What research can tell us about the interaction between dyslexia and bilingualism : an integrative review

Received on: Jun. 26th, 2020. Accepted on: Jan. 08th, 2021. Abstract: Dyslexia is a reading difficulty of neurological basis which is often associated to a deficit at the phonological level of reading – but not restricted to it. In this review, we sought to identify relevant research conducted on the interaction between dyslexia and bilingualism in the last twenty years. For this purpose, eleven studies were selected from CAPES Journals Portal. Results of these studies reached contrasting conclusions, some indicating a continuum of dyslexic traces across the languages, while others emphasized the role of interacting factors such as orthographic depth and language granularity modulating the manifestations of dyslexia in each language. The implications of these findings for education are also discussed.

Resumen: La Dislexia es una dificultad en la lectura que presenta orígenes neurológicos, comúnmente asociada a una discapacidad en el nivel fonológico de la lectura -pero no limitada a eso. En esta revisión, buscamos identificar investigaciones relevantes a lo que toca la interacción entre la dislexia y el bilingüismo en los últimos veinte años. Para alcanzarlo, fueron seleccionados once estudios en el Portal de Periódicos de la Capes. Los resultados presentan conclusiones contrastantes en las cuales algunas enseñan un continuo de huellas de la dislexia en las dos lenguas mientras otras enfatizan el papel de factores relacionados, como niveles de profundidad de las ortografías y la granularidad de las lenguas disminuyendo la manifestación de la dislexia en cada lengua. Las implicaciones de estos hallazgos para la enseñanza y el aprendizaje también son discutidas. One topic that has gathered considerable attention of both Brazilian and worldwide researchers is reading difficulties 2 , and in this paper we focus on one specific difficulty namely dyslexia. Simply put, dyslexia is a reading and writing difficulty of neurobiological origin, as Kuerten, Mota and Segaert (2019)  With these issues in mind, we felt the need to carry out a more thorough investigation of the relationship (if any) of bilingualism and dyslexia.

Introduction
In this paper, we seek to review recent studies that investigated a possible interaction between dyslexia and bilingualism. More importantly, we focus on the implications of these findings for teaching and learning to read, especially in the context of language acquisition. This paper is structured in the following manner: first, we bring a brief theoretical background where we define dyslexia and bring four main theories that attempt to explain its causes and describe its characteristics. Second, we describe the methodological procedures for selecting the articles that compose our review. Third, we summarize the studies on dyslexia and bilingualism.
Last, we sum up the main findings attempting to draw some major considerations on how these findings might contribute to teaching and learning.

