Rachel Heath
Cardiff University
Double the work: The unanticipated labours of bilingual research
Paper presented at the British Educational Research Association Annual Conference, Cardiff University, September 7-10 2000
This paper asks whether researchers should make the effort more often to work with their informants through their first language when this is different to English. It is so tempting to say ‘well can they speak English?’ and then turn the study into a monoglot cut and dried case. The discussion is a reflexive account of a bilingual study conducted in both English and Welsh in three contrasting regions of Wales. It explores the potential for multicultural and multilingual research by acknowledging the growing scope for working with informants in their first language when this is not English. The paper argues that informants should probably have the RIGHT to participate in their first language and that there are many benefits as well as costs for the research when their ‘right to choose’ is exercised. From experiencing bilingual fieldwork, the researcher reflects on the additional dimension of the language medium of the research study. It is argued that there are additional dimensions to consider when the fieldwork involves such settings as schools and colleges. These include field roles and language identity as well as the rapport that can be built when working with informants expressing themselves in their first language. As well as arguably yielding ‘better data’ by working bilingually, the life of a research project can be a very different experience.
The paper explores the many unanticipated human and physical resource costs incurred as well as the numerous problems and pitfalls at each stage of the research project’s life. Issues surrounding courtesy and competence with a language are outlined in the discussion on interviews and their sensitive negotiation and renegotiation phases. The problems of recruiting appropriate assistance when the language of transcription is other than English are outlined. One of the areas discussed is the use of “Mac World” computer software that does not recognise minority languages or allow a recognised system for analysing bilingually.
The paper also discusses the issues surrounding the dissemination of findings when there are audiences of different language competence and preference. This account concludes that bilingual research is more than double the workload and in the context of the current project it has been worthwhile on two counts. Firstly, informants have had the opportunity to work with the language with which they are most comfortable. Secondly, by expressing themselves in their first language, the costs are justified in the rich data that has been collected.
It is now widely accepted that we live in a multicultural society and there is research work that documents such diverse cultures. However, even when the research questions are not ‘cultural’ per se, it is still reasonable to expect that when entering the fieldwork setting, a number of the informant sample may also be from ethnic minority groups. There are around two million adults in Britain whose first language is not English and one third of these can speak little or no English (Edwards, 1998; see also Baker and Jones, 1998). Paradoxical to the growth in studies on multicultural issues, very little literature describes how data are collected when working with such a diverse sample of informants. In the multicultural society in which we find ourselves, it is increasingly possible to work with informants whose first language may not be English, hence, there is an argument for working in a whole host of different languages when conducting research. There is therefore a growing potential to work multilingually and/or multiculturally. Yet, having noted this potential for researching through several languages, there are key issues and practical concerns that the researcher needs to address in the research design and strategies in order for the research timetable to be realistic and productive. This paper notes that when undertaking research projects involving more than one language, there are many practical challenges that a lone researcher or research team should consider.
This paper draws upon fieldwork, data preparation and the analysis and presentational issues deriving from my own doctoral research on language and post-16 markets[1]. The PhD explores the post-16 choices of students in Welsh medium schools. One hundred and one semi-structured interviews were conducted with students in Year 11, Year 12 of sixth form, and at first years of FE College to investigate their perceptions of their post-16 choices, pre- and post making that choice. The study focused on investigating the underlying reasons for choosing key institutions in sampled markets and whether the language medium of that provision was also a reason. The study also probes into the preconceptions to choices and the influences on students, namely parents, career guidance, school teachers and friends, for example. This research project will highlight a new field of research on language medium provision and offer a new angle to the choice making process research field. Also, the paper contributes towards a general methodological commentary and highlights the potential for a growing strand of bilingual research, or researching in multiple languages.
