Dynamics of visual feature analysis and object-level processing in face versus letter-string perception

Clinical neurophysiology of language: The MEG approach

Clinical evaluation of language function and basic neuroscience research into the neurophysiology of language are tied together. Whole-head MEG systems readily facilitate detailed spatiotemporal characterization of language processes. A fair amount of information is available about the cortical sequence of word perception and comprehension in the auditory and visual domain, which can be applied for clinical use. Language production remains, at present, somewhat less well charted. In clinical practice, the most obvious needs are noninvasive evaluation of the language-dominant hemisphere and mapping of areas involved in language performance to assist surgery. Multiple experimental designs and analysis approaches have been proposed for estimation of language lateralization. Some of them have been compared with the invasive Wada test and need to be tested further. Development of approaches for more comprehensive pre-surgical characterization of language cortex should build on basic neuroscience research, making use of parametric designs that allow functional mapping. Studies of the neural basis of developmental and acquired language disorders, such as dyslexia, stuttering, and aphasia can currently be regarded more as clinical or basic neuroscience research rather than as clinical routine. Such investigations may eventually provide tools for development of individually targeted training procedures and their objective evaluation.

Neural representation of language: activation versus long-range connectivity

Cognitive functions are thought to build on connectivity within large-scale neuronal networks, rather than on strictly localized processes. Yet, present understanding of neural mechanisms of language function, as derived from neuroimaging, is based on mapping brain areas that are more active during specific linguistic tasks than in control conditions. Connectivity can then be evaluated among those areas. However, network nodes should ideally be determined based on their correlated time series of activity. Recent developments in analysis methods now facilitate localization and characterization of functionally connected neural networks directly from real-time magnetoencephalography data. Analysis of long-range connectivity might clarify and expand the view provided by traditional neurophysiological and hemodynamic activation studies. Here, we use silent reading as the example process.

Perceptual contributions to problem solving: Chunk decomposition of Chinese characters

Chunk decomposition is the decomposing of familiar patterns into their component elements so that they can be regrouped in another meaningful manner. Such a regrouping is sometimes critically required in problem solving because during initial encoding the problem elements become automatically grouped into familiar chunks and this may prohibit finding a novel or efficient solution to problems [G. Knoblich, S. Ohlsson, H. Haider, D. Rhenius, Constraint relaxation and chunk decomposition in insight problem solving, J. Exp. Psychol. Learn. Mem. Cogn. 25 (1999) 1534-1556]. In order to elucidate the brain mechanisms underlying the process of chunk decomposition, we developed a task that uses Chinese character as materials. Chinese characters are ideal examples of perceptual chunks. They are composed of radicals, which in turn, are composed of strokes. Because radicals are meaningful chunks themselves but strokes are not meaningful in isolation, it is much easier to separate a character by its radicals than to separate a character by its strokes. By comparing the stroke-level decomposition and the radical-level decomposition, we observed activities in occipital, frontal, and parietal lobes. Most importantly, during the moment of chunk decomposition, we found the early visual cortex showed a tendency of negative activation whereas the higher visual cortex showed a tendency of positive activation. This suggests that in order to successfully decompose a chunk, the higher visual areas must at least partly be 'disconnected' from the input provided by early visual processing in order to allow simple features to be rearranged into a different perceptual chunk. We conclude that early perceptual processes can crucially affect thinking and problem solving.

Attentional shifting and the role of the dorsal pathway in visual word recognition

A substantial amount of evidence has been collected to propose an exclusive role for the dorsal visual pathway in the control of guided visual search mechanisms, specifically in the preattentive direction of spatial selection [Vidyasagar, T. R. (1999). A neuronal model of attentional spotlight: Parietal guiding the temporal. Brain Research and Reviews, 30, 66-76; Vidyasagar, T. R. (2001). From attentional gating in macaque primary visual cortex to dyslexia in humans. Progress in Brain Research, 134, 297-312]. Moreover, it has been suggested recently that the dorsal visual pathway is specifically involved in the spatial selection and sequencing required for orthographic processing in visual word recognition. In this experiment we manipulate the demands for spatial processing in a word recognition, lexical decision task by presenting target words in a normal spatial configuration, or where the constituent letters of each word are spatially shifted relative to each other. Accurate word recognition in the Shifted-words condition should demand higher spatial encoding requirements, thereby making greater demands on the dorsal visual stream. Magnetoencephalographic (MEG) neuroimaging revealed a high frequency (35-40Hz) right posterior parietal activation consistent with dorsal stream involvement occurring between 100 and 300ms post-stimulus onset, and then again at 200-400ms. Moreover, this signal was stronger in the shifted word condition, compared to the normal word condition. This result provides neurophysiological evidence that the dorsal visual stream may play an important role in visual word recognition and reading. These results further provide a plausible link between early stage theories of reading, and the magnocellular-deficit theory of dyslexia, which characterises many types of reading difficulty.

