Papers, cards, and perilous encounters with the state 23 to flee with just the clothes on his back, drinking water, and the iron pipe containing his document. At the time of the interview, Gulam and many of those around him were worrying about the risk of being forcibly returned to Myanmar again. He told us that he would use the same method again to keep hold of his document—leave it behind and then find a ‘sneaky way’ to retrieve it. The discussion then moved to the risks of repatriation, their strategies of resistance, and their demands for restoration of their citizenship as a pre-condition of return. For many Rohingya participants, then, treasured registration documents— such as NRCs and family lists—were also often imbued with an ethnographic significance relating to their group identity, the salience of which outranked their legal and administrative functions of providing access to rights, welfare, and/or forms of citizenship. Rohingya refugees in the camps in Bangladesh had collected together filed copies of outdated or expired identity documents in one community collection with the purpose of showing them to others as a historical record.5 Through these self-organised archival collections, it was possible, for example, to identify patterns relating to the dates and under which regimes the ethnic, religious, and nationality categorisations changed from Rohingya or Muslim or from citizen to foreigner (see Chapter 5). As Gulam’s account demonstrates, whilst such systems organise, categorise, and ‘other’ people through hegemonic means, such processes are also contested by the documented and the undocumented. Registration and identity documents do not only plot the state’s bureaucratic and administrative approaches to citizenship. They are also anthropological objects around which identities are organised, produced, reproduced, destroyed, resisted, contested, mediated, and negotiated (Navaro-Yashin 2007; McConnell 2009; Hull 2012; Reddy 2015; Chhotray and McConnell 2018). For Rohingya, being enumerated, registered, and re-categorised became a hugely important site of resistance, most visibly in the case of the national census in 2014 and the National Verification Cards (NVCs) as explored in Chapter 7. Thus, IDs and registration were points at which Rohingya encountered the violence of the state and so provided useful sites for understanding resistance and contestation. Narratives relating to IDs, other documents, and their absences, then, can be researched as evidence, as artefacts, and as anthropological objects. All these areas provide ways to explore events, experiences, and affect relating to the production of statelessness and identity destruction and reorganisation. Narrative research and oral histories In this section, I explain why narrative research and oral histories provide an effective research approach in critical statelessness and state crime research. Oral history research most often falls within a broader body of narrative research (Portelli 2006:35–36; Czarniawska 2010). Narratives, like oral histories, are accounts or descriptions of a series of connected events. The gaps between state narratives and the narratives of marginalised groups