My Masters Dissertation – A short plain-language summary of my findings
Why I did this study
Heading a Speech and Language Resource Base at a large, multilingual London school, I noticed the high number of EAL children who were on the NHS caseload for Speech and Language Impairment (SLI). Looking through case notes, I noticed the lack of contextualised background gathered on those pupils. I felt an investigation into the methods used to diagnose and norm would be useful to my setting.
I asked, what role should the L1 play in the assessment of SLI in bilingual children? What problems can arise when assessing SLI in the absence of complete ethno-linguistic information about the L1 use of the bilingual child and family?
In addition, the 2008 release of the DCSF Bercow Report had highlighted links between poverty and a higher occurrence of Speech, Language and Communication Needs (SCLN) which risked poorer educational attainment and life outcomes with attendant mental health challenges. NHS referrals to Speech and Language Therapists had doubled in the 3-year period between the report’s release and my dissertation in our borough. Considering the increased immigration to the London borough I worked in, it seemed an important investigation which directly affected our ability to allocate either quality EAL provision or SEN referrals. EAL professionals were also critiquing the inconsistent SEN referrals of EAL pupils. The accurate diagnosis and effective treatment of bilingual children with suspected Speech and Language disorders was emerging as a challenging issue.
What I did
I chose two pupils who were old enough to participate with informed consent: 19-year-old Vietnamese ‘V’ and 15-year-old Congolese ‘H’. Both had early medical diagnoses of SLI as an SEN. Sifting through more than ten years of medical notes and reports clearly showed a lack of or incorrect use of first-language assessment. This had played a key role in each diagnosis evolving from ‘delayed’ (normal trajectory but below age expectation) to ‘disordered’ (distinct from normal developmental sequence) to ‘impaired’ (cognitive difficulty associated with language processing impairment). After interviewing teachers, parents, the pupils themselves and the medical personnel for contextualising info on the medical documents, I then examined the assessment battery itself and the normed populations on which they derived construct validity. Finally, I looked at theories of language development, acquisition, dominance and attrition as it applied to both monolingual and bilingual populations.
What I Found
Although NHS SALTs assessing in multi-ethnic London were aware of techniques like first-language assessments, they did not appear to have in-depth training on key issues which could skew results and produce a ‘false positive’. Obtaining complete ethno-linguistic background, training on second-language acquisition norms and phenomena like language attrition and code-switching might completely change the interpretation of assessments, profoundly affecting the diagnostic process. Trauma and its effect on refugee families and children would also be key when deciding whether a child had a language delay, disorder or impairment.
In H’s case, the parents spoke a dialect mix of French and Lingala, but told the SALT that French was the main language. H was tested in continental French to which he did not respond. He also spoke back in his family dialect which the SALT didn’t understand. Hence, the conclusion was that H had disordered receptive ability in his ‘home language’ and that he spoke ‘jargon’, as it was reported.
H’s mother when interviewed expressed the idea that speaking both Lingala and French with H had ‘confused’ him and had contributed to his ‘impairment’. It is likely that a SALT untrained in the cognitive aspects of bilingualism may have given her this impression. H’s mother reacted by changing her language interactions with H. As he grew and began to experience language attrition, where the first language atrophies, his diagnosis changed to reflect what was then assessed as a progressive language processing difficulty in his home language.
In V’s case, the family spoke through an interpreter, reporting his lack of Vietnamese use at home and limited vocabulary. But little else on his background was shared. The family’s deprived and traumatic refugee background was not discussed with outside professionals nor was his family’s use of Cantonese. It may have changed the whole view of his need.
V was born in a refugee camp for Vietnamese ‘boat people’: part of the post-Vietnam War ethnic cleansing of Chinese-heritage Vietnamese. He did not crawl or become toilet trained until well beyond those early milestones. This was noted as a developmental delay. The professionals assessing could not have known that V would not have been able to do so in the tiny space allotted each refugee family. The SALT noted he knew the word ‘rat’ but could not name primary colours, something which speaks volumes about his early deprivation. He also didn’t speak until much later, but, dyadic, language-building interaction between mother and child assumes the mother is not traumatised and able to attend.
V also began to experience language attrition of Vietnamese (or Cantonese). His actual home language was never determined; he reportedly did not respond to Vietnamese interpreters. This, as in H’s case, was interpreted as a progressive language processing difficulty in his ‘home language’. His mother’s mental health was never considered as a contributing factor.
Finally, the assessments used did not take into account the grammatical features and markers of the assumed home languages nor of their dialects. For example, the lack of verb tenses in Vietnamese or tones in Cantonese were not accounted for in assessments meant to isolate and measure ‘normal’ syntactic items of English language. Using an interpreter did not help because certain English language features have no equivalent in many languages. The experiences of refugees make them an exception in another way. The assessment of ‘normal’ age-appropriate vocabulary items assumes the universal experience of children in a home curriculum that is echoed in and recognised by the national educational system. If, as in a refugee camp, linguistic input from traumatised family is depressed and the environment is non-stimulating, then conceptual building blocks for language acquisition remain weak especially if it occurs in the child’s second year.
What it Means
Does all this mean that H and V did not have a Speech and Language Impairment as diagnosed? Possibly. They never received EAL input which would have helped them build proficiency. They were instead given therapeutic inputs which did not address their second-language learner status. Oddly, once they were on the NHS caseload, their ethnic and linguistic identities were uncoupled: yes, they were ethnic minorities, but, no, they weren’t EAL.
Educating or diagnosing EAL and refugee pupils requires specific knowledge about how language is acquired, develops and is sometimes eroded. Critically, aspects of normal second-language acquisition mimic disorder or impairment in monolingual pupils. This is a key training need for Speech and Language specialists and EAL specialists practicing within diverse, urban areas in the UK. Complete ethno-linguistic background is crucial to deciding upon either a pedagogical or diagnostic course of input. This lack of information clearly led to a minor diagnosis of language delay—something easily remedied by extra classroom input—becoming an SEN with attached negative assumptions about cognitive ability and not providing the EAL input that might have helped.