Interplay between computational models and cognitive electrophysiology in visual word recognition

In this article, we discuss the relevance of electrophysiological data to the enterprise of analyzing and understanding the reading process. Specifically, we detail how the event-related brain potential (ERP) technique (and its magnetic counterpart) can aid in development of models of visual word recognition. Any viable and accurate account of reading must take into account the temporal and anatomical constraints imposed by the fact that reading is a human brain function. We believe that neurophysiological (especially, although not limited to electrophysiological) data can serve an essential reference in the development of biologically realistic models of reading. We assess just how well extant electrophysiological data comport with specific predictions of existing computational models and offer some suggestions for the kinds of research that can address some of the remaining open questions.

Cortical sequence of word perception in beginning readers

Efficient analysis of written words in normal reading is likely to reflect use of neural circuits formed by experience during childhood rather than an innate process. We investigated the cortical sequence of word perception in first-graders (7-8 years old), with special emphasis on occipitotemporal cortex in which, in adults, letter-string-sensitive responses are detected at 150 ms after stimulus. To identify neural activation that is sensitive to either the amount of basic visual features or specifically to letter strings, we recorded whole-head magnetoencephalography responses to words embedded in three different levels of noise and to symbol strings. As was shown previously in adults, activation reflecting stimulus nonspecific visual feature analysis was localized to occipital cortex in children. It was followed by letter-string-sensitive activation in the left occipitotemporal cortex and, subsequently, in the temporal cortex. These processing stages were correlated in timing and activation strength. Compared with adults, however, the timing of activation was clearly delayed in children, and the delay was progressively increased from occipital to occipitotemporal and further to temporal areas. This finding is likely to reflect increasing immaturity of the underlying neural generators when advancing from low-level visual analysis to higher-order areas involved in written word perception. When a salient occipitotemporal letter-string-sensitive activation was detected (10 of 18 children), its strength was correlated with phonological skills, in line with the known relevance of phonological awareness in reading acquisition.

Multiple Levels of Stimulus Representation in Visual Working Memory

Object recognition presumably involves activation of multiple levels of representation. Here we use the encoding-related lateralization (ERL) method [Gratton, G. The contralateral organization of visual memory: A theoretical concept and a research tool. Psychophysiology, 35, 638–647, 1998] to describe the sequential activation of several of these levels. The ERL uses divided-field encoding to generate contralaterally biased representations in the brain. The presence and nature of these representations can be demonstrated by examining the event-related potentials (ERPs) elicited by centrally presented test probes for lateralized activity corresponding to the encoding side. We recorded ERPs during a memory-search task. Memory sets were composed of two or four uppercase letters displayed half to the left and half to the right of fixation. Probe stimuli were composed of one letter presented foveally in either upper- or lowercase. Letter case was manipulated to differentiate the time course of physical and symbolic levels of letter representation. Memory set size was manipulated to examine a relational level of letter representation. We found multiple ERLs in response to the probes: (1) An early (peak = 170 msec) case-dependent (but set size independent) ERL, most evident at P7/P8, indexing the availability of a physical level of letter representation; (2) a later (200–400 msec) more diffusedly distributed ERL, independent of both letter case and set size, indexing a symbolic level of letter representation; (3) a long-latency (400–600 msec) ERL occurring at posterior sites, larger for the case match, Set Size 2 condition, indexing competition for neural representation across multiple letters. By assuming that these ERL activities track the progression of letter representation over time, we propose a model of letter processing in the context of visual working memory.

Oscillatory activity in the occipitotemporal area related to the visual perception of letters of a first/second language and pseudoletters

The objective of this study was to reveal the oscillatory activity in the occipitotemporal area related to the visual perception of the letters of first (L1) and second languages (L2) and pseudoletters. We recorded neuromagnetic signals while Korean native speakers were exposed to a phonogram of Korean, acquired at school age as their L1 (Hangul), that of Japanese, learned in adulthood as a L2 (Kana) and pseudoletter (Pseudo), and quantified the event-related desynchronization (ERD) and synchronization (ERS). In all conditions, sustained ERDs in the alpha band were observed in both hemispheres. ERD for Pseudo was gradually attenuated after approximately 400-500 ms after stimulus onset, whereas both Hangul and Kana produced stronger and longer-lasting ERD. ERD for Kana showed a broader alpha band than Hangul. Furthermore, transient ERSs in the gamma band around 70 Hz were observed between 100 and 400 ms in the bilateral occipitotemporal areas. In the left hemisphere, gamma band oscillations showed similar enhancement in all conditions, suggesting that gamma band activity in the left occipitotemporal area might be enhanced not only by the bottom-up process as visual perception but also by the top-down process as attention to prelexical visual stimuli. In the right hemisphere, gamma band ERS was stronger for Hangul than Pseudo and no differences were shown between Kana and Pseudo. The differences of oscillatory activity in the alpha and gamma bands suggest that neuronal networks, including the occipitotemporal area, are related to the visual perception of letters differing between L1 and L2.

