Opinion

Dyslexia: a deficit in visuo-spatial

attention, not in phonological
processing
Trichur R. Vidyasagar1 and Kristen Pammer2
1
Department of Optometry & Vision Sciences, University of Melbourne, Parkville, Vic 3010, Australia
2
Department of Psychology, The Australian National University, Canberra A.C.T., Australia

Developmental dyslexia affects up to 10 per cent of the related to poor reading. Pre-reading phonological skills
population and it is important to understand its causes. predict later reading failure, and phonological interven-
It is widely assumed that phonological deficits, that is, tions have been demonstrated to be successful [2]. How-
deficits in how words are sounded out, cause the reading ever, there are a number of reasons why poor phonological
difficulties in dyslexia. However, there is emerging evi- coding might not be the whole story, or even be an etiolo-
dence that phonological problems and the reading gical factor in dyslexia. Some cases of dyslexia are clearly
impairment both arise from poor visual (i.e., ortho- not phonological, for example where the reading errors are
graphic) coding. We argue that attentional mechanisms for irregular words, not non-words, and impairments in
controlled by the dorsal visual stream help in serial reading non-words are not always matched by deficits in
scanning of letters and any deficits in this process will phonological awareness [4,5]. There are reports of children
cause a cascade of effects, including impairments in [5,6] and adults with brain damage [7,8], who have diffi-
visual processing of graphemes, their translation into culties in non-word reading but nevertheless exhibit good
phonemes and the development of phonemic aware- phonological awareness. Such evidence should, at the very
ness. This view of dyslexia localizes the core deficit least, lead us to question the causal link between a pho-
within the visual system and paves the way for new nological deficit and dyslexia. Phonological representa-
strategies for early diagnosis and treatment. tions have themselves been shown to be normal in
dyslexics, but a task-specific deficit might occur in the
Misreading dyslexia access to such representations [9]. Moreover phonological
Developmental dyslexia, a specific difficulty in reading sensitivity might occur partially as a consequence of read-
despite adequate learning opportunities, affects from 5 ing instruction [10], and children who are explicitly taught
to 10 per cent of the population [1], but there are many the alphabetic code before reading instruction are better at
manifestations of reading failure (Box 1). The etiology of phonological awareness tasks than children who are not
dyslexia itself has been hotly debated for a long time. taught letters and sounds prior to formal reading instruc-
Theories have ranged from the reading disorder being a tion [11]. Such findings should prompt us to explore
high-level learning disability, to a visual perceptual defect. additional cognitive/neurophysiological possibilities to
Although it is generally recognized that a majority of poor provide a more coherent account for causal factors in
readers have severe problems in phonological awareness dyslexia. We propose that the poor phonological sensitivity
[2,3], it is still an open issue whether the causal deficit in that is characteristic of many forms of dyslexia could be a
dyslexia is necessarily phonological. The debate is particu- failure of the sensory/cognitive system to fine-tune phono-
larly timely in light of recent reevaluations – especially in logical representations through reciprocal cortical feed-
the USA, Australia and UK – of how children learn to read. back in the way that occurs in normal reading.
In this paper, we present evidence for an alternative Misordering of letters and reversal of letters in a word
possibility, namely that a fundamental defect in the visual are common complaints from dyslexic readers, and sensi-
pathways, with or without a corresponding defect in the tivity to spatial sequencing of the constituent components
auditory system, can potentially cause a cascade of effects of text-like object arrays predicts reading in adults [12] and
that can ultimately manifest as a reading problem, in- children [13]. Such difficulties in ascertaining the sequence
cluding phonological impairments. Understanding the of letters in words cannot be easily explained by phonolo-
mechanisms underlying dyslexia allows better informed gical deficits. Studies of neurophysiological bases of pat-
policy making with regard to teaching and remedial inter- tern and object recognition indicate that such sequencing
vention. of letters is a non-trivial problem for the brain. Neurones in
the inferotemporal neocortical areas that presumably med-
Phonological theory of dyslexia – controversies iate pattern recognition typically have large receptive
Poor phonological processing is a core component of read- fields [14]. Such position invariance helps to recognize
ing failure. Poor phonological skills are consistently an object irrespective of its location in the visual field,
but this property, in the case of identifying letters in
Corresponding author: Vidyasagar, T.R. ([email protected]). reading, inevitably leads to loss of information about the
1364-6613/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.tics.2009.12.003 57
Opinion Trends in Cognitive Sciences Vol.14 No.2

Box 1. Reading errors in dyslexia

We do not really understand what it is that a dyslexic reader sees. We know that some poor readers experience distorted text:

Ambiguity regarding the invariance of some letters can make decoding difficult. We can simulate the decoding difficulties of poor readers
where d’s, p’s, q’s, and g’s are interchangeable:

Even just reorganizing the internal letters of words can slow down decoding and make it difficult to read:

We all konw aobut the Camrbdirge Unviresity efefcet, we can raed it, but it is mcuh solwer and we mkae ltos mroe scacades and rgreesressoins.

