ABSTRACT
In this article it is argued that interventions in pupil-pupil interactions have to be grounded in a thorough understanding of pupil meanings and the social processes which generate them. The argument is elaborated via a critique of some existing studies of such interventions, like those of peer tutoring, and in discussion of the findings of an interview study. In general, studies of peer tutoring have been psychology-based and have failed to consider the significance of pupils' understanding of deviance derived from pupils' social location. An alternative approach is explored utilizing the findings of ethnographic studies, and the findings of an interview study with fourth year pupils in a secondary school. The data is organised using the constructs of consensual and non-consensual deviance. Pupils formally designated ‘examination’ pupils revealed different views from those designated ‘non-examination’ and ‘less able’ although they also held views in common. The implications of these findings for peer tutoring interventions are discussed.