Social Research update logo
Issue 31 Winter 2000

Social Research Update is published quarterly by the Department of Sociology, University of Surrey, Guildford GU7 5XH, England. Subscriptions for the hardcopy version are free to researchers with addresses in the UK. Apply to SRU subscriptions at the address above, or email sru@soc.surrey.ac.uk.

Translating from one language to another

Maria Birbili

Maria Birbili is currently a lecturer in Educational Research Methodology at the department of Educational Studies, University of Oxford. After finishing her DPhil on Greek higher education, she worked as a research officer for 18 months. Her research interests include issues involved in the socialization of academics and institutional differentiation, and the training of researchers.

Once practised almost exclusively by anthropologists, collecting data in one language and presenting the findings in another is now increasingly common among social researchers. As student and staff mobility increases, a considerable number of theses, dissertations and funded-research projects concern studies which involve moving between languages, sometimes even from the very first steps of the research endeavour. As the need for mutual understanding on an international scale increases, more and more organisations and individuals are seeking comparable information across national and cultural boundaries using research instruments prepared in one language and culture for use in others.

Moving between languages can take different forms. For example (taking English as the language in which the research is to be reported): an English-speaking researcher might conduct an interview in a language other than English; a researcher interviews in her primary language which in not English; researcher and participants are ‘fully and fluently bilingual - they slip between the two languages during the interview’ (Rossman and Rallis, 1998:161). In cases of international assessments or cross-cultural research test items, questionnaires or interview schedules are translated from English into the language of different countries and vice versa.

Although not all studies which involve different languages are equally dependent on the precision of linguistic data (for instance, in studies that are predominantly observational), researchers need to be aware that the translation process usually requires both time and effort on their part and can present various types of problems, some of which may not be completely overcome (Phillips, 1960; Lewin, 1990; Broadfoot and Osborn, 1993; Ercikan, 1998). More importantly, however, researchers need to keep in mind that translation-related decisions have a direct impact on the validity of the research and its report.

Factors influencing the quality of translation

The quality of translation depends on a number of factors, some of which, as Phillips (1960:290) says, may be beyond the researcher’s control. In those cases where the researcher and the translator are the same person the quality of translation is influenced by factors such as: the autobiography of the researcher-translator; the researcher’s knowledge of the language and the culture of the people under study (Vulliamy, 1990:166); and the researcher’s fluency in the language of the write-up. When the researcher and the translator are not the same person, the quality of translation is influenced mainly by three factors: the competence, the autobiography and what Temple (1997:610) calls ‘the material circumstances’ of the translator, that is the position the translator holds in relation to the researcher.

Different dimensions of potential translation-related problems

Gaining conceptual equivalence

One of the major difficulties of any kind of research in which the language of the people under study is different from that of the write-up is gaining conceptual equivalence or comparability of meaning (Deutscher, 1968; Whyte and Braun, 1968; Sechrest et al., 1972; Temple, 1997). Phillips (1960:291) sees this ‘in absolute terms an unsolvable problem’ which results from the fact that ‘almost any utterance in any language carries with it a set of assumptions, feelings, and values that the speaker may or may not be aware of but that the field worker, as an outsider, usually is not’. Whether one is trying to translate a survey instrument, an interview schedule or a test, as several researchers caution us, even an apparently familiar term or expression for which there is direct lexical equivalence might carry ‘emotional connotations’ in one language that will not necessarily occur in another. A good example of such a case is the expression ‘civil service mentality’. Although, as one might argue, this expression conjures up a similar ‘image’ in several cultures (for example, that people who have a civil service mentality are, as Moses and Ramsden (1992:102) say, ‘very observant of their rights’), it might not be easy for the English reader to pick-up the full implications the term carries for a Greek unless it is accompanied by more ‘cultural’ information on the (negative) associations and connotations that the term ‘civil mentality’ has in a Greek context.

On those occasions where two languages do not offer direct lexical equivalence several researchers and linguists suggest that one’s efforts should be directed ‘towards obtaining conceptual equivalence without concern for lexical comparability’ (Deutscher, 1968:337; Whyte and Braun, 1968:121; Bassnett-McGuire, 1980; Overing, 1987; Broadfoot and Osborn, 1993; Temple, 1997:610). For many researchers (Sechrest et al., 1972; Brislin et al., 1973; Warwick and Osherson, 1973) the process of gaining comparability of meanings is greatly facilitated by the researcher (or the translator) having not only ‘a proficient understanding of a language’ but also, as Frey (1970) puts it, an ‘intimate’ knowledge of the culture. Only then can the researcher pick up the full implications that a term carries for the people under study and make sure that the cultural connotations of a word are made explicit to the readers of the research report.