Castles and Coltheart [5] made the distinction between ‘Surface’ and ‘Phonological’ dyslexia. In Phonological dyslexia, an impairment in
reading pronounceable non-words (e.g. ‘plimp’), indicates a difficulty in implementing grapheme-to-phoneme correspondence rules, or
‘sounding-out’ a word. Non-words have no representation in our mental dictionary, or lexicon. Thus in order to read such words, we must sound
them out. The same challenge is faced when presented with a real word for the first time – as occurs when learning to read. Over repeated
exposure, we might develop a visual representation of the word and we can access directly its lexical representation. Surface dyslexia then, is
characterized by difficulties using the lexical route, with a corresponding problem in developing direct visual representations of words.
Difficulties manifest in reading irregular words such as ‘meringue’ and regularization errors are typical, for example ‘meringoo’, since the surface
dyslexics tend to over-rely on the phonological route. In fact most dyslexic readers have some degree of both Phonological and Surface dyslexia.
Thus some dyslexic readers have exclusive difficulties in developing a visual representation of words and have ‘pure’ Surface dyslexia, but most
dyslexic readers also experience some form of surface dyslexia. In addition, other poor readers suffer from a disorder called Visual Discomfort, in
which sufferers describe very visual problems such as the words ‘shimmering’, or ‘floating’, and ‘letters moving over each other’. The
relationships between visual discomfort and the more cognitive aspects of reading remain unclear.

sequence of letters within a word. However, such infor- each letter in a text [16]. The time required by children to
mation is clearly preserved to an excellent degree in nor- learn to read effectively could be due to the need for training
mal reading. How is this achieved? the visual search mechanism, which is usually randomly
and not systematically deployed across the visual field [18],
Visuo-spatial attention and dyslexia to proceed in a sequential manner from left to right at a
There are two possible solutions to the above conundrum: fine enough spatial scale across the letters of each word.
(i) information of the relative location of each letter is Contrary to earlier beliefs, there is also now evidence that
somehow tagged to the process of recognizing that particu-
lar letter or (ii) letters are recognized sequentially, with
only one or a few letters being processed at a time by the
object recognition system and this temporal sequence pre-
serves the spatial sequence of the letters. Although there is
no direct evidence supporting the former possibility, the
latter is consistent with many neurophysiological and
psychophysical studies as will be described below.
Reading is a recent cultural trait in human history,
unlikely to have been a direct product of evolution. Learning
to read is thus an opportunistic training of neural mechan-
isms underlying a range of cognitive, perceptual and motor
skills that had evolved for other purposes. One of the most
relevant, neuro-cognitive functions used for reading might
be what has been termed in the visual search literature as ‘a
spotlight of attention’ [15]. In our usual cluttered world,
targets of interest rarely possess unique features that help
them to pop out from among distracting elements in a scene.
Thus a serial search is usually undertaken that supposedly
sweeps a spotlight of attention across a scene in a random Figure 1. Role of visuo-spatial attention in reading. During reading, both normal
fashion. The process helps to recognize one item at a time and dyslexic readers make a number of saccades of about 25 milliseconds each,
and also to ‘bind’ the different attributes of each object such with intervening fixations each lasting about 250 milliseconds, although the
number of saccades and the fixation durations are highly dependent upon the
as its form, color, depth, motion and size. We have suggested reading material and the skill of the reader, with poor readers making far more
[16,17] that in reading, the same top-down search mechan- regressions, shorter saccades and longer fixations. During each fixation,
isms are used to sweep the spotlight of attention serially approximately 7 or 8 letters are read [65], but they have to be processed in the
appropriate sequence. We propose that in reading the brain uses essentially the
over the letters of a word during the periods of fixation (see same circuitry and attentional mechanisms that are used for serial visual search,
Figure 1 and animation in Supplementary Materials). The but scans the spotlight of attention sequentially along the letters of the text during
speed of serial visual search processes varies between 15 those periods of fixation. The letters are then processed in that temporal order to
yield the spatial order of letters in a word. (See also the animation in
and 44 millisecond per item [18] and it has been noted that supplementary data of how attention might shift during each fixation, shown at
this is in the same ballpark as the average speed of reading approximately one fourth the real speed).