A brief theoretical background
The field of reading difficulties research has been divided into two domains: a) difficulties at the decoding level and b) problems in the construction of meaning, at the comprehension level (HULME; SNOWLING, 2014; KRONBICHLER; KRONBICHLER, 2018). Difficulty with decoding seems to stem from phonological deficits (e.g. spelling), since "reading is parasitic on speech" (MATTINGLY, 1972, p. 133); that means, there is a strong relationship between learning to read and being able to associate sounds and letters; in other words, difficulty in the grapheme-phoneme correspondence. In contrast, comprehension difficulties are associated with vocabulary knowledge, morphological and pragmatic skills.
Dyslexic reading is situated within the first realm.
In fact, this view of dyslexia is commonly the most cited one, but as we shall depict along this paper, not all types of dyslexia have to do with phonological processing (RAMUS, 2003;KUERTEN;SEGAERT, 2019 Under a neurocognitive perspective, the magnocellular theory defends that dyslexia is associated with abnormalities in the neural visual system, more specifically, in the magnocellular pathway, which rapidly transmits visual input to the occipital and parietal brain regions. Neuroimaging studies indicate that magnocells are smaller in dyslexics, causing a decrease in visual and auditory sensitivity (LOVEGROVE et al., 1980;LIVINGSTONE et al., 1991;SKOTTUN, 2000;STEIN;TALCOTT;WALSH, 2000;STEIN, 2001).
The last theory brought by Kuerten, Mota and Segaert (2019) assumes that the magnocellular system carries information to the cerebellum; thus, abnormalities in this system also affect that part of the brain. This is the cerebellar deficit theory (NICOLSON; FAWCETT, 2008), according to which a dysfunction in the cerebellum has been found to be linked with dyslexic reading.
Since decoding problems are a salient characteristic of dyslexia, orthographic depth has been referred to as a linguistic factor associated to this condition. In a nutshell, the concept refers to how reliable print-to-speech correspondences are (SCHMALZ et al., 2015).
For instance, in consistent orthographies, also known as transparent or shallow orthographies, the grapheme-phoneme correspondence is clear, such as in Italian and Spanish. Differently, in inconsistent orthographies, also known as opaque or deep orthographies, the correspondence between graphemes-phonemes is less straightforward. In English, for example, the same grapheme ch is pronounced differently in "chocolate", "choir" and "chef" (CHUNG; HO, 2010; TAINTURIER; ROBERTS; LEEK, 2011).
Attempts have been made to explain how cross-linguistic influences such as orthographic depth interact with dyslexic reading. In their Psycholinguistic Grain Size Theory, Ziegler and Goswami (2005) proposed that reading in more More recently, studies on dyslexia have investigated its interaction with the bilingual mind (CHUNG; HO, 2010;LALLIER et al., 2018), given the assumption underlying several studies is that there might be a bilingual advantage for dyslexics. For instance, Lallier and colleagues (2018) proposed that the reading difficulties posed by dyslexia should be ameliorated due to the bilingual experience, which was further investigated in Lallier and collaborators (2018).
In order to achieve the goal of tracing the current state of studies in this field, we carried out an online research in Periódicos Capes (www.periodicos. capes.gov.br) from late May to early July, 2019.
This website is a Brazilian online journal portal which makes available more than 45 thousand national and international articles from over 21.500 journals. The relevance of this database relies on its coverage, encompassing studies from a wide range of countries and areas. The former aspect is important since the present review includes works both in the areas of Psycholinguistics and education. Yet, a more thorough search, including different databases, is advised in future studies to provide a more comprehensive picture of the field. Finally, it is also important to mention that this paper was initially written as a requisite for the Graduate course "The implementation of reading in the brain" offered in 2019 at the Graduate Program in English (PPGI-UFSC).
The search was conducted under the searchstrings "dyslexia" and "bilingualism", both only in the English language. As for a temporal criterium, only works published within the last twenty years were selected. Among the results obtained, the first eleven articles that appeared in the search engine were selected to form the scope of the present analysis. Default screen position was used as criteria because, in this website, results are ordered by relevance.
Among these, one article deals with brain data,  and Butterworth (1999); and Wydell and Kondo (2003). Interestingly, the majority of the studies are very recent, a fact which highlights the current interest in this topic. The fact that only one neuroimaging study was found also points to lack of evidence on the interaction between dyslexia and bilingualism -an issue we discuss in the concluding remarks, extending to its relationship with education. Finally, in order to analyze the data from these studies, we bring supporting literature from likely recent Brazilian and foreign publications.

What research can tell us about dyslexia and bilingualism
As aforestated, we carried out a search at the most reliable scientific database available in Brazil in order to investigate what research has shown, in the last twenty years, in terms of the relationship between dyslexia and bilingualism.
The studies reviewed are summarized in Table 1. We first approach the work of Park and colleagues (2012) which deals with brain data, and then proceed to the behavioral studies in chronological order. In addition, the studies conducted by the same research group are presented in sequence in order to facilitate the observation of their progress.
Park, Badzakova-Trajkov and Waldie (2012) investigated the neural bases of dyslexia and bilingualism using a widely known neuroimaging technique known as Functional Magnetic Resonance Imaging (hereafter fMRI) which consists of measuring the oxygenation of specific brain areas, taken that "neural activity leads to changes in the amount of oxygen nearby in the brain" (WILLEMS; CRISTIA, 2018, p.266). According to the authors, there is plenty of evidence that language processing in bilinguals is more widely spread in brain in relation to monolinguals. Similarly, dyslexics also have more spread activation in the brain when compared to typical readers.
Such findings have recently led researchers to investigate the brain of dyslexic bilinguals, as shall be discussed as it follows.
As stated, Park, Badzakova-Trajkov and Waldie investigated the neural basis of dyslexia, which A few years later, Wydell and Kondo (2003) further analyzed AS's case, by replicating tests with AS in order to check for the persistency of phonological impairment. It is important to highlight that during this study, AS had been an undergraduate student in Science in an English-speaking country, and he had been taking remedial classes in reading and writing in English a year before his application and throughout his first year in the course. In spite of these remedial lessons, Wydell and Butterworth (1999) suspected that his phonological ability remained faulty.
For this experiment, eight control participants aged 20 (same as AS) at the time underwent Orthographic/Phonological lexical decision tasks and the Spoonerism Test. For the first task, the words were extracted from textbooks AS had used in junior-high school, in order to guarantee that he had had previous exposure to stimuli. The second tasks were the same used in Wydell and Butterworth (1999). Results confirmed that AS was a phonological dyslexic, since the types of errors made enabled researchers to identify the use of larger-grain strategies. The fact that this deficit does not affect Japanese was justified by the fact that this language cannot be decomposed phonemically (like English) neither in Kanji nor in Kana. In addition, this study provided evidence for the persistence of dyslexia over time. Despite the authors' claim that "AS has a core phonological deficit which led to his dyslexia", it is interesting to note that "AS had successfully completed a BSc in science in an English-speaking country" (WYDELL; KONDO, 2003, p. 43), which suggests that despite this reading difficulty, with the appropriate intervention people are able to function as a reader in society. (2010)  and that these skills could be transferred across two languages" (CHUNG; HO, 2010, p. 199).

