ASYLUM INTERVIEWS IN THE UK 57 The interviewer decided to write down the translator’s partly broken and abbreviated English verbatim. This suggests authenticity, as if one were ‘close’ to what was actually said. However, it is the words of the translator that are reproduced, not those of the interviewee. Regarding the content of the questions asked in this sequence, it is noteworthy that the interviewer asks three times if the applicant had made an effort to find out ‘more’ information about ‘the country [he was] born in’. But when the applicant talks about what he has heard (Muslims being burnt alive), the interviewer does not follow up on this point – he or she either lacks knowledge of the cultural setting to be able to contextualise the applicant’s statements or lacks interest in engaging with the actual statements provided. Arguably, it is also not their job to do so. They are only supposed to generate data for others to work with. Their duty, in other words, is to help create a case, not to understand. Fittingly, the next question (on possession of a passport) has no connection with the previous sequence of questions and had been answered in the negative already in the screening interview. In the context of cross-examination in American courtrooms, Paul Drew has argued that the production of contrasting accounts ‘aims to provide the jurors with the material from which they can work out for themselves what to make of the facts’ (1992: 516). In asylum interviews, however, caseworkers do not have a conflicting version of an account to offer or to elicit from an applicant, nor do they have the means to scrutinise the applicant’s descriptions. The asylum seeker and the interviewer therefore share what in ethnomethodology is called an ‘interactional dilemma’ (see Garfinkel 1967, 2002; Solberg 2011), and they solve it by trying to avoid face-threatening acts. For example, the applicant does not answer ‘I don’t know’ or ‘I don’t understand you’, but tries to accommodate the interviewer’s repeated demands for more information by saying that he will ‘find out’. In this way, both contribute to the creation of accountability in their interaction, regardless of the fact that their positions of power and their interests are unevenly distributed. The alienation process from the actual story of a human life that the inherently uncertain knowledge undergoes in the course of such interviews due to the different participants, their respective knowledge horizons and language barriers, and the formality of their interaction is also clear from the sequence reproduced above. Strictly speaking, documents created in this way are not biographical accounts. The interview transcripts rather showcase ‘hint-and-guess sequences’ (see Holstein and Gubrium 1997) in which ‘suggestions’ are made by the caseworkers regarding the asylum seeker’s origins, language, culture and flight history that are to be taken up by the applicant. In their ‘suggestions’, the caseworkers refer, on the one hand, to the information available in the COI databases and, on the other hand, to the expertise they have acquired in the course of their professional activities (see Kolanoski 2018). Such ‘suggestion’ prompts are either accepted or rejected by the asylum seekers, although the rejection of ‘suggestions’ is much rarer, as it puts the rejecting person in the position of a ‘repair initiator’, in which they would have to make a counter-suggestion in order to keep the interaction going (Schegloff et al 1977). In such interactions, the dialogue partners therefore attempt to constantly update their accounts. What needs to be scrutinised is whether they work towards a meaningful social outcome even when their interests are potentially opposed and a common understanding cannot be assumed.