Category specificity in the processing of color-related and form-related words: An ERP study

In this study, we investigated the spatio-temporal patterns of category-specific cortical activation elicited by the visual presentation of words whose meaning relates to a color or to a visual form or shape. We recorded the event-related potentials (ERPs) of ten healthy, right-handed volunteers while they passively read words presented tachistoscopically. As early as 150 ms after stimulus onset, the ERPs revealed significant neurophysiological differences between words and strings of hash marks. Around 200 ms after stimulus onset, we found significant differences in the ERPs elicited by color- and form-related words. We used minimum norm current estimates to investigate the spatial location of these differences. This revealed that, at the 150 ms peak, the activation advantage of words over sequences of hash marks was located in a left posterior area, proximal to what has previously been called the visual word form area. At the 200 ms peak, the advantage of words related to colors over words related to forms seemed to reside in temporal cortical areas, whereas the form-related words elicited greater activation in frontal areas than color-related words. These results provide evidence for early access to detailed category-specific representations of word meaning, with subtle differences in meaning being reflected in the activation of different cortical areas, as early as 200 ms after stimulus presentation. In line with previous studies, these differences can be related to the areas involved in the conceptual processing of sensory (visual) and action-related information.

Hemispheric asymmetry emerges at distinct parts of the occipitotemporal cortex for objects, logograms and phonograms: A functional MRI study

Behavioral and neuropsychological studies have suggested that the right hemisphere has a special advantage in the visual recognition of logograms. While this long-standing 'right hemisphere hypothesis' has never been investigated systematically by previous neuroimaging studies, a candidate neural substrate of such asymmetry might be found within the occipitotemporal cortex that is known to exhibit lateralized response to a certain class of stimuli, such as letters and faces. The present study examined the hemispheric specialization of brain activation during naming of objects, logograms and phonograms using functional magnetic resonance imaging. The three types of stimuli overall produced left-predominant activation of the perisylvian and inferior parietal regions relative to the resting baseline. This inter-hemispheric difference was significant irrespective of the stimuli type. In the occipitotemporal cortex, six subregions showing lateralized response were identified. That is, the three stimuli commonly produced left-lateralized response in the posterior fusiform and superior temporal gyri and right-lateralized response in the extrastriate cortex. Only logograms and objects produced a distinct cluster showing right-lateralized activation in the medial anterior fusiform gyrus associated with semantic knowledge, whereas only phonograms produced a left-lateralized activation in the posterior middle temporal cortex close to the site associated with visual perception of alphabetical letters. These findings suggest that while these stimuli similarly recruit the left perisylvian language area as a common neural component for naming, processing of objects and logograms becomes left-lateralized only in the downstream of the occipitotemporal cortex. By contrast, visual processing of phonograms is specialized to the left hemisphere in earlier stages of the area. The present data provide further evidence suggesting that both the left-right and anterior-posterior axes of the occipitotemporal cortex are differentially tuned according to the specific features of visual stimuli.

When meaningless symbols become letters: Neural activity change in learning new phonograms

Left fusiform gyrus and left angular gyrus are considered to be respectively involved with visual form processing and associating visual and auditory (phonological) information in reading. However, there are a number of studies that fail to show the contribution of these regions in carrying out these aspects of reading. Considerable differences in the type of stimuli and tasks used in the various studies may account for the discrepancy in results. This functional magnetic resonance imaging (fMRI) study attempts to control aspects of experimental stimuli and tasks to specifically investigate brain regions involved with visual form processing and character-to-phonological (i.e., simple grapheme-to-phonological) conversion processing for single letters. Subjects performed a two-back identification task using known Japanese, and previously unknown Korean, and Thai phonograms before and after training on one of the unknown language orthographies. Japanese subjects learned either five Korean or five Thai phonograms. Brain regions related to visual form processing were assessed by comparing activity related to native (Japanese) phonograms with that of non-native (Korean and Thai) phonograms. There was no significant differential brain activity for visual form processing. Brain regions related to character-to-phonological conversion processing were assessed by comparing pre- and post-tests of trained non-native phonograms with that of native phonograms and non-trained non-native phonograms. Significant differential activation post-relative to pre-training exclusively for the trained non-native phonograms was found in left angular gyrus. In addition, psychophysiologic interaction (PPI) analysis revealed greater integration of left angular gyrus with primary visual cortex as well as with superior temporal gyrus for the trained phonograms post-relative to pre-training. The results suggest that left angular gyrus is involved with character-to-phonological conversion in letter perception.

