Interpretation, Translation, and Confusion in Refugee Status Determination Procedures

Abstract

This article examines the impact on legal processes of the need to use interpreters, drawing examples from refugee status determination procedures in the United Kingdom. It describes the roles played by interpreters in facilitating intercultural communication between asylum applicants and the administrative and legal actors responsible for assessing or defending their claims at the various stages of those procedures. The UK authorities' somewhat naïve expectations about the nature of the interpretation process display little understanding of the practical dilemmas that interpreters face. Much of the confusion and many of the barriers to communication created by the involvement of interpreters reflect the inherent untranslatability of particular notions, and so arise irrespective of the technical competence of the interpreters themselves. For example, dates may be reckoned using non-Gregorian calendars; terminologies for family relationships and parts of the body may be incongruent between the two languages; and there may be no exact indigenous legal equivalents to UK notions such as 'detention' or 'rape'. Different interpreters may therefore give different, though equally legitimate, translations of such terms, creating apparent 'inconsistencies' in the resulting translated accounts. Given the centrality of notions of credibility in asylum decision-making, even quite trivial divergencies over such matters may prove crucial.

1 Introduction
2 The Refugee Status Determination Process in the United Kingdom
3 The Roles of the Interpreters
4 Institutional Expectations and Codes of Conduct
5 Fragmented Narratives, Interpreting Dilemmas, and Barriers to Communication
6 Untranslatability
7 Complex Problems of Translation
8 Conclusions
Notes
References

1 Introduction

Article 1A(2) of the 1951 Refugee Convention defines a 'refugee' as someone who has a 'well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion' [42: para. 34]. Deciding whether a person satisfies this definition involves complex processes of cultural and linguistic translation. This article examines the problems posed by the use of interpreters to facilitate communication between asylum applicants and the civil servants, lawyers and immigration judges who assess, present or decide their claims. Similar issues are likely to arise in all legal proceedings involving interpreters, but they are especially crucial here because assessments of credibility are so central. ^[1]

The UNHCR Handbook says that an asylum applicant 'should be given … the services of a competent interpreter, for submitting [their] case to the authorities' [42: para. 192(iv)]. Similarly, Art. 15.3(c) of the recast EU 'Procedures Directive' [13] requires member states to:

select an interpreter who is able to ensure appropriate communication between the applicant and the person who conducts the interview. The communication shall take place in the language preferred by the applicant unless there is another language ... in which he or she is able to communicate clearly. Wherever possible, [they] shall provide an interpreter of the same sex if the applicant so requests. ^[2]

But how is the competence of interpreters defined, and what constitutes 'appropriate communication'? This article summarises refugee status determination procedures in the United Kingdom, focusing on the main contexts in which interpreters operate and the restrictions placed on their behaviour by administrative and legal expectations. It then considers some of the practical dilemmas they face and the barriers to communication created by their involvement, many of which arise irrespective of their technical competence.

2 The Refugee Status Determination Process in the United Kingdom

Asylum claims in the UK are administered by UK Visas and Immigration (UKVI), a branch of the Home Office. ^[3] UKVI holds initial screening interviews with applicants to collect basic personal information. A more detailed substantive interview then takes place, usually a few months or even years later. This aims to establish the chronology of the applicant's narrative and test its credibility. If the asylum claim is then refused, the case officer writes a Reasons for Refusal Letter (RFRL) explaining this decision. Most RFRLs claim that the applicant's story lacks credibility because of alleged inconsistencies, or because it conflicts with available country of origin information (COI). ^[4] Credibility decisions are particularly important here because there is often no direct evidence other than the applicant's personal narrative.

Most refusals can be appealed before an Immigration Judge (IJ) in the Immigration and Asylum Chamber of the First-tier Tribunal. A solicitor assembles the documents for the appeal and helps draft the appellant's own witness statement. In Scotland the solicitor usually participates in the hearing too, but in England the advocate who actually represents the appellant in court is normally a barrister. UKVI is represented by a Home Office Presenting Officer (HOPO), who is a civil servant rather than a trained lawyer.

Typically, hearings begin with a short 'examination-in-chief': the appellant is simply asked by their barrister to confirm that the contents of their asylum interview and witness statement are true and that they wish to submit them as evidence. The HOPO then cross-examines the appellant, seeking largely to uncover inconsistencies in their account. This is the longest part of the hearing, typically lasting one or two hours. The barrister may then re-examine their client, and occasionally other witnesses are called. Closing submissions by the HOPO generally involve attacks on the credibility of the appellant's narrative and references to the Home Office's own sources of COI. The barrister then seeks to rebut these credibility points, offers alternative analyses of the COI (usually drawn from a wider range of sources), and makes other arguments in support of their client's case. The IJ almost invariably reserves their decision, and subsequently produces a written determination giving the outcome of the appeal, the reasons for their decision, and an indication of how much weight was given to each piece of evidence.

Subsequent appeals to higher tribunals or courts are limited to matters of law, although the boundary between law and fact is often hazy in asylum claims. There is mostly no fresh evidence at these later hearings, apart perhaps from updates to the COI, so appellants themselves rarely attend and no provision is normally made for interpretation. A detailed discussion of this entire process as things stood on 1 May 2023, describing best practice from the legal representative's perspective, has been produced by Mark Henderson et al. under the auspices of the Immigration Law Practitioners Association (ILPA) [25].

3 The Roles of the Interpreters

In most situations a distinction is conventionally drawn between 'interpretation', as 'the oral transfer of meaning between languages'; and 'translation', as the equivalent process for written text. In legal contexts, however, as Colin and Morris point out [10: 16], the term 'interpretation' refers primarily to the process of determining the 'true meaning' of a legal text, so lawyers often describe what interpreters do in court as 'translation' rather than 'interpretation'.

At both asylum interviews interpreters are present if requested; they are hired by UKVI on a freelance basis. They translate the questions and answers, but are no longer routinely required to read the transcript back to the applicant at the end. These two kinds of interview are described in detail, from an interpreter's perspective, by Rycroft [35: 227–232]. In brief, the interpreter is quite an active participant in the screening interview, involved in clarifying answers or circumscribing them so that they fit into small boxes on the screening form, whereas their role is far more constrained at the substantive interview:

It is acceptable to ask the speaker to repeat, to ask for permission to clarify an answer or question, and it is acceptable to step in if there are misunderstandings. Other than that the interpreter must not intervene, even if they suspect that the meaning of the question or answer has escaped the other party. [35: 231]

It occasionally happens that the applicant's solicitor also attends this substantive interview, and they may even bring their own interpreter with them. Both of them are restricted to making comments at the end, however; neither is permitted to intervene during the actual interview, even if there seem to be problems associated with the interpreting. Further barriers to their participation are the cost, especially in the context of progressive cuts in legal aid funding, and various arbitrary obstacles created by the Home Office: for example, if the solicitor and interpreter do not arrive at the same time as the applicant they may not be allowed into the building.

The taking of the applicant's witness statement by their solicitor may occur at any stage between the initial asylum application and (if the claim is refused) the ensuing appeal hearing. Solicitors differ over the best time to do this, and when best to submit that statement to the Home Office or the court. There is surprisingly little discussion in the legal literature of the process of taking witness statements, apart from the particular instance of statements taken by police in criminal cases [24], and many solicitors told me that they had received little if any training. In any case, statement taking usually requires the aid of an interpreter. Most such interpreters are freelancers, but a few immigration law firms have interpreters on their payroll if large numbers of their clients speak particular languages such as Somali or Tamil. Elsewhere I have discussed the statement-taking process in detail [19: 100–102, 106–112; 20], showing how immigration lawyers employ far more open-ended and flexible questioning strategies than Home Office case-owners. ^[5] The interpreter is granted more freedom too, to clarify or explain what the client is intending to say, or to supply missing context.