58
Opinion Trends in Cognitive Sciences Vol.14 No.2

words are not usually read as wholes, but as letters or small strings predicts reading proficiency in children [12] and
groups of them in a sequence [19]. adults [13].
Recent neurophysiological and brain imaging studies The idea we have proposed [16,17] of a serial allocation
provide support for the existence of a neural system that of attention across the text in reading is also an essential
could mediate a gating function that could have been co- part of the SERIOL model [36]. However the latter differs
opted for reading. Such modulation is most likely driven from ours in that it postulates a gradient of attention only
from areas of the dorsal stream, in particular from the during learning to read, but not in skilled reading, where
posterior parietal cortex, which has been implicated in the gradient is supposed to be automatically activated.
spatial attention [20]. The visual information coming into Further descriptions of SERIOL and other models that
the striate cortex via the three major parallel pathways – seek to explain how letter order is encoded are beyond the
magnocellular, parvocellular and koniocellular – is further scope of this paper.
channeled via two major streams, dorsal and ventral,
projecting into the parietal and temporal cortices respect- Magnocellular deficits in dyslexia – controversies
ively [16,21]. The dorsal stream is believed to be concerned Because intuitively one would expect the visual sensitivity
with spatial localization, movement, depth and visually in reading to depend upon the fine pattern recognition
guided action such as reaching and saccades, whereas the skills of the ventral stream, it is not surprising that
ventral stream is thought to be concerned with object schemes of a visual deficit in dyslexia restricted to the
recognition and perception [22]. The dorsal stream receives magnocellular pathway [26,27,37] have been controversial
a fast input, ostensibly via the magnocellular pathways, [38,39]. However, our scheme puts any such defect in
which can potentially provide feedback to the primary perspective. The defect need not be with the magnocellular
visual cortex to gate the inputs entering the ventral stream cells themselves, but it could be anywhere along the dorsal
areas (Figure 2). Consistent with this, serial visual search stream. Furthermore, any defect in the magnocellular
is more efficient with stimuli that involve the magnocel- pathway need only reduce the density of magnocellular
lular channel, which provides the dominant visual input to cells in a few critical regions of the visual field, say for a
the dorsal stream [23,24]. These gating inputs are pre- degree or two to one side of the fovea, to cause a reading
sumably the means by which serial visual search functions: deficit. This can cause serious difficulties in the sweep of
the dorsal stream using its cruder spatial resolution and the focus of attention along the length of each word. Such a
information of locations of objects to sequentially select defect is not likely to be picked up by classical tests of
specific items in the visual field to be processed in the magnocellular function. By the same token, unless a mag-
ventral stream. nocellular deficit affects sampling at the critical visual field
If our scheme is a description of the crucial processing locations, widespread M (magnocellular) cell losses can co-
steps in reading that lead to grapheme identification and to exist with normal reading abilities. The reality about
subsequent matching of graphemes with phonemes, it dyslexia is that it is not a unitary disease and there are
follows that reading can be affected by a deficit at any step probably a variety of manifestations that relate to the
along the neural pathway, such as: actual site of impairment. A recent study suggests that
(a) Poorer sampling density of the magnocellular system deficiencies at different levels of the pathway from retina to
in the retina [25]. parietal cortex could relate to poorer performance in differ-
(b) Specific deficit anywhere along the visual pathway to ent aspects of reading [40]. The bottom-line of our scheme
the dorsal stream [26,27]. has been that dyslexia is a visual processing deficit in the
(c) Concomitant damage of other parallel pathways (eg. proper recognition of the sequence of letters by an atten-
koniocellular) that could otherwise compensate for any tional mechanism, a suggestion that has been supported by
magnocellular damage [28]. a host of other studies [31–35,41,42]. A recent study that
(d) Damage in the dorsal cortical stream – to area V5/MT teased apart the correlation with reading ability of motion
or the posterior parietal cortex [28–30]. perception (in Ternus task) and attentional mechanisms
(e) Damage to feedback pathways from the dorsal stream (in a visual search task), found that a significant visual
areas to visual area V1 or the ventral stream. deficit in the dyslexic group was found only when the task
(f) Lesion at the site where the attentional modulation required visual search, but not when it involved motion
occurs. perception alone [43]. This again underscores the point
that visual search and reading both exert considerable
A relationship to dyslexia has been demonstrated in demands on the dorsal stream, but possibly in a very
some of the above cases (a, b and d), but in others (c, e and f) restricted region of the visual field [44].
the link remains hypothetical. Nevertheless, a common
outcome from all of the above is a poorer allocation of Phonological deficits in dyslexia – cause or effect?
spatial attention. Indeed, visual search is impaired in Despite providing a physiological framework for reading
dyslexic children [17,31]], and many studies have ident- and its impairments, the scheme proposed here also has to
ified a range of problems with visuo-spatial attention in confront and explain the severe phonological deficits seen
children with dyslexia [31–35]. Consistent with the notion in many dyslexics. It is possible that the development of
that reading utilizes the same mechanisms that are essen- phonological awareness itself might be at least partially
tial for selecting and stringing together small local dependent upon a normal input from the visual system into
elements in the visual scene, it has been demonstrated brain regions subserving grapheme–phoneme correspon-
that sensitivity to the spatial sequence of word-like symbol dence. Indeed there is substantial evidence indicating that