An early electrophysiological response associated with expertise in letter perception

Expertise with print is likely to optimize visual processes for recognizing characters of a familiar writing system. Although brain activations have been identified for words and letter strings in contrast with other stimuli, relatively little work has focused on the neural basis of single-letter perception. English readers and Chinese-English bilinguals participated in an ERP study and performed a 1-back identity judgment on Roman letters, Chinese characters, pseudofonts, and their string versions. The Chinese-English bilinguals showed an enhanced N170 for both Roman letters and Chinese characters relative to pseudofonts. For the non-Chinese readers, the N170 amplitude was larger for Roman letters relative to Chinese characters and pseudofonts. Our results suggest that changes in relatively early visual processes underlie expert letter perception.

Spatiotemporal brain maps of delayed word repetition and recognition

Whole-head magnetoencephalography (MEG) was used to spatiotemporally map the brain response underlying episodic retrieval of words studied a single time following a long delay (approximately 40 min). Recognition following a long delay occurs as a strong, sustained, differential response, within bilateral, ventral, and lateral prefrontal cortex, anterior temporal and medial parietal regions from approximately 500 ms onward, as well as ventral occipitotemporal regions from approximately 700 ms onward. In comparison with previous tasks using multiple repetitions at short delays, these effects were centered within the same areas (anteroventral temporal and ventral prefrontal) but were shifted to longer latencies (approximately 500 ms vs. approximately 200 ms), were less left-lateralized, and appear more in anterolateral prefrontal regions and less in lateral temporal cortex. Furthermore, comparison of correctly classified words with misclassified, novel and repeated words, suggests that these frontotemporal-parietocingulate responses are sensitive to actual as well as perceived repetition. The results also suggest that lateral prefrontal regions may participate more in controlled effortful retrieval, while left ventral frontal and anterior temporal responses may support sustained lexicosemantic processing. Additionally, left ventromedial temporal sites may be relatively more involved in episodic retrieval, while lateral temporal sites may participate more in automatic priming.

Cortical mechanisms of attention in time: neural correlates of the Lag-1-sparing phenomenon

If humans monitor streams of rapidly presented (approximately 100-ms intervals) visual stimuli, which are typically specific single letters of the alphabet, for two targets (T1 and T2), they often miss T2 if it follows T1 within an interval of 200-500 ms. If T2 follows T1 directly (within 100 ms; described as occurring at 'Lag 1'), however, performance is often excellent: the so-called 'Lag-1 sparing' phenomenon. Lag-1 sparing might result from the integration of the two targets into the same 'event representation', which fits with the observation that sparing is often accompanied by a loss of T1-T2 order information. Alternatively, this might point to competition between the two targets (implying a trade-off between performance on T1 and T2) and Lag-1 sparing might solely emerge from conditional data analysis (i.e. T2 performance given T1 correct). We investigated the neural correlates of Lag-1 sparing by carrying out magnetoencephalography (MEG) recordings during an attentional blink (AB) task, by presenting two targets with a temporal lag of either 1 or 2 and, in the case of Lag 2, with a nontarget or a blank intervening between T1 and T2. In contrast to Lag 2, where two distinct neural responses were observed, at Lag 1 the two targets produced one common neural response in the left temporo-parieto-frontal (TPF) area but not in the right TPF or prefrontal areas. We discuss the implications of this result with respect to competition and integration hypotheses, and with respect to the different functional roles of the cortical areas considered. We suggest that more than one target can be identified in parallel in left TPF, at least in the absence of intervening nontarget information (i.e. masks), yet identified targets are processed and consolidated as two separate events by other cortical areas (right TPF and PFC, respectively).

Cerebral asymmetry in children when reading Chinese characters

This study examined cerebral asymmetry, especially in the hierarchical visual system, when reading Chinese characters. Twelve right-handed Chinese children (mean age = 11.6 years) were scanned while performing semantic and phonological tasks. Strong leftward asymmetry was found in the left inferior frontal cortex (BA44/45/47), the parietal lobule (BA40), and the cingulate cortex (BA24/32). In the visual system, we found significant left-hemispheric dominance in the fusiform cortex (BA19/37), but no asymmetry was found in the primary visual cortex (BA17/18). The differential results for the primary visual cortex versus high-order visual cortex (i.e., the fusiform cortex) are discussed in terms of the contribution of the logographic nature of Chinese characters to the asymmetry pattern in the hierarchical visual system.

