Bringing together researchers, practitioners and policy makers working with migrants, refugees and asylum seekers in Scotland
Language analysis is a tool used by Government decision-makers to “test” the credibility of an asylum seeker’s claim, and to assess whether the applicant is being honest about the country they claim to be from. It involves the consideration of a recording of an asylum seeker’s speech in order to judge their country of origin, and although language analysis has proved popular with the Home Office in recent years, its use has caused concern in several countries in Europe and elsewhere for over a decade (K. Zwaan et al (eds) Language and Origin: the Role of Language in European Asylum Procedures: Linguistic and Legal Perspectives, (Wolf, 2010)).
When language analysis is used, it indicates that a decision-maker is sceptical about the asylum claim they are deciding. Sceptical approaches remind us of the association between asylum decision-making and a “culture of disbelief”, and of the reputation which the asylum process has of being unfair to applicants. Looked at this way, language analysis is simply a manifestation of that culture.
To try and understand why language analysis is used, we need to think about the position of the decision-maker, and to recognise that credibility is the key determinant in asylum claims: their success or failure almost always turns on whether the applicant is believed. Faced with the task of determining credibility, Government decision-makers have turned to language analysis, and while the occasional use of reliable forms of language analysis may be a justifiable way to “test” a particular claim when there is a good reason to be sceptical, the problem is that the extent of its use is not clear.
Aim to reduce the number of successful claims
Partly in order to respond to some of these concerns, an expert group of linguists published what are now known as the “2005 Guidelines” on the use of language analysis (“Guidelines for the Use of Language Analysis in Relation to Questions of National Origin In Refugee Cases” (D Eades, “Applied Linguistics and Language Analysis in Asylum Seeker Cases” (2005) Applied Linguistics 26/4 503-526)). More recently, the work of John Campbell and others have shown Governments’ use of language analysis to be politically motivated, aimed at “managing” -and discrediting – claims to be from particular war-torn countries, such as Somalia and Afghanistan, and now Syria, without proper regard for the quality or reliability of the methods used, and with the aim of reducing the number of successful claims from those countries, even if it means refusing potentially valid claims. (J. Campbell, “Language Analysis in the United Kingdom’s refugee status determination system: seeing though policy claims about “expert knowledge” 2012 Ethnic and Racial Studies, pp1-21; Ministerial statement 25 February 2013.)
The practices of SPRAKAB, the Swedish company which has provided language analysis services to the UK Home Office in recent years, do not appear to be governed by the 2005 Guidelines, and they involve the applicant being put on the phone during a screening interview, and being subjected to a brief interview, after which a report can be issued that the applicant is “with certainty” not speaking the language of their place of origin, but “with certainty” that spoken elsewhere. Not surprisingly, these practices have raised important questions about the competence of the analysts who carry out the assessments, the quality of the resulting reports, and the fairness of the way in which decision-makers treat those reports. These questions were considered by the Court of Appeal (England and Wales) in 2012 and, just a few weeks ago, they were also addressed by its Scottish equivalent, the Inner House of the Court of Session.
The English case (RB (Somalia) v SSHD  EWCA Civ 277 (CA (Civ Div) concerned an applicant who claimed to be Somali, and the Court of Appeal deferred to the opinion of the body which determines asylum appeals: the Immigration and Asylum Chamber of the Tribunals Service. That Chamber’s Upper Tribunal had decided that if the SPRAKAB report concluded that the person’s native language was “with certainty not” that of their claimed country of origin, then “little more” than that opinion was required to conclude that the person was not Somali. Although the Court of Appeal had been uncomfortable with the company’s insistence that their analysts should be granted anonymity, the Upper Tribunal had heard in person from the head of SPRAKAB, although not from the personnel who had carried out the analysis, and the Court of Appeal concluded that it was the Upper Tribunal’s task to decide how to manage the conduct of the experts whose evidence they heard. The Court of Appeal therefore did not think it was appropriate to interfere with the Upper Tribunal’s conclusions.
Reports found to be unlawful
It seemed that a negative language analysis report could seal the fate of an asylum applicant’s credibility, and of their whole claim, and concerns about the use of language analysis continued to be voiced after this case was decided (S. Craig “The use of language analysis in asylum decision-making in the UK- a discussion” JIANL 2012 26(3) 255-268). As noted on the Scottish Courts website and elsewhere (see also here), two appeals were recently heard together in the Court of Session (Inner House) M ABN and KASY v Adv. Gen for Scotland  CSIH 68.
In these appeals, the Court of Session found the Government’s use of anonymous “expert” reports provided by SPRAKAB to be unlawful because the reports failed to demonstrate sufficient expertise, since their authors were anonymous, and gave views which went beyond their expertise. This case is good news for applicants firstly because of the even-handed approach which the Inner House took, treating the Home Office reports in the same way that an “expert” report for an appellant – such as a doctor’s report – would be treated. In one of the reports which the court considered, the appellant was described as speaking “with certainty” a form of Somali spoken only in Kenya, when there was no indication that the personnel who spoke to the appellant had ever been to Kenya or had made any study of Somali as spoken there.
Secondly, the case is good news because it rejects the practice of offering assessment of country knowledge or of credibility in a language analysis report. The author of the report claimed that the appellant lacked knowledge of their home area, without identifying what they understood that area to be, and the report purported to do the decision-maker’s job of assessing credibility for them, by saying that the applicant’s answers sounded “rehearsed”.
A UK-wide approach
Taking a different approach in these cases is significant because, although Scotland has a separate legal system, and the Inner House of the Court of Session, as its Scottish equivalent, is not bound to follow the Court of Appeal’s decisions, Court of Session judges are usually keen to take the same approach as the rest of the UK in relation to immigration and asylum cases. The reasons for this include administrative consistency, since the Home Office decides asylum applications for the whole of the UK, and appeals against Home Office asylum refusals are also determined by a UK-wide body, the Immigration and Asylum Chamber of the First Tier and Upper Tribunals.
It is only in those few cases that reach the higher courts – the Inner House of the Court of Session in Scotland, or the Court of Appeal in England, Wales and Northern Ireland- that decision-making in UK immigration and asylum cases has the potential to diverge. The aim of administrative consistency therefore encourages the higher courts to promote a common approach to similar cases among decision-makers across the UK. There are limits to consistency, however, and when cases raise procedural issues – about how cases should be treated in court, or about the role of the court itself – then the Court of Session has, at times, been willing to take a separate approach, perhaps recognising that when questions about its own role are being raised, then a separate approach can be appropriate. (Eba v Adv General for Scotland  UKSC 29; S. Craig, M. Fletcher and K. Goodall, Challenging Asylum and Immigration Tribunal decisions in Scotland: An evaluation of Onward Appeals and Reconsiderations 2008; S. Craig, The Edinburgh Law Review 16.2 (2012): 210-223). In this context, therefore, it is not so surprising that the Court of Session diverged from the Court of Appeal when deciding a question about how the court should treat “expert” evidence.
And the discussion is not over yet, because the Home Office have indicated that they intend to appeal, which means that the UK’s Supreme Court will have to resolve these questions for the UK as a whole. Hopefully the Supreme Court will endorse the approach of the Court of Session and recognise that because refugee cases deserve to be determined fairly and individually, this means firstly that where “expert” evidence is offered in an asylum case, it should meet the same standards regardless of whether that evidence is offered by the Home Office or the applicant. Secondly, the task of determining credibility belongs to the individual decision-maker and cannot be delegated to an “expert”, however strong the institutional pressures to do so may be.