Neural correlates of foveal splitting in reading: evidence from an ERP study of Chinese character recognition

Recent research on foveal structure and reading suggests that the two halves of a centrally fixated word seem to be initially projected to, and processed in, different hemispheres. In the current study, we utilize two contrasting structures in Chinese orthography, “SP” (the semantic radical on the left and the phonetic radical on the right) and “PS” characters (the opposite structure), to examine foveal splitting effects in event-related potential (ERP) recordings. We showed that when participants silently named centrally presented characters, there was a significant interaction between character type and hemisphere in N1 amplitude: SP characters elicited larger N1 compared with PS characters in the left hemisphere, whereas the right hemisphere had the opposite pattern. This effect is consistent with the split fovea claim, suggesting that the two halves of a character may be initially projected to and processed in different hemispheres. There was no such interaction observed in an earlier component P1. Also, there was an interaction between character type and sex of the reader in N350 amplitude. This result is consistent with Hsiao and Shillcock's [Hsiao, J. H., & Shillcock, R. (2005b). Foveal splitting causes differential processing of Chinese orthography in the male and female brain. Cognitive Brain Research, 25, 531-536] behavioural study, which showed a similar interaction in naming response time. They argued that this effect was due to a more left-lateralized network for phonological processing in the male brain compared with the female brain. The results hence showed that foveal splitting effects in visual word recognition were observed in N1 the earliest, and could extend far enough to interact with the sex of the reader as revealed in N350.     

Foveal splitting causes differential processing of Chinese orthography in the male and female brain

Chinese characters contain separate phonetic and semantic radicals. A dominant character type exists in which the semantic radical is on the left and the phonetic radical on the right; an opposite, minority structure also exists, with the semantic radical on the right and the phonetic radical on the left. We show that, when asked to pronounce isolated tokens of these two character types, males responded significantly faster when the phonetic information was on the right, whereas females showed a non-significant tendency in the opposite direction. Recent research on foveal structure and reading suggests that the two halves of a centrally fixated character are initially processed in different hemispheres. The male brain typically relies more on the left hemisphere for phonological processing compared with the female brain, causing this gender difference to emerge. This interaction is predicted by an implemented computational model. This study supports the existence of a gender difference in phonological processing, and shows that the effects of foveal splitting in reading extend far enough into word recognition to interact with the gender of the reader in a naturalistic reading task.     

A magnetic stimulation examination of orthographic neighborhood effects in visual word recognition

The split-fovea theory proposes that visual word recognition is mediated by the splitting of the foveal image, with letters to the left of fixation projected to the right hemisphere (RH) and letters to the right of fixation projected to the left hemisphere (LH). We applied repetitive transcranial magnetic stimulation (rTMS) over the left and right occipital cortex during a lexical decision task to investigate the extent to which word recognition processes could be accounted for according to the split-fovea theory. Unilateral rTMS significantly impaired lexical decision latencies to centrally presented words, supporting the suggestion that foveal representation of words is split between the cerebral hemispheres rather than bilateral. Behaviorally, we showed that words that have many orthographic neighbors sharing the same initial letters ("lead neighbors") facilitated lexical decision more than words with few lead neighbors. This effect did not apply to end neighbors (orthographic neighbors sharing the same final letters). Crucially, rTMS over the RH impaired lead-, but not end-neighborhood facilitation. The results support the split-fovea theory, where the RH has primacy in representing lead neighbors of a written word.     

The cortical representation of foveal stimuli: evidence from quadrantanopia and TMS-induced suppression

To address the extent to which the visual foveal representation is split, we examined a 29-year-old patient with a lower right quadrantanopia following surgical removal of the left occipital cortex above the calcarine sulcus and compared her performance with subjects receiving transcranial magnetic stimulation (TMS) over the occipital lobes. In a letter/digit classification task, the patient responded accurately to targets presented in the upper visual field, for all horizontal eccentricities. In the lower visual field, she failed to discriminate letters from digits when targets were presented in the right, but not the left visual field (RVF and LVF, respectively). This pattern was also true for the foveal targets, with poor performance to foveal-RVF (0.5 degrees to the right of fixation) but not foveal-LVF (0.5 degrees to the left of fixation) targets. Similar patterns of normal performance to LVF but not RVF or foveal-RVF targets were observed in a group of nine normal observers when TMS was applied over their left occipital cortex. Complementary impairments to LVF and foveal-LVF target classification were induced with TMS over the right occipital cortex. Thus, we have induced an hemianopic pattern in normal observers contralateral to the magnetically stimulated hemisphere. This correspondence between real and TMS-induced visual field defects is further evidence, in neurologically intact subjects, that the cortical representation of the fovea is split between the two hemispheres along the vertical meridian.     

