The use and abuse of language analysis in asylum cases
July 21, 2005 — Comment
Written by Tim Cleary
The recent practice of language analysis in asylum cases has met with criticism from concerned professionals. Linguists have responded with a set of professional guidelines with the aim of preventing potential abuses of this method of analysis.
Professional linguists and law professionals are becoming increasingly concerned about the use and abuse of language analysis in the examination of asylum applications. In essence, this involves the use of a recording and an analysis of an applicant’s speech to determine his or her national origin, with a view to identifying whether an asylum claim is genuine or ‘bogus’.
As discovered in recent collaborative research conducted by linguist Professor Peter L Patrick of the University of Essex and barrister Nick Oakeshott of the Refugee Legal Centre, many issues have been raised about the practice of language analysis in a number of countries, most notably in relation to the degree to which many language analysts are qualified – or not – to conduct such analyses, and the methods being used to form judgements on an applicant’s geographical origin or nationality. And although the practice is not as widespread here as it is in other countries such as Sweden, Switzerland, the Netherlands and Australia, language analysis has been used on a smaller scale in the United Kingdom since at least 2001. Sri Lanka, Afghanistan, Iraq and Somalia are among the countries from where asylum applicants subjected to such scrutiny have claimed to come.
A further concern has been the controversial use of commercial analysis companies in the investigation of United Kingdom asylum claims. One fear is that the analysts being employed – often translators and interpreters without a formal linguistic grounding – are not sufficiently qualified to form a reliable linguistic judgement, which can be at best only tentative even where a linguist is sufficiently qualified and experienced. Furthermore, language analysts’ identities have remained anonymous, which raises questions about the accountability of analysts and the transparency of such methods. Barrister Nick Oakeshott states that these issues have ‘led to questions being asked as to whether this can be used as expert evidence in immigration courts’.
Although language – a term used here to cover the particular language spoken, as well as such phenomena as accent, grammar and vocabulary – can often be suggestive of a person’s geographical origin, this is not always the case and the use of this form of analysis needs to be approached with extreme caution. Professor Patrick warns that such an analysis can only ‘suggest where somebody is likely to have been socialised’, whereas commercial analysis companies have in the past aimed to establish ‘certainty rather than likelihood’; and unlike the usual evidence and counterevidence expected to be employed in linguistic research, language analysts have been criticised for presenting ‘evidence that tends to point towards only one conclusion’, added Professor Patrick.
In the areas of the world where many United Kingdom asylum claimants come from, the relationship between language and geographical origin is highly complex and difficult to establish. For example, there are probably almost twenty distinctive language varieties spoken in Iraq, many of which, including Arabic dialects and Kurdish, are also spoken in neighbouring countries. Thus, the use of language analysis on applicants claiming to come from Iraq is extremely complicated and fraught with difficulties, especially since many people will be multilingual and may even mix languages quite freely. As for Somalia, Nick Oakeshott highlights that language analysis has been used in attempts to determine whether asylum applicants are members of at-risk minority clans. Although the practice may also be used to protect applicants who are at risk, the use of what are possibly unqualified analysts, as well as the absence of other evidence to be used alongside linguistic evidence, poses the risk that this could be an unreliable method for judging a person’s origin.
Despite this need for caution, language analysis companies have been called upon to provide language experts in asylum cases in the United Kingdom. This was highlighted in an article in the Guardian on 30 October 2001 (‘Asylum seekers face language test’), which brought attention to a government trial in which so-called language experts were to be consulted in United Kingdom asylum cases between November 2001 and June 2002. Since then, in March 2003, there was a further one-month trial period for would-be Iraqi asylum applicants (Guardian, 11 March 2003, ”Iraqi’ refugees to face language checks’). In October 2003 there was a report presented to Parliament by the then Minister for Citizenship and Immigration, Beverley Hughes, on the outcome of the trial periods. Despite recommendations in the report that a United Kingdom language analysis bureau be set up, there appears to have been no steps taken towards this, which means that foreign commercial analysis companies continue to provide their services to the Home Office. Linguists such as Professor Patrick are clear that any such analysis should be based on evidence presented by academically-trained experts.
With a view to addressing the serious issues posed by language analysis, a set of professional guidelines was developed and signed in 2004 by a group of linguists, including Professor Patrick, under the name of the Language and National Origin Group. The guidelines appeared in the International Journal of Speech, Language and the Law: Forensic Linguistics, vol. 11 no. 2 (2004), and have been endorsed internationally by various bodies of professional linguists, including the Linguistic Association of Great Britain (LAGB). Some key issues raised in the Guidelines for the Use of Language Analysis in Relation to Questions of National Origin in Refugee Cases include:
- Linguistic evidence should be used critically only alongside other evidence to determine national origin;
- Language analysis does not suggest national origin, nationality or citizenship, but rather the context in which a person has been socialised;
- Language analysis should be conducted by suitably qualified professional linguists, and not necessarily native speakers of the language in question, since native speakers do not necessarily have a sufficient level of expertise;
- Any analysis should be conducted using good quality, useful and reliable data, and should take account of the context of the interview.
The guidelines are especially relevant to language experts, government agencies and NGOs, for they raise awareness about the complex issues involved in language analysis. They constitute an important step towards revealing the inaccurate and potentially disastrous conclusions that can be drawn if language analysis is used uncritically and without due care. It is hoped that more bodies will endorse these guidelines and more support and information surrounding these issues will be shared between concerned professionals.
Tim Cleary is studying for an MA in Language Documentation and Description at the School of Oriental and African Studies in London, where he is looking at the sociolinguistics of the Tamazight (Berber) language of North Africa. He is also a voluntary translator for the Institute of Race Relations.
The Institute of Race Relations is precluded from expressing a corporate view: any opinions expressed are therefore those of the authors.