Top-down modulation of ventral occipito-temporal responses during visual word recognition

Semantic dementia in Arabic: An assessment of Arabic word reading within sentences

INTRODUCTION

Language impairments have not yet been fully explored in native Arabic speakers with semantic dementia (SD). The aim of this paper is to describe the impairments in language in two Saudi Arabians with SD and to determine if their word reading within a sentence context would result in incorrect responses.

METHODS

Two patients with semantic dementia (one with right > left and the other with left > right temporal involvement) underwent a reading assessment in Arabic. Patients were asked to read a series of words within a sentence context in which the correct reading of the word was dependent on the context of the sentence. Thirty-four sentences were designed in which 17 Arabic homographs were used. The same homograph would occur in two separate sentences, in which the pronunciation and meaning would differ between sentences. Patients were also assessed using five other sentences that contained irregular pronouns of high frequency. Eighteen healthy controls were used as reference.

RESULTS

Both patients made errors in reading the target Arabic homographs; this was more pronounced in the patient with left > right variant of SD. The patient with right > left variant of SD also suffered from prosopagnosia.

CONCLUSION

Correct reading of Arabic words within the sentence context may be impaired from semantic language impairments in semantic dementia. The role of comprehension in the correct reading of words in Arabic sentences is important.

Sensory modality and spoken language shape reading network in blind readers of Braille

The neural basis of reading is highly consistent across many languages and scripts. Are there alternative neural routes to reading? How does the sensory modality of symbols (tactile vs. visual) influence their neural representations? We examined these questions by comparing reading of visual print (sighted group, n = 19) and tactile Braille (congenitally blind group, n = 19). Blind and sighted readers were presented with written (words, consonant strings, non-letter shapes) and spoken stimuli (words, backward speech) that varied in word-likeness. Consistent with prior work, the ventral occipitotemporal cortex (vOTC) was active during Braille and visual reading. A posterior/anterior vOTC word-form gradient was observed only in sighted readers with more anterior regions preferring larger orthographic units (words). No such gradient was observed in blind readers. Consistent with connectivity predictions, in blind compared to sighted readers, posterior parietal cortices were recruited to a greater degree and contained word-preferring patches. Lateralization of Braille in blind readers was predicted by laterality of spoken language and reading hand. The effect of spoken language increased along a cortical hierarchy, whereas effect of reading hand waned. These results suggested that the neural basis of reading is influenced by symbol modality and spoken language and support connectivity-based views of cortical function.

Task modulates the orthographic and phonological representations in the bilateral ventral Occipitotemporal cortex

As a key area in word reading, the left ventral occipitotemporal cortex is proposed for abstract orthographic processing, and its middle part has even been labeled as the visual word form area. Because the definition of the VWFA largely varies and the reading task differs across studies, the function of the left ventral occipitotemporal cortex in word reading is continuingly debated on whether this region is specific for orthographic processing or be involved in an interactive framework. By using representational similarity analysis (RSA), this study examined information representation in the VWFA at the individual level and the modulatory effect of reading task. Twenty-four subjects were scanned while performing the explicit (i.e., the naming task) and implicit (i.e., the perceptual task) reading tasks. Activation analysis showed that the naming task elicited greater activation in regions related to phonological processing (e.g., the bilateral prefrontal cortex and temporoparietal cortex), while the perceptual task recruited greater activation in visual cortex and default mode network (e.g., the bilateral middle frontal gyrus, angular gyrus, and the right middle temporal gyrus). More importantly, RSA also showed that task modulated information representation in the bilateral anterior occipitotemporal cortex and VWFA. Specifically, ROI-based RSA revealed enhanced orthographic and phonological representations in the bilateral anterior fusiform cortex and VWFA in the naming task relative to the perceptual task. These results suggest that lexical representation in the VWFA is influenced by the demand of phonological processing, which supports the interactive account of the VWFA.

Level of Orthographic Knowledge Helps to Reveal Automatic Predictions in Visual Word Processing

The brain generates predictions about visual word forms to support efficient reading. The “interactive account” suggests that the predictions in visual word processing can be strategic or automatic (non-strategic). Strategic predictions are frequently demonstrated in studies that manipulated task demands, however, few studies have investigated automatic predictions. Orthographic knowledge varies greatly among individuals and it offers a unique opportunity in revealing automatic predictions. The present study grouped the participants by level of orthographic knowledge and recorded EEGs in a non-linguistic color matching task. The visual word-selective N170 response was much stronger to pseudo than to real characters in participants with low orthographic knowledge, but not in those with high orthographic knowledge. Previous work on predictive coding has demonstrated that N170 is a good index for prediction errors, i.e., the mismatches between predictions and visual inputs. The present findings provide unambiguous evidence that automatic predictions modulate the early stage of visual word processing.

Task-Relevant Representations and Cognitive Control Demands Modulate Functional Connectivity from Ventral Occipito-Temporal Cortex During Object Recognition Tasks

The left ventral occipito-temporal cortex (vOTC) supports extraction and processing of visual features. However, it has remained unclear whether left vOTC-based functional connectivity (FC) differs according to task-relevant representations (e.g., lexical, visual) and control demands imposed by the task, even when similar visual-semantic processing is required for object identification. Here, neural responses to the same set of pictures of meaningful objects were measured, while the type of task that participants had to perform (picture naming versus size-judgment task), and the level of cognitive control required by the picture naming task (high versus low interference contexts) were manipulated. Explicit retrieval of lexical representations in the picture naming task facilitated activation of lexical/phonological representations, modulating FC between left vOTC and dorsal anterior-cingulate-cortex/pre-supplementary-motor-area. This effect was not observed in the size-judgment task, which did not require explicit word-retrieval of object names. Furthermore, retrieving the very same lexical/phonological representation in the high versus low interference contexts during picture naming increased FC between left vOTC and left caudate. These findings support the proposal that vOTC functional specialization emerges from interactions with task-relevant brain regions.

Emergence of a compositional neural code for written words: Recycling of a convolutional neural network for reading

The visual word form area (VWFA) is a region of human inferotemporal cortex that emerges at a fixed location in the occipitotemporal cortex during reading acquisition and systematically responds to written words in literate individuals. According to the neuronal recycling hypothesis, this region arises through the repurposing, for letter recognition, of a subpart of the ventral visual pathway initially involved in face and object recognition. Furthermore, according to the biased connectivity hypothesis, its reproducible localization is due to preexisting connections from this subregion to areas involved in spoken-language processing. Here, we evaluate those hypotheses in an explicit computational model. We trained a deep convolutional neural network of the ventral visual pathway, first to categorize pictures and then to recognize written words invariantly for case, font, and size. We show that the model can account for many properties of the VWFA, particularly when a subset of units possesses a biased connectivity to word output units. The network develops a sparse, invariant representation of written words, based on a restricted set of reading-selective units. Their activation mimics several properties of the VWFA, and their lesioning causes a reading-specific deficit. The model predicts that, in literate brains, written words are encoded by a compositional neural code with neurons tuned either to individual letters and their ordinal position relative to word start or word ending or to pairs of letters (bigrams).

