Content uploaded by Tina Cambier-Langeveld
Author content
All content in this area was uploaded by Tina Cambier-Langeveld on Aug 11, 2016
Content may be subject to copyright.
IJSLL (PRINT) ISSN -
IJSLL (ONLINE) ISSN -
Commentaries and responses
Afliation
Immigration and Naturalisation Service, the Netherlands
email: gm.cambier.langeveld@ind.minbzk.nl
IJSLL VOL . –08 doi : 10.1558/ijsll.v19i1.95
©, EQUINOX PUBLISHING
The International
Journal of
Speech,
Language
and the Law
Clarication of the issues in language analysis:
a rejoinder to Fraser and Verrips
Tina Cambier-Langeveld
Abstract
Volume 18.1 (2011) of this journal presented two responses to my article ‘The role of
linguists and native speakers in language analysis for the determination of speaker
origin’, published in volume 17.1 (2010). Both Helen Fraser and Maaike Verrips pick up
on eight cases presented in my article, and question whether these cases have any
implications for the ‘Guidelines’.1 The responses compel me to bring forward a direct
connection between the eight cases and the ‘Guidelines’. In this rejoinder, the impact of
the eight cases on the ‘Guidelines’ is spelled out. Secondly, various arguments relevant to
the ‘linguist vs native speaker debate’ are distinguished, in order to increase under-
standing of the controversy. Finally, this rejoinder reects on how this debate could have
become polarised.
KEYWORDS
LANGUAGE ANALYSIS
,
LADO
,
GUIDELINES
,
ASYLUM SEEKERS
,
SPEAKER
ORIGIN
96
THE INTERNATIONAL JOURNAL OF SPEECH
,
LANGUAGE AND THE LAW
1 Introduction: contents of my original article and two
responses
Some eight years ago, a group of linguists called the ‘Language and National
Origin Group’ (2004) published so-called guidelines for the use of language
analysis in asylum procedures. These ‘Guidelines’ consist mainly of basic lin-
guistic, sociolinguistic and general observations which clearly show why
linguistic expertise must be involved. The ‘Guidelines’ also prescribe that Lan-
guage Analysis for the Determination of the Origin of asylum seekers (LADO)
must be done by ‘qualied linguists with recognized and up-to-date expertise,
both in linguistics and in the language in question, including how this language
diers from neighboring language varieties’ (p. 262).
In volume 17.1 (2010) of this journal, I published an article entitled ‘The role
of linguists and native speakers in language analysis for the determination of
speaker origin’ (henceforth referred to as ‘my article’). In my article, several
shortcomings of the ‘Guidelines’ were addressed; below is a brief summary of
the main points of my article.
First, my article specied the criticism of the ‘Guidelines’, as expressed on
various earlier occasions by myself (e.g. Cambier-Langeveld 2007) and also by
Eriksson (2008). In short, it was argued that the ‘Guidelines’ fail to recognise
that there is more to LADO than linguistic expertise. Also, there is no scientic
basis for the assertion (implicit in the ‘Guidelines’) that specialised linguists
perform better at the task of determining speaker origin than native speakers.
Second, my article documented a range of problems that may arise when
non-forensically trained linguists are engaged to perform this type of forensic
analysis. For this purpose, eight cases were presented in which the correct
outcome is now known; these will henceforth be referred to as ‘the eight cases’.
In each of these eight cases, analyses had been carried out by a supervised
native speaker working for the Dutch Immigration and Naturalisation Service
(INS), and counter analyses were done by a linguist engaged by the asylum
seeker as a counter expert. Several factors that could have caused the incorrect
conclusions by linguists were discussed.
Third, my article aimed to formulate additional criteria for proper conduct of
LADO which are not listed in the ‘Guidelines’. I discussed three ways of
combining native-speaker competence and linguistic expertise, and stated
explicitly that there is no valid reason to declare that any one of these is in
principle superior to the other.
My article answered the call in Fraser (2009:131): ‘those who are aware of
problems with the Guidelines should publish clear, evidence-based docu-
mentation of the issues’.
CLARIFICATION OF THE ISSUES IN LANGUAGE ANALYSIS
97
Volume 18.1 (2011) of this journal presented two responses to my article:
one by Helen Fraser and one by Maaike Verrips. Both authors belong to the
‘Language and National Origin Group’. Both authors contend that the eight
cases do not provide solid evidence for the conclusion that native-speaker
competence should be involved in LADO. In the same vein, they argue that
these cases are insucient to legitimise a research method that is not recog-
nised by the ‘Guidelines’. Also, both authors point out that the incorrect
conclusions in these cases come from counter experts who might not have the
qualications required by the ‘Guidelines’.
