Transcribing and Translating Data
1 Issues around transcribing data
Many projects will need to decide what to do with recorded interviews or focus-group discussions. Should interviews be transcribed in the full sense (that is, to provide an as accurate as possible, word-for-word written equivalent of the recording)? If so, should the whole interview be transcribed? These questions raise important issues. Some of the issues are technical; others have to do with translation; others have to do with resources; and others depend heavily on the purposes of the research. See the handout on transcribing issues for more discussion on this.
Time: 30-45 minutes
- to clarify the meanings of ‘note-taking’ and ‘transcription’ through a worked example
- to help participants plan their data collections and research design in the knowledge of the issues of equipment, soft-ware, translation, costs and purposes that will arise in the management of recorded data
If you have a data projector and computer, add a pair of loudspeakers to make sure that participants can hear the audio clip. Otherwise you may need to copy the clip onto a CD to play in a CD-player.
Open the handout on three versions of accounts of the interview on your computer so that participants can hear the interview and read each version in turn. Alternatively, prepare the three versions as a handout, to be given to participants before they listen to the clip.
The 4-minute audio clip is from an interview of SS by RJ, as part of a project on pharmaceuticals that includes attempts to understand how drugs are used by medical personnel who attend women who give birth in their own homes in rural north India. You might want to play this clip (or parts of it) three times while showing the different levels of report and transcription. This will probably generate a lively discussion. If you want to take the opportunity to make decisions about your own project, the discussion might need 30 minutes; otherwise, 15 minutes might be sufficient.
Note the use of italics for the interviewer’s contributions; and the occasional insertion of the time in the interview (so that researchers can go back and listen again if they want to check on the accuracy of the transcription).
2 Issues around translation of data
All social research involves translation, if only from the ‘language of the streets’ into formal academic prose, but more issues arise if the data collection is in a different language from the language in which the results are analysed or reported. In many qualitative social research projects in the South, researchers carry out data collection in their native language, or in another local language, and record field-notes (or responses to interview questions) in either English or (in some cases) in yet a further language.
Note: For the sake of simplicity, in what follows we assume that the language of analysis will be English. If you will be working in some other language, please change to suit your local circumstances.
Time: 30 or 45 minutes
- To make participants aware of key issues with respect to the trade off between getting almost exact translations of individual words or phrases, on the one hand, and conveying the full sense of the clause, sentence of paragraph, on the other
- To make participants aware of the importance of careful thought about who should translate what parts of the research fieldnotes and transcripts, when this should be done, and with what kinds of quality controls
- (If this training takes place near the beginning of a research project) To help develop routines to ensure adapt these general considerations to the specific requirements of a particular project
Provide a short extract from an interview or focus group discussion transcript (or recording) in a language (other than English) understood by all participants. The extract should include some sections where translation is not straightforward, and should be provided either as a handout or as an overhead slide.
- Start by asking participants about their own experiences of translation, and move into an individual or small group exercise in which all attempt a translation of the extract. (15 minutes)
- Facilitate a whole-group discussion of the results (15 minutes)
- If appropriate, facilitate a whole group discussion of procedures to be adopted in a particular project (15 minutes)
There is a handout on issues of translation which could be given out after the session.
References for Transcribing Data
Gibbs, G. R. (2007). Analysing Qualitative Data. London: Sage, especially Chapter 2, 'Data Preparation'.
References for Translation of Data
Birbili, M. (2000). Translating from one language to another. Social Research Update 31, University of Surrey, http://sru.soc.surrey.ac.uk/SRU31.html
- Collecting data in one language and presenting the findings in another involves researchers taking translation-related decisions that have a direct impact on the validity of the research and its report.
- Factors which affect the quality of translation in social research include: the linguistic competence of the translator/s; the translator’s knowledge of the culture of the people under study; the autobiography of those involved in the translation; and the circumstances in which the translation takes place.
- There is a need for social researchers who have to translate data from one language to another to be explicit in describing their choices and decisions, translation procedures and the resources used.
Bradby, H. (2002). Translating culture and language: a research note on multilingual settings. Sociology of Health and Illness 24(6): 842-855.
Author’s Abstract: Language and translation are not treated as part of the problematic in sociological research when compared with that of social anthropology, and this relative inattention can be related to the parallel development of the two disciplines. Ethnographic evidence from studies of identity, social support and wellbeing among Glasgow Punjabis suggests that the complex and strategic blending and switching of vocabulary, tone and accent is one means through which identities and support networks are negotiated and affirmed. The hybrid use of language can and should be reflected in the way that research is conducted with multilingual communities and some preliminary suggestions are made.
Pitchforth, E. and E. v. Teijlingen (2005). "International public health research involving interpreters: a case study from Bangladesh." BMC Public Health 5(71).
Author’s Abstract: Cross-cultural and international research are important components of public health research, but the challenges of language barriers and working with interpreters are often overlooked, particularly in the case of qualitative research.
Temple, B. (1997). Watch Your Tongue: Issues in Translation and Cross-Cultural Research. Sociology 31(3): 607-618.
Author’s Abstract: Considering the large amount of research being undertaken in Eastern Europe and elsewhere involving the use of more than one language, there has been a remarkable silence in sociological debate about the status of this research. In this article I argue that such issues should be of concern to social scientists generally as well as to linguists. Using my own research with British-Polish communities, I raise some concerns surrounding the translation of concepts. I suggest one way of beginning to address these problematics: an opening of discussion on an analytical level with those who are often seen as mere technicians, translators and interpreters.
Temple, B. and A. Young (2004). Qualitative Research and Translation Dilemmas. Qualitative Research 4(2): 161-178.
Author’s Abstract: The focus of this article is an examination of translation dilemmas in qualitative research. Specifically it explores three questions: whether methodologically it matters if the act of translation is identified or not; the epistemological implications of who does translation; and the consequences for the final product of how far the researcher chooses to involve a translator in research. Some of the ways in which researchers have tackled language difference are discussed. The medium of spoken and written language is itself critically challenged by considering the implications of similar ‘problems of method’ but in situations where the translation and interpretation issues are those associated with a visual spatial medium, in this case Sign Language. The authors argue that centring translation and how it is dealt with raises issues of representation that should be of concern to all researchers.
- Kamler B & Threadgold T (2003) 'Translating Difference: questions of representation in cross-cultural research encounters' Journal of Intercultural Studies Vol.24, No. 2 pp. 137-151
This paper addresses questions of cross-cultural communication and represen tation as they arose in a longitudinal research project which sought to learn about the lives and concerns of older women. It focuses on the translations and mistranslations that occurred in narrative workshops where Australian researchers, who did not speak Vietnamese, worked with Australian Vietnamese women aged 55-74 and a translator to produce video diaries of the older women's everyday life. A number of workshop interactions around storytelling are examined to document the complexities that can arise when communities meet and interact across cultures. The aim is to 'come clean' about the problems of trying to conduct research without a common language and to suggest just how difficult translations and representations of culture really are and how easily preconceptions and cultural positionings interfere with the process of communication that is actually occurring.
Singal, N., and Jeffery, R. (2008). Qualitative Research Skills Workshop: A Facilitator's Reference Manual, http://oer.educ.cam.ac.uk/wiki/RECOUP, Cambridge: RECOUP (Research Consortium on Educational Outcomes and Poverty, http://recoup.educ.cam.ac.uk/). CC BY-NC-SA 4.0. (original page)