A Representational Similarity Analysis of Cognitive Control during Color-Word Stroop

Progress in understanding the neural bases of cognitive control has been supported by the paradigmatic color-word Stroop task, in which a target response (color name) must be selected over a more automatic, yet potentially incongruent, distractor response (word). For this paradigm, models have postulated complementary coding schemes: dorsomedial frontal cortex (DMFC) is proposed to evaluate the demand for control via incongruency-related coding, whereas dorsolateral PFC (DLPFC) is proposed to implement control via goal and target-related coding. Yet, mapping these theorized schemes to measured neural activity within this task has been challenging. Here, we tested for these coding schemes relatively directly, by decomposing an event-related color-word Stroop task via representational similarity analysis. Three neural coding models were fit to the similarity structure of multivoxel patterns of human fMRI activity, acquired from 65 healthy, young-adult males and females. Incongruency coding was predominant in DMFC, whereas both target and incongruency coding were present with indistinguishable strength in DLPFC. In contrast, distractor information was strongly encoded within early visual cortex. Further, these coding schemes were differentially related to behavior: individuals with stronger DLPFC (and lateral posterior parietal cortex) target coding, but weaker DMFC incongruency coding, exhibited less behavioral Stroop interference. These results highlight the utility of the representational similarity analysis framework for investigating neural mechanisms of cognitive control and point to several promising directions to extend the Stroop paradigm.SIGNIFICANCE STATEMENT How the human brain enables cognitive control - the ability to override behavioral habits to pursue internal goals - has been a major focus of neuroscience research. This ability has been frequently investigated by using the Stroop color-word naming task. With the Stroop as a test-bed, many theories have proposed specific neuroanatomical dissociations, in which medial and lateral frontal brain regions underlie cognitive control by encoding distinct types of information. Yet providing a direct confirmation of these claims has been challenging. Here, we demonstrate that representational similarity analysis, which estimates and models the similarity structure of brain activity patterns, can successfully establish the hypothesized functional dissociations within the Stroop task. Representational similarity analysis may provide a useful approach for investigating cognitive control mechanisms.

Reading and alexia

Alexia refers to a reading disorder caused by some form of acquired brain pathology, most commonly a stroke or tumor, in a previously literate subject. In neuropsychology, a distinction is made between central alexia (commonly seen in aphasia) and peripheral alexia (a perceptual or attentional deficit). The prototypical peripheral alexia is alexia without agraphia (pure alexia), where patients can write but are impaired in reading words and letters. Pure alexia is associated with damage to the left ventral occipitotemporal cortex (vOT) or its connections. Hemianopic alexia is associated with less extensive occipital damage and is caused by a visual field defect, which creates problems reading longer words and passages of text. Reading impairment can also arise due to attentional deficits, most commonly following right hemisphere or bilateral lesions. Studying patients with alexia, along with functional imaging studies of normal readers, has improved our understanding of the neurobiological processes involved in reading. A key question is whether an area in the left ventral occipitotemporal cortex is specialized for or selectively involved in word processing, or whether reading relies on tuning of more general purpose perceptual areas. Reading deficits may also be observed in dementia and traumatic brain injury, but often with less consistent deficit patterns than in patients with focal lesions.

Distinct mechanisms drive hemispheric lateralization of object recognition in the visual word form and fusiform face areas

The Visual Word Form Area (VWFA) and the Fusiform Face Area (FFA) represent classical examples of functional lateralization. The known hypothesis that lateralization of the VWFA and FFA are related remains controversial. We hypothesized that lateralization is independent and might be associated with lateralized high-level top-down mechanisms. For the VWFA this could emerge from left-lateralized language regions. This driving force might modulate local reorganization/recycling of function. Using an fMRI recognition paradigm, we quantified lateralization and investigated effective connectivity to examine mechanisms associated with lateralization in these regions (n = 58). Laterality patterns were more pronounced for VWFA than for FFA. Granger Causality Analysis found top-down effects only for the VWFA (left-lateralized, stemming from Broca's area). FFA exerted top-down effects on low-level visual areas. These findings suggest that distinct mechanisms are associated with hemispheric lateralization in object recognition: left lateralized top-down for VWFA and only early visual top-down effects concerning the FFA.

The contribution of phonological information to visual word recognition: Evidence from Chinese phonetic radicals

Lateralization is a critical characteristic of language production and also plays a role in visual word recognition. However, the neural mechanisms underlying the interactions between visual input and spoken word representations are still unclear. We investigated the contribution of sub-lexical phonological information in visual word processing by exploiting the fact that Chinese characters can contain phonetic radicals in either the left or right half of the character. FMRI data were collected while 39 Chinese participants read words in search of target color words. On the basis of whole-brain analysis and three laterality analyses of regions of interest, we argue that visual information from centrally presented Chinese characters is split in the fovea and projected to the contralateral visual cortex, from which phonological information can be extracted rapidly if the character contains a phonetic radical. Extra activation, suggestive of more effortful processing, is observed when the phonetic radical is situated in the left half of the character and therefore initially sent to the visual cortex in the right hemisphere that is less specialized for language processing. Our results are in line with the proposal that phonological information helps written word processing by means of top-down feedback.

Neural Components of Reading Revealed by Distributed and Symbolic Computational Models

Determining how the cognitive components of reading - orthographic, phonological, and semantic representations - are instantiated in the brain has been a longstanding goal of psychology and human cognitive neuroscience. The two most prominent computational models of reading instantiate different cognitive processes, implying different neural processes. Artificial neural network (ANN) models of reading posit non-symbolic, distributed representations. The dual-route cascaded (DRC) model instead suggests two routes of processing, one representing symbolic rules of spelling-sound correspondence, the other representing orthographic and phonological lexicons. These models are not adjudicated by behavioral data and have never before been directly compared in terms of neural plausibility. We used representational similarity analysis to compare the predictions of these models to neural data from participants reading aloud. Both the ANN and DRC model representations corresponded with neural activity. However, ANN model representations correlated to more reading-relevant areas of cortex. When contributions from the DRC model were statistically controlled, partial correlations revealed that the ANN model accounted for significant variance in the neural data. The opposite analysis, examining the variance explained by the DRC model with contributions from the ANN model factored out, revealed no correspondence to neural activity. Our results suggest that ANNs trained using distributed representations provide a better correspondence between cognitive and neural coding. Additionally, this framework provides a principled approach for comparing computational models of cognitive function to gain insight into neural representations.