Split fovea theory and the role of the two cerebral hemispheres in reading: A review of the evidence

Split fovea theory proposes that when the eyes are fixated within a written word, visual information about the letters falling to the left of fixation is projected initially to the right cerebral hemisphere while visual information about the letters falling to the right of fixation is projected to the left cerebral hemisphere. The two parts of the word must be re-united before the word can be recognised. Bilateral projection theory proposes instead that visual information is projected simultaneously to both hemispheres provided that it falls within the fovea (defined as the central 2-3°). On this more traditional account, no interhemispheric transfer would be required in order to read a word presented within the fovea. We review the evidence in support of split fovea theory and consider some of the objections that have been raised. We argue that a split fovea affects the reading of words at fixation, something that must be recognised and accounted for by cognitive, computational and neural models of reading.     

Evaluating a split processing model of visual word recognition: effects of word length

A new theory of visual word recognition is based on the fact that the fovea is split in humans. When a reader fixates the center of a written word, the initial letters of the word that are to the left of fixation are projected first to the right cerebral hemisphere (RH) while the final letters are projected to the left cerebral hemisphere (LH). This paper explores the possibility that this has consequences for the early processing of the beginning and ends of centrally fixated words: specifically that lexical decision RTs are affected by the number of letters to the left of fixation but not by the number of letters to the right of fixation. For centrally presented five- and eight-letter words, we manipulated number of letters presented to the right or to the left of a fixation point (Experiment 1). We found that longer latencies to longer letter strings characterised the processing of the initial letters of words while LH word recognition features characterised the ends of words. Experiment 2 was a lateralized version of Experiment 1, and revealed the well established visual field and word length interaction. The results supported the split fovea theory.     

Magnetic stimulation of the right visual cortex impairs form-specific priming

Recent evidence suggests that priming of objects across different images (abstract priming) and priming of specific images of an object (form-specific priming) are mediated by dissociable neural processing subsystems that operate in parallel and are predominantly linked to left and right hemispheric processing, respectively [Marsolek, C. J. Dissociable neural subsystems underlie abstract and specific object recognition. Psychological Science, 10, 111-118, 1999]. Previous brain imaging studies have provided important information about the neuroanatomical regions that are involved in form-specific and abstract priming; however, these techniques did not fully establish the functional significance of priming-related changes in cortical brain activity. Here, we used repetitive transcranial magnetic stimulation (rTMS) in order to establish the functional role of the right occipital cortex in form-specific priming [Kroll, N. E. A., Yonelinas, A. P., Kishiyama, M. M., Baynes, K., Knight, R. T., & Gazzaniga, M. S. The neural substrates of visual implicit memory: Do the two hemispheres play different roles? Journal of Cognitive Neuroscience, 15, 833-842, 2003]. Compared to no TMS and sham TMS, rTMS of the right occipital cortex disrupted immediate form-specific priming in a semantic categorization task. Left occipital rTMS, on the other hand, had no converse effect on abstractive priming. Abstract priming may involve deeper semantic processing and may be unresponsive to magnetic stimulation of a single cortical locus. Our TMS results show that form-specific priming relies on a visual word-form system localized in the right occipital lobe, in line with the predictions from divided visual field behavioral studies [Marsolek, 1999].     

Language and speech capacity of the right hemisphere

Right hemisphere language and speech capacity was further analyzed in brain-bisected patients. The results indicate that little or no syntactic capability exists in the right hemisphere. The only semantic dimension that was comprehended in a series of pictorial-verbal matching tests was the affirmative-negative. Moreover, earlier indications of a right hemisphere speech capacity could not be confirmed. Differences in verbal reaction time to visual stimuli projected to right and left hemispheres were alternatively interpreted as consequences of subcortical transfer mechanisms or cross-cuing strategies.     

