Parts I, II and III of the book all carry conclusions of their own, and so I would like to end with some methodological reflections. The research that I have reported can be described as ‘linguistic ethnography’, and as such, by taking a close look at situated language use, it has tried to provide insights into everyday social and cultural production that are both fundamental and – methodologically – relatively distinctive. But exactly what kinds of ‘insight’ are these, and how were they actually put together? If books such as this want to be seen as a contribution to social science, if students are to learn how to do it themselves, and if researchers in other disciplines are to weigh up the potential value of the procedures at work in analyses like mine, it is important to be clear about (a) the ways in which this book makes claims to wider relevance, and (b) the steps and elements involved in the production of these knowledge claims. So in this final chapter, I shall try to formulate an explicit account of the analytic procedures, focusing in particular on the interplay between data and theory.

It has often been said that, in ethnography, researchers allow their data to speak for itself, and that once the researcher has come to understand a local culture from within, the meanings of activity can be captured in its own terms (Hammersley 1992:19–20,22; Hammersley and Atkinson 1995:Chapter 1).