Taking the sublexical route: brain dynamics of reading in the semantic variant of primary progressive aphasia

Reading aloud requires mapping an orthographic form to a phonological one. The mapping process relies on sublexical statistical regularities (e.g. 'oo' to |uː|) or on learned lexical associations between a specific visual form and a series of sounds (e.g. yacht to/jɑt/). Computational, neuroimaging, and neuropsychological evidence suggest that sublexical, phonological and lexico-semantic processes rely on partially distinct neural substrates: a dorsal (occipito-parietal) and a ventral (occipito-temporal) route, respectively. Here, we investigated the spatiotemporal features of orthography-to-phonology mapping, capitalizing on the time resolution of magnetoencephalography and the unique clinical model offered by patients with semantic variant of primary progressive aphasia (svPPA). Behaviourally, patients with svPPA manifest marked lexico-semantic impairments including difficulties in reading words with exceptional orthographic to phonological correspondence (irregular words). Moreover, they present with focal neurodegeneration in the anterior temporal lobe, affecting primarily the ventral, occipito-temporal, lexical route. Therefore, this clinical population allows for testing of specific hypotheses on the neural implementation of the dual-route model for reading, such as whether damage to one route can be compensated by over-reliance on the other. To this end, we reconstructed and analysed time-resolved whole-brain activity in 12 svPPA patients and 12 healthy age-matched control subjects while reading irregular words (e.g. yacht) and pseudowords (e.g. pook). Consistent with previous findings that the dorsal route is involved in sublexical, phonological processes, in control participants we observed enhanced neural activity over dorsal occipito-parietal cortices for pseudowords, when compared to irregular words. This activation was manifested in the beta-band (12-30 Hz), ramping up slowly over 500 ms after stimulus onset and peaking at ∼800 ms, around response selection and production. Consistent with our prediction, svPPA patients did not exhibit this temporal pattern of neural activity observed in controls this contrast. Furthermore, a direct comparison of neural activity between patients and controls revealed a dorsal spatiotemporal cluster during irregular word reading. These findings suggest that the sublexical/phonological route is involved in processing both irregular and pseudowords in svPPA. Together these results provide further evidence supporting a dual-route model for reading aloud mediated by the interplay between lexico-semantic and sublexical/phonological neurocognitive systems. When the ventral route is damaged, as in the case of neurodegeneration affecting the anterior temporal lobe, partial compensation appears to be possible by over-recruitment of the slower, serial attention-dependent, dorsal one.

Top-down activation of the visuo-orthographic system during spoken sentence processing

The left ventral occipitotemporal cortex (vOT) is considered the key area of the visuo-orthographic system. However, some studies reported that the area is also involved in speech processing tasks, especially those that require activation of orthographic knowledge. These findings suggest the existence of a top-down activation mechanism allowing such cross-modal activation. Yet, little is known about the involvement of the vOT in more natural speech processing situations like spoken sentence processing. Here, we addressed this issue in a functional Magnetic Resonance Imaging (fMRI) study while manipulating the impacts of two factors, i.e., task demands (semantic vs. low-level perceptual task) and the quality of speech signals (sentences presented against clear vs. noisy background). Analyses were performed at the levels of whole brain and region-of-interest (ROI) focusing on the vOT voxels individually identified through a reading task. Whole brain analysis showed that processing spoken sentences induced activity in a large network including the regions typically involved in phonological, articulatory, semantic and orthographic processing. ROI analysis further specified that a significant part of the vOT voxels that responded to written words also responded to spoken sentences, thus, suggesting that the same area within the left occipitotemporal pathway contributes to both reading and speech processing. Interestingly, both analyses provided converging evidence that vOT responses to speech were sensitive to both task demands and quality of speech signals: Compared to the low-level perceptual task, activity of the area increased when efforts on comprehension were required. The impact of background noise depended on task demands. It led to a decrease of vOT activity in the semantic task but not in the low-level perceptual task. Our results provide new insights into the function of this key area of the reading network, notably by showing that its speech-induced top-down activation also generalizes to ecological speech processing situations.

Deaf readers benefit from lexical feedback during orthographic processing

It has been proposed that poor reading abilities in deaf readers might be related to weak connections between the orthographic and lexical-semantic levels of processing. Here we used event related potentials (ERPs), known for their excellent time resolution, to examine whether lexical feedback modulates early orthographic processing. Twenty congenitally deaf readers made lexical decisions to target words and pseudowords. Each of those target stimuli could be preceded by a briefly presented matched-case or mismatched-case identity prime (e.g., ALTAR-ALTAR vs. altar- ALTAR). Results showed an early effect of case overlap at the N/P150 for all targets. Critically, this effect disappeared for words but not for pseudowords, at the N250—an ERP component sensitive to orthographic processing. This dissociation in the effect of case for word and pseudowords targets provides strong evidence of early automatic lexical-semantic feedback modulating orthographic processing in deaf readers. Interestingly, despite the dissociation found in the ERP data, behavioural responses to words still benefited from the physical overlap between prime and target, particularly in less skilled readers and those with less experience with words. Overall, our results support the idea that skilled deaf readers have a stronger connection between the orthographic and the lexical-semantic levels of processing.

Modulating Effects of Contextual Emotions on the Neural Plasticity Induced by Word Learning

Recently, numerous studies have investigated the neurocognitive mechanism of learning words in isolation or in semantic contexts. However, emotion as an important influencing factor on novel word learning has not been fully considered in the previous studies. In addition, the effects of emotion on word learning and the underlying neural mechanism have not been systematically investigated. Sixteen participants were trained to learn novel concrete or abstract words under negative, neutral, and positive contextual emotions over 3 days; then, fMRI scanning was done during the testing sessions on day 1 and day 3. We compared the brain activations in day 1 and day 3 to investigate the role of contextual emotions in learning different types of words and the corresponding neural plasticity changes. Behaviorally, the performance of the words learned in the negative context was lower than those in the neutral and positive contexts, which indicated that contextual emotions had a significant impact on novel word learning. Correspondingly, the functional plasticity changes of the right angular gyrus (AG), bilateral insula, and anterior cingulate cortex (ACC) induced by word learning were modulated by the contextual emotions. The insula also was sensitive to the concreteness of the learned words. More importantly, the functional plasticity changes of the left inferior frontal gyrus (left IFG) and left fusiform gyrus (left FG) were interactively influenced by the contextual emotions and concreteness, suggesting that the contextual emotional information had a discriminable effect on different types of words in the neural mechanism level. These results demonstrate that emotional information in contexts is inevitably involved in word learning. The role of contextual emotions in brain plasticity for learning is discussed.