Functional dissociation of the hemispheres using foveal visual input

A consistent 10 (± 2.2) msec visual field superiority in vocal RT to letters was found using both foveal and extra-foveal input (0.5°, 2° and 4° eccentricities), when typical procedures in the paradigm were modified. This consistency occurred in both the right visual field (for extreme right-handers) and left visual field (for extreme left-handers). A second experiment using a recognition memory task with manual RT also revealed concurrent opposite hemisphere superiorities for letters and geometric forms presented only 5.2–20.7' left or right of fixation. Acuity differences, directional scanning, and asymmetric cerebal activation could not account for the results. A substantial splitting of cortical connections from the innermost fovea is thereby indicated, as well as the viability of using foveal-only input in studies of cerebral functional asymmetries.     

Magnetic stimulation of the left visual cortex impairs expert word recognition

One of the hallmarks of expert reading is the ability to identify arrays of several letters quickly and in parallel. Such length-independent reading has only been found for word stimuli appearing in the right visual hemifield (RVF). With left hemifield presentation (LVF), response times increase as a function of word length. Here we investigated the comparative efficiency with which the two hemispheres are able to recognize visually presented words, as measured by word length effects. Repetitive transcranial magnetic stimulation (rTMS) of the left occipital cortex disrupted expert processing of the RVF such that a length effect was created (Experiment 1). Right occipital rTMS, on the other hand, had no such effect on RVF words and nor did it modulate the length effect already present in the LVF. Experiment 2 explored the time course of these TMS-induced effects by applying single pulses of TMS at various stimulus-onset asynchronies for the same task. We replicated the TMS-induced length effect for RVF words, but only when a single pulse was applied to the left visual cortex 80 msec after target presentation. This is the first demonstration of TMS-induced impairment producing a word length effect, and as such confirms the specialization of the left hemisphere in word recognition. It is likely that anatomical differences in the pathway linking retinal input to higher level cortical processing drive this effect.     

Semantic and phonemic priming in the cerebral hemispheres

Representation of semantic and phonemic codes in the cerebral hemispheres was investigated in two priming experiments where prime and target words were independently projected to the left or right visual fields. The first experiment, using phonemic primes, confirmed the view that phonological information is not accessible to the right hemisphere. Priming effects were obtained only when the prime and target were both projected to the right visual field. The second experiment, employing category exemplars as primes, again found the left hemisphere to be the principal locus of the priming effects. The right hemisphere was unable, by itself, to activate words related to the exemplar prime. However, projection of the prime to the right visual field significantly facilitated responses to left visual field targets. The present findings support the view advanced by Drews (Neuropsychologia 25, 419-427, 1987) and Levy and Trevarthen (J. exp. Psychol., Hum. Percep. Perform. 2,299-312, 1976) that the left lexicon is structured in accordance with an hierarchy of logical semantic relationships, while the right lexicon is organized on the basis of simple associations between concepts. It is suggested, furthermore, that the patterns of semantic, but not phonological, activation invoked by a prime may be relayed between the two lexicons.     

Impaired interhemispheric interactions in patients with major depression

Previous studies have shown that patients with major depression have an interhemispheric imbalance between right and left prefrontal and motor cortex. We aimed to investigate the interhemispheric interactions in patients with major depression using repetitive transcranial magnetic stimulation (rTMS). Thirteen patients with major depression and 14 age-matched healthy subjects participated in this study. Corticospinal excitability before and after 1 Hz rTMS (applied to the left primary motor cortex) was assessed in the left and right motor cortex and these results were compared with those in healthy subjects. There was a significant difference in the interhemispheric effects between patients with depression and healthy subjects. In healthy subjects, 1 Hz rTMS significantly decreased corticospinal excitability in the stimulated, left hemisphere and increased it in the contralateral, right hemisphere. In depressed subjects, 1 Hz rTMS also decreased corticospinal excitability in the left hemisphere; however, it induced no significant changes in corticospinal excitability in the contralateral, right hemisphere. In addition, there was a significant correlation between the degree of interhemispheric modulation and the severity of the depression as indexed by the Beck Depression Inventory scores. Our findings showing a decreased interhemispheric modulation in patients with major depression are consistent with the notion that mood disorders are associated with slow interhemispheric switching mechanisms.     

The Split Fovea Theory and the Leicester critique: What do the data say?