While some points made by Fraser and Verrips are legitimate, their res-
ponses do not invalidate my conclusions. The eight cases were not intended as
legitimisation or defence of the ‘supervised native speaker’ approach, as Fraser
and Verrips seem to have understood.
2
I just argued that there is no valid basis
for
a priori
exclusion of the ‘supervised native speaker’ approach. Neither
Fraser nor Verrips respond adequately to this point.
I have in fact argued for the validity of a process that would seem to fall
outside the ‘Guidelines’, but in a dierent article, where appropriate evidence is
put forward.
3
2 Contents and limitations of this rejoinder
The responses by Fraser and Verrips justify elaboration on a number of issues.
First of all, it appears that the implications of the eight cases for the ‘Guidelines’
must be specied. In section 3.1, I will discuss how these cases undermine a
basic assumption underlying the
‘Guidelines’. In section 3.2, I will list the
problem factors evident in the eight cases. It will thus become clear that the
‘Guidelines’ were not an appropriate response to the state of aairs in the
Netherlands as witnessed in these cases; see section 3.3.
In section 4, the basis for the view that native-speaker competence is relevant
and should be included in LADO will be claried once more. It seems essential
to separate certain arguments here, so that the real points of controversy can be
localised.
The transparency of LADO reports is addressed in section 5, in order to
increase understanding of how the discussion about LADO might have become
polarised (section 6).
This rejoinder concludes with a take-home message for the authors of the
‘Guidelines’.
In what follows below I will not respond to some other issues that are raised
by Verrips. Fixing the holes in her portrayal of casework by the Dutch INS is
beyond the scope of this rejoinder.
4
98
THE INTERNATIONAL JOURNAL OF SPEECH
,
LANGUAGE AND THE LAW
3 The eight cases and their relevance to the ‘Guidelines’
The eight cases serve to underpin criticism of the ‘Guidelines’, but
not
by
showing that supervised native speakers performed better than expert linguists
in these few cases. The implications of the eight cases are laid out in this section.
The eight cases were not handpicked, but form the complete subset of cases
in which both the identity of the asylum seeker has been established and in
which a counter report was submitted to the INS. Since counter reports are
only requested when the government analysis did not conrm the claimed
origin, the dataset by denition contains only cases with a negative result by the
INS. The dataset could have included cases in which the established identity
conrmed the claimed country of origin, for instance if the date of birth (i.e.
age), place of birth or name was established by documents rather than nation-
ality (this is not just a theoretical possibility; see the data in Cambier-Langeveld
2010b). Such cases would then have proven the INS wrong. However, such
cases did not occur.
In order to protect the identity of the counter experts, not much information
about their academic background was given in my article. As a result, both
Fraser and Verrips question whether the counter experts involved full the
criteria set by the ‘Guidelines’, and if not, whether the problems in the eight
counter reports are rightfully used to criticise the ‘Guidelines’.
First, let it be clear that these counter experts were trained in linguistics and
had some specialisation that made them seem eligible (to a layman at least) for
acting as an expert. Of course some training in linguistics and some special-
isation does not automatically indicate competence for the task at hand, but
these counter experts themselves claimed to have sucient expertise in the
relevant language(s) to perform a counter analysis. Second, whether these
counter experts t the criteria of the ‘Guidelines’ or not is hard to say due to lack
of precise denitions, but is not crucial to my argument anyway; see below.
3.1 The direct connection between the eight cases and the ‘Guidelines’
The discussion has reached a point where the scientic community should
learn about the direct connection between the eight cases and the ‘Guidelines’.
Two of the counter analyses with an incorrect conclusion were submitted by a
linguist who had a key role in the ‘Language and National Origin Group’, which
authored the ‘Guidelines’. There are even indications that at least one of the
eight cases discussed in my article belongs to the
very same set of cases
that led
to the conception of the ‘Guidelines’.
