Over time, my silence, though became a great source of anxiety. I soon realised that my reactions in the field would be the lens through which the data would be generated (Woodthorpe, 2009) and through which my supervisors’ interpretation of border control and immigration detention in the country would come. Reflections by research assistants are rarely found in social science literature and the impact the research may have on their lives is hardly acknowledged (Carretta, 2015; Caretta and Cheptum - Area, 2017). As the interviews were coming to a close, I wondered what implications instances like the ones mentioned above would have in terms of me succeeding further in academia. Would I live up to my supervisors’ expectations? Was I being careful enough in the interview process? More importantly, if I did not speak up to defend detained women, why was I there? In raising these questions, I confronted my limitations as a research assistant and saw my own powerlessness to speak for those I interviewed.24 Interviewing women migrants inside the police control office, a theatre of state power (Hasselberg, 2016), was ethically challenging. Women seemed very suspicious of my presence at first. ‘I don’t want to be deported’ they would say to me, implying that either I was police or that I had the powers to get them out of there. Issues of informed consent arose here. As Christine Halse and Anne Honey reveal about their research with anorexic teenage girls in a clinical setting, what is written in ethics applications about the voluntarism in consent is often an ‘illusion of certainty’ that cannot always be guaranteed (2005; 2148). To what extent could detained women exercise their agency when asked to participate in a project by the police? Could they perceive the space where interviews took place as coercive? And more importantly, could they speak freely about their experiences in close proximity to the police? While all women gave their consent (either verbal or written) and, thus, satisfied the ethics committees’ requirements, this cannot be considered a panacea for the ethical difficulties of consent with vulnerable individuals (Halse and Honey, 2005). It seemed that throughout my time in detention, the most important thing was to clarify who I was not (Aitken, 2019). In response to their fears, I would explain my role and repeat several times that I was not a police officer. This, though, often seemed sharply disjointed from their fear of police and deportation. When police officers entered the room, I stopped the interview and recording, until they left. This meant that in some cases our conversations were interrupted multiple times. To my surprise, the police never actively 24 For more on my reflections upon the issue of representation in criminal justice research see Fili (2019). 52