ASYLUM INTERVIEWS IN THE UK 61 paradigm, the category of ‘ethnicity’ or ‘race’ is often maintained in an essentialised understanding when the judges overturn the rejection of the asylum decision. In the case discussed above, for example, the wording of the judge’s decision was as follows: ‘I am satisfied on the evidence before me that the appellant is Rohingya.’ However, I had not written anything about the man’s identity being Rohingya in any part of my expert report. Instead, I had argued that I considered it likely that the applicant will be categorised as Rohingya if sent to Bangladesh. With regard to this risk, the judge also found that ‘[t]here is a real risk that his true ethnicity will be exposed. If he is asked about his true ethnicity, he cannot be expected to lie about it. If he tells the authorities that he is Rohingya he is at real risk of forcible repatriation to Burma.’ Here the judge had included risk-related passages from my expert report, but added his understanding to it, namely ‘true ethnicity’ (as opposed to a merely claimed one?). As soon as judicial decisions are argued in writing, it seems, the anthropological knowledge consulted for the judgment ‘disappears’ behind the established legal categories. Whereas Bruno Latour has formulated in The Making of Law that ‘[j]ustice only writes law through winding paths . . . For her to speak justly, she must have hesitated’ (2010: 151–152), in writing their asylum decisions, judges need to demonstrate that they are ‘satisfied’ by the evidence they have seen. It is through drawing on or re-invoking essentialised notions of identity, ethnicity and culture that they demonstrate that knowledge has been secured. It remains the task of an applied legal anthropology to shed light on this practice and understand why this is the case. The position I have advocated for thus lies between the two groups I have outlined in the beginning: I acknowledge that there will likely always be limits to judges’ capacity to take anthropological expert knowledge into account, but I do insist that there is a positive way for anthropological experts to apply their knowledge without falling back to the ‘cultural defence’. Similarly, for example, Susan Bibler Coutin and Véronique Fortin write that ‘[p]racticing this craft with care is one way to counter the otherwise alienating and state-centric nature of bureaucratic inscription’ (2023: 24). There are also cases in which an anthropological concept of ‘culture’ was explicitly acknowledged by the judges (see Thuen 2004). However, my impression after ten years of working as a COI expert is that even when judges manage to take account of anthropological reasoning, the law rarely ‘liberates itself from some of the positivism’ as Trond Thuen could argue from his material (2004: 283). Rather, judges translate their newly acquired knowledge into new bureaucratic inscriptions. Conclusion It is necessary to understand the textual documentation of an asylum interview as an unequal encounter and yet a co-constructive practice. In this article I have set out to elaborate the context in which asylum interviews are conducted, highlighting the fact that all involved parties remain ‘other’ to one another as there is no shared knowledge they can draw on. This is, at the same time, the reason why interviewer and interviewee resort to co-constructing ‘evidence’, as my ethnomethodological interpretation has shown. As the emphasis rests on co-constructing a ‘case’, I have also argued