Analyzing documents produced under extreme historical violence poses methodological, epistemological and ethical challenges that remain insufficiently theorized in qualitative inquiry. This article examines these challenges through an IPSE-inspired qualitative document analysis of 300 letters written by Jewish internees in the Drancy camp (France, 1941–1944), conducted as part of the LettresCamps project. Originally designed for long, co-constructed interviews, the IPSE approach was combined with Ethnographic Document Analysis to dwell with short, censored, pre-existing texts shaped by surveillance, coercion and the looming threat of deportation. The study produced three intertwined sets of results. First, an IPSE-inspired structure of experience was constructed, organized around three axes: telling the daily story of internment, expressing horror and suffering, reassuring and staying connected. Second, the analytic process itself became a result, as we encountered a series of methodological impossibilities: the unbearable nature of literal description, the tension between phenomenological and historical truth, the need to move toward a hybrid model combining descriptive, testimonial and dialogical/reception-oriented readings. Third, an unintended collective autoethnography emerged, documenting the fragmentation of the analytic we, the ethical asymmetry between the dead and the living and the mirrored fractures of subjectivity between the internees’ I and the researchers’ I. The study offers a phenomenology of what censored Shoah letters do to qualitative methods. We argue that working with archival correspondence produced under genocidal terror requires a situated, abductive, ethically grounded epistemology that acknowledges the limits of subjectivity, the impossibility of bracketing history and the necessity of reading such letters as dialogical gestures under terror. We conclude by proposing methodological principles arising from this specific corpus and by conceptualizing these tensions as a form of productive impossibility. This article does not propose a general theory of qualitative trauma research but offers a situated methodological reflection grounded in a singular corpus.
Mathé et al. (Mon,) studied this question.