Language analysis still under strain

Jim Hoskin, Dr. Tina Cambier-Langeveld and Prof. Paul Foulkes have endeavoured to provide an objective outline of language analysis in a recently published article, entitled Improving objectivity, balance and forensic fitness in LAAP: a response to Matras, as a result of Prof. Yaron Matras’s Duly verified? Language analysis in UK asylum applications of Syrian refugees.

To provide some context, Prof. Yaron Matras raises a series of issues in his article, pertaining to the LADO (Language Analysis for the Determination of Origin) method. Among the disparate concerns of his is the use of native speakers. While the IAFPA has endorsed the use of native speakers for LADO, professor Matras does not report any merits, claiming instead that the native speakers are underqualified, lack training and fail to take linguistic variation into account.

Hoskin, Cambier-Langeveld and Foulkes disagree fundamentally with professor Matras: “Matras’s approach is problematic in a number of ways. First, Matras argues de facto in favour of testing the evidence against a single hypothesis; yet this may lead to inadvertent bias and betrays logical flaws that have serious forensic implications, as is well known in other branches of forensic science. Second, there are potential problems with the formulation of Matras’s initial hypothesis, based as it is on the applicant’s story. We stress that claims of secondary influences cannot be tested since their effect is inherently unknowable. Neither can they be relied upon, since they may be false. Third, it appears that Matras’s view of the question to be addressed is not consistent with the trier of fact’s requirements, given that the burden of proof is on the applicant.”

The one-eyedness of professor Matras is in line with his understanding of the role of an expert witness. As underlined in the response article, “…ideas on the nature of the task vary, depending on whether the analyst is positioned as working for an asylum seeker or an asylum agency. This appears to lead some counter-experts to act as a witness for the defense and, in some cases, to see it as their task not to supply an objective second opinion but to counter previous negative findings (…)”. While professor Matras denies the usefulness of working with a native speaker, the fact that he assesses Arabic, Kurmanji and Sorani without truly being a native speaker of the dialect or even language in question goes on to show that he ignores a crucial aspect of language analysis.

The article by Hoskin, Cambier-Langeveld and Foulkes does a very good job at describing the shortcomings of the method preferred by Matras, while also not being over-appreciative of the method where linguists and analysts work together, admitting that no method is perfect and that there is still room for improvement.

A full version can be found here:

https://journals.equinoxpub.com/IJSLL/issue/current

IAFPA Code of Practice: 

https://www.iafpa.net/the-association/code-of-practice/