Processing of Japanese morphogram and syllabogram in the left basal temporal area: electrical cortical stimulation studies

Language functions in the left basal temporal area (LBTA) were investigated using electrical cortical stimulation during functional mapping in six Japanese patients with refractory epilepsy. This study provides the first direct evidence that kana (Japanese syllabogram) is processed in the LBTA. Electrical stimulation of some areas within LBTA induced disturbance in overt reading of kana words only in the first trials, with no errors in the subsequent trials. By contrast, stimulation of the same area caused obvious disturbance in kana non-word reading in all trials. Since a kana word carries both meaning and sound while a kana non-word carries only sounds of a letter string, the contrasting results of partial and complete disturbance imply a possibility that there are two distinct pathways for kana reading: one dealing with both phonological and semantic aspects of the words and the other dealing only with phonological aspect. Kanji words (Japanese morphogram) and objects/pictures were found to be processed in an area different from the area for the kana non-word processing. Furthermore, the present study also identified the common area for processing kanji reading and object/picture naming. There were no errors in matching pictures with kanji words, indicating that concepts of pictures and meanings of kanji words were not interfered by the electrical stimulation of that area. The new insight provides a clue for partial description of processing pathways for language-related visual information in LBTA. Three types of information (morphological, phonological, and semantic) are conveyed together at some stages and are separated into different routes at some other stages.

Target consolidation under high temporal processing demands as revealed by MEG

We investigated the nature of resource limitations during visual target processing by imposing high temporal processing demands on the cognitive system. This was achieved by embedding target stimuli into rapid-serial-visual-presentation-streams (RSVP). In RSVP streams, it is difficult to report the second of two targets (T2) if the second follows the first (T1) within 500 ms. This effect is known as the attentional blink (AB). For the AB to occur, it is essential that T1 is followed by a mask, as without such a stimulus, the AB is significantly attenuated. Usually, it is thought that T1 processing is delayed by the mask, which in turn delays T2 processing, increasing the likelihood for T2 failures (AB). Predictions regarding amplitudes and latencies of cortical responses (M300, the magnetic counterpart to the P300) to targets were tested by investigating the neurophysiological effects of the post-T1 item (mask) by means of magnetoencephalography (MEG). Cortical M300 responses to targets drawn from prefrontal sources--areas associated with working memory--revealed accelerated T1 yet delayed T2 processing with an intervening mask. The explanation we are proposing assumes that "protection" of ongoing T1 processing necessitated by the occurrence of the mask suppresses other activation patterns, which boosts T1 yet also hinders further processing. Our data shed light on the mechanisms employed by the human brain for ensuring visual target processing under high temporal processing demands, which is hypothesized to occur at the expense of subsequently presented information.

The time and space of lexicality: a neuromagnetic view

Illuminating the neural mechanisms subserving lexico-semantic processing is requisite to further understanding the neurophysiological basis of the dyslexias. Yet, despite numerous functional neuroimaging experiments, the location and temporal behavior of brain regions mediating word-level language processing remain an area of debate. Such investigations typically utilize the word/pseudoword contrast within hemodynamic measurements, and report several left hemisphere regions that respond more strongly to pseudowords but fail to replicate neural areas unique to real word processing. The present experiment addressed this problem from a different perspective. Mainly, we hypothesized that the time course, but not the neuroanatomy, would show within-subject across-condition disparities. For that purpose, we applied dipole-modeling techniques to high-density magnetoencephalographic recordings of healthy subjects, and utilized excellent spatiotemporal accuracy to demonstrate significant across-condition differences in the time domain, along with indistinguishable neural correlates within-subject. In all participants, both words and pseudowords elicited activity in left perisylvian language areas, with words consistently activating these regions approximately 100 ms earlier than pseudowords. Considerable functional heterogeneity was also observed, and this might underlie the inconsistencies among previous studies. We conclude that the neural distinction in word/pseudoword processing is not in spatial localization, but is better conceptualized as a dynamic difference in processing time.

Visual recognition of faces, objects, and words using degraded stimuli: where and when it occurs

We studied time course and cerebral localisation of word, object, and face recognition using event-related potentials (ERPs) and source localisation techniques. To compare activation rates of these three categories, we used degraded images that easily pop out without any change in the physical features of the stimuli, once the meaning is revealed. Comparisons before and after identification show additional periods of activation beginning at 100 msec for faces and at around 200 msec for objects and words. For faces, this activation occurs predominantly in right temporal areas, whereas for objects, the specific time period gives rise to bilateral posterior but right dominant foci. Finally, words show a maximum area of activation in the left temporooccipital area at their specific time period. These results provide unequivocal evidence that when effects of low-level visual features are circumvented, faces, objects, and words are not only distinct in terms of their anatomic routes, but also in terms of their times of processing.

