During the 20th century , it was commonly assumed that the best way to teach English as a foreign language was through the exclusive use of English as the medium of instruction . Although recent publications have challenged this view , many educational policy-makers in schools and universities still tend to to insist on ?English only" , and decry the practice of codeswitching between the target language and the first languages of learners and teachers . Findings from a recent volume of case studies clearly show that codeswitching is a common practice across a wide range of university EFL classrooms in Asian contexts . In many cases , this codeswitching may be regarded as a ?flexible convergent approach" , where languages are switched more or less spontaneously and at random . Often , however , the teachers reported in Barnard and McLellan ( 2014 ) responsibly alternated between languages in a principled and systematic way . This paper presents and discusses observational data and extracts from interviews with teachers on the rationale for codeswitching . Thus codeswitching is both normal and can be pedagogically justifiable . The paper will then suggest how teachers can reflect in-, on- and for-action ( Farrell , 2007 ) by systematically recording , listening to , and analysing their use of language(s) in their own classrooms , using a set of interactional categories (Bowers , 1980 ) . Reflective practitioners may then develop collaborative action research projects (Burns , 1999 , 2010 ) which could empower them to critique and challenge monolingual language-in-education policies .