriptions does not hold for their translations. Surprisingly, reflection of transcript translation tends to be neglected in literature. This might be due to several reasons: the different language systems and cultural imaging habits make it difficult, if not impossible, to propose an all-fitting translation guideline for transcription. Further, discourse analysts usually are (experienced) non-professional translators and do not often have a theoretical translation background. It is an additional (practical) challenge to align the demands of transcript theory and readability with editorial requirements. While translating transcripts analysts would have to reflect five different aspects: (1) the status of translation in transcription, (2) a (not yet established) standard for transcript translation, (3) the translation‘s readability (recipients should be able to follow it with only little effort), (4) the potentially extremely differing systems of source and target language (representation of syntax and semantics), and (5) translation problems (e.g. cultural aspects) with regard to the given language pair. The article considers the status of translation in transcription. The few theoretical findings are analyzed and contrasted with examples from discourse analysis. The article is complemented with the results of a survey among analysts concerning their translation behavior. Finally, some proposals for solutions are outlined.