Papers, cards, and perilous encounters with the state 25 interactions, the impact of structural factors and the significance of social change or upheaval can be traced. Thereby, orality, narrative form, memory, subjectivities, and the relationship between the researcher and research participants all become resources from which meaning and interpretation can be drawn rather than weaknesses relating to credibility (Perks and Thomson 2006:4). This is not to say that facts do not matter; the large body of human rights documentation relating to Rohingya during this period enabled me to check narratives against other bodies of work to ascertain the relationship between facts and meaning. Narrative research often considers the relationship amongst past, present, and future (Popular Memory Group 2006:51; Squire, Andrews et al. 2014). This was a particular feature of Rohingya oral histories within this research, whereby accounts of past mass expulsions, forced repatriations, and identity card schemes were drawn on as resources through which to decipher their futures in Myanmar by weighing up the risks on both sides of the border in Bangladesh and Myanmar, and their current-day situations were reflected backwards (Brinham 2017). Thus, Rohingya group demands and social actions in the present could be traced through the ways in which their narratives interpreted the past. Third, narrative research is useful in exploring both individual and collective identities. Narrative is understood as providing order to personal and collective identities. Identity in this context is ‘a specific kind of narrative in which people tell themselves and others who they are, who they are not and who/how they would like to/should be’ (Yuval-Davis 2010:279). All identities are fluid, socially constructed, and are in a process in which they are being (re)constructed, (re)interpreted, and (re)negotiated. They are shaped by both the structures of power and individual subjectivities (Brubaker and Cooper 2000; Yuval-Davis 2006; Riessman 2008). Narratives and oral histories are tied up with the performance and negotiation of social identities and the creation of common platforms for identity (Riessman 2008; Squire, Andrews et al. 2014). They are also constrained by and resist larger and more dominant narrative patterns (Gready 2008; Squire, Andrews et al. 2014). Fluid individual identity narratives and subjectivities are produced within a social and political context that increases in significance as identities come under threat (Yuval-Davis 2010:266). Thus, applying narrative inquiry to identity destruction, formation, and reorganisation can tell us about both the state-led identity destruction projects and the collective resistance and action that are central to this research. The fourth useful approach from within narrative research relates to the notion of co-constructed narratives. This notion holds that all narratives are produced and shaped by the social context and by the listeners or audience which may or may not be present (Czarniawska 2004; Clandinin 2007; Riessman 2008; Yuval-Davis 2010). This co-construction takes place not only during the interview or focus group but also in the process of interpreting and retelling (Salmon and Riessman 2008). As such, the researcher does not find narratives that represent external realities but in the process of