After reviewing the purposes of descriptive research on teaching in general, this paper discusses the value of using a system of analysis and a technical language to describe classroom interaction. The need for such descriptive research in language classrooms is then outlined in detail. A critical review of the few systems of analysis designed for the language classroom and of the research studies done in this field is presented. The paper concludes with a call for more objective research that is experimental, as well as descriptive, in nature. The hope is expressed that these studies will provide the empirical evidence that is needed in order to prescribe specific teacher behaviors that will effect desired student outcomes. (Author/GC)