Comparability of grammatical forms

A different kind of translation problem occurs when sentences in the language of data collection involve grammatical and syntactical structures that do not exist in English. Syntactical style, as Ercikan (1998:544) points out, is one of the most difficult features to carry over from one language to another. Where sentences are resistant to translation because of incompatibility between languages in terms of the structure, Bassnett-McGuire (1980:32) suggests that the sense of sentences ‘can be adequately translated into English once the rules of English structure are applied’. However, such a process, as Ervin and Bower (1952:597-598) warn us, inevitably involves ‘the introduction of pseudo-information or the loss of information’.

Although obtaining grammatical and syntactical equivalence is not something that can be taken lightly, it does appear that the more important aim of researchers-translators should be to achieve conceptual equivalence.

Making participants’ words accessible and understandable

Seen as sometimes the only opportunity that readers of research reports have to ‘see for themselves’ what participants ‘look like’ (Wolcott, 1994), the use of direct quotations deserves careful attention in discussions about translation. Decisions about translating quotations are of course dependent on the intended function of the quotation in the research text and whether one perceives translated words as a direct quotation (Rossman and Rallis, 1998:162).

One of the first decisions that researchers are asked to make when translating participants’ words is whether to go for ‘literal’ versus ‘free’ translation of their text. A literal translation (i.e. translating word-by-word) could perhaps be seen as doing more justice to what participants have said and ‘make one’s readers understand the foreign mentality better’ (Honig, 1997:17). At the same time, however, such practice can reduce the readability of the text, which in turn can test readers’ patience and even ability to understand ‘what’s going on’.

Researchers who decide to go for the more ‘elegant’ free translation, on the other hand, need to think of the implications of creating quotations that ‘read well’. Even in one’s own language, editing quotations always involves the risk of misrepresenting the meaning of the conversational partner (Rubin and Rubin, 1995:273). In translated quotations the risk of losing information from the original is greater.

The use of translators or interpreters

In some studies, the researcher and the translator or interpreter are not the same person and there might even be more than one translator involved in a research project. These people might be professional translators, bilingual people with knowledge of the topic under investigation (or not), or native speakers employed to help the researcher communicate with respondents who do not speak English.

As Temple (1997:614) points out, the use of translators and interpreters ‘is not merely a technical matter that has little bearing on the outcome. It is of epistemological consequence as it influences what is “found”’. Kluckhohn (1945) suggests that there are ‘three basic problems which arise from the use of interpreters: a) the interpreter’s effect on the informant; b) the interpreter’s effect on the communicative process; and c) the interpreter’s effect on the translation’ (quoted in Phillips, 1960:297). Focusing on the latter, Temple (1997:608) argues that researchers who use translators need to acknowledge their dependence on them ‘not just for words but to a certain extent for perspective’. In doing so, researchers need to constantly discuss and ‘debate’ conceptual issues with their translators in order to ensure that conceptual equivalence has been achieved (Temple, 1997:616).

Techniques for dealing with translation-related problems

Important techniques for eliminating translation-related problems include back translation, consultation and collaboration with other people during the translation process and pre-testing or piloting (for example, interviews) whenever this is possible.

Back translation, one of the most common techniques used in cross-cultural research, involves looking for equivalents through a) the translation of items from the source language to the target language, b) independent translation of these back into the source language, and c) ‘the comparison of the two versions of items in the source language until ambiguities or discrepancies in meaning are clarified or removed’ (Ercikan, 1998:545; Warwick and Osherson, 1973:30). Although it can be helpful, as Deutscher (1968:321) points out, in identifying semantic errors in translations, some researchers (Phillips, 1960; Sechrest et al., 1972; Broadfoot and Osborn, 1993) argue that back translation is far from the ideal solution and can create new problems. For example, ‘it can ... instil a false sense of security in the investigator by demonstrating a spurious lexical equivalence’ (Deutscher, 1968:322). Back translation can also be a very time-consuming procedure, and might require more than one person (or a dictionary) involved in order to achieve good results.

Consultation with other people, on the other hand, involves discussions about the use and meaning of words identified as problematic with people who are bilingual (Whyte and Braun, 1968; Brislin et al., 1973) or having a number of people sitting around a table jointly making decisions about the best terms to use (Brislin et al., 1973:46). Collaboration with other people can also take the form of researchers from all countries involved in a study, jointly producing the research design and instrument.

Whether interviews or questionnaires or any kind of test are to be used, another way of eliminating translation-related problems is to pre-test or pilot the research instrument in the local culture. When pretesting a research instrument, Warwick and Osherson (1973:33) see it as particularly important to ask respondents not only for their answer but also for their interpretation of the item’s meaning. Once a questionnaire or assessment instrument has been constructed, another way of identifying problems is the application of statistical methods (Hambleton, 1993; Ercikan, 1998).

For many researchers, combining some or all of the above-mentioned techniques is seen as the best and most efficient way to deal with translation-related problems. When using multiple methods, as Brislin et al. (1973:51) argue, the weakness of one method could be offset by the strengths of the other.