Face recognition and cortical responses show similar sensitivity to noise spatial frequency

To find cortical correlates of face recognition, we manipulated the recognizability of face images in a parametric manner by masking them with narrow-band spatial noise. Face recognition performance was best at the lowest and highest noise spatial frequencies (NSFs, 2 and 45 c/image, respectively), and degraded gradually towards central NSFs (11-16 c/image). The strength of the 130-180 ms neuromagnetic response (M170) in the temporo-occipital cortex paralleled the recognition performance, whereas the mid-occipital response at 70-120 ms acted in the opposite manner, being strongest for the central NSFs. To noise stimuli without faces, M170 was small and rather insensitive to NSF, whereas the mid-occipital responses resembled closely the responses to the combined face and noise stimuli. These results suggest that the 100 ms mid-occipital response is sensitive to the central spatial frequencies that are critical for face recognition, whereas the M170 response is sensitive to the visibility of a face and closely related to face recognition.

Visual word recognition: the first half second

We used magnetoencephalography (MEG) to map the spatiotemporal evolution of cortical activity for visual word recognition. We show that for five-letter words, activity in the left hemisphere (LH) fusiform gyrus expands systematically in both the posterior-anterior and medial-lateral directions over the course of the first 500 ms after stimulus presentation. Contrary to what would be expected from cognitive models and hemodynamic studies, the component of this activity that spatially coincides with the visual word form area (VWFA) is not active until around 200 ms post-stimulus, and critically, this activity is preceded by and co-active with activity in parts of the inferior frontal gyrus (IFG, BA44/6). The spread of activity in the VWFA for words does not appear in isolation but is co-active in parallel with spread of activity in anterior middle temporal gyrus (aMTG, BA 21 and 38), posterior middle temporal gyrus (pMTG, BA37/39), and IFG.

What do lateralized displays tell us about visual word perception? A cautionary indication from the word-letter effect

A common assumption underlying laterality research is that visual field asymmetries in lateralized word perception indicate the hemispheric specialisation of processes generally available for the perception of words, including words viewed in a more typical setting (i.e. in the central visual field). We tested the validity of this assumption using a phenomenon (the word-letter effect) frequently reported for displays viewed in the central visual field, where letters in words are perceived more accurately than the same letters in isolation. Words and isolated letters were presented in the left visual field (LVF), right visual field (RVF) and central visual field (CVF), the Reicher-Wheeler task was used to suppress influences of guesswork, and an eye-tracker ensured central fixation. In line with previous findings, lateralized displays revealed a RVF-LVF advantage for words (but not isolated letters) and CVF displays revealed an advantage for words over isolated letters (the word-letter effect). However, RVF and LVF displays both produced an advantage for isolated letters over words (a letter-word effect), indicating that processing subserving the advantage for words when participants viewed stimuli in the central visual field was unavailable for lateralized displays. Implications of these findings for studies of lateralized word perception are discussed.

Specialization within the ventral stream: the case for the visual word form area

Is there specialization for visual word recognition within the visual ventral stream of literate human adults? We review the evidence for a specialized "visual word form area" and critically examine some of the arguments recently placed against this hypothesis. Three distinct forms of specialization must be distinguished: functional specialization, reproducible localization, and regional selectivity. Examination of the literature with this theoretical division in mind indicates that reading activates a precise subpart of the left ventral occipitotemporal sulcus, and that patients with pure alexia consistently exhibit lesions of this region (reproducible localization). Second, this region implements processes adequate for reading in a specific script, such as invariance across upper- and lower-case letters, and its lesion results in the selective loss of reading-specific processes (functional specialization). Third, the issue of regional selectivity, namely, the existence of putative cortical patches dedicated to letter and word recognition, cannot be resolved by positron emission tomography or lesion data, but requires high-resolution neuroimaging techniques. The available evidence from single-subject fMRI and intracranial recordings suggests that some cortical sites respond preferentially to letter strings than to other categories of visual stimuli such as faces or objects, though the preference is often relative rather than absolute. We conclude that learning to read results in the progressive development of an inferotemporal region increasingly responsive to visual words, which is aptly named the visual word form area (VWFA).

Learning letters in adulthood: direct visualization of cortical plasticity for forming a new link between orthography and phonology

To identify which brain regions in adults show plasticity for learning letters, Hangul letters were experimentally associated with either speech sounds (HS condition) or nonspeech sounds (HN condition) in fMRI sessions over two consecutive days. Selective activations under the HS condition were found in several regions including the left posterior inferior temporal gyrus (PITG) and the parieto-occipital cortex (PO), as compared with activations under a condition for familiar letters and speech sounds, and with those under the HN condition. The left PITG showed a selective activation increase under the HS condition over two days, the degree of which predicted individual performance improvement. Further, functional connectivity between the left PITG and the left PO was selectively enhanced under the HS condition. These results demonstrate that a new link between orthography and phonology is formed by the plasticity of a functional system involving the left PITG in association with the left PO.

