Visual attention modulates the transition from fine-grained, serial processing to coarser-grained, more parallel processing: A computational modeling study.

During reading acquisition, beginning readers transition from serial to more parallel processing. The acquisition of word specific knowledge through orthographic learning is critical for this transition. However, the processes by which orthographic representations are acquired and fine-tuned as learning progresses are not well understood. Our aim was to explore the role of visual attention in this transition through computational modeling. We used the BRAID-Learn model, a Bayesian model of visual word recognition, to simulate the orthographic learning of 700 4-to 10-letter English known words and novel words, presented 5 times each to the model. The visual attention quantity available for letter identification was manipulated in the simulations to assess its influence on the learning process. We measured the overall processing time and number of attentional fixations simulated by the model across exposures and their impact on two markers of serial processing, the lexicality and length effects, depending on visual attention quantity. Results showed that the two lexicality and length effects were modulated by visual attention quantity. The quantity of visual attention available for processing further modulated novel word orthographic learning and the evolution of the length effect on processing time and number of attentional fixations across repeated exposures to novel words. The simulated patterns are consistent with behavioral data and the developmental trajectories reported during reading acquisition. Overall, the model predicts that the efficacy of orthographic learning depends on visual attention quantity and that visual attention may be critical to explain the transition from serial to more parallel processing.

Visual attention span as a predictor of reading fluency and reading comprehension in Arabic.

Introduction: Visual attention span is a measure of multielement parallel processing. Individuals with higher visual attention span are expected to allocate more attention to letters within strings, which boosts letter identification and translates into more efficient reading. Given the high visual complexity of the Arabic writing system, we expected visual attention span to be an important predictor of reading in the Arabic language. Methods: Native Arabic readers from Grade 4 and Grade 5 were recruited in Iraqi schools. We assessed the contribution of visual attention span to their reading fluency performance in tasks of fully vowelized word and pseudo-word reading, non-vowelized text reading, and written text comprehension. Their phonological awareness, IQ, and single letter processing speed were further evaluated. Results: Results showed that visual attention span was a significant unique predictor of all the reading measures. Visual attention span and phonological awareness accounted for a similar amount of variance in word and pseudo-word reading fluency. Visual attention span was a far higher predictor than phonological awareness for text reading fluency and the sole predictor of text comprehension. Discussion: The role of visual attention span to reading is discussed by reference to current word recognition models. Higher involvement of visual attention is expected in vowelized script to compensate for increased crowding in the presence of diacritics. Visual attention would thus contribute to sub-lexical orthographic parsing and favor orthography-to-phonology mapping, in particular for the pseudo-words that do not benefit from efficient lexical feedback. In non-vowelized script, higher visual attention would enhance the accurate and fast identification of root letters within words, thus resulting in faster word recognition.

The visual-attention span deficit in developmental dyslexia: Review of evidence for a visual-attention-based deficit.

The visual attention span (VAS) deficit hypothesis in developmental dyslexia posits that a subset of dyslexic individuals shows a multielement parallel processing deficit due to reduced visual attention capacity. However, the attention-based interpretation of poor performance on VAS tasks is hotly debated. The purpose of the present paper is to clarify this issue through a critical review of relevant behavioural and neurobiological findings. We first examine the plausibility of alternative verbal interpretations of VAS performance, evaluating whether performance on VAS tasks might reflect verbal short-term memory, verbal coding or visual-to-verbal mapping skills. We then focus on the visual dimensions of VAS tasks to question whether VAS primarily reflects visuo-attentional rather than more basic visual skills. Scrutiny of the available behavioural and neurobiological findings not only points to a deficit of visual attention in dyslexic individuals with impaired VAS but further suggests a selective endogenous attentional system deficit that relates to atypical functioning of the brain dorsal attentional network. The overview clarifies the debate on what is being measured through VAS tasks and provides insights on how to interpret the VAS deficit in developmental dyslexia.

Probabilistic modeling of orthographic learning based on visuo-attentional dynamics.