Chung and Ho
Eighty-four primary school students who spoke Cantonese as a medium language and Chinese as a primary language took part in the study.
The participants were split into three groups: dyslexics (mean age 9.9 years old), control (chronological age -mean age 9.9 years old) and control (reading level -mean age 7.8 years old). In order to test their hypotheses, several tasks were To be more precise, the study investigated the extent to which phonological processing and decoding skills would differ between dyslexic and non-dyslexic bilinguals. The former group was composed of ten Spanish-Swedish users (mean age 14.9) identified by their teachers with reading and writing problems, while the latter was composed of ten Spanish-Swedish bilinguals with no reading or writing problems (mean age 14.5). The battery of tests included a non-word repetition task to test phonological memory; a digit-span task to measure phonological shortterm memory; a spoonerism task to measure metalinguistic skills; a rapid naming task (since slow naming is indicative of dyslexia); and a phonological mobilization task to measure meta-phonological processing (lexical retrieval).
Hedman (2012)  One important aspect to consider in Hedman's study is that the author briefly mentions that participants read better in Spanish than in Swedish -differences that might be due to orthography.
In concluding, the researcher highlights that the bilingual dyslexia continuum can be used as a tool for assessment of the degrees of reading difficulty regarding, comprising phonological processing and decoding in L1 and L2. In addition, the comparison between participants' reading skills in L1 and L2 demonstrated that L1 reading performance is a strong predictor of L2 reading. To be more precise, the authors suggested that "adolescents with dyslexia had not only a literacy impairment in Dutch, but generally also in ESL" (VAN SETTEN et al., 2017, p. 15). Interestingly, the fact that the English orthography was not consistent did not impair L2 reading, spelling, nor Evidence for a bilingual advantage in dyslexia was found by Vender and colleagues (2018). The researchers assessed participants' ability to generate plural noun inflexions of nonwords -a tool known as the Wug Test, designed by Berko (1958 The tasks with determiners showed that dyslexic readers performed more poorly in the production of plural articles of nonwords. As for the plural inflection of invented nouns, bilinguals, including dyslexics, performed better than the two groups of monolinguals. From these findings, the authors concluded that bilingual dyslexics seem to further develop their morphological and metalinguistic abilities compared to monolingual dyslexics, even surpassing monolingual children with no reading difficulties in some conditions. Prior to the discussion of the results, we would like to draw attention to the methods used across the studies reviewed. Non-word reading tasks were used to assess phonological awareness, departing from the knowledge that familiar words are read via lexical decoding, while reading words that are uncommon or unknown forces phonological decoding. The studies conducted by Butterworth (1999), Joshi, Padakannaya andNishanimath (2010), Tanturier, Roberts andLeek (2011), Park, Badzakova-Trajkov andWaldie (2012), Hedman (2012), andVender et al. (2018) used such method to assess phonological awareness, which is believed to be a central problem in dyslexia. Reading aloud was used by Chung and Ho (2010), Tainturier, Roberts and Leek (2011), Wydell and Kondo (2003), Hedman (2012), Valdois et al. (2014. This method enables access to the reader's long-term memory representation of the spoken words, which allow researchers to infer the availability of the lexical representation (TAINTURIER; ROBERTS; LEEK, 2011). Another recurrent tool was spelling for dictation (JOSHI;PADAKANNAYA;NISHANIMATH, 2010;LALLIER et al., 2018). Lallier et al. (2018), Wydell and Butterworth (1999) and Wydell and Kondo (2003) used the Spoonerisms task, in which the initial phonemes of a pair of words are changed, e.g. car park -par cark. The aforementioned task focuses on testing the phonological component of dyslexic readers, since there is evidence for its impairment. Under a different perspective, the VA span test was used by Lallier et al. (2018) and Valdois et al. (2014) in order to investigate the role of visual attention impairment in dyslexia.
Next, we move to the final remarks of this review.