59
Opinion Trends in Cognitive Sciences Vol.14 No.2

Figure 2. Neural circuitry that is exploited for reading is the system that normally enables serial visual search. (a) Visual pathways of the macaque, showing signals
proceeding from the eyes to the lateral geniculate nucleus (LGN) and on to primary visual cortex (V1), with the faster signals of the magnocellular channel reaching the
dorsal stream areas, namely middle temporal area (MT) and the posterior parietal cortex (PPC) as early as 40 milliseconds [66]. (b) The location information and crude
pattern information in the input is used by the fronto-parietal network to select the object location to be facilitated among the signals reaching V1 via the slower

60
 Opinion Trends in Cognitive Sciences Vol.14 No.2

orthographic training itself seems to influence and/or temporal processing problems [54]. Thus there might be
enhance phonological awareness [45–47]. Thus demon- many reasons for the impaired phonemic awareness in the
strations, that phonological awareness predicts future dyslexic population, whereas reading difficulties them-
reading skills, could be due to orthographic training facil- selves could have their origins primarily in abnormal
itating the development of both phonological awareness visual attentional mechanisms. Poorer phonemic aware-
and reading skills. This might occur in one or both of two ness, due to a variety of reasons including abnormal visual
ways: (i) The segmentation of words into component gra- processing, could make the reading difficulty worse with-
phemes might provide the necessary substrate for improv- out being the major cause of it. This might explain the
ing phonemic awareness through facilitating the process of improvement that has been reported in some cases of
grapheme–phoneme correspondence [48]. (ii) Multimodal phonological remediation [55]. This study’s claim that
integration of visual and auditory inputs during early the computer-based exercises that target auditory
development might be necessary to fine-tune the acoustic temporal processing, as in the commercially marketed Fast
system and its ability to process changing sound inputs ForWord, improve reading skills, has been highly contro-
[16,35,49]. At the crux of this argument is an assumption versial. A number of independent studies that have incorp-
that the development of a ‘healthy’ neural network – such orated extensive controls have been able to show
as in skilled reading – is dependent on connections that improvements in auditory processing from Fast ForWord
build through all components of the network, and viable and similar training regimes, but no comparable improve-
network nodes that allow strengthening and fine-tuning of ments in reading skills [56,57]. Thus, although a general
all components of the network. That pre-readers at risk for temporal processing deficit might cause slower processing
dyslexia demonstrated reduced magnocellular sensitivity of auditory signals, the primary reason for the reading
before they began to read [40], supports this notion of difficulty in dyslexic children might arise within the visual
reciprocal feedback from the dorsal steam and the need system and be exacerbated by any concurrent auditory
for all components of the reading network to be fully deficit.
functioning, because subtle impairments in one part of Finally, the recent finding in 7–12-year-old children that
the network can have profound consequences for the de- contrast responsivity in area MT+ (but not in V1) is sig-
velopment of the whole network. nificantly correlated with phonological awareness [58]
It is conceivable that with our proposed inadequate indicates that magno-dominated dorsal stream activity
parsing of a stream of text into graphemes in dyslexic may be crucial for reading. In the light of our hypothesis,
individuals, areas of the brain that subserve grapheme– it is interesting that this study showed that there was little
phoneme correspondence might show reduced activity with correlation between MT+ responsivity and rapid naming,
compensatory increases in other areas. Thus altered age and IQ and among the various reading related
neural activity in a region in the reading impaired need measures tested, but the strongest correlation was with
not mean that these areas are the sites of the critical defect. phonological awareness. This underscores our suggestion
Brain imaging studies on dyslexia can be seen in this light. that development of grapheme–phoneme correspondence
Posterior regions, particularly the angular gyrus, supra- and phonological awareness benefit from intact dorsal
marginal gyrus, posterior part of the superior temporal stream activity in the visual system.
gyrus and portions of middle occipital and middle temporal
gyri show reduced activity in dyslexics, whereas the left Causality
inferior frontal gyrus shows increased activity [1]. In our As Castles and Coltheart [5] point out in their compre-
view, the reduced activity in the posterior brain regions hensive paper on the role of phonological coding in read-
that are likely to be involved in grapheme–phoneme cor- ing, the burden of test of causality falls upon researchers
respondence could be due simply to a poorer input into being able to demonstrate in longitudinal designs that
these regions. This is supported by evidence that early deficits both precede and predict reading development,
processing stages of the reading network are delayed or and that appropriate training facilitates reading acqui-
absent in dyslexics [50]. sition. Such studies in the context of a dorsal stream deficit
The visual processing deficit could, however, be part of a in dyslexia are in their infancy, but these studies are
general temporal processing deficit, which would explain encouraging. Children who are less sensitive to coherent
the high incidence of deficits in rapid auditory processing motion in preschool subsequently go on to have poorer
among dyslexic children [51]. Recently, learning disabil- literacy skills in grade 1 [59]. Contrast sensitivity in
ities including severe reading impairments, have been kindergarten predicts reading ability two years later
associated also with abnormal brainstem timing, indicat- [60]. Temporal order judgment in preschool children pre-
ing a deficit at a very early level of the sensory pathways dicts their single word reading skill in Grade 1 [61]. Pre-
[52]. However, dyslexics’ auditory deficits might be related reading children at risk for dyslexia – that is, who have a
more to their difficulties in focusing spatial auditory atten- first-degree relative with dyslexia – are significantly less
tion rather than impairments in low-level temporal pro- sensitive to coherent motion and visual frequency dou-
cessing per se [53], a finding also supported by studies that bling compared to a large unselected sample of same-age
could not directly attribute the reading difficulty to children [40].

parvocellular channel. The object in the selected location is then processed in the ventral, ‘what’ pathway and recognized. Neurones in the ventral stream are known to have
longer latencies [67]. (c) Orchestration of serial search by the fronto-parietal network chooses one location after another until the target is found. Consistent with such neural
substrate for a feedback from the dorsal steam, physiological studies have shown both focal modulation of striate cortical activity by attention [68,69] and a neural
mechanism for a fronto-parietal network controlling spatial attention in early visual areas [20,70].