How is orthographic knowledge acquired? In line with the self-teaching hypothesis, most computational models assume that phonological recoding has a pivotal role in orthographic learning. However, these models make simplifying assumptions on the mechanisms involved in visuo-orthographic processing. Against evidence from eye movement data during orthographic learning, they assume that orthographic information on novel words is immediately available and accurately encoded after a single exposure. In this paper, we describe BRAID-Learn, a new computational model of orthographic learning. BRAID-Learn is a probabilistic and hierarchical model that incorporates the mechanisms of visual acuity, lateral interference, and visual attention involved in word recognition. Orthographic learning in the model rests on three main mechanisms: first, visual attention moves over the input string to optimize the gain of information on letter identity at each fixation; second, top-down lexical influence is modulated as a function of stimulus familiarity; third, after exploration, perceived information is used to create a new orthographic representation or stabilize a better-specified representation of the input word. BRAID-Learn was challenged on its capacity to simulate the eye movement patterns reported in humans during incidental orthographic learning. In line with the behavioral data, the model predicts a larger decline with exposures in number of fixations and processing time for novel words than for known words. For novel words, most changes occur between the first and second exposure, that is to say, after creation in memory of a new orthographic representation. Beyond phonological recoding, our results suggest that visuo-attentional exploration is an intrinsic portion of orthographic learning seldom taken into consideration by models or theoretical accounts.

Atypical viewing position effect in developmental dyslexia: A behavioural and modelling investigation.

The probability of recognizing a word depends on the position of fixation during processing. In typical readers, the resulting word-recognition curves are asymmetrical, showing a left-of-centre optimal viewing position (OVP). First, we report behavioural results from dyslexic participants who show atypical word-recognition curves characterized by the OVP being right of centre with recognition probability being higher on the rightmost than on the leftmost letters. Second, we used BRAID, a Bayesian model of word recognition that implements gaze position, an acuity gradient, lateral interference and a visual attention component, to examine how variations in the deployment of visual attention would affect the OVP curves. We show that the atypical dyslexic curves are well simulated assuming a narrow distribution of visual attention and a shifting of visual attention towards the left visual field. These behavioural and modelling findings are discussed in light of current theories of visual attention deficits in developmental dyslexia.

Varieties of Cognitive Profiles in Poor Readers: Evidence for a VAS-Impaired Subtype.

A wide share of secondary school children does not reach the expected competence level in reading. These children could benefit from more efficient intervention responses, providing a better understanding of their cognitive weaknesses/deficits. Our aim was to explore the cognitive heterogeneity of a population of poor readers identified from a large sample of 948 sixth-grade children. We first assessed the contribution of phoneme awareness (PA), rapid automatized naming (RAN), and visual attention span (VAS) to reading performance in a subset of 281 children including poor and average readers/spellers. We show that all three skills are unique and significant predictors of reading fluency. We then restricted the analysis to participants with normal Raven's score (IQ) and oral language skills to focus on 110 children with more specific reading difficulties. A unique VAS deficit was found in 18% of these poor readers while 20% and 15.5% showed a unique PA or RAN deficit. Children with multiple or no deficit were further identified. The overall findings provide evidence for a variety of cognitive profiles in poor readers. They suggest that, in addition to PA interventions, training programs targeting VAS might be useful for the nontrivial share of poor readers who exhibit a VAS deficit.

Visual attention modulates reading acquisition.

The processing of letters within strings is challenging for beginning readers. Letter identification is affected by visual similarity, loss of information with eccentricity and interference from nearby letters. In contrast, visual attention enhances letter identification. We here explored whether visual attention resources for multi-element processing, as measured through tasks of visual attention span prior to literacy instruction, predicted reading fluency performance one year later. One hundred and twenty-four mainstream children were assessed in kindergarten on their visual attention span abilities, phonological awareness, letter-name knowledge, early literacy knowledge, verbal short-term memory and non-verbal IQ. The participants' reading performance was measured at the end of grade 1 using tasks of irregular word, pseudo-word and text reading. Results from regression analyses showed that kindergarteners' VA span predicted reading fluency for text, irregular words and pseudo-words one year later, after controlling for age, non-verbal IQ, phonological skills, letter name knowledge and early literacy skills. Path analyses carried out to estimate the differential contribution of VA span to the different reading skills revealed a stronger contribution for pseudo-word reading than irregular word or text reading at the end of Grade 1. These results suggest that pre-reading visual attention resources contribute to later reading fluency, whatever the reading subskills and whatever the reading context (words in isolation or in sentences), with higher involvement to pseudo-word reading. We propose a new conceptual model of the role of visual attention in reading acquisition and argue that many aspects of the models are already supported by available findings.

Remediation of Allophonic Perception and Visual Attention Span in Developmental Dyslexia: A Joint Assay.

Categorical perception of phonemes and visual attention span are cognitive processes that contribute independently to poor reading skills in developmental dyslexia. We here explored whether training programs specifically targeting one or the other process do improve reading performance in dyslexic children. The dyslexic participants were trained using either the RapDys© program designed to improve phonemic perception or the MAEVA© program targeting visual attention span. Each participant was provided the two programs successively for intensive training. Results show specific effects of RapDys© on phonemic discrimination and pseudo-word reading. MAEVA© specifically improved visual attention span and irregular word reading. Phonemic awareness and regular word reading improved after application of both training programs, suggesting similar positive effects of both methods although effects of concomitant phonic training cannot be ruled out (as there was no control group). The overall findings suggest that both categorical perception and visual attention span remediation contribute to reading.