Conclusion
The present review aimed at tracing the most recent and relevant findings in the field of dyslexia and its interface with bilingualism. In order to do so, eleven articles were selected from Capes journals portal and reviewed in order to identify the main conclusions and underlying theories of each study, which will be brought together to discussion in this section.
In terms of neuroscientific findings, there is still plenty of room to explore how dyslexia and bilingualism interact. Park, Badzakova-Trajkov and Waldie (2012) showed that dyslexic bilinguals resort to alternative cortical resources in the right hemisphere due to their atypical reading, which demonstrates that dyslexia has indeed a neurobiological basis. This finding is relevant in the sense that it helps to deconstruct some commonsense beliefs in attributing dyslexia to low intellectual abilities, poor schooling, poor family structure or even laziness (SHAYWITZ, 2003;PINHEIRO;SCLIAR-CABRAL, 2018). In fact, Pinheiro and Scliar-Cabral (2018)  perceptions and evaluations (HEDMAN, 2012) are necessary steps towards the connection between these two realms, as we further argue in the end of this section.
The discussion on whether there was a bilingual advantage permeated the studies. This claim was supported in some studies with dyslexics (LALLIER et al., 2018;VENDER et al., 2018). In addition to that, these authors put that "bilingual individuals can adapt their resources to the orthographic properties of the target language", which seems to imply that the bilingual brain has some advantage in relation to monolinguals which might be persistent in atypical readers. In fact, the debate on whether there is such bilingual advantage is far from an end. A clear example is a very recent study by Waldie and colleagues (2020) (2011) also found that orthographic depth did not play a prominent role in the reading ability of English-Welsh dyslexic bilinguals, since their performance was similar in the two languages investigated.
Interestingly, Van Setten et al. (2017) found that orthographic consistency did not affect L2 reading in the case of a consistent L1 but inconsistent L2.
Notwithstanding, more studies endorsed the claim of orthographic consistency as an influential factor. Chung and Ho (2010) found evidence of cross-linguistic transfer of reading difficulties from L1 (Chinese) to L2 (English) dyslexic readers, which might be explained with basis on the difference in psycholinguistic units across the two languages.
Yet, reading problems related to phonological processing were greater in English than in Chinese. Hedman (2012) also acknowledged the fact that orthographic depth had played a role in the performance of her participants, since they scored better in reading in Spanish. Recently, Lallier et al. (2018) found evidence for transfer of reading skills from more consistent to less consistent orthographies, a factor which seems to modulate the manifestations of dyslexia.
However, and as stated by Ziegler and Goswami (2005), it would be too assertive to say that consistency and granularity alone explain differences in bilingual atypical reading.
Other put, in all languages, reading relies on phonological processes -including the ones with clear grapheme-phoneme correspondence.
Thus, more research is needed in order to clarify the extent to which orthographic depth is an influential factor in bilingual atypical reading. ROBERTS; LEEK, 2O11; VAN SETTEN et al., 2017) and the fact that it is modulated by language consistency (VALDOIS et al., 2014;LALLIER et al., 2018) has its implications for education. Language teachers need to be aware of the differences in orthographic depth between the learners' L1 and L2, especially in the case of dyslexic readers.
Importantly, and as stated by some researchers, dyslexic bilinguals need guidance to develop metacognitive awareness, making use of strategies to cope with their difficulties in reading.
This review scrutinized the current rationale underlying studies in dyslexia, supporting the claim for its phonological component, the role of the visual attention span, its neurological underpinnings and its relation with bilingualism.
As stated by Joshi, Padakannaya and Nishanimath (2010), this reading difficulty seems to "cut across linguistic boundaries", affecting performance in languages of different consistency and granularity -variables which notwithstanding moderate the manifestations of dyslexia. Still, it remains a complex phenomenon which calls for further discussion and research, especially regarding bilingual dyslexics and how the two -or more -languages known interact and improve or impair reading ability. In addition to that, the implications of these studies for education were explored by few studies (WYDELL; BUTTERWORTH, 1999, WYDELL;KONDO, 2003;CHUNG;HO, 2010;HEDMAN, 2012). We address this issue in the next lines.