61
Opinion Trends in Cognitive Sciences Vol.14 No.2

Box 2. Future directions grapheme–phoneme correspondence and due to a general

temporal processing deficit affecting all modalities.
 In order to provide a comprehensive argument that poor visual
coding is a core deficit in dyslexia, we need more extensive Acknowledgements
studies looking at whether visual training is effective in remediat- The article is partly based upon studies conducted with support to TRV
ing dyslexia. from the Australian National Health and Medical Research Council. We
 Reading is a multi-sensory, multi-cognitive task. Further research also thank Mr. Sivaram Viswanathan for help with figures and the video.
should be targeted at determining which aspects of dorsal coding
are intrinsic to which aspects of the reading process. Appendix A. Supplementary data
 The literature to date suggests that not all dyslexics demonstrate a
Supplementary data associated with this article can be
dorsal deficit. However, the studies conducted might not ade-
quately target the very subtle visual deficits that some dyslexics found, in the online version, at doi:10.1016/j.tics.2009.
might experience. Thus further research is required to identify 12.003.
such specific deficits.
References
1 Shaywitz, S.E. (1996) Dyslexia. Sci. Am. 275, 98–104
2 Goswami, U. and Bryant, P. (1990) Phonological skills and learning to
A small number of training studies have also been read, Erlbaum
conducted. Fischer and Hertnegg [62] trained children 3 Snowling, M.J. (2001) From language to reading and dyslexia. Dyslexia
for three weeks on a saccadic control task. The dyslexic 7, 37–46
children’s saccadic control normalized, to be consistent 4 Castles, A. and Coltheart, M. (1993) Varieties of developmental
dyslexia. Cognition 47, 149–180
with the non-dyslexic control group. However, there was 5 Castles, A. and Coltheart, M. (2004) Is there a causal link from
no measure to indicate if the visual improvements trans- phonological awareness to success in learning to read? Cognition 91,
lated into improved reading. Solan et al. [63] trained poor 77–111
readers on visuo-perceptual tasks. They demonstrated 6 Tree, J.J. and Kay, J. (2006) Phonological dyslexia and phonological
that children improved on reading comprehension and impairment: an exception to the rule? Neuropsychologia 44, 2861–2873
7 Derousné, J. and Beauvois, M.F. (1985) The ‘‘phonemic’’ state in the
word attack measures after the intervention. However, non-lexical reading process: Evidence from a case of phonological
the children were also ‘‘provided the opportunity to develop alexia. In Surface Dyslexia (Patterson, K. et al., eds.), pp. 399–457,
improved cognitive strategies’’ (p. 644), making it difficult Lawrence Erlbaum Associates Ltd.
to evaluate the impact of the visual training on reading 8 Bisiacchi, P.S. et al. (1989) Impairment in processing meaningless
verbal material in several modalities: the relationship between
independently of the increased exposure to words pre-
short-term memory and phonological skills. Q. J. Exp. Psychol. A 41,
sented in a novel way. Similarly, Lorusso et al. [64] used 293–319
a technique in which the children were required to make a 9 Ramus, F. and Szenkovits, G. (2008) What phonological deficit? Q. J.
rapid attentional shift to a stimulus in either the right or Exp. Psychol. (Colchester) 61, 129–141
