Researching Dyslexia in Multilingual Settings

Diverse Perspectives

Deirdre Martin, Deirdre Martin

• Publisher: Multilingual Matters
• Year: 2013
• ISBN:
  9781783090679
• Topic: Médecine
• Subtopic: Audiologie et orthophonie

This volume draws together current research on dyslexia and literacy in multilingual settings across disciplines and methodologies. The contributors, all internationally recognised in the field, address developmental and acquired literacy difficulties and dyslexia in a range of language contexts including EAL/EFL. The book uses theories and analytical frameworks of a critical nature to reveal prejudicial social practices, and suggests future research directions towards a critical re-consideration of current understandings of dyslexia in multilingual settings, with a view to foregrounding the potential for interdisciplinarity. The book also suggests ways forward for evidence-informed practice, and it will be a valuable resource for researchers, practitioners and students alike.

1 Assessment Approaches for multilingual learners With dyslexia

John Everatt, Gavin Reid and Gad Elbeheri

Introduction

Cross-linguistic studies of the diverse language/cultural societies in all parts of the world suggest that the manifestation of dyslexia may vary across languages. This indicates strongly a need for more dyslexia-focused research and recommendations for best practice targeted at individuals from multilingual backgrounds. One of the aims of this chapter is to review briefly current work as it relates specifically to dyslexia, in order that the reader understands some of the challenges for practice and theory in working with children learning several languages. An assumption underlying this review is that appropriate (and early) assessment procedures and tools designed to identify potential areas of difficulties and strengths are essential to practitioners, in both their initial identification of those at risk of learning difficulties and their formation of an appropriate intervention designed for the needs of the individual.

Early identification leads to more effective intervention, particularly in the areas of reading and writing (see Torgesen, 2005), and a failure to recognise difficulties can often lead to the child becoming anxious or depressed and losing self-esteem, confidence and motivation (see Elbeheri et al., 2009; Everatt & Reid, 2010; Miles, 2004). Despite the availability of screening and assessment procedures that have been developed to aid this process in, primarily, monolingual English-speaking contexts, there is a lack of work that can be used to inform best practice within a multilingual context (see Cline & Shamsi, 2000). Two of the main obstacles to this have been the lack of procedures and tools (and suitably trained testers) to assess across different language contexts, and the potential for the effects on literacy development of learning a second language to obscure the identification of dyslexia (see Cline & Reason, 1993; Peer & Reid, 2000). The following pages provide a discussion of these obstacles and the factors that may need to be considered to improve practice, based on our current research understanding.

Dyslexia in Multilingual/Multicultural Societies

One of the potential consequences of the lack of research, and a key issue that needs be raised in relation to dyslexia and multilingualism, is that the prevalence of dyslexia identified among children from dual (or multi) language backgrounds is lower than expected (based on general population samples). For example, Curnyn et al. (1991) found that bilingual learners were significantly under-represented among pupils who were assessed as having dyslexia. Similar research by Landon et al., (2000) investigated 144 Scottish primary schools and confirmed the results of the earlier Curnyn study. The Scottish research revealed that parents from ethnic minority communities are frequently excluded from involvement in the assessment process because of poor provision of interpreters (see also Diniz, 1997; Shah, 1995). The Landon et al. study also revealed a great deal of confusion among teachers about the assessment and support of bilingual learners suspected of being dyslexic. Such work has indicated that, when multilingual learners fail to make progress in the curriculum, those conducting assessments and classroom interventions should not assume that low language proficiency is necessarily the problem. There has been a trend for specialist teachers and psychologists to misdiagnose or ignore dyslexia indicators in multilingual students because of the multiplicity of possible causes for failure to make progress and the risk of a ‘false positive’ result (see discussions in Peer & Reid, 2000). Lack of suitable test materials, cultural perceptions of dyslexia, policies for identification and classroom practices all affect the nature of the support for bilingual learners suspected of being, or diagnosed as being, dyslexic.

Berryman and Wearmouth (2009) argue that dyslexia is commonly understood as an explanation of difficulties in literacy from a cognitive perspective (i.e. as a set of factors within the brain/individual). This level of explanation, they argue, ignores the impact of culture. They suggest that research shows the benefits to literacy learning that accrue when schools work to address the cultural backgrounds of the children they are teaching. Such views are supported by theorists such as McNaughton (1995) who discuss socialisation values that match home culture. This is relevant across the modern world, as virtually every country has individuals who derive from a range of different cultural backgrounds. Dyslexia, therefore, should be seen from a broader, more culturally aware perspective. Both assessment and intervention practices need to take account of these potential differences (see also Mcfarlane et al., 2000).