For many years, accepted practice for validating witness statements in asylum cases was that set out in Njehia ^[6] :

The proper procedure … is for a competent interpreter, in the correct language and dialect, to read back the statement and for the maker of the statement then to sign it, confirming that the document has been read back in his or her own language. The interpreter should then append ... his own short statement that he has read back the contents of the document to the maker of the statement ... That should then be signed and dated by the interpreter.

However, Henderson et al. reported an apparent change in these requirements regarding witness statements [25: paras 12.67–12.69]. Appended to the Civil Procedure Rules (CPR) which govern civil law cases are several Practice Directions which specify in detail how these Rules are to be applied. Practice Direction 32, which covers the general topic of evidence, states that witness statements 'must, if practicable, be in the intended witness's own words and must in any event be drafted in their own language' [29: para 18.1]. Specific courts and tribunals also produce their own Practice Directions, and the updated Practice Direction of the Immigration and Asylum Chamber of the First-tier Tribunal [36], which came into effect on 13 May 2022, largely mirrored para 18.1 of the CPR Practice Direction. It stated:

5.5 The witness statement must, if practicable, be in the intended witness's own words and must in any event be drafted in a language they understand. …

5.12 Where a witness statement is in a language other than English—

(a) the party wishing to rely on it must—

(i) have it translated; and

(ii) file the translation and the foreign language witness statement with the tribunal; and

(b) the translator must sign the original statement and must certify that the translation is accurate.

The substitution of the phrase 'a language they understand' for the CPR's 'their own language' did suggest a greater degree of flexibility in the case of asylum hearings, but it was unclear to practitioners what changes, if any, were required as compared to previous practice. In March 2023, ILPA sought clarification from the Tribunal as to whether the words 'drafted in a language they understand' in para 5.5 were now intended to apply to the statement itself rather than, as hitherto, to the instructions forming the basis of that statement [27]. Such a change would raise several problems, such as the high financial and time costs associated with the translation required by para 5.12(a)(i); possible unfairness to applicants who were not literate or not legally represented; and the scarcity of interpreters with the legal training needed to take and draft witness statements. Henderson et al. suggested that the Practice Direction was likely to be amended as a result [25: para 12.69], and this did indeed happen in a subsequent Practice Direction [37] coming into effect on 1 November 2024, which more-or-less reverted to the procedure set out in Njehia. Thus, its paragraph 8.5 states:

The witness statement must, if practicable, be in the intended witness's own words. The statement need not be in a language that the witness understands. If drafted in English and this is a language not understood by the witness, it must include a signed and dated attestation by both the witness and the person who interpreted it that the statement has been read back to the witness in a language they understand and that it accurately reflects their evidence.

At appeal hearings, interpreters are provided via the Court Service. The training and terms of employment of such interpreters were discussed by Gibb and Good [16], and the current situation is described by Henderson et al. [25: paras 34.1–34.3]. There are more stringent qualification requirements for interpreters of the 41 standard languages listed in the Ministry of Justice's Guide to Language Interpreter and Translation Services [30], as opposed to other languages arising less commonly.

At asylum hearings, court interpreters must provide consecutive interpretation of questions and answers throughout the proceedings and should also provide appellants with simultaneous interpretation of dialogue involving other participants, including the closing submissions. Simultaneous interpretation does not always happen, however, and even when interpreters try to do this, judges sometimes ask them to stop because they find it distracting. It is in any case extremely demanding for interpreters to have to switch to and fro constantly between consecutive and simultaneous modes, over periods of several hours without breaks. To add to the difficulty, court interpreters do not have access to the substantial documentation involved in asylum appeals, so they can easily be left guessing by spoken uses of particular names or technical terms and may not fully understand vague or allusive comments referring to those documents. Moreover, they are not legal experts, yet they are expected to translate British legal terms into equivalent terms in the target language, and vice versa. Not surprisingly, this may cause problems, of which Henderson et al. give several examples:

There may be different interpretations of the word 'detained' in the appellant's language depending upon whether one is referring to arrest for questioning and release on the same day, detention overnight in a police station, or detention for a longer period in a prison. Arrest by a militarised police force may be translated into English by one interpreter as arrest by the police and by another interpreter as arrest by the army. The HOPO will then allege a discrepancy. [25: para 34.6]

If anything this understates the risk, because the asylum procedures described above always involve not just one, but a whole series of interpreters. As a minimum, there will be different interpreters for the screening interview, the substantive interview, and the court hearing. There are also highly likely to be further interpreters involved in the taking of witness statements and at medical examinations, so it will frequently be the case that five or more interpreters have contributed to the presentation of the written and oral evidence before the court. Moreover, such problems are not confined to translations of legal terminology, as the examples below illustrate. As one judge told Robert Thomas:

it just stands to reason that if you get two, three, or four interpreters interpreting the same story at different stages of the process, then they are going to use different words and the different interpreters may make different sense of what the appellant is saying. [41: 144; see also 25: para 34.6]

Despite all these potential difficulties, it is quite unusual for the accuracy of interpretation to be challenged during a court hearing, not least because there is rarely an independent observer present with the necessary linguistic qualifications. Complaints raised after the actual hearing generally get short shrift. ^[7]

Finally, while the non-involvement of appellants and interpreters in appeals to higher courts is understandable from a strictly procedural perspective, one consequence may be that any fluidities or ambiguities arising through the vagaries of the interpretation process at the earlier hearing become congealed, reified, or hidden when those oral exchanges appear in abbreviated written form, as documentary evidence in a debate conducted and assessed solely by lawyers and judges.

4 Institutional Expectations and Codes of Conduct

Both UKVI and the Tribunals Service have codes of conduct for interpreters. The UKVI code is almost entirely concerned with administrative matters, but the actual process of interpretation is dealt with briefly. Key points are as follows:

• you must retain every element of information that was contained in the original message and interpret in as close to verbatim as English allows.

• you must not try to anticipate what the interviewer or claimant/customer is trying to say or give an answer different from the one being provided ….

• you must interpret language which may be offensive—for cultural reasons, obscenities may be difficult to translate or have no direct equivalent—in this case you must look for the closest equivalent and inform the Interviewing Officer that is the case. …

• you must not ask the claimant/customer what they mean by a particular answer—you must ask the Interviewing Officer's permission to ask the claimant/customer to repeat or clarify. [26] ^[8]

Similarly, the Tribunal Service's Handbook for Freelance Interpreters states:

Please do use the witness's exact words. If you cannot make a direct or exact interpretation, interpret it as accurately as possible in the witness's own words and then inform the Judiciary what the phrase means. ... Please do not ... use an English expression or phrase which is not an exact translation of the witness's own words. (Quoted in Henderson et al. [25: para. 34.24])

As these examples illustrate, law courts generally instruct interpreters to provide 'verbatim' or 'exact' translations. This seems to assume that each word in language A has one, and only one, 'exact' equivalent in language B, even though, as Colin and Morris have pointed out, 'word-for-word … translation often produces distorted communication' [4: 17], because words depend for their meaning on how they are combined with other words. ^[9]

Whereas the Ministry of Justice has guidance for criminal court staff on working with interpreters (see also [1]), the Civil Procedure Rules and the Tribunal's own Practice Directions [36, 37] do not discuss how to do this. By contrast, there is very full guidance for immigration lawyers on working with court interpreters in ILPA's Best Practice Guide. Unlike those other documents, this Guide displays awareness of the limitations of even the most competent interpretation [25: paras. 34.4–34.8]; see also Thomas [41: 144].