The ‘Guidelines’ were published in a special issue of the
International
Journal of Speech, Language and the Law
(11.2, 2004) which was guest edited
CLARIFICATION OF THE ISSUES IN LANGUAGE ANALYSIS
99
by Diana Eades and Jacques Arends. In the introduction to this special issue,
the two guest editors describe how the linguistic community came to be so con-
cerned with the topic of language analysis. Some casework is described (Dutch
asylum cases), including the following comment: ‘In at least three instances, the
report claims that the asylum seeker “does not know the national anthem”,
while in all three cases they sing the entire rst stanza virtually without any
mistake both with regard to melody and text’ (Eades and Arends 2004:194).
Note the resemblance with case 6 in my article (p. 80):
The INS analyst comments that the applicant sings the Sierra Leonean
national anthem but with some mistakes in the text, and takes this as
evidence against the claimed origin. CE [counter expert] remarks that
the applicant sings the anthem ‘nearly faultlessly’, with ‘just a few aber-
rations from the ocial text’, and takes this as evidence supporting the
claimed origin.
In case 6, the conclusion of the supervised native speaker working for the INS
turned out to be correct and the conclusion of the counter expert proved
to
be
incorrect: the applicant involved is from Nigeria, not Sierra Leone.
The authors of the ‘Guidelines’, all of them linguists, clearly (if tacitly)
worked on the assumption that, in cases where a counter expert disagreed with
the native-speaker analyst working for the INS, the counter expert (a linguist)
was right. Given the evidence we now have, this assumption underlying the
‘Guidelines’ must be called into question.
It is fair to comment that a case in which a linguist and a native speaker come
to dierent conclusions can in itself not be considered proof that the native
speaker is wrong (or vice versa, of course). Although this may seem obvious,
argumentation along those lines is quite common in the LADO context.
3.2 Identication of the problems in the eight counter reports
In my article, I left it partly to the reader to infer the problem factors involved
from the case descriptions. Perhaps it is helpful to specify the nature of these
factors more precisely. Looking at the analyses of the counter experts, most of
whom failed to come to the correct conclusion, the following problem factors
can be identied:
1. a lack of objectivity or professional distance (see for explicit examples
cases 2 and 7. Perhaps this problem is also underlying to problems 2, 3,
4 and 5);
2. incorrect identication of the language variety (see cases 5 and 6);
3. too lenient a view on what may reasonably be expected from the
100
THE INTERNATIONAL JOURNAL OF SPEECH
,
LANGUAGE AND THE LAW
applicant in terms of language skills (see cases 1, 3 and 4);
4. an unwillingness to draw a conclusion which naturally follows from the
established linguistic facts
(see cases 3 and 4);
5. a readiness to accept unsubstantiated claims (e.g. as explanations for
lack of competence in a language) without critical review (see case 1);
6. a lack of understanding of the burden of proof in the context of asylum
law (see case 3);
5
7. acceptance of a supplementary recording, while there is obvious reason
to question the origin of this recording. In this particular case (case 8),
the supplementary recording contained a sample of uent speech in a
language of which the applicant had previously claimed to have no
knowledge.
These problems are not necessarily related to the fact that seven of these eight
counter analyses were performed without consultation of a native speaker;
various other factors could have played a role. The point with respect to the
‘Guidelines’ is that these problems are not anticipated or covered.
3.3 The ‘Guidelines’ as an appropriate response?
At the time, back in 2003–2004, it was deemed appropriate by a large group of
linguists to set up the ‘Guidelines’. These were explicitly aimed at governments,
insisting that specic linguistic expertise be involved in LADO. This was in itself
a useful endeavour, and the ‘Guidelines’ provide a compelling case for this core
message. No-one disputes that linguists should be involved. Given what we
have witnessed in the eight cases, however, the ‘Guidelines’ as they stand can
hardly be deemed an appropriate response to what was happening in the
Netherlands: the ‘Guidelines’ are not only intended to exclude the method
which – as we now know – gave the correct outcome in these cases, but the
‘Guidelines’ also fail to tackle the problems that were leading to incorrect
conclusions. The ‘Guidelines’ do not consider problem factors such as those
identied above, nor do they provide requirements that would guard against
them.
6
The problem factors evident in the eight counter reports have not dis-
appeared by themselves. All the problem factors listed above are still a cause of
potential error in counter analyses submitted to the INS to this day, even
though they are now generally performed by linguists explicitly selected
according to the ‘Guidelines’.
7
In all, the ‘Guidelines’ are not an eective tool for
quality control in LADO.
