| Zdroj: |
Egbert, M, Hirataka, F, Sadeghi, S & Yufu, M 2013, ' Glossing, transliterating and translating interactional data on languages other than English for international publication ', The 13th International Pragmatics Conference, New Delhi, India, 08/09/2013-13/09/2013 . |
| Popis: |
In the area of pragmatics the numbers of studies on non-English data has been increasing. For the career of most researchers, it is highly desirable to publish internationally, and thus in English. When researchers transcribe and gloss non-English data for publications in English, they often need to make practical decisions that have an impact on the analysis and presentation of results. This paper argues methodologically that non-English data should be treated more faithfully.The paper starts by examining how researchers in different fields of pragmatics have handled this issue and what problems result from the current ‘mainstream’ way. Most authors make their decisions on a case-by-case basis, considering the publisher’s requirements, possibly taking space requirements into account, and most importantly, balancing the trade-off between precision in the data display and the readability of the transcript. Though most studies noted that a perfect translation is impossible, and grammatical glossing or descriptions are added to account for what is lost in the English translation, there are also relevant characteristics that are missing in the glossing process. The paper then addresses three major problems the authors have encountered in publishing in English on conversational data in three linguistically and pragmatically different languages (Japanese, Persian, German), and in conversations between native and nonnative speakers (e.g. Brazilians speaking Japanese).1) Regarding the layout of the transcription it is of concern whether the original writing system should be preserved. The transcript would thus include the original writing system, a transliteration, a gloss and an approximate translation. For Persian, the original right-to-left writing direction collides with English, for example in marking overlap. For Japanese, the Latinized transliteration cannot show variations in pronunciation which are relevant to the analysis. (2) When preparing English glosses, there are specific phenomena such as cut-off, which are problematic to translate. Likewise, it is difficult to represent learner constructions. (3) The most troubling difficulty for the authors lies in the decision of whether, or how much, of the culture-specific context and participants’ perspectives should be included. In particular, validity in conversation analysis is based on showing analytically what aspects of interactional conduct and the participation framework are made relevant by the participants. The paper shows that what is perceived as meaningful aspects of the language and interaction to the participants may depend on the researchers’ cultural knowledge.The authors conclude with a discussion of the researcher's dilemma that a faithful display of the original data with a high degree of precision in preserving the original language as much as possible in transcribing, transliterating, glossing and translating reduces the readability of the transcript, and may even lead to difficulties in publishing the work. The authors argue for more radical solutions in order to increase the integrity of the data.First authors: Mamiko Yufu and Sima SadeghiSecond authors: Maria Egbert and Fumiya Hirataka |