Converging evidence for functional and structural segregation within the left ventral occipitotemporal cortex in reading

The ventral occipitotemporal cortex (vOTC) is crucial for recognizing visual patterns, and previous evidence suggests that there may be different subregions within the vOTC involved in the rapid identification of word forms. Here, we characterize vOTC reading circuitry using a multimodal approach combining functional, structural, and quantitative MRI and behavioral data. Two main word-responsive vOTC areas emerged: a posterior area involved in visual feature extraction, structurally connected to the intraparietal sulcus via the vertical occipital fasciculus; and an anterior area involved in integrating information with other regions of the language network, structurally connected to the angular gyrus via the posterior arcuate fasciculus. Furthermore, functional activation in these vOTC regions predicted reading behavior outside of the scanner. Differences in the microarchitectonic properties of gray-matter cells in these segregated areas were also observed, in line with earlier cytoarchitectonic evidence. These findings advance our understanding of the vOTC circuitry by linking functional responses to anatomical structure, revealing the pathways of distinct reading-related processes.

Orthographic and phonological selectivity across the reading system in deaf skilled readers

People who are born deaf often have difficulty learning to read. Recently, several studies have examined the neural substrates involved in reading in deaf people and found a left lateralized reading system similar to hearing people involving temporo-parietal, inferior frontal, and ventral occipito-temporal cortices. Previous studies in typical hearing readers show that within this reading network there are separate regions that specialize in processing orthography and phonology. We used fMRI rapid adaptation in deaf adults who were skilled readers to examine neural selectivity in three functional ROIs in the left hemisphere: temporoparietal cortex (TPC), inferior frontal gyrus (IFG), and the visual word form area (VWFA). Results show that in deaf skilled readers, the left VWFA showed selectivity for orthography similar to what has been reported for hearing readers, the TPC showed less sensitivity to phonology than previously reported for hearing readers using the same paradigm, and the IFG showed selectivity to orthography, but not phonology (similar to what has been reported previously for hearing readers). These results provide evidence that while skilled deaf readers demonstrate coarsely tuned phonological representations in the TPC, they develop finely tuned representations for the orthography of written words in the VWFA and IFG. This result suggests that phonological tuning in the TPC may have little impact on the neural network associated with skilled reading for deaf adults.

Functional Dynamics of Dorsal and Ventral Reading Networks in Bilinguals

In today's world, bilingualism is increasingly common. However, it is still unclear how left-lateralized dorsal and ventral reading networks are tuned to reading in proficient second-language learners. Here, we investigated differences in functional regional activation and connectivity as a function of L1 and L2 reading, L2 orthographic depth, and task demands. Thirty-seven late bilinguals with the same L1 and either an opaque or transparent L2 performed perceptual and semantic reading tasks in L1 and L2 during functional magnetic resonance imaging (fMRI) scanning. Results revealed stronger regional recruitment for L2 versus L1 reading and stronger connectivity within the dorsal stream during L1 versus L2 reading. Differences in orthographic depth were associated with a segregated profile of left ventral occipitotemporal (vOT) coactivation with dorsal regions for the transparent L2 group and with ventral regions for the opaque L2 group. Finally, semantic versus perceptual demands modulated left vOT engagement, supporting the interactive account of the contribution of vOT to reading, and were associated with stronger coactivation within the ventral network. Our findings support a division of labor between ventral and dorsal reading networks, elucidating the critical role of the language used to read, L2 orthographic depth, and task demands on the functional dynamics of bilingual reading.

Representational similarity analysis reveals task-dependent semantic influence of the visual word form area

Access to semantic information of visual word forms is a key component of reading comprehension. In this study, we examined the involvement of the visual word form area (VWFA) in this process by investigating whether and how the activity patterns of the VWFA are influenced by semantic information during semantic tasks. We asked participants to perform two semantic tasks - taxonomic or thematic categorization - on visual words while obtaining the blood-oxygen-level-dependent (BOLD) fMRI responses to each word. Representational similarity analysis with four types of semantic relations (taxonomic, thematic, subjective semantic rating and word2vec) revealed that neural activity patterns of the VWFA were associated with taxonomic information only in the taxonomic task, with thematic information only in the thematic task and with the composite semantic information measured by word2vec in both semantic tasks. Furthermore, the semantic information in the VWFA cannot be explained by confounding factors including orthographic, low-level visual and phonological information. These findings provide positive evidence for the presence of both orthographic and task-relevant semantic information in the VWFA and have significant implications for the neurobiological basis of reading.

How many deficits in the same dyslexic brains? A behavioural and fMRI assessment of comorbidity in adult dyslexics

Dyslexia can have different manifestations: this has motivated different theories on its nature, on its underlying brain bases and enduring controversies on how to best treat it. The relative weight of the different manifestations has never been evaluated using both behavioural and fMRI measures, a challenge taken here to assess the major systems called into play in dyslexia by different theories. We found that adult well-compensated dyslexics were systematically impaired only in reading and in visuo-phonological tasks, while deficits for other systems (e.g., motor/cerebellar, visual magnocellular/motion perception) were only very occasional. In line with these findings, fMRI showed a reliable hypoactivation only for the task of reading, in the left occipito-temporal cortex (l-OTC). The l-OTC, normally a crossroad between the reading system and other systems, did not show the same level of intersection in dyslexics; yet, it was not totally silent because it responded, in segregated parts, during auditory phonological and visual motion perception tasks. This minimal behavioural and functional anatomical comorbidity demonstrates that a specific deficit of reading is the best description for developmental dyslexia, at least for adult well-compensated cases, with clear implications for rehabilitation strategies. The reduced intersection of multiple systems in the l-OTC suggests that dyslexics suffer from a coarser connectivity, leading to disconnection between the multiple domains that normally interact during reading.