The Tribunal's Guidance Notes [11] do at least provide a script for IJs to follow when establishing that appellant and interpreter do understand one another, but it is a very basic check of mutual comprehension in which the appellant is asked to tell the interpreter how they travelled to the court that morning. What's more, it is put right at the end of this script, after the judge has explained the structure of the hearing and the role of the interpreter, all of which has to be translated even though it is not yet clear that the appellant understands it. Moreover, as Henderson et al. drily comment:

The conversation [is] at a level of sophistication that could be handled in French by the average British school pupil. Having established this level of understanding, the interpreter is then expected to be capable of translating the witness' explanation of Marxist ideology. Not surprisingly, it does not always follow. [25: para 34.23]

Interpreters themselves may be reluctant to acknowledge communication difficulties even when there are legitimate reasons for this (for example, between Kosovans and Albanians, or Arabic speakers from Morocco and Egypt), because doing so may seem to reflect adversely on their competence. Things may become further complicated by the fact that some languages commonly arising in asylum claims, such as Arabic and Tamil, are cited as classic examples of diglossia in Ferguson's definitive study of the phenomenon [15]. This may mean, for example, that different dialects are spoken within the community and when conversing with outsiders; or that written and formal styles are markedly different in syntax, phonology, and morphology from those used in everyday conversation, which may themselves display almost infinite variations between different communities or localities [15: 325, 337].

Differences in syntax pose general problems facing all interpreters to some degree, but their precise forms vary from language to language. For example, whereas the normal word order in English is subject-verb-object, in Tamil narratives the subject typically begins the sentence (though it may sometimes be omitted altogether in favour of an appropriate verb ending), followed by strings of subordinate clauses ending with adverbial participles, with the main verb coming right at the end. In A progressive grammar of the Tamil language, Arden [2: 200–201] gives a rather quaint and old-fashioned example of a Tamil sentence, but it is an appropriate one in that the words are attributed to a judge. When translated word by word (verbatim!) into English by me, and slightly abbreviated, it reads: 'A woman, a chicken having stolen, having eaten, in this assembly having stood, "I don't know about it" having said, went away.' In instances like this, an interpreter must not only attend to the meaning but also drastically rearrange the word order, both of which tasks are made more difficult, one imagines, because the dénouement, in the form of the main verb, does not appear until the very end.

5 Fragmented Narratives, Interpreting Dilemmas, and Barriers to Communication

At all stages of the asylum process the interposition of interpreters creates barriers to communication, irrespective of their competence. For example, all the various codes and guidance manuals stress the need for asylum applicants to answer questions in short sentences, so that they can be fully translated. It is important, of course, that everything the applicant says is communicated to the court, but this fragmentation introduces limitations of its own, as interpreters themselves are well aware. They also know that questioners can easily turn these limitations to their advantage, especially when one considers that everyone else at the Home Office and in court is well-versed in this process, whereas it is entirely new for the asylum applicant. ^[10] As one interpreter told me:

[we are] always made the excuse for that; you know, that you need to give short answers so that the interpreter can translate? That is another constraint brought on by the interpretation process itself ... and it works in the favour of the Home Office because people do not speak like that naturally, and they will lose track; they will say less than they mean to say simply because they have to break it down. ^[11]

The impact of this varies according to context. At substantive Home Office interviews, long answers are often interrupted by the interviewer, with the result—as Rycroft points out—that the applicant may forget where they had got to in their answer, or not realise that some of what they have already said has not been written down, especially when the interviewer does not engage with their answer but simply moves on to the next question on their checklist [35: 230–231]. By contrast, the process of drafting the witness statement usually entails further exploration of initial answers and is more flexible regarding conversational turn-taking, so such inadvertent omissions are far less likely.

The various guidelines also specify that interpreters should, as the Tribunal Handbook puts it, 'endeavour to reflect the type of language that is being used, whether it is simple, formal, colloquial etc.' [quoted in 25: para. 34.24]. As one interpreter told me, they find themselves in a quandary when appellants speak ungrammatically, rudely, or colloquially:

That's a dilemma. I find myself actually brushing up [their] speech because I certainly don't want to come across as a bad interpreter and it can be believed that it is me who is making the grammatical errors … and then I thought, hang on, these people are going to go to appeal and the way they speak is going to come across so disjointed from how they came across in the asylum interview. What's that going to do to their credibility?

This is an example of the use of 'face-saving strategies', as discussed by Pöllabauer in her analysis of interpreters' behaviour in Austrian asylum procedures [33]. Her notion of 'face' derives ultimately from Goffman [17], as adapted by Brown and Levinson [7] in their notion of 'politeness theory'. For them, 'face'—the public self-image that people wish to claim for themselves—has two aspects: 'negative face' is a claim to autonomy, that one's actions be unimpeded by others; while 'positive face' expresses the personality that one seeks to present to the world and for which one desires the approval of others [7: 62]. Both need constant maintenance work, but Brown and Levinson posit that it is generally in everyone's interest to acknowledge, at least to some degree, the face of others with whom one interacts.

There are, however, circumstances, perhaps involving social breakdowns such as violent arguments, or interrogations where the aim is to acquire information as rapidly as possible, where face may in practice largely be ignored. In contexts where actors communicate through the medium of an interpreter—especially if those contexts also involve extreme power imbalances and questioning that goes only in one direction—one might expect face maintenance to become more than usually problematic, particularly for the interpreter who largely controls face-saving in practice. Not surprisingly then, Pöllabauer found that interpreters 'feel obliged to save their own face and mark the source of misunderstandings or threats to their own and/or other participants' face, in order to avoid any blame for communication breakdown' [33: 39–40]. Looking at the comment quoted above from this perspective, for instance, the interpreter can be seen as describing the conflict between maintaining their own face by not seeming to be bad at their job, and maintaining the credibility (and hence, ultimately, the face) of the applicant by consistently and accurately expressing the applicant's ungrammatical mode of speech.

Pöllabauer gives similar examples in her study of interviews between German-speaking Austrian officials and English-speaking asylum applicants, and also shows how interpreters sometimes modify the wording of questions by the examining official, that in their original form might pose a significant threat to the applicant's face. In one instance, the official had by chance discovered a sheet of paper discarded by the applicant, on which his personal data and history of persecution were noted down in abbreviated form, and the official suspected that this was an aid to learning an invented story. The conversation then ran as follows:

Official: Well, I get the impression that his data here obviously… or that this paper obviously serves learning purposes.

Interpreter: The officer has the impression that this paper serves you as a sort of help with particulars to study, this is the officer's impression, the paper contains information, and you need the paper so to learn it.