CLARIFICATION OF THE ISSUES IN LANGUAGE ANALYSIS
101
4 Clarication of the arguments
Both Fraser and Verrips are under the mistaken impression that the eight cases
were supposed to underpin the conclusion that native-speaker competence
should be included in LADO. This was not my intention; the eight cases only
served to illustrate the risks of using a non-native, non-forensically trained
specialised linguist (allowed by the ‘Guidelines’). My position with respect to
native-speaker competence is primarily based on a matter of principle: there is
simply no reason to identify two dierent types of expertise and then to assume
that only one of them is relevant. To understand the controversy, it may be
helpful to separate various sub-arguments.
4.1 The native speaker according to the ‘Guidelines’
In my article, I stated that the ‘Guidelines’ fail to recognise that the competence
of native speakers is also relevant and should be included in LADO. In Fraser’s
response, it is claimed that the ‘Guidelines’ do not oppose using native speakers
in analysis: according to Fraser, what the ‘Guidelines’ oppose is accepting the
conclusion of native speakers without appropriate evaluation.
Before moving on to the real argument, it is important that this issue be
cleared up. Let us go back to the text of the ‘Guidelines’ themselves. Guideline 7
(p. 263) explicitly states that ‘the expertise of native speakers is not the same as
the expertise of linguists’, and that ‘people without training and expertise in
linguistic analysis should not be asked for such expertise, even if they are native
speakers of the language’. The possibility of consulting native speakers is not
mentioned. The literal text of Guideline 7 does not curtail the role of the native
speaker by demanding proper evaluation, but can easily be read as dismissing
the input of native speakers altogether (see also 4.2 below).
This being the case, I stand by my statement that the ‘Guidelines’ fail to
recognise the value of native-speaker competence in LADO.
4.2 The rejection of the native speaker as analyst
As outlined above, the ‘Guidelines’ do not mention the possibility of consulting
a native speaker, or the logic of doing so. It is hard to imagine, however, that the
authors meant to disallow such consultation. They would probably leave it to
the linguist in charge of the analysis to decide whether he/she felt that such
consultation is appropriate.
I expect it to be obvious to any linguist that consulting a native speaker is
advisable whenever one is working and writing on a specic language or lan-
guage variety, and especially when one is working with potentially disputable
102
THE INTERNATIONAL JOURNAL OF SPEECH
,
LANGUAGE AND THE LAW
material. Surely we can agree that native-speaker competence cannot be
replaced by linguistic expertise, or vice versa: they are dierent skill sets.
Given the background of the ‘Guidelines’, it seems more likely that the
authors wanted to ban the native speaker from the position of
analyst
in LADO
cases, as in the ‘supervised native speaker’ approach.
The authors of the ‘Guidelines’ may picture the ‘supervised native speaker’
approach as ‘LADO conducted by a native speaker with no expertise in lin-
guistics supervised by a linguist with no expertise in the language’ (Fraser
2011:128). It is indeed hard to see how this could work. However, this is not a
tting description of the process that I refer to as the ‘supervised native speaker’
approach. What I am referring to is a method in which the supervising linguist
collects linguistic information about the language varieties from all available
sources, and the native speaker receives specic linguistic training required for
the job.
The point is that there is no valid reason to dismiss
all
approaches within
which the native speaker has the role of ‘analyst’ − pointing out the risks of
using native-speaker judgments is one thing, throwing out the baby with the
bathwater is quite another. One of my main criticisms levelled at the authors of
the ‘Guidelines’ is that they have not explored various possible ways of
including linguistic expertise in LADO. In particular, they should have explored
the possibility to provide appropriate evaluation of native-speaker judgments
through a supervising linguist.
8
4.3 The acceptance of the specialised linguist as analyst
The ‘Guidelines’ proclaim that language analysis must be done by qualied
linguists with recognised and up-to-date expertise, both in linguistics and in the
language in question, including how this language diers from neighbouring
language varieties.
In my article, I questioned the validity of the assumption that non-native
linguists can perform this task on their own. In the literature overview by
Fraser (2009), no studies testing non-native linguists’ abilities are discussed; as
far as I am aware, there is no evidence available that supports the notion that
(non-native) specialised linguists perform better than native speakers when it
comes to the determination of speaker origin or language variety. While Fraser
argues that much more information on the ‘supervised native speaker’
approach is required before it can be accepted by the scholarly community, and
perhaps rightfully so, I wonder why this should not apply to the ‘specialised
linguist’ approach.
