AbstractTeaching commonly involves asking questions, in sequences that enable a teacher to control the direction and duration of subject‐matter discussion, while also maintaining attention and order. The form of questions and their role as means of instruction have received more study and discussion than the function of questions and their role in achieving particular ends of instruction. This study examines qualitatively the function of questions in developing arguments that establish scientific knowledge claims on the basis of reasons and evidence, and thereby suggest a rational attitude toward authority. Peters' (1966) distinction between a teacher's (rational) authority of knowledge and (traditional) authority of position is linked with Toulmin's (1958) pattern for rational arguments to establish a qualitative framework for judging the function of questions in arguments. Episodes from three science lessons are presented in verbatim transcription and analyzed to reveal three different ways in which teachers did not achieve the standard of suggesting a rational attitude toward authority. Question sequences such as these have a clear potential for distorting student understanding of the nature of scientific authority, with possible negative consequences for students' attitudes toward scienc