Appellant: No, I throw it away. I wanted to write it so that if I come here I will submit it. [33: 45] ^[12]

Here, instead of translating the official's comment in the first person, which would be correct practice, the interpreter seems to distance herself from the accusation by rendering 'I get the impression' as 'The officer has the impression'; she also softens the accusation that 'this paper obviously serves learning purposes', by way of a lengthy circumlocution.

Very occasionally the reverse happens, and the officer's face is threatened by a comment by the applicant; again the interpreter softens this. Here, the applicant had already explained that he did not know the name of a particular river. Later in the interview the following exchange occurred:

Official: Okay, and this big river, what is it called?

Appellant: The name? I tell you I don't know the name of the river!

Interpreter: I don't know the name. [33: 46]

Differences in the speech registers employed by the various actors in appeal hearings pose further challenges for interpreters. As Rycroft notes, 'the interpreter will interpret what is said by four different parties and must maintain consistency with the tone and demeanour of each one … like an actor playing several roles' [35: 233–234]. All this highlights Berk-Seligson's key point that, despite the institutional expectation that an interpreter will be invisible, she is in practice 'an active verbal participant' [6: 64].

As we saw, the UK codes of conduct stress that the interpreter's task is to provide 'verbatim' or 'exact translation'. They must not, as the UKVI Code of Conduct puts it in the passage quoted above, 'try to anticipate what the interviewer or claimant/customer is trying to say [or] ask the claimant/customer what they mean by a particular answer'. An earlier version added that interpreters should not 'let your own experience or views get in the way of how you interpret the evidence'. This assumes that the interpreter's own understanding will 'get in the way' of the desired 'verbatim' translation. On the other hand, the UK Tribunal's Handbook for Freelance Interpreters provided potentially contradictory guidance:

You may intervene at the hearing for the following reasons: to seek clarification …; to alert the Judiciary that … the question or statement may not have been understood; to alert the court to a possible missed cultural inference—such as when an item of information has not been stated but knowledge of which has been assumed. (Quoted in Henderson et al. [25: para. 34.24]; my italics)

As Morris points out, interpreters commonly feel that the constraints placed on their role in legal contexts limit their ability to facilitate communication [31: 26]. Building on Morris's work, Wadensjö argues that legal systems scarcely acknowledge the differences between monolingual situations and those mediated by interpreters. Instead of recognising that their role is 'truly interpretive', the interpreter is 'defined as a disembodied mechanical device' [43: 74]. This 'mechanisation' may be quite explicit. For example, when my co-researcher Robert Gibb interviewed a senior official in the French asylum bureaucracy, the officer gestured to his desk and likened the interpreter's role to that of his telephone.

Naivety about the interpretation process is not confined to legal contexts of course, but its stress here seems partly to reflect the centrality of language in legal procedures. As performed in court, law depends heavily on the skilled manipulation of language by lawyers and its untrained use by those under cross-examination. As Atkinson and Drew pointed out in their study of courtroom dialogue, cross-examination seeks 'to challenge or blame the witness' by getting them to agree to the 'facts' progressively brought out during the questioning [3: 105–106]. This becomes far more difficult when lawyers have to work through interpreters, and legal efforts to 'mechanise' the interpreter's role serve as attempts to maintain the hegemony of the cross-examining lawyer. As Wadensjö points out:

it would obviously be a challenge to the court if interpreters were ... allowed to clarify an attorney's deliberately ambiguous question. It would be a threat to the system if interpreters were allowed to improve the image of witnesses ... by rendering eloquently and precisely statements which were originally voiced carelessly and imprecisely. [43: 75]

Clearly then, interpreters must constantly use their judgement over whether to intervene and explain that a misunderstanding or missed inference may have occurred. Not surprisingly, they differ greatly in how often they do this. This reflects not only their personal shyness or ebullience, but also their knowledge that if they intervene too often, it is likely to be attributed to incompetence rather than alertness and sensitivity.

Missed social and cultural inferences are rendered more likely by what Rycroft terms the 'silent actors' in asylum hearings, those elements of refugee law that motivate particular questions but of which the appellant is not aware. For instance, to qualify as refugees they must first have exhausted all avenues of domestic protection, so they are often asked whether they had sought to relocate within their own country. Not knowing the significance of this, they simply say that was not possible, whereas awareness of the law might have prompted a fuller explanation. Rycroft [35: 241] gives the example of asylum applicants from Romania (prior to EU enlargement) who fail to explain that moving elsewhere in the country would have required a residence visa, issued by the very police from whom the person was fleeing. As an experienced legal interpreter who is herself Romanian, Rycroft is well aware of both the legal reasons for posing the question and the relevant information missing from the applicant's answer. But should she interpose her own explanation? Some case-owners and judges will accept such behaviour, but they are more likely to rebuke the interpreter for giving evidence as if she were herself a witness.

Despite all this, most interpreters do not wish to go beyond their primary role of facilitating communication. Fenton's survey of court interpreters in New Zealand found that they all 'understood their role to be strictly that of interpreting and were adamant about keeping it this way' [14]. They rejected Barsky's plea for interpreters to be 'active intermediaries between the claimant and the adjudicating body' [4: 46], who could compensate for applicants' unfamiliarity with the law by actively eliciting more information. The dilemma, from an interpreter's perspective, is how to achieve high standards of communication without appearing to lapse into the role of advocate.

Elsewhere I have given a long example involving an exceptionally proactive Farsi interpreter, showing how his own interpolations frequently helped the court by clarifying the appellant's answers, but also sometimes disrupted it, to the extent of 'correcting' the lawyer's questions and even—on one occasion—subverting the normal order of proceedings [18: 167–169, 177–178]. That interpreter did certainly prevent potential confusion over dates, the status of the Iranian gendarmerie, and so on, yet one wondered whether he was doing more than just translate. Because he interfered so openly with both the form and the content of questions and answers, his potential for influencing the outcome became very clear. Even when their behaviour makes it less obvious, interpreters may exert powerful and usually unknowable influences on proceedings, and perhaps on the decisions reached.

6 Untranslatability

There are certain usages, such as dates in non-Western calendars, or kin terms where kinship itself is structured differently, that are inherently impossible to translate 'exactly' or 'verbatim'. Different interpreters may offer different but equally valid English equivalents, thereby creating, purely as an artefact of interpretation, the impression that the appellant is confused, inconsistent, and possibly untruthful. Linguists distinguish between linguistic untranslatability 'when there is no lexical or syntactical equivalent' in the two languages [5: 97]; and cultural untranslatability, where there seem to be equivalent words in the two languages but their cultural significance is quite different [8]. Both situations seem to arise in the examples discussed below. ^[13]

Confusion and ambiguity are common in asylum cases, as Kalin [28] and Spotti [39], among others, have pointed out. To illustrate how easily potentially damaging confusion may arise, consider personal names transliterated from languages with non-Latin scripts. I witnessed one Tamil appeal where the appellant's name, செல்வன், had been transliterated as 'Selvan' in some documents and transcribed phonetically as 'Chelvan' in others. To complicate matters, there were similar issues with some of his other documents translated from Sinhala, which the Tamil interpreter in court was not officially qualified to read. The HOPO tried to use such differences to cast doubt on the appellant's identity.

However, the commonest inconsistencies concern dates. Any discrepancies between dates given by appellants at interview, in their statement, in cross-examination, and so on, will be seized on by the HOPO as damaging the credibility of their account as a whole, even though, as has often been pointed out, it is highly unrealistic to expect people to remember the exact dates of events occurring years ago [9]. One further complication here is that some applicants come from cultures that do not follow the Gregorian calendar. Consider the following exchange during cross-examination of a Tamil appellant:

HOPO: When did you arrive in Colombo?