I agree that any ‘legitimisation’ would require other data and more
CLARIFICATION OF THE ISSUES IN LANGUAGE ANALYSIS
103
information than is provided in my article. Crucially though, these data are also
lacking for the approach that is legitimised by the ‘Guidelines’.
4.4 Testing the abilities of a language analyst
Fraser (2011:127) notes that the ‘supervised native speaker’ approach relies
heavily on testing, and she suggests an analogy with language prociency tests,
such as those used to establish whether international students have sucient
competence in English to be admitted to a university.
Testing is indeed a crucial part of the selection procedure for candidate
analysts. Through testing, native speakers with the required well-developed
‘ear’ can be identied. The need for testing of individual skills was recently
corroborated in a study by Köster, Kehrein, Mastho and Boubaker (2012).
Their study shows that performance in a dialect recognition task predom-
inantly depends on the (innate or trained) skills of the individual expert, rather
than on the methodology or the procedures followed. Note that testing of
individual skills is recommendable not just for native speakers, but also for
linguists who venture into the eld of LADO
(cf. Baltisberger and Hubbuch
2010).
It must be kept in mind though that the ‘tests’ for candidate language
analysts are part of a much broader process, aimed at getting to know the
native speaker, and his/her qualities and limitations. Analysts for the
INS who
have passed the initial tests are not provided with a licence to perform analyses
by themselves. The initial tests do little more than reveal whether the candidate
analyst has enough potential to start an extensive training procedure (Cambier-
Langeveld 2010b). Native-speaker analysts will remain under the constant
supervision of a qualied linguist. The initial testing is followed by continuous
cross-checking of individual cases. Cross-checks, i.e. several analyses of the
same recording, are performed by dierent analysts from the same agency or
from other agencies.
The whole process actually has more resemblance to working with native
speakers in linguistic eldwork than it does to prociency testing. Linguists
with experience in linguistic eldwork will conrm that, as a linguist, one can
tell good consultants from less useful ones by discussing their language with
them. Supervising linguists do this initially by using (old) test cases and, after
that, continuously with actual (new) cases. An excellent description of what a
supervising linguist does was written by Dixon to describe linguistic eldwork:
One strikes up a relationship with each of a small coterie of intelligent,
reliable, interested, and willing language consultants. The consultants
104
THE INTERNATIONAL JOURNAL OF SPEECH
,
LANGUAGE AND THE LAW
will get on ‘the same wavelength’ as the linguist, understanding what the
linguist is trying to do (…). (Dixon 2010:316)
In all, we can conclude that testing is an important aspect of the supervising
linguist’s work, as one step within a more comprehensive process. Further-
more, the importance of testing individual skills and supervising an expert’s
performance is not restricted to the ‘supervised native speaker’ approach, but
also applies to the ‘specialised linguist’ approach.
5 The transparency of LADO reports
Is it surprising that the authors of the ‘Guidelines’ jumped to the conclusion
that the so-called ‘supervised native speaker’ approach was inadequate? Per-
haps not. I have to acknowledge that, at the time when the ‘Guidelines’ were
written, many LADO reports were not very satisfactory from a linguistic point
of view. Although things have improved over the years, today’s reports are still
not likely to satisfy the average academic linguist.
With respect to the desire for transparency, there is a crucial dierence here
with most other types of forensic analyses. Language analyses are not per-
formed after a crime was committed, but are done as part of the asylum
procedure. The language analysis interview and its purpose are announced
beforehand, and asylum seekers are quite aware of the nature of the research
instruments that are being employed. Asylum seekers have been known to
prepare themselves for a language analysis interview; they have been reported
to practise fake language skills with fellow countrymen, and language analysis
reports are circulated by way of study material. Asylum seekers are eager to
nd holes in the asylum policy that would leave them eligible for asylum yet
would leave the immigration service with no testing instruments. It is under
these conditions that our immigration service made it a policy to give away as
little as possible about the information that is used to check what is claimed by
asylum seekers.
None of this means that I am in favour of opaque approaches or conclusions
unsubstantiated by
linguistic evidence. What it means is that the validity of the
method cannot and should not be derived from the transparency of the reports
alone. I realise that this leaves an outsider with little to go by. I also understand
the diculty of separating good practice from poor practice, if the information
that could distinguish between them is not in the reports. This does not change
the fact that many LADO reports hold only a limited amount of information.