Neural correlates of bilingual language control during interlingual homograph processing in a logogram writing system

Bilingual studies using alphabetic languages have shown parallel activation of two languages during word recognition. However, little is known about the brain mechanisms of language control during word comprehension with a logogram writing system. We manipulated the types of words (interlingual homographs (IH), cognates, and language-specific words) and the types of participants (Chinese (L1)-Japanese (L2) bilinguals vs. Japanese monolinguals). Greater activation was found in the bilateral inferior frontal gyri, supplementary motor area, caudate nucleus and left fusiform gyrus, when the bilinguals processed IH, as compared to cognates. These areas were also commonly activated when the bilinguals processed L2 control words during an L1 lexical decision task. The areas function as the task/decision system that plays a role in cognitive control for resolving response conflict. Furthermore, the anterior cingulate cortex, left thalamus, and left middle temporal gyrus were activated during IH processing, suggesting resolution of the semantic conflict at the stimulus level (i.e., one logographic word having different meanings in the two languages).

Seeing the Meaning: Top-Down Effects on Letter Identification

Most models of reading agree that visual word recognition is underpinned by a highly interactive network in which both bottom–up and top–down processes contribute. What remains unknown is whether evidence of top–down effects upon letter processing are restricted to word-form level information, or whether meaning-level information also plays a role. Here we sought to investigate top–down semantic influences upon letter detection using semantic manipulations of real word imageability and semantic priming, as well as a manipulation of nonword orthographic and phonological composition which varied degree of similarity to real words. A continuous adaptive staircase procedure was used, allowing us to assess the exposure duration needed for accurate letter perception in different stimulus types. Results revealed that in terms of both exposure duration and decision reaction times, words showed an advantage over pseudohomophones and pseudowords, which in turn showed advantages over consonant strings. High imageability words were processed more efficiently than low imageability words, both in terms of the exposure duration required for accurate letter identification and also decision reaction times. The presence of a related as opposed to unrelated semantic prime significantly shortened exposure duration, but also lengthened decision reaction times. This inhibitory semantic priming effect in reaction time was attributed to the interference at the decision stage by stronger activation of the prime letters in the case of related relative to unrelated trials. Taken together, the present results establish for the first time that the semantic dimensions of imageability and semantic priming exert significant effects on letter identification, indicating meaning-level influences on the very earliest stages of written word recognition.

Bottom-up and top-down computations in word- and face-selective cortex

The ability to read a page of text or recognize a person's face depends on category-selective visual regions in ventral temporal cortex (VTC). To understand how these regions mediate word and face recognition, it is necessary to characterize how stimuli are represented and how this representation is used in the execution of a cognitive task. Here, we show that the response of a category-selective region in VTC can be computed as the degree to which the low-level properties of the stimulus match a category template. Moreover, we show that during execution of a task, the bottom-up representation is scaled by the intraparietal sulcus (IPS), and that the level of IPS engagement reflects the cognitive demands of the task. These results provide an account of neural processing in VTC in the form of a model that addresses both bottom-up and top-down effects and quantitatively predicts VTC responses.

Automaticity of phonological and semantic processing during visual word recognition

Reading involves activation of phonological and semantic knowledge. Yet, the automaticity of the activation of these representations remains subject to debate. The present study addressed this issue by examining how different brain areas involved in language processing responded to a manipulation of bottom-up (level of visibility) and top-down information (task demands) applied to written words. The analyses showed that the same brain areas were activated in response to written words whether the task was symbol detection, rime detection, or semantic judgment. This network included posterior, temporal and prefrontal regions, which clearly suggests the involvement of orthographic, semantic and phonological/articulatory processing in all tasks. However, we also found interactions between task and stimulus visibility, which reflected the fact that the strength of the neural responses to written words in several high-level language areas varied across tasks. Together, our findings suggest that the involvement of phonological and semantic processing in reading is supported by two complementary mechanisms. First, an automatic mechanism that results from a task-independent spread of activation throughout a network in which orthography is linked to phonology and semantics. Second, a mechanism that further fine-tunes the sensitivity of high-level language areas to the sensory input in a task-dependent manner.

Learning and retrieving holistic and componential visual-verbal associations in reading and object naming

Understanding the neural processes that underlie learning to read can provide a scientific foundation for literacy education but studying these processes in real-world contexts remains challenging. We present behavioural data from adult participants learning to read artificial words and name artificial objects over two days. Learning profiles and generalisation confirmed that componential learning of visual-verbal associations distinguishes reading from object naming. Functional MRI data collected on the second day allowed us to identify the neural systems that support componential reading as distinct from systems supporting holistic visual-verbal associations in object naming. Results showed increased activation in posterior ventral occipitotemporal (vOT), parietal, and frontal cortices when reading an artificial orthography compared to naming artificial objects, and the reverse profile in anterior vOT regions. However, activation differences between trained and untrained words were absent, suggesting a lack of cortical representations for whole words. Despite this, hippocampal responses provided some evidence for overnight consolidation of both words and objects learned on day 1. The comparison between neural activity for artificial words and objects showed extensive overlap with systems differentially engaged for real object naming and English word/pseudoword reading in the same participants. These findings therefore provide evidence that artificial learning paradigms offer an alternative method for studying the neural systems supporting language and literacy. Implications for literacy acquisition are discussed.

•

Generalisation of novel orthography shows componential processing in reading.

•

Real and artificial words and objects rely upon the same neural systems.

•

Different neural systems support reading novel orthography and naming novel objects.

•

No evidence of whole-word cortical representations for artificial written words.

•

Reduced hippocampal responses suggest overnight consolidation of artificial items.

Cerebellar BOLD signal during the acquisition of a new lexicon predicts its early consolidation

Cerebellar contributions to language are presently poorly understood, but it has been argued that the cerebellar role in motor learning can be extended to learning in cognitive and linguistic domains. Here, we used fMRI to investigate whether the cerebellum is recruited in mapping novel words onto existing semantic concepts. On separate days, participants performed a Basque vocabulary learning task and a control English synonym task in the MRI scanner. Learning-related BOLD activity was found in left inferior frontal gyrus, bilateral insula, pre-SMA, left superior parietal cortex, right caudate, the right cerebellar vermis and right cerebellar Crus II. The extent to which the cerebellar regions, but not the cerebral areas, were recruited during learning correlated positively with participants' off-line improvement in performance after the learning task. These data provide evidence for a cerebellar role in lexical learning, and suggest that the right cerebellum may contribute toward consolidation of lexico-semantic associations in the language network.