Appellant: 21st September 99.

HOPO: In your interview, page A4, you say you went to Colombo in August 99?

Appellant: I went in August.

HOPO: Remember the date in August?

Appellant: 21st.

HOPO: So, 21st August 99?

Appellant: I went on 21st September 99.

HOPO: You've just said you went in August?

Appellant: I went in September.

HOPO: So can you explain why in your interview you said you went in August?

Appellant: Must have forgotten.

This confusion can be understood in light of differences between the Gregorian and Tamil luni-solar calendars. According to the latter, the month of Avani begins in mid-August and ends in mid-September, though the exact Western dates differ each year. If one interpreter renders Avani as 'August' at the asylum interview, and another interpreter translates it as 'September' during the court hearing, then the HOPO can triumphantly claim to have uncovered a 'discrepancy' even though the appellant may have been consistent throughout. Under such circumstances, and being unaware of the differing English translations and their consequences, applicants must find cross-examinations of this kind extremely confusing.

Such problems were never in my experience raised as mitigating factors at asylum hearings. Even those British barristers with wide experience of Tamil appeals seemed unaware that Tamils had their own calendrical system. This is even more astonishing in view of the fact that many appellants have Tamil-speaking solicitors. In this case, unusually, the Tamil solicitor was actually present, but did not seem to register the problem. My hypothesis is that because most Tamil solicitors will have spoken both English and Tamil from birth, and constantly switch between the two in their everyday speech, they do not appreciate intuitively why the confusion has arisen. Indeed, one Tamil solicitor even asserted to me that people in Jaffna never used the traditional calendar, even though Tamil dates appear on the mast-heads of Tamil-language newspapers like Virakesari, and I had frequently heard them being used in appeal hearings. The judge here was open to the idea that there had been a misunderstanding, but no one was ready or able to explain it to him. ^[14]

The one such example that many representatives and judges are aware of is that Iranians do not use the Gregorian calendar, and it has become widely recognised that this may cause problems. Many evidential bundles in Iranian cases include a conversion table for the two calendars, originally prepared by an immigration adviser who was himself Iranian. In my experience, however, even this proves no guarantee that confusion will be avoided. ^[15]

Variations in kin relationship terminology also create problems. I most often became aware of such misunderstandings in Tamil cases where I could follow the dialogue. In one instance, the HOPO was asking about the appellant's brothers, because his previous answers at interview had led to the Refusal Letter casting doubt on the appellant's credibility:

HOPO: Your elder brother is not involved with the LTTE because he has a family?

Appellant: Yes.

HOPO: But on pages A5 and 6 you say your elder brother was a member of the LTTE; why do you say that?

Appellant: I meant my cousin-brother. He is my mother's younger sister's son, who grew up with us and is considered to be like my brother.

HOPO: Why not mention him among your family members?

Appellant: I was asked about my own brothers and sisters.

English-speaking Tamils use the term 'cousin-brother' for their male parallel cousins (father's brothers' sons, mother's sisters' sons, etc.). The appellant's barrister in this case, an elderly, highly Westernised Tamil, did not intervene to sort out this confusion, presumably because it was self-evident to him what his client meant. Fortunately, the judge was equally aware and commented to me afterwards that the HOPO was relatively new and must not yet have come across this 'cousin-brother' idiom.

This by no means exhausts the problems potentially posed by the structural differences between Tamil and English relationship terminologies, however. Cross-examination of another Tamil man went as follows:

HOPO: You have one brother in the Tigers?

Appellant: Two brothers.

HOPO: What are their names?

Appellant: K----- and R-----.

HOPO: K----- joined the Tigers in 199-, when did your second brother join?

Appellant: 199-.

HOPO: Why did you not mention in interview that R----- joined?

Appellant: I did say.

The HOPO emphasised this 'discrepancy' in his closing submissions, and it seemed that no one in court was aware of the likely reason for it. Rather than a composite overall term for 'brother', Tamil uses two terms: annan ('elder brother', 'senior parallel cousin'); and tampi ('younger brother', 'junior parallel cousin'). At the hearing the interpreter translated the HOPO's initial question above by asking the appellant whether he had 'an annan or a tampi' in the LTTE. The appellant replied 'one annan and one tampi', which the interpreter translated as 'two brothers'. It is easy to imagine how the interpreter at the asylum interview might have made different translation choices. In a context where everything hangs on the appellant's credibility, even such a trivial 'discrepancy' may prove decisive.

Similar 'discrepancies' regarding parts of the body may be crucial when appellant's statements about how they were tortured are compared with the placing of scars on their bodies. I was involved in a Tamil appeal where an earlier judge had mentioned an ambiguity during cross-examination regarding the location of bayonet wounds on the appellant's lower limbs. Yet again, however, this 'discrepancy' could have been merely an artefact of perfectly valid translation choices made by different interpreters on separate occasions. In everyday speech the basic Tamil word for foot and leg is the same (kal). If more precision is needed, composite terms are used: ankle is kanukkal, heel is kutikal, knee is mulankal, and so on. Again there is clearly considerable scope for confusion, depending on how precise the appellant had been on these different occasions, and how exact the interpreters had been when translating questions into Tamil and rendering the appellant's answers into English.

7 Complex Problems of Translation

The examples in the previous section concern matters of linguistic and cultural usage that could potentially be resolved if the court had the necessary information, but such confusions are important if allowed to stand because HOPOs seize on these apparent inconsistencies so as to call credibility into question. Things are not always so simple, though, and I end with two examples of more complex and convoluted cultural and linguistic confusion, arising in asylum appeals for which I wrote expert reports.

In the first case it is less a matter of confusion stemming from the actions, however blameless, of the interpreter, and more a question of inherent cross-cultural incomprehension. Nonetheless, it illustrates how the very act of translation may obscure the implicit meaning of an initial utterance. The Reasons for Refusal Letter to a young Tamil who I shall call 'CS' argued as follows:

you claim that B----- was your uncle. He lived in M-----. You were asked at your asylum interview for his full name, you stated that you did not know his surname. You were unable to state whether he was a maternal uncle or a paternal uncle. It is considered that if B----- was your uncle who lived near to your [sic] then you would have known his full name and you would have also known whether he was on your mothers [sic] side or your fathers [sic] side. It is not accepted that you had an uncle in Sri Lanka called B-----. It is considered that he was an agent who facilitated your departure from Sri Lanka.

Many points arise from this passage. First, no allowance is made in the RFRL for the applicant's youth: CS was a fourteen-year-old boy, describing events that had occurred when he was ten. Second, he had already said that his departure was facilitated by an agent named R---, so it is not clear why the Home Office was so keen to portray B---- too as an agent. Third, Tamils do not have surnames: 'B-----' would have been the man's full name, save only for a patronymic initial. Fourth, and most relevant for this discussion, there is no Tamil word corresponding in semantic range to the English word 'uncle'; moreover, because Tamils practice cross-cousin marriage, the relevant distinction is not between father's side and mother's side relatives (most relationships can be traced on both sides), but between parallel relatives and cross relatives.

