Much time and effort has gone into trying to demonstrate an empirical link between research activity and teaching performance. In general, the correlations between these factors have been shown to be low. This paper argues that the attempt to find such a link will always be confounded by different conceptions of the two enterprises. The debate about the relationships between teaching and research as presently conceived is not fruitful. It there is a link between the two it operates through that which teaching and research have in common; both are concerned with the act of learning, though in different contexts. Greater emphasis needs to be placed on the ways in which knowledge is generated and communicated. Those aspects of teaching which lead to learning and the learning which occurs through research provide the vital link. This is important if the debate is to progress beyond a political defence of the status quo and be of practical use to considerations of whether, in higher education, teaching without research is to be encouraged.