According to the Split Fovea Theory (SFT) recognition of foveally presented words involves interhemispheric transfer. This is because letters to the left of the fixation location are initially sent to the right hemisphere, whereas letters to the right of the fixation position are projected to the left hemisphere. Both sources of information must be integrated for words to be recognized. Evidence for the SFT comes from the Optimal Viewing Position (OVP) paradigm, in which foveal word recognition is examined as a function of the letter fixated. OVP curves are different for left and right language dominant participants, indicating a time cost when information is presented in the half-field ipsilateral to the dominant hemisphere (Hunter, Brysbaert, & Knecht, 2007). The methodology of the SFT research has recently been questioned, because not enough efforts were made to ensure adequate fixation. The aim of the present study is to test the validity of this argument. Experiment 1 replicated the OVP effect in a naming task by presenting words at different fixation positions, with the experimental settings applied in previous OVP research. Experiment 2 monitored and controlled eye fixations of the participants and presented the stimuli within the boundaries of the fovea. Exactly the same OVP curve was obtained. In Experiment 3, the eyes were also tracked and monocular viewing was used. Results again revealed the same OVP effect, although latencies were remarkably higher than in the previous experiments. From these results we can conclude that although noise is present in classical SFT studies without eye-tracking, this does not change the OVP effect observed with left dominant individuals.     

Re-evaluating split-fovea processing in word recognition: Effects of fixation location within words

It has been claimed that word recognition is affected fundamentally by the precise location at which a word is fixated because a precise split in hemispheric processing at the point of fixation causes all letters to the left and right of fixation to project to different, contralateral hemispheres. To assess this claim, 5-letter words (and nonwords) were presented for lexical decision when participants fixated the space immediately to the left (location 1) or right (location 6) of each stimulus, or one of the four possible inter-letter spaces (locations 2-5). Fixation location was controlled using an eye-tracker linked to a fixation-contingent display and all stimuli were presented entirely within foveal vision to avoid confounding influences of extrafoveal hemispheric projections. Performance was equally poorest when fixating locations 1 and 6 (when words were shown entirely to either the right and left of fixation), intermediate for location 5, and equally superior for locations 2, 3, and 4. Additional word-specific analyses also showed no evidence of the effects of fixation location on optimal word recognition predicted by split-fovea processing. These findings suggest that, while fixation location influences word recognition, word recognition is apparently not affected by a split in hemispheric processing at the point of fixation and does not depend critically on the precise location at which a word is fixated. Implications of these findings for the role of fixation location in word recognition are discussed.     

Evaluating a split fovea model of visual word recognition: effects of case alternation in the two visual fields and in the left and right halves of words presented at the fovea

Two experiments are reported exploring the effect of cAsE aLtErNaTiOn on lexical decisions to words and nonwords presented laterally or centrally. In line with previous research, Experiment 1 found that case alternation slowed lexical decision responses to words more in the right visual field (RVF) than in the left visual field (LVF). In Experiment 2, the words and nonwords were all presented centrally. There were three conditions, a condition in which the word and nonwords were presented in lower case letters, a condition in which the letters to the left of the central fixation were case alternated (e.g., aMbItion, mOdLants) and a condition in which the letters to the right of fixation were case alternated (e.g., collApSe, pireNtOl). Alternating the case of letters to the right of fixation slowed lexical decision responses more than alternating letter case to the left of fixation. The results provide further support for a split fovea account of visual word recognition according to which those letters of a centrally-fixated word that fall to the left of fixation are processed initially by the right cerebral hemisphere while those letters that fall to the right of fixation are processed initially by the left cerebral hemisphere, with the characteristics of the left and right hemispheres being revealed in the processing of initial and final letters in centrally presented words.     

Hemispheric balance in coding speech and non-speech sounds in Chinese participants

To study the role of neuromagnetic auditory approximately 100-ms responses (N100m) in phonetic processing, we recorded N100m in 24 right-handed Chinese participants using a whole-head neuromagnetometer. The stimuli included vowel /a/ and consonant-vowels /ba/ and /da/, spoken by one Chinese speaker, and a 1-kHz tone. N100m to tones was larger in the right hemisphere, whereas that to speech sounds was bilaterally similar. The amplitude ratio of speech to non-speech N100m was larger in the left hemisphere. N100m dipoles in the left hemisphere were approximately 2 mm more anterior for speech than for tone stimuli. The results suggest that N100m reflects both acoustics and phonetic processing. Moreover, the ratio of speech to non-speech activation in individual hemispheres may be useful for language lateralization.     