Whilst one FE College in my study sample, Coleg Teifi, recruited 90 per cent of its full-time students from ethnic minority backgrounds, the research work I wish to address in this paper explains the particular complexities faced when researching current and ex-Welsh medium school students. Though the number of ethnic minority students sampled was minimal, many of the informants perceived English as their second language. Welsh speakers as well as Welsh medium school students could be viewed as a ‘numerical minority’ (Mason, 2000), because their numbers are small relative to the total population of Wales and the UK. However, as Mason (2000) argues, a group of people labelled as a ‘minority’ is often too simplistically taken as a numerical minority. Whilst my informants might be a numerical minority over a large geographical area, there are regions of Wales where Welsh speakers are in the majority in their community (e.g. Gwynedd) (Baker and Jones, 1999). Mason (2000) explains the concept of numerical minorities, by saying that black people in South Africa are a majority, just as women are a numerical majority. It is therefore more appropriate to define ‘minority’ in terms of power:
Minorities being those groups in subordinate positions irrespective of their relative size” (Mason, 2000: 15)
Hence, in taking the power definition of minority, it is relevant and appropriate to include the under-represented voices of Welsh speakers and Welsh medium school students as a ‘minority’. It is therefore extremely important that researchers allow ‘voice’ to minority groups, especially those who can better express themselves and their cultures through their first language. Baker and Jones (1998) argue that a culture is often best expressed through its language, and so informants can better express notions and concepts concerning that culture via the language in which the culture manifests itself.
The two languages in which I worked for this study were, of course, Welsh and English. This paper therefore notes the challenges and practical difficulties of embarking upon bilingual research, but any reader studying multiple ethnic groups and working with multilingual informants may find the issues noted here as interesting and even consoling.
No Stopping to Think or Blink: A Right to Choose
As a first language Welsh speaker, it was an automatic decision for me to provide my student informants with the choice of language medium in which to conduct our informal interviews. At the outset of fieldwork, I perceived giving informants the choice of language in which to participate as the ‘ethical thing to do.’ In the interests of rapport, it was hoped they would feel more comfortable in the research setting if they could contribute using their first and most natural language. The interview agenda was framed to explore the Welsh medium post-16 provision (or lack of it) and what choices each student perceived were available in the market for them. In explaining their choice of language for the interview, their justification and rationales often provided insights into their decisions for choosing one or the other languages (English or Welsh) for their studies. Furthermore, in arguing for students to have the ‘right of choice’ and access to Welsh medium post-16 provision, I felt strongly that students had the right to choose the language medium of their interview. However, in making this operational and ‘ethical’ decision at the beginning, I paved the way for myself to work bilingually. The following sections of the paper will outline what this important decision meant in terms of data collection, transcription, analysis and other phases during the life of the bilingual research project.
Factors shaping the ‘language’ of a research project
The informant’s right to a choice of study medium as well as a choice in the research situation does depend on many key factors. One key variable is the availability of confident and competent researchers that are willing and able to conduct research in the first language of the informants sampled. Karseras and Hopkins (1987) and Edwards (1998) are examples of researchers who feel it right and proper to offer their informants the choice of language, but they require interpreters to perform bilingual research. These authors describe the problems of recruiting appropriate interpreters and discuss how to manage a triangular relationship when interviewing. Hence whilst there may be a recognised right to choice of language, identifying adequately skilled linguists may provide a contradiction between right, the ethical thing to do and practicalities of research work. As a lone researcher involved in a single handed study the informant’s right to choice has to be balanced with the costs of hiring key personnel with the expertise in such minority languages and cultures. Indeed as numerous writers have pointed out (Walford, 1991 and 1987, Burgess, 1984, Van Maanen, 1988, Delamont, 1984 and Atkinson, 1984) single-handed researchers repeatedly report their ‘heroic tales’ as ‘loners’ (Van Maanen, 1988: 2). It is also tempting to work monolingually and arguably through the English, because so many minority language speakers can also speak English to some degree. In the case of Welsh as a minority language, monoglot speakers are now rarely found and so potentially, research can in general, always be conducted in English (for those participants that are willing to comply!). Interestingly, Williams and Raybould (1991) argue that Welsh, as a minority language would have received more recognition and status if Welsh speakers had remained monoglot. Indeed, the profile of published research in Welsh and written in Welsh would have been greater by now. Arguably, the ‘crisis status’ of Welsh and other minority languages is linked to the linguistic dexterity that speakers demonstrate.
Having attempted a bilingual research approach, the relative ‘straightforwardness’ of monolingual research is temptingly attractive. Despite the complexities arising from working bilingually for the doctoral investigation, I do feel there have been benefits in terms of the data collected and the access granted to certain settings. A later section argues that a closer rapport was built with students who took the opportunity to speak in their first language, as they felt more comfortable.