Uncovering phonological and orthographic selectivity across the reading network using fMRI-RA

Reading has been shown to rely on a dorsal brain circuit involving the temporoparietal cortex (TPC) for grapheme-to-phoneme conversion of novel words (Pugh et al., 2001), and a ventral stream involving left occipitotemporal cortex (OTC) (in particular in the so-called "visual word form area", VWFA) for visual identification of familiar words. In addition, portions of the inferior frontal cortex (IFC) have been posited to be an output of the dorsal reading pathway involved in phonology. While this dorsal versus ventral dichotomy for phonological and orthographic processing of words is widely accepted, it is not known if these brain areas are actually strictly sensitive to orthographic or phonological information. Using an fMRI rapid adaptation technique we probed the selectivity of the TPC, OTC, and IFC to orthographic and phonological features during single word reading. We found in two independent experiments using different task conditions in adult normal readers, that the TPC is exclusively sensitive to phonology and the VWFA in the OTC is exclusively sensitive to orthography. The dorsal IFC (BA 44), however, showed orthographic but not phonological selectivity. These results support the theory that reading involves a specific phonological-based temporoparietal region and a specific orthographic-based ventral occipitotemporal region. The dorsal IFC, however, was not sensitive to phonological processing, suggesting a more complex role for this region.

Emergence of representations through repeated training on pronouncing novel letter combinations leads to efficient reading

Printed text can be decoded by utilizing different processing routes depending on the familiarity of the script. A predominant use of word-level decoding strategies can be expected in the case of a familiar script, and an almost exclusive use of letter-level decoding strategies for unfamiliar scripts. Behavioural studies have revealed that frequently occurring words are read more efficiently, suggesting that these words are read in a more holistic way at the word-level, than infrequent and unfamiliar words. To test whether repeated exposure to specific letter combinations leads to holistic reading, we monitored both behavioural and neural responses during novel script decoding and examined changes related to repeated exposure. We trained a group of Dutch university students to decode pseudowords written in an unfamiliar script, i.e., Korean Hangul characters. We compared behavioural and neural responses to pronouncing trained versus untrained two-character pseudowords (equivalent to two-syllable pseudowords). We tested once shortly after the initial training and again after a four days' delay that included another training session. We found that trained pseudowords were pronounced faster and more accurately than novel combinations of radicals (equivalent to letters). Imaging data revealed that pronunciation of trained pseudowords engaged the posterior temporo-parietal region, and engagement of this network was predictive of reading efficiency a month later. The results imply that repeated exposure to specific combinations of graphemes can lead to emergence of holistic representations that result in efficient reading. Furthermore, inter-individual differences revealed that good learners retained efficiency more than bad learners one month later.

Lexical enhancement during prime-target integration: ERP evidence from matched-case identity priming

A number of experiments have revealed that matched-case identity PRIME-TARGET pairs are responded to faster than mismatched-case identity prime-TARGET pairs for pseudowords (e.g., JUDPE-JUDPE < judpe-JUDPE), but not for words (JUDGE-JUDGE = judge-JUDGE). These findings suggest that prime-target integration processes are enhanced when the stimuli tap onto lexical representations, overriding physical differences between the stimuli (e.g., case). To track the time course of this phenomenon, we conducted an event-related potential (ERP) masked-priming lexical decision experiment that manipulated matched versus mismatched case identity in words and pseudowords. The behavioral results replicated previous research. The ERP waves revealed that matched-case identity-priming effects were found at a very early time epoch (N/P150 effects) for words and pseudowords. Importantly, around 200 ms after target onset (N250), these differences disappeared for words but not for pseudowords. These findings suggest that different-case word forms (lower- and uppercase) tap into the same abstract representation, leading to prime-target integration very early in processing. In contrast, different-case pseudoword forms are processed as two different representations. This word-pseudoword dissociation has important implications for neural accounts of visual-word recognition.

Identification of the regions involved in phonological assembly using a novel paradigm

Here we adopt a novel strategy to investigate phonological assembly. Participants performed a visual lexical decision task in English in which the letters in words and letterstrings were delivered either sequentially (promoting phonological assembly) or simultaneously (not promoting phonological assembly). A region of interest analysis confirmed that regions previously associated with phonological assembly, in studies contrasting different word types (e.g. words versus pseudowords), were also identified using our novel task that controls for a number of confounding variables. Specifically, the left pars opercularis, the superior part of the ventral precentral gyrus and the supramarginal gyrus were all recruited more during sequential delivery than simultaneous delivery, even when various psycholinguistic characteristics of the stimuli were controlled. This suggests that sequential delivery of orthographic stimuli is a useful tool to explore how readers, with various levels of proficiency, use sublexical phonological processing during visual word recognition.

Laterality and unilateral deafness: Patients with congenital right ear deafness do not develop atypical language dominance

Auditory speech perception, speech production and reading lateralize to the left hemisphere in the majority of healthy right-handers. In this study, we investigated to what extent sensory input underlies the side of language dominance. We measured the lateralization of the three core subprocesses of language in patients who had profound hearing loss in the right ear from birth and in matched control subjects. They took part in a semantic decision listening task involving speech and sound stimuli (auditory perception), a word generation task (speech production) and a passive reading task (reading). The results show that a lack of sensory auditory input on the right side, which is strongly connected to the contralateral left hemisphere, does not lead to atypical lateralization of speech perception. Speech production and reading were also typically left lateralized in all but one patient, contradicting previous small scale studies. Other factors such as genetic constraints presumably overrule the role of sensory input in the development of (a)typical language lateralization.

Reprint of: Visual processing of words in a patient with visual form agnosia: A behavioural and fMRI study

Patient D.F. has a profound and enduring visual form agnosia due to a carbon monoxide poisoning episode suffered in 1988. Her inability to distinguish simple geometric shapes or single alphanumeric characters can be attributed to a bilateral loss of cortical area LO, a loss that has been well established through structural and functional fMRI. Yet despite this severe perceptual deficit, D.F. is able to "guess" remarkably well the identity of whole words. This paradoxical finding, which we were able to replicate more than 20 years following her initial testing, raises the question as to whether D.F. has retained specialized brain circuitry for word recognition that is able to function to some degree without the benefit of inputs from area LO. We used fMRI to investigate this, and found regions in the left fusiform gyrus, left inferior frontal gyrus, and left middle temporal cortex that responded selectively to words. A group of healthy control subjects showed similar activations. The left fusiform activations appear to coincide with the area commonly named the visual word form area (VWFA) in studies of healthy individuals, and appear to be quite separate from the fusiform face area (FFA). We hypothesize that there is a route to this area that lies outside area LO, and which remains relatively unscathed in D.F.

