Research methodology in contextually- based second language research Larry Selinker University of Michigan and Dan Douglas lowa State University Addresses for correspondence: Larry Selinker, ELI and Linguistics Program, The University of Michigan, Ann Arbor, MI 48109, USA: Dan Douglas, English Department, Iowa State Univer- sity, Ames, IA 50010, USA. We suggest a combined research methodology for studying SLA in real-life and important contexts, a methodology sensitive to the demands of such contexts. For studying IL learning in context, our suggested methodology combines and integrates aspects from three fields: grounded ethnography in ethnomethodol- ogy ; subject-specialist informant procedures in language for specific purposes; and rhetorical/grammatical strategies in discourse analysis. We first present evidence for the importance of devising a research approach to contexually- based SLA. Then we sketch the suggested research methodology and present two extended case studies which illustrate the methodology. We view such research as complimentary to universal approaches to core IL grammar. Finally, we sketch our theoretical approach, showing a possible link between research in universal and contextually-based SLA. I Background This paper suggests a methodology for studying second language acquisition (SLA), use and fossilization when the research concern is understanding interlanguage (IL) particularities in real-life and important contexts. Our suggested approach for this purpose is to create a combined research methodology demanded by circum- stances. It is our working hypothesis that this methodology reduces problems of extrapolation by more directly relating empirical results to such contexts. The contexts we are most interested in are the academic settings in which we work and in which a lot of SLA takes place, called language for specific purposes (LSP) contexts. In our approach, one thing we are deeply concerned with is gaining insight into the use and development and possible fossilization of IL in these important real-life work contexts. For example, we are interested in understanding the ability of a nonnative speaker (NNS) to use English