This paper grows out of research into the way people in local churches read and interpret the Bible. It considers a particular Bible-study group, paying attention to the kind of talk that occurs there, and the way in which meaning is achieved in the interaction between participants.

Initial data extracts are taken to suggest a tension between deference to the authority of the Bible, and an alternative approach, concerned with ‘what the text says to me/us’. This latter approach is inclusive of different viewpoints.

This initial analysis gives rise to the hypothesis that there might exist within this Bible-study talk two ‘interpretative repertoires’. In comparison with scientific talk, as considered by Gilbert and Mulkay (1984), these are labelled as ‘canonical’ and ‘contingent’ repertoires.

This hypothesis is tested through examination of further data. The paper concludes that the hypothesis does highlight a distinction between more canonical and more contingent talk, and enables an examination of the construction of these ‘repertoires’. However, the term ‘interpretative repertoire’ tends to overemphasis the boundaries round these constructions, and to downplay the lexical and other connections between them. The paper concludes that the term ‘discursive node’ provides for a better characterisation of these patterns of discourse.