Eyes on words: A fixation-related fMRI study of the left occipito-temporal cortex during self-paced silent reading of words and pseudowords

The predominant finding of studies assessing the response of the left ventral occipito-temporal cortex (vOT) to familiar words and to unfamiliar, but pronounceable letter strings (pseudowords) is higher activation for pseudowords. One explanation for this finding is that readers automatically generate predictions about a letter string's identity - pseudowords mismatch these predictions and the higher vOT activation is interpreted as reflecting the resultant prediction errors. The majority of studies, however, administered tasks which imposed demands above and beyond the intrinsic requirements of visual word recognition. The present study assessed the response of the left vOT to words and pseudowords by using the onset of the first fixation on a stimulus as time point for modeling the BOLD signal (fixation-related fMRI). This method allowed us to assess the neural correlates of self-paced silent reading with minimal task demands and natural exposure durations. In contrast to the predominantly reported higher vOT activation for pseudowords, we found higher activation for words. This finding is at odds with the expectation of higher vOT activation for pseudowords due to automatically generated predictions and the accompanying elevation of prediction errors. Our finding conforms to an alternative explanation which considers such top-down processing to be non-automatic and task-dependent.

Neuroimaging the short- and long-term effects of repeated picture naming in healthy older adults

Repeated attempts to name pictures can improve subsequent naming for aphasic individuals with anomia, however, the neurocognitive mechanisms responsible for such improvements are unknown. This study investigated repeated picture naming in healthy older adults over a period of minutes (short-term) after one repetition and a period of days (long-term) after multiple repetitions. Compared to unprimed pictures, both repeated conditions showed faster naming latencies with the fastest latencies evident for the short-term condition. Neuroimaging results identified repetition suppression effects across three left inferior frontal gyrus regions of interest: for both the short- and long-term conditions in the pars orbitalis, and for long-term items in the pars triangularis and pars opercularis regions. The whole brain analysis also showed a repetition suppression effect in bilateral pars triangularis regions for the long-term condition. These findings within the inferior frontal gyrus suggest that effects of repeated naming may be driven by a mapping mechanism across multiple levels of representation, possibly reflecting different levels of learning, and lend support to the idea that processing may be hierarchically organised in the left inferior frontal gyrus. The whole brain analysis also revealed repetition suppression for the long-term condition within the posterior portion of bilateral inferior temporal gyri, which may reflect attenuation of integration processes within this region following the learning of task-relevant information.

Variable Left-hemisphere Language and Orthographic Lateralization Reduces Right-hemisphere Face Lateralization

It is commonly believed that, in right-handed individuals, words and faces are processed by distinct neural systems: one in the left hemisphere (LH) for words and the other in the right hemisphere (RH) for faces. Emerging evidence suggests, however, that hemispheric selectivity for words and for faces may not be independent of each other. One recent account suggests that words become lateralized to the LH to interact more effectively with language regions, and subsequently, as a result of competition with words for representational space, faces become lateralized to the RH. On this interactive account, left-handed individuals, who as a group show greater variability with respect to hemispheric language dominance, might be expected to show greater variability in their degree of RH lateralization of faces as well. The current study uses behavioral measures and ERPs to compare the hemispheric specialization for both words and faces in right- and left-handed adult individuals. Although both right- and left-handed groups demonstrated LH over RH superiority in discrimination accuracy for words, only the right-handed group demonstrated RH over LH advantage in discrimination accuracy for faces. Consistent with this, increased right-handedness was related to an increase in RH superiority for face processing, as measured by the strength of the N170 ERP component. Interestingly, the degree of RH behavioral superiority for face processing and the amplitude of the RH N170 for faces could be predicted by the magnitude of the N170 ERP response to words in the LH. These results are discussed in terms of a theoretical account in which the typical RH face lateralization fails to emerge in individuals with atypical language lateralization because of weakened competition from the LH representation of words.

Lexical is as lexical does: computational approaches to lexical representation

In much of neuroimaging and neuropsychology, regions of the brain have been associated with 'lexical representation', with little consideration as to what this cognitive construct actually denotes. Within current computational models of word recognition, there are a number of different approaches to the representation of lexical knowledge. Structural lexical representations, found in original theories of word recognition, have been instantiated in modern localist models. However, such a representational scheme lacks neural plausibility in terms of economy and flexibility. Connectionist models have therefore adopted distributed representations of form and meaning. Semantic representations in connectionist models necessarily encode lexical knowledge. Yet when equipped with recurrent connections, connectionist models can also develop attractors for familiar forms that function as lexical representations. Current behavioural, neuropsychological and neuroimaging evidence shows a clear role for semantic information, but also suggests some modality- and task-specific lexical representations. A variety of connectionist architectures could implement these distributed functional representations, and further experimental and simulation work is required to discriminate between these alternatives. Future conceptualisations of lexical representations will therefore emerge from a synergy between modelling and neuroscience.

Early Visual Word Processing Is Flexible: Evidence from Spatiotemporal Brain Dynamics

Visual word recognition is often described as automatic, but the functional locus of top-down effects is still a matter of debate. Do task demands modulate how information is retrieved, or only how it is used? We used EEG/MEG recordings to assess whether, when, and how task contexts modify early retrieval of specific psycholinguistic information in occipitotemporal cortex, an area likely to contribute to early stages of visual word processing. Using a parametric approach, we analyzed the spatiotemporal response patterns of occipitotemporal cortex for orthographic, lexical, and semantic variables in three psycholinguistic tasks: silent reading, lexical decision, and semantic decision. Task modulation of word frequency and imageability effects occurred simultaneously in ventral occipitotemporal regions-in the vicinity of the putative visual word form area-around 160 msec, following task effects on orthographic typicality around 100 msec. Frequency and typicality also produced task-independent effects in anterior temporal lobe regions after 200 msec. The early task modulation for several specific psycholinguistic variables indicates that occipitotemporal areas integrate perceptual input with prior knowledge in a task-dependent manner. Still, later task-independent effects in anterior temporal lobes suggest that word recognition eventually leads to retrieval of semantic information irrespective of task demands. We conclude that even a highly overlearned visual task like word recognition should be described as flexible rather than automatic.

New insights into the neural network mediating reading processes provided by cortico-subcortical electrical mapping

OBJECTIVES

To ascertain the neural network mediating reading using intraoperative electrostimulation.

