Dr Anwei Feng

As a supervisor of many international students in the past decade, like many members of the research team, I have also been frequently asked how language issues can be dealt with in empirical research and in data analysis. My answer to this question is normally a pragmatic one based on my previous experience in dealing with multilingual research students. Because of heavy workload and tight schedules most of academics in HEIs face these days, I never got around to acting upon the issues, even though I was aware of the relevance of the issues to research quality.

In the last three year, I have been passionately engaged in a nation-wide research project titled trilingualism in China. This project has taken me to many fascinating regions where minority groups dominate. Language issues have constantly popped up during the whole process of the research study, including designing of research tools and methods for field work, analysing data, writing up papers and disseminating research findings. Some researchers in the network in China have expressed vastly different views. Talking about the general principles, some, for example, state firmly that to research in minority dominated communities, the researcher should first of all be a (balanced) trilingual the minority home language (L1), the majority Han language (L2), and English (L3) him/herself. Otherwise, they argue subtle cues and vital information would be lost at any stage of research. Some, on the other hand, are complacent for the ‘fact’ that they have never experienced any problem in conducting research using only L2, the dominant language. They further argue that there is clearly an imbalance between the research that needs to be done and qualified researchers from minority background. Hence, most research in minority education has indeed been conducted by researchers who don’t speak L1. As the PI of the project, however, although no rigid rules were made, I did insist at the initial stage that issues on which language to use in research, whether for questionnaire surveys, in interviews in the real world or in data analysis, should not be ignored. Decisions should be made according to clearly-defined criteria such as the degree of ethnolinguistic vitality in each context.

Through my own site visits to different regions and ad hoc observations of the work done by different teams in the past three years, I have noticed that the language issues are more acute than expected on many occasions. In order to find out first-hand information about language use and language allocation in a school, for example, I sometimes found myself sitting in an office facing some school officials who occasionally talked to each other in their L1, or in a classroom where the medium of instruction was in L1 most of the time. Even though I could observe the class and the school from a non-L1 speaker’s point of view and still found the experience interesting, there could indeed be much that was lost in the process. Affectively, if I learned a few words such as those for greeting, thanking and joking in L1 and used them, I would often win surprise smiles and get more favourable responses in return. Linguistically, some researchers including me noticed that some minority-language speakers often struggled to find suitable words in L2 to express themselves. Not often but interestingly, some teacher informants would prefer to use L3 in the chats. However, on most occasions, most informants did not show problems in conversing in L2. How would all these mean as far as the quality of research data is concerned? How could and should the issues surrounding language(s) used in research be better addressed? … These beg research and answers.