Modeling the length effect for words in lexical decision: The role of visual attention.

The word length effect in Lexical Decision (LD) has been studied in many behavioral experiments but no computational models has yet simulated this effect. We use a new Bayesian model of visual word recognition, the BRAID model, that simulates expert readers' performance. BRAID integrates an attentional component modeled by a Gaussian probability distribution, a mechanism of lateral interference between adjacent letters and an acuity gradient, but no phonological component. We explored the role of visual attention on the word length effect using 1,200 French words from 4 to 11 letters. A series of five simulations was carried out to assess (a) the impact of a single attentional focus versus multiple shifts of attention on the word length effect and (b) how this effect is modulated by variations in the distribution of attention. Results show that the model successfully simulates the word length effect reported for humans in the French Lexicon Project when allowing multiple shifts of attention for longer words. The magnitude and direction of the effect can be modulated depending on the use of a uniform or narrow distribution of attention. The present study provides evidence that visual attention is critical for the recognition of single words and that a narrowing of the attention distribution might account for the exaggerated length effect reported in some reading disorders.

Cross-Language Modulation of Visual Attention Span: An Arabic-French-Spanish Comparison in Skilled Adult Readers.

In delineating the amount of orthographic information that can be processed in parallel during a single fixation, the visual attention (VA) span acts as a key component of the reading system. Previous studies focused on the contribution of VA span to normal and pathological reading in monolingual and bilingual children from different European languages, without direct cross-language comparison. In the current paper, we explored modulations of VA span abilities in three languages -French, Spanish, and Arabic- that differ in transparency, reading direction and writing systems. The participants were skilled adult readers who were native speakers of French, Spanish or Arabic. They were administered tasks of global and partial letter report, single letter identification and text reading. Their VA span abilities were assessed using tasks that require the processing of briefly presented five consonant strings (e.g., R S H F T). All five consonants had to be reported in global report but a single cued letter in partial report. Results showed that VA span was reduced in Arabic readers as compared to French or Spanish readers who otherwise show a similar high performance in the two report tasks. The analysis of VA span response patterns in global report showed a left-right asymmetry in all three languages. A leftward letter advantage was found in French and Spanish but a rightward advantage in Arabic. The response patterns were symmetric in partial report, regardless of the language. Last, a significant relationship was found between VA span abilities and reading speed but only for French. The overall findings suggest that the size of VA span, the shape of VA span response patterns and the VA Span-reading relationship are modulated by language-specific features.

Relationships between Categorical Perception of Phonemes, Phoneme Awareness, and Visual Attention Span in Developmental Dyslexia.

We tested the hypothesis that the categorical perception deficit of speech sounds in developmental dyslexia is related to phoneme awareness skills, whereas a visual attention (VA) span deficit constitutes an independent deficit. Phoneme awareness tasks, VA span tasks and categorical perception tasks of phoneme identification and discrimination using a d/t voicing continuum were administered to 63 dyslexic children and 63 control children matched on chronological age. Results showed significant differences in categorical perception between the dyslexic and control children. Significant correlations were found between categorical perception skills, phoneme awareness and reading. Although VA span correlated with reading, no significant correlations were found between either categorical perception or phoneme awareness and VA span. Mediation analyses performed on the whole dyslexic sample suggested that the effect of categorical perception on reading might be mediated by phoneme awareness. This relationship was independent of the participants' VA span abilities. Two groups of dyslexic children with a single phoneme awareness or a single VA span deficit were then identified. The phonologically impaired group showed lower categorical perception skills than the control group but categorical perception was similar in the VA span impaired dyslexic and control children. The overall findings suggest that the link between categorical perception, phoneme awareness and reading is independent from VA span skills. These findings provide new insights on the heterogeneity of developmental dyslexia. They suggest that phonological processes and VA span independently affect reading acquisition.

Auditory Stream Segregation in Autism Spectrum Disorder: Benefits and Downsides of Superior Perceptual Processes.

