`Dangerous fieldwork' re-examined: the question of researcher subject position

This article takes two examples of trying to collect fieldwork data in dangerous or difficult circumstances in Bali and uses them to explore some issues central to qualitative research. These issues include shifting researcher subject positions in qualitative sociology approaches, and the coherence and usefulness of data collected in chaotic or risky circumstances. Methodological practices such as reflexivity are considered, as well as the task of writing research accounts up from messy and chaotic data sets. It is concluded that data collected at moments of fieldwork crisis may not be particularly useful, except as a cultural reminder of the insider/outsider status of the researcher, and to inform more productive factual data collected after the event.