Dyslexia and Research

From Reid, G. (ed.) TheRoutledge Companion Book of Dyslexia, Routledge Publications 2009


There have been significant advances in research in dyslexia over the last twenty years. This has aided explanations of dyslexia and supported policy and practice. The impact has been considerable, but yet there is still no clear explanation that is universally accepted of what exactly constitutes dyslexia. Identification is still riddled with controversies despite the emergence of a number of new tests to identify dyslexia, or sub-components of dyslexia. Indeed, there is still an ongoing debate on the value of dyslexia as an identifiable syndrome.

Neurobiological Factors

The advances in MRI and other forms of brain imagery have been of great benefit to neuroscientists investigating factors relating to dyslexia. From these studies a number of different factors have emerged focusing on structural and functional brain-related factors. Some of these will be discussed here.

Procedural timing

The lateral zone of the cerebellum is an area that has generated interest over its apparent role in cognitive processes and for recent theoretical positions that argue for its relationship with dyslexia. In terms of its formation, the cerebellum is one of the first brain structures to begin to differentiate, yet it is one of the last to achieve maturity as the cellular organization of the cerebellum continues to change for many months after birth. According to Fawcett and Nicolson (2008) there is now extensive evidence that the cerebellum is a brain structure particularly susceptible to insult in the case of premature birth, and that such insults can lead to a range of motor, language and cognitive problems subsequently. Fawcett and Nicolson (2008) argue that the cerebellar deficit hypothesis may provide close to a single coherent explanation of the three criterial difficulties in dyslexia – reading, writing and spelling. They argue this can place dyslexia research within a meaningful context in terms of the cognitive neuroscience of learning while maintaining its position as a key educational issue. They also suggest the cerebellur deficit hypothesis provides an explanation for the overlapping factors between dyslexia and the other developmental disorders (an area that will be returned to later in this chapter).

One of the hypothesised functions of the cerebellum is in the precise timing of procedures (e.g., several motor movements) that accomplish some sort of behavioural response or task performance. This timing of sequences may play a critical role in making task accomplishment or behavioural skills automatic. Indeed, a critical aspect of learning a skill may be to make its accomplishment automatic. This means that the skill can be carried out without the individual giving it too much thought – and resources can be used to undertake other behaviours or processes. For most adults and children, the ability to walk, talk and possibly read and write may be partially or completely automatic. Consequently Fawcett and Nicolson (2008) put forward the hypothesis that dyslexic children would have difficulty in automatising any skill (cognitive or motor). They suggest that reading is subject to automaticity and since all dyslexia hypotheses predict poor reading as a factor in dyslexia then the automatisation deficit hypothesis would be valid in relation to dyslexia. Fawcett (1989) and Fawcett and Nicolson (1992) argue there is clear support stemming from a set of experiments in which they asked dyslexic children to do two things at once. If a skill is automatic, then the children should have been able to do two tasks at the same time. These findings strongly suggested that dyslexic children were not automatic, even at the fundamental skill of balance. For some reason, dyslexic children had difficulty automatising skills, and had therefore to concentrate harder to achieve normal levels of performance. This has clear implications for teaching and learning in that there will be a significant need for over learning to be utilised with children with dyslexia in the classroom.

Cognitive Processes

Researchers/theorists who concentrate on the processes that go into reading and writing typically look at these from a cognitive perspective, although many of the ideas related to cognitive theories of dyslexia also consider neurological factors and hence there is often overlap between these two general areas. For example, automaticity can be considered from the point of view of a process in learning or as a function of a particular brain area (as in the arguments of Fawcett and Nicolson, 2008). Speed of processing can be seen as an aspect of cognition (as in the views of Wolf and colleagues: Wolf & Bowers, 2000; Wolf & O’Brien, 2001) or of brain transmission (as in the views of Breznitz, 2008, and Stein, 2008, discussed above). Of the cognitive focussed research that looks at processing factors, work on memory and learning may be considered as the most likely to link research with classroom practices, with a fair number of recent work either directly or indirectly relating findings to intervention (see discussions in Singleton, 2008). For example, work on cognitive style links to learning style in classrooms (Entwistle, 1981; Given and Reid, 1999; Riding and Cheema, 1991; Schmeck, 1988) and multisensory teaching methods have often been linked to the way information is stored in memory (Broomfield & Combley, 1997; Clark & Uhry, 1995). Similarly, Working Memory research has begun to inform ideas about special educational needs (Gathercole & Pickering, 2000; Jeffries & Everatt, 2004) and meta-cognitive work has focused on strategies for learning. However, perhaps the area that has linked most clearly to the teaching of literacy has been the work related to language processes.