Auditory stream segregation allows us to organize our sound environment, by focusing on specific information and ignoring what is unimportant. One previous study reported difficulty in stream segregation ability in children with Asperger syndrome. In order to investigate this question further, we used an interleaved melody recognition task with children in the autism spectrum disorder (ASD). In this task, a probe melody is followed by a mixed sequence, made up of a target melody interleaved with a distractor melody. These two melodies have either the same [0 semitone (ST)] or a different mean frequency (6, 12 or 24 ST separation conditions). Children have to identify if the probe melody is present in the mixed sequence. Children with ASD performed better than typical children when melodies were completely embedded. Conversely, they were impaired in the ST separation conditions. Our results confirm the difficulty of children with ASD in using a frequency cue to organize auditory perceptual information. However, superior performance in the completely embedded condition may result from superior perceptual processes in autism. We propose that this atypical pattern of results might reflect the expression of a single cognitive feature in autism.

An Analysis of Reading Skill Development using E-Z Reader.

Previously reported simulations using the E-Z Reader model of eye-movement control suggest that the patterns of eye movements observed with children versus adult readers reflect differences in lexical processing proficiency (Reichle et al., 2013). However, these simulations fail to specify precisely what aspect(s) of lexical processing (e.g., orthographic processing) account for the concurrent changes in eye movements and reading skill. To examine this issue, the E-Z Reader model was first used to simulate the aggregate eye-movement data from 15 adults and 75 children to replicate the finding that gross differences in reading skill can be accounted for by differences in lexical processing proficiency. The model was then used to simulate the eye-movement data of individual children so that the best-fitting lexical-processing parameters could be correlated to measures of orthographic knowledge, phonological-processing skill, sentence comprehension, and general intelligence. These analyses suggest that orthographic knowledge accounts for variance in the eye-movement measures that is observed with between-individual differences in reading skill. The theoretical implications of this conclusion will be discussed in relation to computational models of reading and our understanding of reading skill development.

The phonological and visual basis of developmental dyslexia in Brazilian Portuguese reading children.

Evidence from opaque languages suggests that visual attention processing abilities in addition to phonological skills may act as cognitive underpinnings of developmental dyslexia. We explored the role of these two cognitive abilities on reading fluency in Brazilian Portuguese, a more transparent orthography than French or English. Sixty-six children with developmental dyslexia and normal Brazilian Portuguese children participated. They were administered three tasks of phonological skills (phoneme identification, phoneme, and syllable blending) and three visual tasks (a letter global report task and two non-verbal tasks of visual closure and visual constancy). Results show that Brazilian Portuguese children with developmental dyslexia are impaired not only in phonological processing but further in visual processing. The phonological and visual processing abilities significantly and independently contribute to reading fluency in the whole population. Last, different cognitively homogeneous subtypes can be identified in the Brazilian Portuguese population of children with developmental dyslexia. Two subsets of children with developmental dyslexia were identified as having a single cognitive disorder, phonological or visual; another group exhibited a double deficit and a few children showed no visual or phonological disorder. Thus the current findings extend previous data from more opaque orthographies as French and English, in showing the importance of investigating visual processing skills in addition to phonological skills in children with developmental dyslexia whatever their language orthography transparency.

Visual processing of multiple elements in the dyslexic brain: evidence for a superior parietal dysfunction.

The visual attention (VA) span deficit hypothesis of developmental dyslexia posits that impaired multiple element processing can be responsible for poor reading outcomes. In VA span impaired dyslexic children, poor performance on letter report tasks is associated with reduced parietal activations for multiple letter processing. While this hints towards a non-specific, attention-based dysfunction, it is still unclear whether reduced parietal activity generalizes to other types of stimuli. Furthermore, putative links between reduced parietal activity and reduced ventral occipito-temporal (vOT) in dyslexia have yet to be explored. Using functional magnetic resonance imaging, we measured brain activity in 12 VA span impaired dyslexic adults and 12 adult skilled readers while they carried out a categorization task on single or multiple alphanumeric or non-alphanumeric characters. While healthy readers activated parietal areas more strongly for multiple than single element processing (right-sided for alphanumeric and bilateral for non-alphanumeric), similar stronger multiple element right parietal activations were absent for dyslexic participants. Contrasts between skilled and dyslexic readers revealed significantly reduced right superior parietal lobule (SPL) activity for dyslexic readers regardless of stimuli type. Using a priori anatomically defined regions of interest, we showed that neural activity was reduced for dyslexic participants in both SPL and vOT bilaterally. Finally, we used multiple regressions to test whether SPL activity was related to vOT activity in each group. In the left hemisphere, SPL activity covaried with vOT activity for both normal and dyslexic readers. In contrast, in the right hemisphere, SPL activity covaried with vOT activity only for dyslexic readers. These results bring critical support to the VA interpretation of the VA Span deficit. In addition, they offer a new insight on how deficits in automatic vOT based word recognition could arise in developmental dyslexia.

New insights on developmental dyslexia subtypes: heterogeneity of mixed reading profiles.