In a report by a working party of the Division of Educational and Child Psychology of the British Psychological Society (BPS) on Dyslexia, Literacy and Psychological Assessment (BPS, 1999), it was argued that an emphasis must be placed on culture-fair assessment, and that indicators such as phonological difficulties and letter-naming speed in the language of tuition should be the focus of a diagnosis. Although culture-fair assessment is crucial (Cline, 1998), analysis of test performance within specific cultural and linguistic groups can itself also help to identify those items which consistently lead to cultural confusion or misperception. Usmani (2000) suggests that the bilingual/bicultural child may have a broad range of thinking skills that can go undetected if the professional is unaware of associated cultural values or fails to understand them in relation to the assessment. Usmani further suggests that the ‘big dip’ in the performance of some bilingual children noted in later primary education may be explained by a failure of professionals to understand and appreciate these cultural values and the actual level of competence of the bilingual child in relation to conceptual development and competence in thinking skills. As suggested by Palincsar and Brown (1984), better thinking skills should lead to improved comprehension in readers who are struggling with acquiring literacy.

Although assessments need to be culturally appropriate, another reason for the under-representation of dyslexia in multilingual groups may be the view that such learners have processing difficulties that are due to poor language proficiency rather than an educationally based learning difficulty, such as dyslexia (see Dal, 2008). Hence, appropriate assessments need to take into account any language factors that may be related to alternative causes of reading problems.

Language, phonology and orthography

Well designed assessment procedures are particularly important where an observed behavioural problem may be due to a number of underlying reasons, and optimal effective intervention may depend on distinguishing between them. A problem with learning to read English could be due to an underlying English-language processing deficit (e.g. a phonological weakness – the dominant causal viewpoint in research on dyslexia: see Snowling, 2000), poor educational experience (e.g. lack of appropriate schooling) or incomplete acquisition of English as a second language (the individual may be new to learning the English language). Each of these may require specific interventions, such as a phonologically based intervention, an increase in school experience or greater language experience. Although good educational practice (pedagogy) should support learning in all cases, an intensive phonological intervention that supports literacy acquisition among dyslexic students may be inappropriate for English-language learners who have appropriately developed phonological awareness skills.

In most cases, an assessment procedure targeted at identifying literacy learning difficulties would include determination of the child’s (or adult’s) language skills, given that the written form is used to represent language and those with weak language would be expected to struggle with literacy learning. If the individual has difficulty understanding language, then comprehending or producing meaningful written text may also suffer. In addition, phonological skills have been considered to support the translation of writing into a form that can be used by already developed language processes. Such phonological skills refer to those used to process basic sounds within the language, that is, to recognise that /dog/ begins with a /d/ sound. There is a large body of research that is consistent with phonological processing skills being a vital component in effective literacy learning and with a phonological deficit being related to the literacy problems faced by children with dyslexia (e.g. Gillon, 2004; Snowling, 2000; Stanovich, 1988). Poor scores on measures of phonological processing are good predictors of early problems with literacy learning (e.g. Puolakanaho et al., 2008) and are associated with literacy weaknesses, and dyslexia, throughout development and into adulthood (Beaton et al., 1997).

Data relating literacy difficulties to phonological deficits have mainly been derived from studies of English-speaking individuals. However, relationships between phonological skills and literacy learning have been found across a number of languages (e.g. Smythe et al., 2008; Ziegler et al., 2010). Despite this, there is still a need for a phonological deficit to be confirmed as the main factor that can be used to identify dyslexia across different orthographies (see discussions in Smythe et al., 2004). One of the reasons for caution here is that children learning a script (an orthography in the present context) with a more consistent relationship between written symbols (letters/graphemes) and language sounds (phonemes) than that found in English seem to show faster progress in literacy, and process language at the level of the phoneme earlier, than those learning a less regular orthography, such as English (see Goswami, 2000; Seymour et al., 2003).

The view that English is less transparent (i.e. the relation between written form and language sounds is less clear in English) than most other orthographies has led some to question the universality of current theories about literacy learning. Basically, the argument is that most theories about literacy have been derived from research on English, which is an atypical orthography in that most other orthographies are more transparent than English (see discussions in Share, 2008). Therefore, research is needed on more typical orthographies to confirm the findings derived from research on English before we can consider them generalisable to other languages.