left visual field. They demonstrated increases in reading 10 Ehri, L.C. (1989) The development of spelling knowledge and its role in
reading acquisition and reading disability. J. Learng. Disabil. 22, 356–
accuracy and speed compared to a control group that was
365
exposed to more traditional remediation methods. How- 11 Mann, V.A. and Wimmer, H. (2002) Phoneme awareness and pathways
ever, again the stimuli in the task were words that the to literacy: a comparison of German and American children. Reading
children had to identify that ‘‘became increasingly difficult and Writing 15, 653–682
in terms of word length and complexity of spelling’’ (p. 201). 12 Pammer, K. et al. (2005) Symbol-string sensitivity and adult
performance in lexical decision. Brain Lang. 94, 278–296
Thus, although such results seem promising, there is a dire 13 Pammer, K. et al. (2004) Symbol-string sensitivity and children’s
need for controlled studies that investigate the impact of reading. Brain Lang. 89, 601–610
visual training on dorsally mediated tasks in reading de- 14 Boussaoud, D. et al. (1991) Visual topography of area TEO in the
velopment and remediation. Moreover, studies need to be macaque. J. Comp. Neurol. 306, 554–575
conducted to determine for example, whether dorsal train- 15 Treisman, A. (1988) Features and objects: the fourteenth Bartlett
memorial lecture. Q. J. Exp. Psychol. A 40, 201–237
ing facilitates reading acquisition in normally developing 16 Vidyasagar, T.R. (1999) A neuronal model of attentional spotlight:
readers, and the resilience of such dorsal visual training on parietal guiding the temporal. Brain Res. Brain Res. Rev. 30, 66–76
reading development over time (Box 2). 17 Vidyasagar, T.R. and Pammer, K. (1999) Impaired visual search in
Consistent with the need for further research is the dyslexia relates to the role of the magnocellular pathway in attention.
Neuroreport 10, 1283–1287
notion of reciprocal feedback, that is, the degree to which
18 Horowitz, T.S. and Wolfe, J.M. (1998) Visual search has no memory.
reading per se facilitates the development of visual coding Nature 394, 575–577
in the same way that it seems to do in phonological coding. 19 Peli, D.G. and Tillman, K.A. (2007) Parts, wholes and context in
We are in the process of investigating whether individuals reading: A triple dissociation. Plos One 2, e860
who have never learned to read also demonstrate less 20 Saalmann, Y.B. et al. (2007) Neural mechanisms of visual attention:
How top-down feedback highlights relevant locations. Science 316,
sensitive dorsal functioning.
1612–1615
21 Bullier, J. (2001) Integrated model of visual processing. Brain Res.
Concluding remarks Brain Res. Rev. 36, 96–107
The critical deficit in developmental dyslexia might be one 22 Goodale, M.A. and Milner, A.D. (1992) Separate visual pathways for
that affects the focal visual attentional mechanisms essen- perception and action. Trends Neurosci. 15, 20–25
23 Cheng, A. et al. (2004) The role of the magnocellular pathway in serial
tial for efficient reading. The poor phonological awareness deployment of visual attention. Eur. J. Neurosci. 20, 2188–2192
that is seen in most dyslexics might not be the cause of the 24 Li, J.C.H. et al. (2007) Interactions between luminance and colour
reading difficulty, but could be the result of the poor channels in visual search and their relationship to parallel neural
orthographic inputs feeding into the regions mediating channels in vision. Exp. Brain Res. 176, 510–518