Phonological deficit viewpoints

At present, the dominant causal viewpoint about dyslexia is the phonological deficit hypothesis. This perspective has been derived from the substantial evidence that difficulties in phonological processing, particularly when related to phonological decoding, have been a major distinguishing factor between dyslexics and non-dyslexics from early literacy learning to adulthood (see Beaton, McDougall & Singleton, 1997; Bruck, 1993; Elbro, Nielsen & Petersen, 1994; Rack, Snowling & Olsen, 1992; Snowling, 2000; Stanovich, 1988) and that early phonological training (together with suitable linkage to orthography and literacy experience) improves word literacy and reduces the likelihood of literacy difficulties (see Bryant and Bradley, 1985; Cunningham, 1990; Elbro, Rasmussen & Spelling, 1996; Olofsson and Lundberg, 1985; Schneider, Küspert, Roth, Visé & Marx, 1997). Children who find it difficult to distinguish sounds within verbally presented words would be predicted to have problems learning the alphabetic principle that letters represent sounds and, hence, should be those children who are most likely to be dyslexic based on the phonological deficit position. If this causal linkage is correct, then the manifestation of dyslexia may vary across languages, since languages vary in the way their orthography represents phonology. Therefore, recent research has attempted to investigate the manifestation of dyslexia across languages to assess the universality of the phonological position as well as to inform international assessment practices.

Orthographic transparency

In some languages, orthography represents phonology relatively simply: there is a close correspondence between the written symbol (grapheme) and the basic sound (or phoneme) that it represents. In other orthographies, this correspondence is less transparent. In these languages, a letter may represent several sounds and a sound may be represented by different letters. The English orthography is the best example of this less than transparent relationship between letters and sounds: consider ‘t’ in ‘thus’ versus ‘talk’, or ‘c’ in ‘chord’ versus ‘chore’ or even ‘receive’. There are many English words that may be considered irregular or exceptions based on the typically taught correspondence between graphemes and phonemes (e.g., ‘have’, ‘said’, ‘pint’, ‘monk’, ‘yacht’). However, this level of divergence from the alphabetic principle (see Adams, 1990; Gillon, 2004) is not universal. Although most languages have some peculiarities, or complexities, in the relationship between graphemes and phonemes, most have rules that connect letters with pronunciation that are more consistent and, potentially, simpler to learn than is the case for English. Indeed, the peculiarities of the English orthography has led some theorists in the field to view English as a ‘dyslexic’ orthography (Spencer, 2000) or, perhaps less controversially, as an outlier in comparison to other alphabetic-type orthographies (Share, 2008).

Dyslexia in different languages

The potential importance of orthographic transparency can be seen in cross-language comparisons of reading ability that contrast scripts varying on the transparency dimension. In the majority of such studies, the rate of literacy learning, particularly word reading/decoding, has been found to increase with the level of orthographic transparency. This has been found in comparisons of different language groups (see the Cost A8 work reported in Seymour, Aro & Erskine, 2003), although differences in terms of the cultural importance of literacy learning or educational practice could also explain these effects. However, similar results have been found amongst bilinguals learning two orthographies of differing transparency (Everatt, Smythe, Ocampo & Veii, 2002; GevaSeigel, 2000; Veii & Everatt, 2005). Typically, these findings point to word recognition and non-word decoding processes developing faster in the more transparent orthography. This relationship suggests that there may be fewer problems for learners of a more transparent orthography than a less transparent one, which might mean that dyslexia as a word-level literacy learning difficulty may be less evident in languages that use a relatively simple relationship between letters and sounds – i.e., the behavioural manifestation of dyslexia (such as literacy deficits) may vary across languages (see discussions in Goswami, 2000; SymtheEveratt, 2004; Zeigler & Goswami, 2005). From a practical perspective, assessment measures used to identify dyslexia may have to vary across languages. For example, Everatt, Smythe, OcampoGyarmathy (2004) found that although alliteration and rhyme phonological awareness tasks could distinguish groups of grade 3 children with and without literacy deficits in English, they were less reliable at distinguishing similar groups of Hungarian children. The same reduction in the ability to identify poor literacy learners from their peers has been found for decoding skills amongst German learners (see Wimmer, 1993), a measure that has often been used in English language dyslexia assessment procedures. These findings suggest the potential need to consider different tests measures in dyslexia assessments across languages, particularly those that vary on the orthographic transparency dimension. Though the same reduction in the relationship between literacy levels and pseudoword decoding can be found in Chinese character reader (Smythe et al, 2008), which is not as easily explained as due to the level of letter-sound regularity.