The relevant terms here are periyappa, cittappa, and maman: the first two apply to parallel relatives, the third to cross relatives. Each of them covers some of those people who would be 'uncles' in English: for example, the genealogically closest periyappa is one's father's older brother; the closest cittappa is one's father's younger brother; and the closest maman is one's mother's brother. However, these terms are also used with reference to many other, more distant, relatives on both the father's and mother's sides. CS must have used one of these three terms in his interview, which was then translated as 'uncle', so his untranslated answer would inevitably have actually conveyed the nature of the relationship. His supposed ignorance on this point was entirely an artefact of translation, but given the limitations on their role at substantive Home Office interviews, as discussed above, there would have been no scope for the interpreter to point this out even if they had understood the issue and its implications, which might well have not been fully apparent at the time.

Seen in retrospect, my last example, the appeal of Mr KM, ^[16] raises even more complex issues than I had appreciated at the time of my involvement. As usual, many interpreters were employed during these proceedings: at KM's screening and substantive interviews; at the taking of his witness statement; and during the initial court hearing. He was also interviewed by a psychiatrist, but the latter was also a Tamil and the resulting medical report did not make clear whether an interpreter was involved.

At KM's substantive asylum interview the following exchange took place:

Case owner: What happened in 19--?

KM: I took food for the workers at M-----. At A----- a claymore [mine] went off. They surrounded the area and arrested me, suspecting I was taking food for them.

Case owner: Do you remember the date in 19--?

KM: I think it was May or June 19--.

Case owner: Where did they take you when you were arrested?

KM: To A----- camp.

Case owner: What happened there?

KM: I was beaten and raped.

Case owner: How long were you held on that occasion?

KM: Five days.

That reference to rape was not followed up at all by the interviewer, even though the transcript noted, on the request of his solicitor, that KM became distressed at this point. As I have documented elsewhere [19: 111; 20: 88–89], it is often the case that Home Office case-owners do not follow up on mentions of rape. This may be partly out of embarrassment, but as noted above there is also a general tendency for them not to engage with answers, but instead to move on to the next question on their pre-determined checklist. So there was no exploration whatever of the rape claim at this interview, yet it was later cited as a reason to disbelieve his account.

KM gave much more detail in his witness statement. It is far too graphic to include here, but briefly, one drunken soldier forced him to perform oral sex and another forced him into oral and anal sex. These events were mentioned in the psychiatric report too, and the RFRL focused on alleged discrepancies between this medical report and the account in Mr KM's own witness statement:

Your detailed witness statement refers to one incident of rape. ... It is further noted that the Psychiatric Report... states, "he was also repeatedly raped which was a shattering experience for him." It is considered reasonable to expect that a person who alleges to have been raped would remain consistent in their detailed account of what happened, including the number of occasions this occurred. It is considered that your witness statement account to have been raped on one occasion is contracted [sic! 'contradicted'] by your account given in the Psychiatric Report.

I addressed this argument at para 46 of my own expert report, noting that the alleged discrepancy hinged on the distinction between sexual assault and rape:

According to the Tamil Lexicon, which plays a similar role for Tamil to that played by the Oxford English Dictionary for English, there is no distinct Tamil word for 'rape'. It gives only one word, which transliterates as palavantam, for which the English meanings given, transliterated where relevant, are '1. To use force, compel... Colloq. 2. To commit rape; palavantappuNarcci seytal'. In other words, the only Tamil term for 'to commit rape' is a colloquialism meaning, literally, 'to use force'. Entering 'sexual assault' into the online Lexicon produces no results. Another word in use, karpazhittal, means 'to violate a woman's chastity' (my italics) so would not be appropriate here. I do not think therefore that any significance can be attached to the apparent rape/sexual assault distinction in the translations of Mr PK's various accounts.

In his determination, the judge addressed this issue as follows:

47. The refusal letter suggests that the appellant's account is inconsistent in relation to his alleged rape during his detention in June ----. His (main) witness statement refers to one incident of rape yet in the psychiatric report of Dr K----- it refers to his having been repeatedly raped. The witness statement is said clearly to differentiate between the sexual assault by one army officer and the rape by the second. ...

49. I note what Professor Good says about this issue in his report at [46]. At [29] he states that he reads Tamil. However, with respect to Professor Good, regardless of his otherwise undoubted expertise, I am not satisfied that it is established that he is qualified to give expert evidence in relation to the Tamil language and the use or otherwise of the word 'rape'.

50. Nevertheless, it is important to bear in mind the evidence in the psychiatric report of the difficulty that the appellant had in providing an account of his "trauma-related experiences" because, in part, of his emotional distress.

The judge's response to my argument was no doubt entirely proper in legal terms insofar as it focused on my own legitimate breadth of expertise ^[17] but, typically, he did not address the translation point, which remained potentially valid whether or not my own evidence on the matter was accepted. Nonetheless I felt that my submission might have had some effect on his overall assessment. That may be wishful thinking, but in any case, the outcome was that Mr KM's asylum appeal was allowed.

There are, however, a number of other issues arising from this example. Firstly, there are two unjustified conflations in the RFRL's phrases 'your witness statement account' and 'your account given in the Psychiatric Report'. The first concerns the authorship of the witness statement. Such statements are written in the first person, giving them the appearance of direct accounts narrated by the statement givers, but of course—and whichever Practice Direction procedure is involved—they are not actually compiled in that way, as we have seen. There is, however, a process of reading back and—if necessary—rephrasing, so that in the end the statement can broadly be said to convey the perspective, but not of course the actual words, of its nominal author. The authorship issue arises even more strongly with respect to the psychiatric report, however, described in the RFRL as 'your account', but not in fact written by or even read back to the appellant. In both cases, UKVI's attribution of direct responsibility for these narratives to the applicant is at best a gross oversimplification.

Secondly, the multiplicity of expert reports raises similar issues to those raised by the multiplicity of interpreters. The more versions of an appellant's narrative that are produced, and the more authors—lawyers, experts, and others—that are involved in their production, the greater the risk that apparent discrepancies, which may really only amount to different word choices by different expert witnesses, may arise. These are then attributed by UKVI to inconsistencies by the applicants themselves, through the conflations just described. That is what seems to have happened here.

Unlike medical experts, who are often accused by the Home Office of having been hoodwinked by an appellant's convincing performance, I never interviewed and rarely even met the appellants for whom I wrote reports, basing my opinions wholly on the case documents supplied to me. Nonetheless, in the hope of avoiding such problems I routinely included this disclaimer in my reports:

In this section I summarise X's account based on the documents listed above. Any apparent discrepancies between this summary and X's own account are my responsibility and not attributable to inconsistency on the part of X herself.

Thirdly, whereas my report had focused on the question of whether 'rape' and 'sexual assault' were clearly distinguished in everyday Tamil, I now recognise that there is another equally fundamental and more precise question here, regarding whether or how they are distinguished in Sri Lankan and English law.

In Sri Lanka sexual offences are addressed in Sects. 363-365C of the Penal Code, which dates from 1885. Though last modified in 2006, its provisions and language still appear notably archaic, and have been roundly criticised on many counts over the years by local human rights organisations such as the Women and Media Collective [44]. Section 363 of the Code defines rape in purely heterosexual terms:

A man is said to commit "rape" who … has sexual intercourse with a woman under circumstances falling under any of the following descriptions: (a) without her consent…. [22]

It goes on to explain that 'Penetration is sufficient to constitute the sexual intercourse necessary to the offence of rape'. In the Tamil version of the Penal Code, the word used for 'rape' is கற்பழிப்பு (karpazhippu) [23: 149]. Like the term karpazhittal given in my expert report quoted above, this is a derivative of karpu, 'female chastity'.