CLARIFICATION OF THE ISSUES IN LANGUAGE ANALYSIS
105
6 Factors leading to a polarised debate
I addressed the conciseness of LADO reports in the preceding paragraph
because it could explain the dierent views that have led to a polarised dis-
cussion. An insider’s view on the quality of analyses and the quality control can
be quite dierent from an outsider’s view, simply because they are not based on
the same information.
Another factor leading to the dierent views and a sometimes polarised
discussion is that some contributors to this discussion may have encountered
poor practice within one approach, while others have witnessed poor practice
within the other approach. The saying is that one should not throw out the
baby with the bathwater. In addition to that, one might say that we should not
judge the baby by its bathwater. I hope I made it clear in my article that I did
not intend to judge all specialised linguists by the poor practice of some.
7 Conclusion
I am not asking anyone to accept my assurance that all is in order, as Fraser
suggests. What I am calling for, is treatment on an equal footing of various
LADO methods that have the potential to lead to useful and valid results, and
testing and (re-)evaluating them all. A prociency test across agencies is pres-
ently in preparation.
Both approaches in wide use today have their risks and may be open to mis-
use, and both should incorporate linguistic expertise, native-speaker compe-
tence and a forensic awareness to best advantage. Both should be further
studied and described, and setting up minimal requirements or new guidelines
for both may be appropriate. Testing of individual skills is recommendable in
either approach.
The overall message of my article is that there is no reason to regard the
‘specialised linguist’ approach as in principle superior to the ‘supervised native
speaker’ approach, or to set requirements for one but not for the other.
The authors of the ‘Guidelines’ would do well to reconsider the basic
assumptions underlying their rejection of one approach and their legitimisation
of the other, taking into account the evidence presented in my article. Their
argument that linguistic expertise must be included in LADO remains stand-
ing.
About the author
Tina Cambier-Langeveld studied linguistics and holds a PhD in phonetics. She
worked as an expert in forensic speech science at the Netherlands Forensic
106
THE INTERNATIONAL JOURNAL OF SPEECH
,
LANGUAGE AND THE LAW
Institute (NFI) for six years. Since 2005 she has been employed by the Dutch
Immigration and Naturalisation Service, where she is in charge of language
analyses as a means of investigating the claims of asylum seekers on their
national/regional origin, ethnicity and language background.
Notes
1 The term ‘Guidelines’ is used as an abbreviation for the ‘Guidelines for the
use of language analysis in relation to questions of national origin in
refugee cases’, drawn up by the Language and National Origin Group and
published in this journal in 2004, vol. 11.2.
2 A case in point where Fraser has misinterpreted my argument is the
reference to Kim Wilson’s thesis (2009). I agree that Wilson’s research does
not allow any conclusions with respect to the debate on the linguist versus
the native speaker as the most preferable analyst. I did not mean to suggest
this, and neither did Wilson. Nevertheless, Wilson’s research does lend
support to the view that LADO may prot from the skills of native
speakers, and that we should look at more than linguistic academic
qualications. Foulkes and Wilson (2011:694) again support my view
when they conclude: ‘Our initial view is that NSs [= native speakers] may
have a valuable role to play in the process, provided, of course, that they are
proven to have good analytical skills and are supervised by competent
linguists’.
3 In Cambier-Langeveld (2010b), the tasks of the linguist are described
within a ‘supervised native speaker’ approach, and the results of 124 cases
are presented. These are the results of the only validation research to date.
These results were also presented in Cambier-Langeveld (2010c).
4 For accurate information on the methodology used by the Dutch INS, the
reader is referred to Cambier-Langeveld (2010b).
5 In asylum law, the burden of proof is in principle on the applicant (the
asylum seeker), not on the government.
6 The linguistic training advocated by the ‘Guidelines’ is not likely to solve
these problems. If it is hoped that academic linguistic training ensures
objectivity, I can say from extensive experience that this is not (always) the
case. The instilling of forensic awareness, on the other hand, is more likely
to tackle some of the problems at hand.
7 Witness, for example, the following comment from a linguist explicitly
selected according to the ‘Guidelines’ by Verrips’s company De Taalstudio:
‘If she was a Kenyan Bajuni, why would she seek refuge anywhere?’