The need to make translation-related decisions explicit

As is common to all scientific enquiry, reports of research which involves the use of more than one language need to include a thorough description of the translation-related issues, problems and decisions involved in the different stages of the research process (Temple, 1997:613). In addition, researchers need to describe the circumstances within which translation took place and discuss the techniques they used during the translation process. Similarly, if the research involved the use of translators, readers need to be informed about who those people were and what kind of role they played at all stages of the research endeavour.

Conclusion

When collecting data in one language and presenting the findings in another, researchers have to make a number of translation-related decisions. Words which exist in one language but not in another, concepts which are not equivalent in different cultures, idiomatic expressions and/or differences among languages in grammatical and syntactical structures are issues which call for very specific decisions. These decisions along with factors such as, for example, who the researcher or her translators are and what they ‘know’ have a direct impact on the quality of the findings of the research and the resulting reports.

References

Bassnett-McGuire, S. (1980) Translation Studies. London: Methuen

Brislin, R. W. (1976) Introduction. In R. W. Brislin (ed) Translation: Approaches and Research. New York: Gardner Press

Brislin, R. W., Lonner, W. and Thorndike, R. M. (1973) Cross-Cultural Research Methods. New York: John Wiley & Sons

Broadfoot, P. and Osborn, M. (1993) Perceptions of Teaching: Primary School Teachers in England and France. London: Cassell

Deutscher, I. (1968) Asking Questions Cross-Culturally: Some Problems of Linguistic Comparability. In H. S. Becker, B. Geer, D. Riesman and R. S. Weiss (eds) Institutions and the Person. Papers presented to Everett C. Hughes. Chicago: Aldine Publishing Company

Ercikan, K. (1998) Translation Effects in International Assessments. International Journal of Educational Research, 29, 543-553

Ervin, S. and Bower, R. T. (1952) Translation Problems in International Surveys. Public Opinion Quarterly, 16, 595-604

Frey, F. (1970) Cross-Cultural Survey Research in Political Science. In R. Holt and J. Turner (eds) The methodology of comparative research. New York: The Free Press

Hambleton, R. K. (1993) Translating Achievement Tests for Use in Cross-National Studies. ERIC Report, ED 358128

Hatim, B. and Mason, I. (1990) Discourse and the Translator. London: Longman

Honig, H. (1997) Positions, Power and Practice: Functionalist Approaches and Translation Quality Assessment. Current Issues in Language and Society, 4(1)

Lewin, K. (1990) Data collection and analysis in Malaysia and Sri Lanka. In G. Vulliamy, K. Lewin and D. Stephens (1990) Doing Educational Research in Developing Countries London: The Falmer Press

May, T. (1993) Social Research. Issues, Methods and Process. Buckingham: Open University Press

Moses, I. and Ramsden, P. (1992) Academic Values and Practice in New Universities. Higher Education Research and Development, 11(2), 101-118

Overing, J. (1987) Translation as a creative process: the power of the name. In L. Holy (ed), Comparative Anthropology. Oxford: Basil Blackwell

Phillips, H. P. (1960) Problems of translation and meaning in field work. In R. N. Adams and J. J. Preiss (eds) Human Organisation Research: Field Relations and Techniques. Homewood, ILL: Dorsey Press Inc.

Rossman, G. B. and Rallis, S. F. (1998) Learning in the Field. An Introduction to Qualitative Research. Thousand Oaks, CA: Sage

Rubin, H. and Rubin, I. (1995) Qualitative Interviewing. The Art of Hearing Data. Thousand Oaks, CA: Sage

Sechrest, L., Fay, T. L. and Zaidi, S. M. H. (1972) Problems of Translation in Cross-cultural research. Journal of Cross-Cultural Psychology, 3(1), 41-56

Temple, B. (1997) Watch your tongue: issues in translation and cross-cultural research. Sociology, 31(3), 607-618

Vulliamy, G. (1990) Research Outcomes: PostScript. In G. Vulliamy, K. Lewin and D. Stephens, Doing Educational Research in Developing Countries. London: The Falmer Press

Warwick, D. P. and Osherson, S. (1973) Comparative Analysis in the Social Sciences. In D. P. Warwick and S. Osherson (eds) Comparative Research Methods: An Overview. Englewood Cliffs, NJ: Prentice-Hall

Whyte, W. F. and Braun, R. R. (1968) On Language and Culture. In H. S. Becker, B. Geer, D. Riesman and R. Weiss, (eds) Institutions and the Person. Papers presented to Everett C. Hughes. Chicago: Aldine Publishing Company

Wolcott, H. F. (1994) Transforming Qualitative Data. Description, Analysis, and Interpretation. Thousand Oaks: Sage

Department of Sociology
University of Surrey
Guildford GU7 5XH
United Kingdom.

Telephone: +44 (0) 1 483 300800
Fax: +44 (0) 1 483 306290

Edited by Nigel Gilbert.

Winter 2000 © University of Surrey

Permission is granted to reproduce this issue of Social Research Update provided that no charge is made other than for the cost of reproduction and this panel acknowledging copyright is included with all copies.