Re-evaluating split-fovea processing in word recognition: a critical assessment of recent research

In recent years, some researchers have proposed that a fundamental component of the word recognition process is that each fovea is divided precisely at its vertical midline and that information either side of this midline projects to different, contralateral hemispheres. Thus, when a word is fixated, all letters to the left of the point of fixation project only to the right hemisphere whereas all letters to the right of the point of fixation project only to the left hemisphere. An informed assessment of research in this area requires an accurate understanding of the nature of the evidence and arguments that have been used to develop this “split-fovea theory” of word recognition (SFT). The purpose of this article is to facilitate this understanding by assessing recent published support for SFT. In particular, we assess (i) the precision with which experiments have been conducted, (ii) the assumptions made about human visual ability, and (iii) the accuracy with which earlier research has been reported. The assessment reveals shortcomings and errors that are likely to impact on an accurate understanding of research in this area and, therefore, on an accurate understanding of the viability of SFT.     

Right hemisphere contributions to the comprehension of low-imagery words

A priming experiment, with normal university students as subjects, was used to investigate whether the right cerebral hemisphere contributes to the comprehension of low-imagery words. Each hemisphere's access to semantic representations of low-imagery words was gauged by comparing responses to low-imagery targets preceded by associated low-imagery primes (e.g., BELIEF-IDEAL) with responses to the same targets when they were preceded by unrelated primes (e.g., FATE-IDEAL). All primes and targets were independently projected to the left or right visual fields (LVF or RVF), and temporally separated by a stimulus onset asynchrony of 250 ms. There was a clear RVF advantage in response speed and accuracy measures, confirming the left hemisphere's advantage in processing low-imagery words. Nonetheless, the priming effects provided evidence that the right hemisphere contributes to the comprehension of low-imagery words, as primes projected to the RVF equally facilitated responses to associated targets subsequently appearing in either visual field. In contrast, primes directed to the LVF did not facilitate responses to associated targets projected to the LVF or RVF. The results suggest that low-imagery words projected to the left hemisphere activated low-imagery associates in both hemispheres to an equivalent degree, whereas low-imagery primes directed to the right hemisphere failed to activate low-imagery associates in either hemisphere. Like Kounios and Holcomb's (1994) study of event-related response potentials evoked by abstract and concrete words, the findings indicate that while the left hemisphere is the primary processor of low-imagery/abstract words, the right hemisphere plays a subsidiary role in the comprehension of these words.     

Elaboration of representations in a left-handed subject with callosal damage

Four experiments were carried out in a familial left-handed patient who, following a vascular accident, underwent a callosal lesion and had a right hemianopia. Previous observations by Poncet et al. (1978) demonstrated that both hemispheres controlled expressive language. In the first experiment haptic unimanual exploration showed that each system (right hand/left hemisphere; left hand/right hemisphere) elaborated very different mental representations of the stimulus and gave different types of verbal report of it (naming and talking about). In the second experiment, manual exploration of a maze in the search of a target was performed randomly with the right hand and following a competent spatial strategy with the left hand. The third experiment investigated the ability of the right hemisphere to discriminate the phonetic features of the names of objects that were identified but not named. In the fourth experiment, recognition memory of objects was shown to be better when information was processed by the left hand - right hemisphere system than by the right hand - left hemisphere system. Findings are discussed in terms of verbal vs. non-verbal processes and of perceptual awareness.     

Differences in semantic category priming in the left and right cerebral hemispheres under automatic and controlled processing conditions.

The contribution of each cerebral hemisphere to the generation of semantic category meanings at automatic and strategic levels of processing was investigated in a priming experiment where prime and target words were independently projected to the left or right visual fields (LVF or RVF). Non-associated category exemplars were employed as related pairs in a lexical decision task and presented in two experimental conditions. The first condition was designed to elicit automatic processing, so related pairs comprised 20% of the positive set, stimulus pairs were temporally separated by a stimulus onset asynchrony (SOA) of 250 ms, and there was no allusion to the presence of related pairs in the instructions to subjects. The second condition, designed to invoke controlled processing, incorporated a relatedness proportion of 50%, stimulus pairs separated by an SOA of 750 ms, and instructions which informed subjects of the presence and use of category exemplar pairs in the stimulus set. In the first condition, a prime directed to either visual field facilitated responses to categorically related targets subsequently projected to the RVF, while in the second condition a prime directed to either visual field facilitated responses to related targets projected to the LVF. The facilitation effects obtained in both conditions appeared to reflect automatic processes, while strategic processes were invoked in the left, but not the right hemisphere in the second condition. The results suggest that both hemispheres have automatic access to semantic category meanings, although the timecourse of activation of semantic category meanings is slower in the right hemisphere than in the left.