Spatiotemporal dynamics of word processing in the human cortex

Understanding language relies on concurrent activation of multiple areas within a distributed neural network. Hemodynamic measures (fMRI and PET) indicate their location, and electromagnetic measures (magnetoencephalography and electroencephalography) reveal the timing of brain activity during language processing. Their combination can show the spatiotemporal characteristics (where and when) of the underlying neural network. Activity to written and spoken words starts in sensory-specific areas and progresses anteriorly via respective ventral ("what") processing streams toward the simultaneously active supramodal regions. The process of understanding a word in its current context peaks about 400 ms after word onset. It is carried out mainly through interactions of the temporal and inferior prefrontal areas on the left during word reading and bilateral temporo-prefrontal areas during speech processing. Neurophysiological evidence suggests that lexical access, semantic associations, and contextual integration may be simultaneous as the brain uses available information in a concurrent manner, with the final goal of rapidly comprehending verbal input. Because the same areas may participate in multiple stages of semantic or syntactic processing, it is crucial to consider both spatial and temporal aspects of their interactions to appreciate how the brain understands words.

The neurobiology of adaptive learning in reading: A contrast of different training conditions

fMRI was used to investigate the separate influences of orthographic, phonological, and semantic processing on the ability to learn new words and the cortical circuitry recruited to subsequently read those words. In a behavioral session, subjects acquired familiarity for three sets of pseudowords, attending to orthographic, phonological, or (learned) semantic features. Transfer effects were measured in an event-related fMRI session as the subjects named trained pseudowords, untrained pseudowords, and real words. Behaviorally, phonological and semantic training resulted in better learning than did orthographic training. Neurobiologically, orthographic training did not modulate activation in the main reading regions. Phonological and semantic training yielded equivalent behavioral facilitation but distinct functional activation patterns, suggesting that the learning resulting from these two training conditions was driven by different underlying processes. The findings indicate that the putative ventral visual word form area is sensitive to the phonological structure of words, with phonologically analytic processing contributing to the specialization of this region.

The development of reading impairment: a cognitive neuroscience model

This review discusses recent cognitive neuroscience investigations into the biological bases of developmental dyslexia, a common disorder impacting approximately 5 to 17 percent of the population. Our aim is to summarize central findings from several lines of evidence that converge on pivotal aspects of the brain bases of developmental dyslexia. We highlight ways in which the approaches and methodologies of developmental cognitive neuroscience that are addressed in this special issue-including neuroimaging, human genetics, refinement of cognitive and biological phenotypes, neural plasticity and computational model-can be employed in uncovering the biological bases of this disorder. Taking a developmental perspective on the biological bases of dyslexia, we propose a simple cascading model for the developmental progression of this disorder, in which individual differences in brain areas associated with phonological processing might influence the specialization of visual areas involved in the rapid processing of written words. We also discuss recent efforts to understand the impact of successful reading interventions in terms of changes within cortical circuits associated with reading ability.

Normal and pathological reading: converging data from lesion and imaging studies

In this paper we discuss cognitive and anatomical models of reading that have emerged from behavioral and lesion studies of dyslexia and functional neuroimaging studies of normal subjects. We then suggest that discrepancies in their findings can partly be overcome by functional neuroimaging studies of patients with acquired dyslexia. We present two such studies. One patient had a large left temporoparietal lesion which limited his reading to words with high semantic associations. When he read these words aloud, activation was observed in all areas of the normal reading system with the exception of the damaged left superior temporal lobe. The second patient had anterior temporal lobe atrophy with semantic dementia and a deficit in reading words that rely on lexical or semantic mediation. When asked to read aloud words on which she was likely to succeed, she activated all the normal areas, with increased activation in a left sensorimotor area associated with phonological processing and decreased activation in several areas associated with semantic processing. By relating these findings to those from lesion studies and imaging studies of normals, we propose that the translation of orthography to phonology is mediated semantically by the anterior part of the left midfusiform gyrus. In contrast, when semantic processing is compromised, the translation of orthography to phonology will be more reliant on the posterior part of the left midfusiform and the left frontal areas associated with phonology. Future studies are required to examine the connectivity between these areas during normal and abnormal reading.

Reverse sequencing syllables of spoken words activates primary visual cortex

Using fMRI, we investigated the neural correlates for sequencing the individual syllables of spoken words in reverse order. The comparison of this task to a control task requiring subjects to repeat identical syllables given acoustically revealed the activation of the primary visual cortex. Because one syllable is generally expressed by one kana character (Japanese phonogram), most subjects used a strategy in which the kana character string corresponding to the word was imagined visually and then read mentally in reverse order to perform the task effectively. Such strategy was not used during a control condition. These results suggest that the primary visual cortex plays a role in the generation of an imagined string.