EXPERIMENTAL DESIGN

A cortical and axonal intraoperative electrical mapping of reading processes was achieved in seven patients who underwent awake surgery for a left occipitotemporal glioma. We performed resection cavity overlapping and superimposition with a diffusion tensor imaging-based white matter atlas. We assessed the relationship between the location of resection cavities and the occurrence of reading impairments of regular, irregular, and pseudowords.

PRINCIPAL OBSERVATIONS

Intraoperative stimulation of the left posterior inferior temporal cortex (ITCp) elicited reading disturbances. Subcortical stimulation at the anterior portion of the visual word form area (VWFA) induced addressed phonology (irregular words reading) disturbances. Subcortical stimulation of the connection between VWFA and the posterior segment of the arcuate fascicle (AFp) induced both addressed and assembled phonology (irregular and pseudowords reading) disturbances. Postoperative assessment showed that resection of the posterior portion of the inferior longitudinal fascicle (ILFp), connecting the visual cortex to VWFA, induced long-term and global reading impairment. Resection of the terminations of left AFp in the ITCp-induced irregular and pseudowords reading disturbances with no impairment of regular words reading. Resection of the anterior portion of ILF did not induce reading impairment.

CONCLUSIONS

Our data support an inner posterior-to-anterior hierarchical coding of letter strings in the VWFA and a crucial role of the left ILFp to provide visual inputs to the VWFA. Furthermore, we suggest that the AFp is involved in an interactive feedback system between visual and nonvisual information, recruited when reading irregular and pseudowords.

Accessing orthographic representations from speech: the role of left ventral occipitotemporal cortex in spelling

The present fMRI study used a spelling task to investigate the hypothesis that the left ventral occipitotemporal cortex (vOT) hosts neuronal representations of whole written words. Such an orthographic word lexicon is posited by cognitive dual‐route theories of reading and spelling. In the scanner, participants performed a spelling task in which they had to indicate if a visually presented letter is present in the written form of an auditorily presented word. The main experimental manipulation distinguished between an orthographic word spelling condition in which correct spelling decisions had to be based on orthographic whole‐word representations, a word spelling condition in which reliance on orthographic whole‐word representations was optional and a phonological pseudoword spelling condition in which no reliance on such representations was possible. To evaluate spelling‐specific activations the spelling conditions were contrasted with control conditions that also presented auditory words and pseudowords, but participants had to indicate if a visually presented letter corresponded to the gender of the speaker. We identified a left vOT cluster activated for the critical orthographic word spelling condition relative to both the control condition and the phonological pseudoword spelling condition. Our results suggest that activation of left vOT during spelling can be attributed to the retrieval of orthographic whole‐word representations and, thus, support the position that the left vOT potentially represents the neuronal equivalent of the cognitive orthographic word lexicon. Hum Brain Mapp, 36:1393–1406, 2015. © 2014 The Authors Human Brain Mapping Published by Wiley Periodicals, Inc.

A robust index of lexical representation in the left occipito-temporal cortex as evidenced by EEG responses to fast periodic visual stimulation

Despite decades of research on reading, including the relatively recent contributions of neuroimaging and electrophysiology, identifying selective representations of whole visual words (in contrast to pseudowords) in the human brain remains challenging, in particular without an explicit linguistic task. Here we measured discrimination responses to written words by means of electroencephalography (EEG) during fast periodic visual stimulation. Sequences of pseudofonts, nonwords, or pseudowords were presented through sinusoidal contrast modulation at a periodic 10 Hz frequency rate (F), in which words were interspersed at regular intervals of every fifth item (i.e., F/5, 2 Hz). Participants monitored a central cross color change and had no linguistic task to perform. Within only 3 min of stimulation, a robust discrimination response for words at 2 Hz (and its harmonics, i.e., 4 and 6 Hz) was observed in all conditions, located predominantly over the left occipito-temporal cortex. The magnitude of the response was largest for words embedded in pseudofonts, and larger in nonwords than in pseudowords, showing that list context effects classically reported in behavioral lexical decision tasks are due to visual discrimination rather than decisional processes. Remarkably, the oddball response was significant even for the critical words/pseudowords discrimination condition in every individual participant. A second experiment replicated this words/pseudowords discrimination, and showed that this effect is not accounted for by a higher bigram frequency of words than pseudowords. Without any explicit task, our results highlight the potential of an EEG fast periodic visual stimulation approach for understanding the representation of written language. Its development in the scientific community might be valuable to rapidly and objectively measure sensitivity to word processing in different human populations, including neuropsychological patients with dyslexia and other reading difficulties.

Hemispheric differences in orthographic and semantic processing as revealed by event-related potentials

Differences in how the right and left hemispheres (RH, LH) apprehend visual words were examined using event-related potentials (ERPs) in a repetition paradigm with visual half-field (VF) presentation. In both hemispheres (RH/LVF, LH/RVF), initial presentation of items elicited similar and typical effects of orthographic neighborhood size, with larger N400s for orthographically regular items (words and pseudowords) than for irregular items (acronyms and meaningless illegal strings). However, hemispheric differences emerged on repetition effects. When items were repeated in the LH/RVF, orthographically regular items, relative to irregular items, elicited larger repetition effects on both the N250, a component reflecting processing at the level of visual form (orthography), and on the N400, which has been linked to semantic access. In contrast, in the RH/LVF, repetition effects were biased toward irregular items on the N250 and were similar in size across item types for the N400. The results suggest that processing in the LH is more strongly affected by wordform regularity than in the RH, either due to enhanced processing of familiar orthographic patterns or due to the fact that regular forms can be more readily mapped onto phonology.

An ERP investigation of the co-development of hemispheric lateralization of face and word recognition

The adult human brain would appear to have specialized and independent neural systems for the visual processing of words and faces. Extensive evidence has demonstrated greater selectivity for written words in the left over right hemisphere, and, conversely, greater selectivity for faces in the right over left hemisphere. This study examines the emergence of these complementary neural profiles, as well as the possible relationship between them. Using behavioral and neurophysiological measures, in adults, we observed the standard finding of greater accuracy and a larger N170 ERP component in the left over right hemisphere for words, and conversely, greater accuracy and a larger N170 in the right over the left hemisphere for faces. We also found that although children aged 7-12 years revealed the adult hemispheric pattern for words, they showed neither a behavioral nor a neural hemispheric superiority for faces. Of particular interest, the magnitude of their N170 for faces in the right hemisphere was related to that of the N170 for words in their left hemisphere. These findings suggest that the hemispheric organization of face recognition and of word recognition does not develop independently, and that word lateralization may precede and drive later face lateralization. A theoretical account for the findings, in which competition for visual representations unfolds over the course of development, is discussed.