Sec. 365, under the heading 'Unnatural Offences and Grave Sexual Abuse', then deals with 'Whoever voluntarily has carnal intercourse against the order of nature with any man, woman or animal…'. Again there is an explanatory note to the effect that: 'Penetration is sufficient to constitute the carnal intercourse necessary to the offence described in this section.' Sec. 365A then deals with 'Acts of gross indecency between persons'— ஆட்களுக்கிடையில் மிக இழிவான செயல்கள் or āṭkalukkiṭaiyil mikka iḻivāṉa ceyalkaḷ in the Tamil version [23: 153]—i.e., homosexual acts. Sec. 365B covers 'Grave sexual abuse'— , பாதுகாப்பான பாலியல் துஷ்பிரயோகம் or pāratūramāna pāliyal tuṣpirayōkam in the Tamil version (ibid.)—which is defined as follows:

Grave sexual abuse is committed by any person who, for sexual gratification, does any act, by the use of his genitals or any other part of the human body or any Instrument on any orifice or part of the body of any other person, being an act which does not amount to rape under section 363 … (a) without the consent of the other person.…

'Penetration' is not further defined in either section, but an investigation by the feminist NGO Equality Now reported that.

While the rape offence is narrowly applied to include peno-vaginal penetration only, other forms of sexual penetration are covered under the crime of '"grave sexual abuse" which carries the same penalties as rape.[12: 19]

While all this remains true, the legal position has recently shown signs of potential change. In November 2016 the Sri Lanka Supreme Court ruled that it was inappropriate to impose custodial sentences on two persons who had engaged in consensual homosexual intercourse, ^[18] while in May 2023 the Court ruled that a proposed private member's bill—seeking to repeal 'provisions that make sexual orientation a punishable offence' by rewording sec. 365 to apply only to 'carnal intercourse… with an animal' and entirely repealing sec. 365A—would not be inconsistent with the Constitution [40]. No actual change in the law has yet occurred, however.

In English law by contrast, under sec. 1 of the Sexual Offences Act 2003, 'rape' is defined more broadly, to cover all forms of non-consensual penile penetration irrespective of the victim's gender or sexuality:

1. Rape

(1). A person (A) commits an offence if—

(a) he intentionally penetrates the vagina, anus or mouth of another person (B) with his penis,

(b) B does not consent to the penetration, and

(c) A does not reasonably believe that B consents.

Penetration other than with the penis is then defined as 'assault by penetration' (sec. 2), and non-penetrative sexual touching is characterised by sec. 3 as 'sexual assault'. ^[19]

As far as Mr KM's experience was concerned, therefore, he was not raped by either soldier in terms of Sri Lankan law, but was subjected to acts of gross indecency or grave sexual abuse by both of them. Under English law, by contrast, he was raped by both. If this difference is not recognised by those in court, as it was not in this case, then the distinction the Home Office was seeking to make becomes highly problematic, and Mr KM's 'discrepancies' seem even more debatable. As a lay witness I could not have legitimately commented on such legal matters even if I had been aware of them at the time, but it may seem surprising that neither his legal representatives nor the IJ addressed this issue.

Be that as it may, the key point for the present discussion is as follows. As the court functionary least likely to have legal training or knowledge, yet the very one tasked with making questions based upon English legal terminology intelligible to a Tamil-speaking Sri Lankan applicant with no knowledge of English law, and with making this applicant's responses intelligible to specialists in English law with no knowledge of Sri Lankan law, the interpreter is placed in an unfair and almost impossible position.

8 Conclusions

Interpreters play a crucial role in facilitating intercultural communication during the asylum process, but their task is a complex one. The institutions in which they work have naïve and unrealistic expectations about the nature of their activity, such as the stress on 'exact' or 'verbatim' translation. In reality, interpreters are not passive translation machines but active participants in the multi-lingual and intercultural exchanges taking place during asylum procedures. Moreover, institutional codes of conduct require interpreters to exercise judgement on whether to intervene to explain a cultural misunderstanding, and on how to negotiate different registers of speech without damaging the credibility of an applicant's account or their own perceived competence. The need to fragment appellants' narratives into short sentences introduces further barriers to effective communication. Even the most competent interpreters may inadvertently generate confusion or apparent discrepancies because of the indeterminacy of the translation process itself [34], and because alternative, but equally legitimate, translation choices are often available. Added to these problems, finally, are the incongruences between national legal systems that interpreters, despite their lack of formal legal training, are expected to mediate.

Notes

[19] See https://www.legislation.gov.uk/ukpga/2003/42/contents. Accessed 5 November 2024. If initial hearings are held in Glasgow, appeals are heard by the Court of Session in Edinburgh rather than the English courts. The Sexual Offences (Scotland) Act 2009 is worded differently, though the principles are similar. See https://www.legislation.gov.uk/asp/2009/9/contents. Accessed 8 February 2025.

References

1. Aliverti, Ana, and Rachel Seoighe. 2017. Lost in translation? Examining the role of court interpreters in cases involving foreign national defendants in England and Wales. New Criminal Law Review 20 (1): 130–156. https://doi.org/10.1525/nclr.2017.20.1.130.

2. Arden, A. H. 1969. A progressive grammar of the Tamil language. (5th edn, rev. A.C. Clayton). Madras: Christian Literature Society.

3. Atkinson, J. Maxwell., and Paul Drew. 1979. Order in court: The organisation of verbal interaction in court settings. London: Macmillan.

4. Barsky, Robert F. 1996. The interpreter as intercultural agent in Convention refugee hearings. The Translator 2 (1): 45–63. https://doi.org/10.1080/13556509.1996.10798963.

5. Basnett, Susan. 2011. The translator as cross-cultural mediator. In The Oxford handbook of translation studies, ed. Kirsten Malmkjaer and Kevin Windle, 94–107. Oxford: Oxford University Press.

6. Berk-Seligson, Susan. 2002. The bilingual courtroom: Court interpreters in the judicial process. Chicago: Chicago University Press.

7. Brown, Penelope, Stephen C. Levinson, and John J. Gumperz. 1987. Politeness: Some universals in language usage. Cambridge University Press. https://doi.org/10.1017/CBO9780511813085.

8. Catford, John Cunnison. 1965. A linguistic theory of translation; an essay in applied linguistics. Oxford: Oxford University Press.

9. Cohen, Juliet. 2001. Questions of credibility: Omissions, discrepancies and errors of recall in the testimony of asylum seekers. International Journal of Refugee Law 13:293–330. https://doi.org/10.1093/ijrl/13.3.293.

10. Colin, Joan, and Ruth Morris. 1996. Interpreters and the legal process. Winchester: Waterside Press.

11. Courts and Tribunals Judiciary. 2002. Adjudicator Guidance Note, No. 3: Pre-Hearing Introduction. https://www.judiciary.uk/wp-content/uploads/2022/08/GuideNoteNo3.pdf. Accessed 28 August 2023.

12. Equality Now. 2021. Sexual violence in South Asia: legal and other barriers to justice for survivors. https://www.equalitynow.org/resource/sexualviolencesouthasia/. Accessed 13 December 2023.