We examined whether classifications based on reading performance are relevant to identify cognitively homogeneous subgroups of dyslexic children. Each of the 71 dyslexic participants was selected to have a mixed reading profile, i.e. poor irregular word and pseudo-word reading performance (accuracy and speed). Despite their homogeneous reading profile, the participants were found to split into four distinct cognitive subgroups, characterized by a single phonological disorder, a single visual attention span disorder, a double deficit or none of these disorders. The two subgroups characterized by single and contrasted cognitive disorders were found to exhibit a very similar reading pattern but more contrasted spelling performance (quantitative analysis). A qualitative analysis of the error types produced in reading and spelling provided some cues about the participants' underlying cognitive deficit. The overall findings disqualify subtyping based on reading profiles as a classification method to identify cognitively homogeneous subgroups of dyslexic children. They rather show an opaque relationship between the cognitive underpinnings of developmental dyslexia and their behavioral manifestations in reading and spelling. Future neuroimaging and genetic studies should take this issue into account since synthesizing over cognitively heterogeneous children would entail potential pitfalls.

Veridical mapping in savant abilities, absolute pitch, and synesthesia: an autism case study.

An enhanced role and autonomy of perception are prominent in autism. Furthermore, savant abilities, absolute pitch, and synesthesia are all more commonly found in autistic individuals than in the typical population. The mechanism of veridical mapping has been proposed to account for how enhanced perception in autism leads to the high prevalence of these three phenomena and their structural similarity. Veridical mapping entails functional rededication of perceptual brain regions to higher order cognitive operations, allowing the enhanced detection and memorization of isomorphisms between perceptual and non-perceptual structures across multiple scales. In this paper, we present FC, an autistic individual who possesses several savant abilities in addition to both absolute pitch and synesthesia-like associations. The co-occurrence in FC of abilities, some of them rare, which share the same structure, as well as FC's own accounts of their development, together suggest the importance of veridical mapping in the atypical range and nature of abilities displayed by autistic people.

Impact of orthographic transparency on typical and atypical reading development: evidence in French-Spanish bilingual children.

The present study aimed to quantify cross-linguistic modulations of the contribution of phonemic awareness skills and visual attention span (VA Span) skills (number of visual elements that can be processed simultaneously) to reading speed and accuracy in 18 Spanish-French balanced bilingual children with and without developmental dyslexia. The children were administered two similar reading batteries in French and Spanish. The deficits of the dyslexic children in reading accuracy were mainly visible in their opaque orthography (French) whereas difficulties indexed by reading speed were observed in both their opaque and transparent orthographies. Dyslexic children did not exhibit any phonemic awareness problems in French or in Spanish, but showed poor VA Span skills compared to their control peers. VA span skills correlated with reading accuracy and speed measures in both Spanish and French, whereas phonemic awareness correlated with reading accuracy only. Overall, the present results show that the VA Span is tightly related to reading speed regardless of orthographic transparency, and that it accounts for differences in reading performance between good and poor readers across languages. The present findings further suggest that VA Span skills may play a particularly important role in building-up specific word knowledge which is critical for lexical reading strategies.

Lexical orthography acquisition: Is handwriting better than spelling aloud?

Lexical orthography acquisition is currently described as the building of links between the visual forms and the auditory forms of whole words. However, a growing body of data suggests that a motor component could further be involved in orthographic acquisition. A few studies support the idea that reading plus handwriting is a better lexical orthographic learning situation than reading alone. However, these studies did not explore which of the cognitive processes involved in handwriting enhanced lexical orthographic acquisition. Some findings suggest that the specific movements memorized when learning to write may participate in the establishment of orthographic representations in memory. The aim of the present study was to assess this hypothesis using handwriting and spelling aloud as two learning conditions. In two experiments, fifth graders were asked to read complex pseudo-words embedded in short sentences. Immediately after reading, participants had to recall the pseudo-words' spellings either by spelling them aloud or by handwriting them down. One week later, orthographic acquisition was tested using two post-tests: a pseudo-word production task (spelling by hand in Experiment 1 or spelling aloud in Experiment 2) and a pseudo-word recognition task. Results showed no significant difference in pseudo-word recognition between the two learning conditions. In the pseudo-word production task, orthography learning improved when the learning and post-test conditions were similar, thus showing a massive encoding-retrieval match effect in the two experiments. However, a mixed model analysis of the pseudo-word production results revealed a significant learning condition effect which remained after control of the encoding-retrieval match effect. This later finding suggests that orthography learning is more efficient when mediated by handwriting than by spelling aloud, whatever the post-test production task.