62
Opinion Trends in Cognitive Sciences Vol.14 No.2

25 Pammer, K. and Wheatley, C. (2001) Isolating the M(y)-cell response in 48 Castles, A. et al. (2003) How does orthographic knowledge influence
dyslexia using the spatial frequency doubling illusion. Vision Res. 41, performance on phonological awareness tasks? Q. J. Exptl. Psychol
2139–2147 65A, 445–467
26 Lovegrove, W.J. et al. (1980) Specific reading disability: differences in 49 Pammer, K. and Vidyasagar, T.R. (2005) Integration of the visual and
contrast sensitivity as a function of spatial frequency. Science 210, auditory networks in dyslexia: a theoretical perspective. J. Res.
439–440 Reading 3, 320–331
27 Livingstone, M.S. et al. (1991) Physiological and anatomical evidence 50 Salmelin, R. et al. (1996) Impaired visual word processing in dyslexia
for a magnocellular defect in developmental dyslexia. Proc. Natl. Acad. revealed with magnetoencephalography. Ann. Neurol. 40, 157–162
Sci. U S A 88, 7943–7947 51 Tallal, P. (1980) Auditory temporal perception, phonics, and reading
28 Vidyasagar, T.R. (2005) Attentional gating in primary visual cortex: A disabilities in children. Brain Lang. 9, 182–198
physiological basis for dyslexia. Perception 34, 903–911 52 Benai, K. et al. (2005) Brainstem timing: Implications for cortical
29 Cornelissen, P. et al. (1995) Contrast sensitivity and coherent motion processing and literacy. J. Neurosci. 25, 9850–9857
detection measured at photopic luminance levels in dyslexics and 53 Petkov, C.I. (2005) Auditory perceptual grouping and attention in
controls. Vision Res. 35, 1483–1494 dyslexia. Cogn. Brain Res. 24, 343–354
30 Eden, G.F. et al. (1996) Abnormal processing of visual motion in 54 Nittrouer, S. (1999) Do temporal processing deficits cause phonological
dyslexia revealed by functional brain imaging. Nature 382, 66–69 processing problems? J. Speech Lang. Hear. Res. 42, 925–942
31 Casco, C. and Prunetti, E. (1996) Visual search of good and poor 55 Temple, E. et al. (2003) Neural deficits in children with dyslexia
readers: effects with targets having single and combined features. ameliorated by behavioral remediation: evidence from functional
Percept. Mot. Skills 82, 1155–1167 MRI. Proc. Natl. Acad. Sci. U. S. A. 100, 2860–2865
32 Facoetti, A. et al. (2000) Visual-spatial attention in developmental 56 Pokorni, J.L. et al. (2004) Phonological awareness intervention:
dyslexia. Cortex 36, 109–123 Comparison of Fast ForWord, Earobics, and LiPS. J. Educ. Res. 97,
33 Hari, R. et al. (1999) Prolonged attentional dwell time in dyslexic 147–157
adults. Neurosci. Lett. 271, 202–204 57 McArthur, G.M. et al. (2008) Auditory processing deficits in children
34 Bosse, M.L. et al. (2007) Developmental dyslexia: the visual attention with reading and language impairments: Can they (and should they) be
span deficit hypothesis. Cognition 104, 198–230 treated? Cognition 107, 946–977
35 Facoetti, A. et al. (2009 Apr 14) Multisensory spatial attention deficits 58 Ben-Shachar, M. et al. (2007) Contrast responsivity in MT+ correlates
are predictive of phonological decoding skills in developmental with phonological awareness and reading measures in children.
dyslexia. J. Cogn. Neursoci [Epub ahead of print] Neuroimage 37, 1396–1406
36 Whitney, C. (2001) How the brain encodes the order of letters in a 59 Boets, B. et al. (2006) Auditory temporal information processing in