The visual word form area: expertise for reading in the fusiform gyrus

Brain imaging studies reliably localize a region of visual cortex that is especially responsive to visual words. This brain specialization is essential to rapid reading ability because it enhances perception of words by becoming specifically tuned to recurring properties of a writing system. The origin of this specialization poses a challenge for evolutionary accounts involving innate mechanisms for functional brain organization. We propose an alternative account, based on studies of other forms of visual expertise (i.e. bird and car experts) that lead to functional reorganization. We argue that the interplay between the unique demands of word reading and the structural constraints of the visual system lead to the emergence of the Visual Word Form Area.

Category-specific occipitotemporal activation during face perception in dyslexic individuals: an MEG study

In dyslexia, it is consistently found that letter strings produce an abnormally weak or no response in the left occipitotemporal cortex. Time-sensitive imaging techniques have located this deficit to the category-specific processing stage at about 150 ms after stimulus presentation. The typically reported behavioral impairments in dyslexia suggest that the lack of occipitotemporal activation is specific to reading. It could, however, also reflect a more general dysfunction in the left inferior occipitotemporal cortex or in the time window of category-specific activation (150 to 200 ms). As early cortical processing of faces follows a sequence practically identical to that for letter strings, both in location and in timing, we investigated these possibilities by comparing face-specific occipitotemporal activations in dyslexic and non-reading-impaired subjects. We found that both the stage of general visual feature analysis at about 100 ms and the earliest face-specific activation at about 150 ms were essentially normal in the dyslexic individuals. The present results emphasize the special nature of the occipitotemporal abnormality to letter strings in dyslexia. However, in behavioral tests dyslexic subjects were slower and more error-prone than non-reading-impaired subjects in judging the similarity of faces and geometrical shapes. This effect may be related to reduced activation of the right parietotemporal cortex at about 250 ms after stimulus onset.

Neural timing of visual implicit categorization

Most of the neuroimaging studies that have shown visual category-specific activations or categorization effects have been based on a subtractive approach. In the present study, we investigated, by means of EEG, not only the net result of the categorization but also the dynamics of the process. Subjects had to perform a target detection task throughout an image set of distractors belonging to six categories: letters, geometrical figures, faces, tools, structured textures and Asiatic characters. Multivariate analyses were performed on the responses to the non-target stimuli according to their category. Categorical neural responses were only obtained on P2 latencies and N2 amplitudes. This result suggests that there are no differences in the first stage of the implicit categorization of the distractors (visual analysis and proximal stimulus representation elaboration from 100 to 220 ms) and that differences appear between 220 and 280 ms (matching to structural representations). Over-learned stimuli (e.g. letters) elicited the shortest P2 latency, contrasting with unknown categories (e.g. Asiatic characters) that revealed the longest P2 latencies and flattened N2 waves. Categorical differences indicate that the more a subject knows about an object, the less cognitive resources are used. In conclusion, our results suggest that a reduction in neural activity could reflect an improved accuracy in cognitive and cortical processing.

Shared and dissociated cortical regions for object and letter processing

The present study determined the extent to which object and letter recognition recruit similar or dissociated neural resources. Participants passively viewed and silently named line drawings of objects, single letters, and visual noise patterns and centrally fixated an asterisk. We used whole-brain functional MRI and a very conservative approach to hypothesis testing that distinguished among brain regions that were selectively activated by different experimental conditions and those that were conjointly activated. The left fusiform gyrus (BA 19 & 37) and left inferior frontal cortex BA(44/6) showed a greater degree of conjoined activation for objects and letters than selective activation for either category, whereas left inferior parietal cortex (BA 40) and the left insula showed a strong letter-selective response. Equal recruitment of left fusiform and inferior frontal regions by objects and letters reflects similar demands on cognitive processing by these two categories and argues against category-specific modules in these regions. However, cortical systems for object and letter processing are not completely shared given the exclusive activation of left inferior parietal cortex by letters.

Cortical Effects of Shifting Letter Position in Letter Strings of Varying Length

Neuroimaging and lesion studies suggest that occipitotemporal brain areas play a necessary role in recognizing a wide variety of objects, be they faces, letters, numbers, or household items. However, many questions remain regarding the details of exactly what kinds of information are processed by the occipito-temporal cortex. Here, we address this question with respect to reading. Ten healthy adult subjects performed a single word reading task. We used whole-head magnetoencephalography to measure the spatio-temporal dynamics of brain responses, and investigated their sensitivity to: (1) lexicality (defined here as the difference between words and consonant strings), (2) word length, and (3) variation in letter position. Analysis revealed that midline occipital activity around 100 msec, consistent with low-level visual feature analysis, was insensitive to lexicality and variation in letter position, but was slightly affected by string length. Bilateral occipito-temporal activations around 150 msec were insensitive to lexicality and reacted to word length only in the timing (and not strength) of activation. However, vertical shifts in letter position revealed a hemispheric imbalance: The right hemisphere activation increased with the shifts, whereas the opposite pattern was evident in the left hemisphere. The results are discussed in the light of Caramazza and Hillis's (1990) model of early reading.