13. European Union. 2013. Directive 2013/32/EU of the European Parliament and of the Council of 26 June 2013 on common procedures for granting and withdrawing international protection (recast). https://eur-lex.europa.eu/legal-content/en/TXT/?uri=celex%3A32013L0032. Accessed 11 September 2023.

14. Fenton, Sabine. 2004. Expressing a well-founded fear; interpreting in convention refugee hearings. https://www.refugee.org.nz/Reference/Sabine.html. Accessed 21 February 2025.

15. Ferguson, Charles A. 1959. Diglossia. Word 15 (2): 325–340. https://doi.org/10.1080/00437956.1959.11659702.

16. Gibb, Robert, and Anthony Good. 2014. Interpretation, translation and intercultural communication in refugee status determination procedures in the UK and France. Language and Intercultural Communication 14 (3): 385–399. https://doi.org/10.1080/14708477.2014.918314.

17. Goffman, Erving. 1967. Interaction rituals: Essays on face to face behavior. Garden City, New York: Anchor Books.

18. Good, Anthony. 2007. Anthropology and expertise in the asylum courts. London: Routledge/Clarendon.

19. Good, Anthony. 2011. Witness statements and credibility assessments in the British asylum courts. In Cultural expertise and litigation: Patterns, conflicts, narratives, ed. Livia Holden, 94–122. London & New York: Routledge.

20. Good, Anthony. 2011. Tales of suffering: Asylum narratives in the refugee status determination process. West Coast Line 68:80–89.

21. Good, Anthony. 2015. The benefit of the doubt in British asylum claims and international cricket. In Of doubt and proof: Ritual and legal practices of judgment, ed. Daniela Berti, Anthony Good, and Gilles Tarabout, 119–140. Farnham: Ashgate.

22. Government of Sri Lanka. 2006. Sri Lanka: Penal Code, 1 January 1885 (Consolidated version up to Act No. 16 of 2006). https://www.refworld.org/docid/4c03e2af2.html. Accessed 12 December 2023.

23. Government of Sri Lanka. 2015. Taṇṭaṉai saṭṭakkōvai (Penal code). Colombo: Sri Lankan Government Printing Office.

24. Heaton-Armstrong, Anthony, and David Wolchover. 1992. Recording witness statements. In Analysing witness testimony: A guide for legal practitioners and other professionals, ed. Anthony Heaton-Armstrong, Eric Shepherd, and David Wolchover, 222–250. London: Blackstone Press.

25. Henderson, Mark, Rowena Moffatt, Alison Pickup and Monika Nollet. 2023. Best practice guide to asylum and human rights appeals. https://www.ein.org.uk/bpg/contents. Accessed 1 September 2023.

26. Home Office. 2024. Interpreters Code of Conduct, Version 5.0. Accessed 4 November 2024.

27. ILPA (Immigration Law Practitioners Association). 2023. RE: Practice Direction of the Immigration and Asylum Chamber of the First-tier Tribunal (Letter). https://ilpa.org.uk/wp-content/uploads/2023/03/ILPA-Letter-Regarding-Witness-Evidence-Requirements-in-FtTIAC-Practice-Direction-10.03.23.pdf. Accessed 7 December 2023.

28. Kalin, Walter. 1986. Troubled communication: Cross-cultural misunderstandings in the asylum hearing. International Migration Review 20:230–241. https://doi.org/10.1177/019791838602000206.

29. Ministry of Justice. 2023. Civil Procedure Rules: Practice Direction 32 – Evidence. https://www.justice.gov.uk/courts/procedure-rules/civil/rules/part32/pd_part32. Accessed 23 August 2023.

30. Ministry of Justice. 2024. Guide to Language Interpreter and Translation Services in Courts and Tribunals. https://assets.publishing.service.gov.uk/media/6604492be8c442001a2203db/A_Guide_to_language_interpreter_and_translation_services_statistics.pdf. Accessed 4 November 2024.

31. Morris, Ruth. 1995. The moral dilemmas of court interpreting. The Translator 1:25–46. https://doi.org/10.1080/13556509.1995.10798948.

32. Nikolaidou, Zoe, Hanna Sofia Rehnberg, and Cecilia Wadensjö. 2023. "Do I have to say exactly word by word?" (Re)producing and negotiating asymmetrical relations in asylum interviews. Journal of International Migration and Integration 24:745–788. https://doi.org/10.1007/s12134-022-00945-2.

33. Pöllabauer, Sonja. 2007. Interpreting in asylum hearings: Issues of saving face. In Critical Link 4: Professionalisation of interpreting in the community, ed. Cecilia Wadensjö, Birgitta Englund Dimitrova and Anna-Lena Nilsson, 39–52. Amsterdam/Philadelphia: John Benjamins Publishing Company.

34. Quine, Willard V. 2013 [1960]. Word and object (new edn.). Cambridge MA & London: MIT Press.

35. Rycroft, Roxana. 2005. Communicative barriers in the asylum account. In The challenge of asylum to legal systems, ed. Prakash Shah, 223–244. London: Cavendish.

36. Senior President of Tribunals. 2022. Practice direction of the Immigration and Asylum Chamber of the First-tier Tribunal. https://www.judiciary.uk/wp-content/uploads/2022/09/2022.05.13-Practice-Direction-FtT-IAC-1.pdf. Accessed 6 December 2023.

37. Senior President of Tribunals. 2024. Practice direction of the Immigration and Asylum Chamber of the First-tier Tribunal. https://www.judiciary.uk/wp-content/uploads/2024/10/Practice-Direction-F-tT-IAC-01.11.24.pdf. Accessed 3 November 2024.

38. Smith-Khan, Laura. 2023. Incorporating sociolinguistic perspectives in Australian refugee credibility assessments: The case of CRL18. Journal of International Migration and Integration. 24 (Suppl 4): S727–S743. https://doi.org/10.1007/s12134-022-00937-2.

39. Spotti, Massimiliano. 2019. "It's all about naming things right": the paradox of web truths in the Belgian asylum-seeking procedure. In Asylum determination in Europe: ethnographic perspectives, eds. Nick Gill and Anthony Good, 69–90. Cham: Palgrave Macmillan, Springer Nature Switzerland.

40. Supreme Court of Sri Lanka. 2023. Penal Code (Amendment) Bill. SC SD No. 13/2023. https://supremecourtdemo-cert.melstasoft.com/wp-content/uploads/2023/12/sc_sd_13_2023.pdf. Accessed 5 November 2024.

41. Thomas, Robert. 2011. Administrative justice and asylum appeals: A study of tribunal adjudication. London: Bloomsbury.

42. UNHCR (United Nations High Commissioner for Refugees). 2019 [1979]. Handbook on procedures and criteria for determining refugee status under the 1951 Convention and the 1967 Protocol relating to the status of refugees (HCR/1P/4/Eng/REV.2). https://www.unhcr.org/uk/media/handbook-procedures-and-criteria-determining-refugee-status-under-1951-convention-and-1967. Accessed 7 November 2024.

43. Wadensjö, Cecilia. 1998. Interpreting as interaction. London & New York: Longman.

44. Women and Media Collective. 2021. Policy brief: Legal framework on sexual violence in Sri Lanka: the need for reforms. https://womenandmedia.org/wp-content/uploads/2021/11/FINAL-LAYOUT-POLICY-BRIEF_Legal-Framework-on-Sexual-Violence-in-Sri-Lanka-English.pdf. Accessed 11 December 2023.