Ethnography has always been subject to criticism from quantitative soc
iologists, who accord it a minimal role, but it has recently come unde
r attack from sociologists sympathetic to the method, who themselves h
ave considerable experience in its use. I call this the ethnographic c
ritique of ethnography. This critique questions the reliability of eth
nographic descriptions, and shows ethnographic texts to be artefacts,
skilfully manufactured in order to construct their persuasive force. T
his paper offers a defence of ethnography. It identifies the strengths
and weaknesses of ethnographic data, and explores some of the procedu
res an ethnographer must adopt in order to give authority to the data.
Some of these procedures are applied to the highly controversial ques
tion of sectarianism in the RUC. However, this substantive topic is al
so useful for illustrating the limits of rules of method to adjudicate
those differences between ethnographers and readers of their texts wh
ich extend beyond technical matters.