Additionally, compared with children learning English literacy skills, phonological awareness deficits may create fewer problems when learning a regular orthography since the simple rules or correspondences between letters and sounds will be less tasking of a weak phonological system. That is, even if a phonological deficit does lead to literacy learning problems, these may be less severe when learning a more transparent orthography. As such, the transparency of the script is a factor that may lead to variations between languages in the ease of literacy acquisition, the manifestation of literacy deficits and the appropriateness of particular assessment procedures. For example, whereas assessments of accuracy in word reading tasks seem appropriate to assess literacy skills in English, such tools may be less effective in more transparent orthographies, where, instead, literacy assessment procedures might more appropriately focus on measuring fluency, or speed, of processing (see Smythe et al., 2004). Determining the speed at which a child can produce a certain level of accuracy (fluency of performance) has been found to identify those children who may be able to achieve reasonable levels of accuracy, but whose slow speeds are indicative of finding word processing effortful. These children may struggle in literacy classes and when text understanding or production is required, and also may show poor acquisition of literacy-related language skills (see also Everatt et al., 2010).

Dyslexia and Biliteracy

Research on children learning to read and write in two orthographies has confirmed some of these concerns regarding orthographic transparency. For example, in research on bilingual children in the Philippines and Namibia, Everatt et al. (2010) found that word reading and non-word decoding developed at different rates for Filipino versus English and for Herero versus English, respectively. Both Filipino and Herero have relatively simple relationships between graphemes and phonemes, making it fairly easy to learn conversion rules for decoding. As might be expected, these more transparent orthographies showed good learning rates for word decoding and, hence, most of the bilingual children studied acquired good levels of accuracy in word-level literacy tasks in Filipino or Herero. Yet, despite these clear differences in acquisition, phonological skills were still the main predictor of a child’s literacy learning in both languages. These findings suggest that literacy learning is dependent on phonological skills irrespective of orthographic transparency. Indeed, in the Filipino/English data, equivalent amounts of the variability in reading among the grade 1 children tested could be predicted by phonological processing measures of decoding, phonological awareness, rapid naming and auditory short-term memory. Furthermore, research by Everatt et al. (2002) has suggested that the assessment of underlying phonological skills affords the potential to distinguish individuals with dyslexia from those who are reading in an additional language, despite equally poor literacy skills being presented by both groups. Therefore, assessments of phonological skills may be a useful tool to identify dyslexia in many language/literacy contexts.

These findings may be interpreted as meaning that an assessment in English can replace an assessment in the other language of the bilingual learner, since the English assessment measures are highly predictive of skills in the other language. However, again, these conclusions need to be treated with caution. Data from the Filipino/English work suggest that measures of Filipino phonological awareness were better at identifying those with poor English literacy skills than tests of English phonological awareness skills. When Filipino and English measures were combined in a bilingual assessment, profiles of difficulties better mirrored profiles typically found with assessments of monolingual English children; that is, there was evidence of deficits in most areas of phonological processing (see Everatt et al., 2010).

Hence, there seems to be a complex relationship between phonologically based language skills and literacy learning. This may not be surprising given the evidence for a reciprocal relationship between language and literacy (Perfetti et al., 1987). The development of phonemic awareness (i.e. recognising sounds at the level of the phoneme) comes with increased linguistic experience and formal reading instruction. Awareness of sounds at the level of the phoneme seems to be a consequence of literacy learning as much as it is a predictor of future literacy levels. Such a relationship may well cross languages in bilingual and/or second language development and go beyond phonologically based language skills. The Filipino/English bilingual data show how phonologically based language skills can be a good predictor of literacy in another language. Similarly, in the Herero/English data, second (English) language measures of phonological awareness predicted more variability in both first (Herero) and second (English) language literacy than first (Herero) language phonological awareness measures. In a longitudinal study by Sparks et al. (2009), language 1 literacy skills in elementary school predicted language 2 language levels in later schooling and adulthood, consistent with literacy skills in one language influencing language competence in another.

Greater exposure to print can lead to improvements in vocabulary, general knowledge and other skills assessed in commonly used language tests. Vocabulary size may be increased by literacy experience because rare words are more likely to be encountered in print than in oral language (Nagy & Anderson, 1984). With increased reading, an individual’s general knowledge also should improve, leading to better performance in verbal ability tests. The average person will encounter more complex syntactic structures in print than they would in speech, which again should lead to improved development of language skills. Exposure to different forms of language (including the written form) may also serve to increase metalinguistic skills (see Koda, 2007), that is, to increase the individual’s ability to understand language in a non-language-specific sense.

Such cross-language effects will, most likely, interact with environmental (educational) factors experienced by the child. Learning the relationship between letters and sounds in a more transparent orthography may make decoding processes sufficiently well practised for them to support literacy learning even when it is less effective in a lower-transparency second orthography. In the context studied in the Philippines, where sound–symbol relationships were explicitly taught as part of literacy learning, there were relatively few children with poor English-language decoding skills. In contrast, the Namibian school system did not focus on the relationship between graphemes and phonemes, even when learning Herero, and many children showed poor levels of English word reading and spelling. Under the right circumstances, learning of a second language may strengthen...