This is not to say that assessments in one language may not be informative of problems in another. Research with bilingual children (e.g., Veii & Everatt, 2005) suggests that skills in one language, particularly those that focus on phonological processing, can be highly predictive of skills in the other, and that (phonological) measures developed for one language may be a useful guide for the development of appropriately translated tools in another language (see also Smythe et al, 2008). However, this translation needs to be carefully considered. For example, consistent with English language work, research in Arabic has indicated that phonological processes are predictive of reading levels amongst Arabic children and that poor Arabic readers show weak phonological decoding and low levels of phonological awareness in comparison to matched normal readers (see Abu-Rabia, Share & Mansour, 2003; Al-MannaiEveratt, 2004; ElbeheriEveratt, 2007). However, further work is needed to determine whether assessment procedures would be better supplemented by additional measures that focus more on the specific features of the writing system, such as an awareness of morphemic roots and patterns (see Elbeheri, Everatt, Reid & Al-Mannai, 2006). In terms of theory, these data are consistent with the phonological deficit viewpoint – i.e., a child with a phonological deficit is likely to show problems with learning the Arabic orthography. However, they point to a need to further investigate hypotheses derived from such causal viewpoints across languages given that the specific features of the language can lead to variations in the relationship between phonological skills and literacy levels, as well as how literacy deficits will manifest in the language.


If dyslexia is to be considered as an educational problem, with difficulties that focus on weaknesses in the acquisition of literacy skills, then the main focus of intervention will be educational and concentrate at improving literacy skills. On the other hand, some see dyslexia as more than a weakness in literacy acquisition and this may require intervention in other areas not directly related to literacy, which may involve non-educational interventions.

At one extreme, dyslexia simply represents the lower end of a normal distribution of reading (and spelling) ability. Therefore, teaching methods should be based on the same ideas as for all children. Dyslexics may take longer to acquire literacy skills, but learning should be based on the usual teaching methods used in schools. Hence, an effective way of teaching literacy to any child should be appropriate for the dyslexic too. Norwich and Lewis (2005) propose the view that teaching approaches themselves should be seen as on a continuum and that children with dyslexia do not need anything different, or special, but they can be placed at a different point in the continuum. This would imply that the pedagogy and the curriculum are the same for all although there may be a greater degree of adaptation in some cases.

However, this view is controversial as many of the specialised approaches that have been developed for children with dyslexia rely to a great extent on the opportunity to pull the child out of class and implement approaches on a one on one basis (Henry 2004).

Some relatively recent remediation methods that have moved away from the literacy-based interventions discussed above have focused on visual- or motor-related problems. The visual deficit hypothesis has been the traditional alternative to the phonological position (for reviews see: Everatt, 2002; Stein, 2001, 2008). As such there have been a great many remediation programmes developed around the idea of changing the way the dyslexic sees the world or the written page: from training eye movements through text to covering one eye. However, few have lasted long as viable remediation procedures. The one that has last some time and for which there is still research activity is the use of coloured overlays or tinted lenses (see Wilkins, 2004). Research has suggested that visual filters (coloured lenses that are worn, or coloured overlays that are placed over the page, when reading) may be effective for alleviating (at least some) reading difficulties (see Wilkins et al, 1994). However, despite the research and practitioner evidence, the theoretical rationale for such visual filters lacks specification and is often contradictory (eg, the requirement for precise hew specification when making lenses, but not for changes in lenses when moving between lighting conditions, such as a normal bulb versus the sun, that will change the perceived colour of the same object). Some of the original ideas for this method were derived from the view that dyslexia may be associated with Scotopic Sensitivity Syndrome (Irlen, 1991). However, the diagnosis of Scotopic Sensitivity Syndrome is controversial (see Lopez et al, 1994) and it is unclear what mechanism is responsible for the hypothesised sensitivity (see discussions in Wilkins, 2004). Additionally, the normal intervention practice of allowing dyslexics to choose the filter that they feel improves their reading is open to placebo effects. Indeed, the one study that has attempted to control potential placebo effects (Wilkins et al, 1994) did not find reliable improvements in reading amongst the individuals provided with correct filters compared to a placebo filter. However, dramatic improvements have been claimed for this treatment and it is a pity that this is one of the weakest visual-based areas in terms of research and theory.

Another intervention procedure that has developed from the research of those investigating visual-related deficits amongst dyslexics, and for which there are still research programmes ongoing, is that related to the use of food supplements that contain appropriate levels of complex (long chain or polyunsaturated) fatty acids (see StordyNicholl, 2000, for a review of this intervention procedure). The use of such supplements has been argued to improve visual processing, particularly hand-eye coordination, motion perception and the processing of low contrast visual stimuli (ie, those areas of visual processing often associated with magnocellular pathway functioning). Supplementation is argued to be important due to the lack of these fatty acids in the modern diet and their hypothesised importance in the rapid transmission of ions across cell membranes. This may slow down processing, leading to many of the features associated with speed of processing deficits amongst learning disabled children. However, the deficits in fatty acid uptake described also argue for dyslexics showing the physical features of such a deficiency. These include skin and hair problems that are not typically associated with dyslexic individuals (though see arguments in Richardson et al, 2000). Therefore this particular intervention procedure has yet to be shown to be appropriate for use with dyslexic individuals (see also the evidence reported by Voight et al, 2002, for a lack of efficacy of such supplementation procedures for children with behavioural difficulties).