printed word: The SERIOL model and selective literature review. preschool children at family risk for dyslexia: Relations with
Psychon. Bull. Rev. 8, 221–243 phonological abilities and developing literacy skills. Brain Lang. 97,
37 Stein, J.F. and Walsh, V. (1997) To see but not to read: the 64–79
magnocellular theory of dyslexia. Trends Neurosci. 20, 147–152 60 Kevan, A. and Pammer, K. (2009) Predicting early reading skills from
38 Skottun, B.C. and Skoyles, J. (2006) Attention, reading, and dyslexia. pre-reading measures of dorsal stream functioning. Neuropsychologia
Clin. Exp. Optom. 89, 241–245 47, 3174–3181
39 Ram-Tsur, R. et al. (2006) Sequential processing deficits of reading 61 Hood, M. and Conlon, E. (2004) Visual and auditory temporal
disabled persons is independent of stimulus interval. Vis. Res. 46, processing and early reading development. Dyslexia 10, 234–252
3949–3960 62 Fischer, B. and Hartnegg, K. (2000) Effects of visual training on
40 Kevan, A. and Pammer, K. (2008) Visual processing deficits in preliterate saccade control in dyslexia. Perception 29, 531–542
children at familial risk for dyslexia. Vis. Res. 48, 2835–2839 63 Solan, H.A. et al. (2004) M-cell deficit and reading disability: a
41 Whitney, C. and Cornelissen, P. (2005) Letter-position encoding and preliminary study of the effects of temporal vision-processing
dyslexia. J. Res. Reading 28, 274–301 therapy. Optometry 75, 640–650
42 Solan, H.A. et al. (2001) Role of visual attention in cognitive control of 64 Lorusso, M.L. et al. (2006) Effects of visual hemisphere-specific
oculomotor readiness in students with reading disabilities. J. Learn. stimulation versus reading-focused training in dyslexic children.
Disabil. 34, 107–118 Neuropsychol. Rehabil. 16, 194–212
43 Jones, M.W. et al. (2008) Visual deficits in developmental dyslexia: 65 Blais, C. et al. (2009) Reading between eye saccades. Plos One 4, e6448
relationships between non-linguistic visual tasks and their 66 Bisley, J.W. et al. (2004) A rapid and precise on-response in posterior
contribution to components of reading. Dyslexia 14, 95–115 parietal cortex. J. Neurosci. 25, 1833–1838
44 Vidyasagar, T.R. (2004) Neural underpinnings of dyslexia as a disorder 67 Tamura, H. and Tanaka, K. (2001) Visual response properties of cells in
of visuo-spatial attention. Clin. Exp. Optom. 87, 4–10 the ventral and dorsal parts of the macaque inferotremporal cortex.
45 Ehri, L. and Wilce, L.S. (1980) The influence of orthography on readers’ Cerebr. Cortex 11, 384–399
conceptualisation of the phonemic structure of words. Applied 68 Vidyasagar, T.R. (1998) Gating of neuronal responses in macaque
Psycholinguistics 1, 371–385 primary visual cortex by an attentional spotlight. Neuroreport 9,
46 Wimmer, H. et al. (1991) The relationship of phonemic awareness to 1947–1952
reading acquisition: more consequence than precondition but still 69 Simola, J. et al. (2009) Topography of attention in the primary visual
important. Cognition 40, 219–249 cortex. Eur. J. Neurosci. 29, 188–196
47 Johnston, R. et al. (1996) Knowledge of the alphabet and explicit 70 Buschman, T.J. and Miller, E.K. (2007) Top-down and bottom-up
awareness of phonemes in pre-readers: the nature of the control of attention in the prefrontal and posterior parietal cortices.
relationship. Reading and writing 8, 217–234 Science 315, 1860–1862

63