Ultimately then, it can be argued that the status that the English language enjoys partly prevents more research being conducted through several languages. Hopefully this paper begins to highlight aspects of research processes in projects that attempt to work in minority languages. Of course, there is a draw back to arguing that most people can manage to speak English. Such a view fails to acknowledge that many people perceive English as a second language and would indeed be more comfortable speaking (and more lucid even) through their first language, be it Welsh, or another minority language. In the interests of good practice interviewing, the researcher may wish to consider one way of building rapport and relationships by communicating with their informants through their mother tongue. Furthermore, in the elicitation of honest, accurate data in the vocabularies and contexts of the informants- the use of Mother tongue is even more important.
With hindsight, the impulse decision to provide a choice of English or Welsh interviews, was largely automatic and of course linked to my abilities to speak both languages fluently. Looking back through my research diary and comparing my doctoral project workload with those of other research scholars has enabled me to see that my initial decision had cost me dearly, in terms of time, cost, energy and many other already thinly dispersed resources. In the section which follows, the practical issues that arose from deciding to provide this language medium choice to my informants will be discussed.
The practical problems and pitfalls that arose at each corner of the doctoral study are highlighted across this paper. Many of the issues raised in detail in the paper were unanticipated labours of bilingual researching, because the decision to work bilingually stemmed from that initial intuitive decision to provide a choice of language to informants. Starting with writing letters of permission and access and preparing interview schedules and focus group games, right through to the dissemination phase, this paper draws attention to the more than ‘double the work’ of bilingual research. Not only does this paper outline the time pressure for the lone researcher, but also the following section outlines the expense and strain that bilingual research put on resources and my finances.
The data collection phase is discussed with many facets highlighted including the problems with communicating across different dialects and with second language speakers, and the barriers to rapport caused by the variety in language accents. Qualitative researchers recognise the importance of rapport, but this researcher was unprepared for the challenges that bilingual research provided when attempting to build rapport. The paper unpacks some of the several assumptions behind the first instinctive decision, one being that it was believed that first language Welsh speakers would choose Welsh interviews and be more comfortable and competent doing so, resulting in richer data and closer rapport/ relationships. One section argues that rapport was often more successfully built when the language of the interview was Welsh, however this is balanced, if not contradicted by another section outlining the problems with understanding each other even when a choice of language had been provided for all interviewees. Finally the practical problems of working with computer software during transcription and analysis is described. These problems stem from a lack of awareness that people do work in languages other than English on computer and that there is a lack of development of alternative, even ‘native’ packages with which bilingual researchers can work. To complete the cycle of a research project’s life, the paper outlines the additional dissemination work undertaken for this project. To summarise, this paper argues that there is a careful trade-off to be considered between the ‘right to choose’ for informants and the rights of the researcher to work a reasonable amount of hours per week!
Preparing to enter the field
‘TIME HUNGRY’
In preparing to enter the fieldwork setting, interview schedules and access letters were drafted. For the English medium institutions in Wales, I produced one letter, and for the Welsh medium or bilingual providers, I composed a Welsh letter. Some providers were sent both versions, as I was not sure in every case, which language some preferred. As readers will appreciate, this operation alone involved additional time for writing, proof reading, and printing etc. I am fortunate to have access to friends who are professional Welsh translators and one friend kindly checked over initial letters to ensure these were grammatically flawless. This fortunately avoided one additional cost. Despite being educated through the Welsh language myself, it was prudent for a professional eye to take a glance over the draft, as this would provide a professionalism which I wanted to characterise my fieldwork. Institutional gatekeepers were likely to be Welsh graduates, or teachers, or other professionals where administrative errors could have determined whether or not access to the research sites and settings were granted.
Of course this meant double the time in preparing to enter the field. Indeed, as well as access documentation being bilingual, numerous other materials were prepared for the planned fieldwork. Figure 1 lists those tasks that a lone researcher undertakes prior to entering the field. What needs to be recognised is that in each of the documents prepared for fieldwork, two copies were produced from draft form, through to final copies. These two copies were in Welsh and English and therefore took double the time, double the work, and double the effort: all in all, making it more than double the work! Figure 2, compiled from my research diary, provides an idea of the additional phases within this for the bilingual researcher preparing to enter the field (see next section).