CLARIFICATION OF THE ISSUES IN LANGUAGE ANALYSIS
107
(counter report from 2004 in the case of an asylum seeker claiming to be a
Bajuni from South Somalia, suspected to come from Kenya). And, more
recently, from the same counter expert: ‘Why would he disguise his
English?’ (counter report from 2009 in a similar case; note that English is
an established language of wider communication in Kenya, but not in
South Somalia). While an expert should be testing one hypothesis (to the
eect that the claimed origin is true) against an alternative hypothesis,
these comments suggest that this particular counter expert does not give
much thought to the existence of any alternative hypotheses.
8 Fraser (2011) suggests that I claim to have developed a ‘new process’ in
LADO, years after the ‘Guidelines’ appeared. This would explain why this
process could not be evaluated in the ‘Guidelines’. However, I do not claim
to have developed any ‘new process’. The cases showing the validity of the
analyses performed by my agency (Cambier-Langeveld 2010b), of which
the eight cases are a subset, were analysed before my arrival in this eld.
References
Baltisberger, E. and Hubbuch, P. (2010) LADO with specialised linguists – The development of
LINGUA’s working method. In K. Zwaan, M. Verrips and P. Muysken (eds.) Language and
Origin: The Role of Language in European Asylum Procedure: A Linguistic and Legal
Survey 9–19. Nijmegen: Wolf Legal Publishers.
Cambier-Langeveld, T. (2007) Hot issues in the eld of ‘Language Analysis’. IAFPA 16th
Annual Conference, Plymouth, UK, 22–25 July 2007.
Cambier-Langeveld, T. (2010a) The role of linguists and native speakers in language analysis for
the determination of speaker origin. International Journal of Speech, Language and the
Law 17(1): 67–93.
Cambier-Langeveld, T. (2010b) The validity of language analysis in the Netherlands. In
K. Zwaan, M. Verrips and P. Muysken (eds.) Language and Origin: The Role of
Language in European Asylum Procedure: A Linguistic and Legal Survey 21–33.
Nijmegen: Wolf Legal Publishers.
Cambier-Langeveld, T. (2010c) Performance of native speakers and linguists in LADO cases
with the true origin established. IAFPA 19th Annual Conference, Trier, Germany, 18–21
July 2010.
Dixon, R.M.W. (2010) Basic Linguistic Theory. Vol. 1. Methodology. Oxford: Oxford
University Press.
Eades, D. and Arends, J. (2004) Using language analysis in the determination of national origin
of asylum seekers: an introduction. International Journal of Speech, Language and the
Law 11(2): 179–199. http://dx.doi.org/10.1558/sll.2004.11.2.179
Eriksson, A. (2008) Guidelines? What guidelines? IAFPA 17th Annual Conference, Lausanne,
Switzerland, 20–23 July 2008.
108
THE INTERNATIONAL JOURNAL OF SPEECH
,
LANGUAGE AND THE LAW
Foulkes, P. and Wilson, K. (2011) Language analysis for the determination of origin: an
empirical study. In Proceedings of the 17th International Congress of Phonetic
Sciences, Hong Kong, 17–21 August 2011, 691–694.
www.icphs2011.hk/resources/OnlineProceedings/RegularSession/Foulkes/Foulkes.pdf
Fraser, H. (2009) The role of ‘educated native speakers’ in providing language analysis for the
determination of the origin of asylum seekers. International Journal of Speech, Language
and the Law 16(1): 113–138.
Fraser, H. (2011) The role of linguists and native speakers in language analysis for the
determination of speaker origin: a response to Tina Cambier-Langeveld. International
Journal of Speech, Language and the Law 18(1): 121–130.
Köster, O., Kehrein, R., Mastho, K. and Boubaker, Y.H. (2012) The tell-tale accent:
identication of regionally marked speech in German telephone conversations by forensic
phoneticians. International Journal of Speech, Language and the Law 19(1): 51–71.
Language and National Origin Group (2004) Guidelines for the use of language analysis in
relation to questions of national origin in refugee cases. International Journal of Speech,
Language and the Law 11(2): 261–266. http://dx.doi.org/10.1558/sll.2004.11.2.261
Verrips, M. (2011) LADO and the pressure to draw strong conclusions: a response to Tina
Cambier-Langeveld. International Journal of Speech, Language and the Law 18(1):
131–143.
Wilson, K. (2009) Language analysis for the determination of origin: native speakers vs. trained
linguists. MSc dissertation, University of York.
