COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3), 243 – 259 Examination of the split fovea theory in a case of pure left hemialexia Eric Siéroff Laboratoire de Psychologie et Neurosciences Cognitive, Université René Descartes/CNRS, Paris, France Michal Lavidor University of Hull, UK To address the extent to which the visual foveal representation is split, we examine the case of a patient, M.B., suffering from a left mesial occipital lesion and presenting a pure left hemialexia and a right hemianopia with a spared area of the macula. Reading performance on tachistoscopically presented four-letter words and pseudowords in the spared area of the right visual field was significantly better than reading performance in the intact left visual field. Reading performance in the spared area of the right visual field was also significantly better than reading performance of stimuli centred on the fovea. Moreover, a length effect was found only in the left half of pseudowords centred on the fovea, but not in the right half (up to five letters). These differences in reading efficiency between the left and right halves of the foveal region militates in favour of the split fovea theory and cannot be explained by the bilateral projection theory. When a person is fixating, information presented to the right of fixation (in the right visual field, RVF) is projected to the visual cortex of the left cerebral hemisphere (LH), while information presented to the left of fixation (in the left visual field, LVF) is projected to the visual cortex of the right cerebral hemisphere (RH). However, it is not entirely clear how the visual half-fields flank one another along the vertical meridian to represent the foveal area (Brysbaert, 1994), which occupies the central 2 degrees of the retina (Choplin & Edwards, 1998). One suggestion is that the representation of foveal stimuli is projected to both hemispheres (with various degrees of overlap); we call this model the bilateral projection theory (BPT). Some influential animal studies have reported anatomical evidence for a hemispheric overlap along the vertical meridian, supporting the BPT (Bunt & Minckler, 1977; Leventhal, Thompson, & Liu, 1993; Stone, 1966; Stone, Leicester, & Sherman, 1973). Alternatively, the foveal representation may be split between the two hemispheres (the split fovea theory, SFT). According to the SFT, visual Correspondence should be addressed to Eric Siéroff, Laboratoire de Psychologie et Neurosciences Cognitive, Université René Descartes/CNRS–FRE 2987, Centre Henri Piéron, 92774 Boulogne-Billancourt Cedex, France. (E-mail: eric.sieroff@univ-paris5.fr). We thank M.B. and J.S. for their patience. We also thank Zofia Laubitz for her help. M. Lavidor is a member of the Research Training Network on Language and Brain, funded by the European Community. # 2007 Psychology Press, an imprint of the Taylor & Francis Group, an Informa business http://www.psypress.com/cogneuropsychology 243 DOI:10.1080/02643290601161926 SIÉROFF AND LAVIDOR information in the right and left halves of the foveal area are initially projected to the visual cortex of the left and right cerebral hemispheres, respectively (Brysbaert, 1994, 2004; Brysbaert, Vitu, & Schroyens, 1996; Lavidor & Walsh, 2004). One prediction of the SFT is that the hemispheric differences that have been found with parafoveal presentation will also be found in the processing of the left and right segments of centrally presented targets (Lavidor & Walsh, 2003). Reading performance is different in the LVF and RVF, with better performance in the RVF, when words are parafoveally presented (Mishkin & Forgays, 1952). Also, hemispheric differences in word length effects have been reported for words presented parafoveally (Bouma, 1973; Chiarello, 1988; Ellis, Young, & Anderson, 1988; Gill & McKeever, 1974; Iacoboni & Zaidel, 1996; Weekes, 1997; Young & Ellis, 1985). Two broadly different approaches have been taken: One focuses on the idea of qualitatively different “modes of processing” available to the two hemispheres (Ellis, 2004; Lavidor & Ellis, 2002a, 2002b); the other argues that the apparent qualitative effects may arise as a result of more quantitative differences in the way the hemispheres process words (Whitney, 2001). Some recent findings show that under careful presentation conditions, differences occurred between the left and right segments of centrally presented words, which mimic parafoveal differences (Lavidor, Ellis, Shillcock, & Bland, 2001). These results were interpreted as supporting the SFT (Lavidor et al., 2001; Shillcock, Ellison, & Monaghan, 2000). To further the exploration of the cortical representation of foveally presented words in the two cerebral hemispheres, we report here a study of a case of left hemialexia. Left hemialexia is invariably observed following posterior callosal lesions (Abe, Nakamura, Sugishita, Kato, & Iwata, 1986; Binder et al., 1992; Cohen & Dehaene, 1996; Damasio, Chui, Corbett, & Kassel, 1980; Degos et al., 1987; Habib, Ceccaldi, & Poncet, 1990; Lausberg, Göttert, Münßinger, Boegner, & Marx, 1999; Levine & Calvanio, 1980; Maspes, 1948; Sidtis, Volpe, Holtzman, Wilson, & Gazzaniga, 1981; Sugishita, Iwata, Toyokura, 244 COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) Yoshioka, & Yamada, 1978; Sugishita, Yoshioka, & Kawamura, 1986; Suzuki et al., 1998; Trescher & Ford, 1937; Wechsler, 1972). Patients with left hemialexia typically show impaired reading of LVF words, contrasting with a perfectly normal performance with RVF words. Cohen et al. (2000) argued that this asymmetric hemispheric pattern is a direct consequence of the inability of visual stimuli initially perceived by the RH to reach the LH’s visual word form area (VWFA) through the splenium of the corpus callosum. Due to the selective impairment of LVF words, the phenomenon of left hemialexia allows one to test the predictions of the SFT. According to the SFT, left hemialexia should be characterized by difficulties in reading words when they are presented in the LVF—even in the foveal portion of the LVF—without difficulties in reading words in the RVF—even in the foveal portion of the RVF. Alternatively, in line with the BPT, left hemialexia should be characterized by difficulties in reading words in the LVF, but reading should be unimpaired in the foveal portion thereof. Case history M.B., a retired 67-year-old university professor, right-handed, suffered from a left mesial occipitotemporal lesion, interrupting the splenial fibres (Figure 1). He was tested one year after the ictus and complained about reading difficulties. He also complained about an amputation of his RVF, and visual perimetry revealed a right homonymous hemianopia sparing the fovea and parafovea. The neurological examination revealed no other sensory, motor, or reflex abnormalities. At the time of testing, reading of running text was correct, even though it was relatively slow. Neuropsychological examination revealed no other deficit. Further testing of his language functions revealed normal speech, auditory comprehension, and repetition. Writing and spelling were normal. Confrontation naming of objects and line drawings was normal, at least when objects were presented without time limits. M.B. showed no signs of neglect (copy, line cancellation PURE LEFT HEMIALEXIA Figure 1. CT scan of M.B., showing the left mesial occipital lesion. test, bisection). Finally, no clinical signs of callosal disconnection in the auditory (normal dichotic test) and somaesthetic modalities (no left tactile anomia) were found. The control participant, J.S., is a 63-year-old right-handed male, also a university professor, without any known neuropsychological deficit. He was still working when tested. EXPERIMENT 1: READING SINGLE LETTERS AT DIFFERENT ECCENTRICITIES In order to explore M.B.’s reading, we first presented him with isolated letters. Moreover, because M.B. has an incomplete right hemianopia, we were interested in defining the spared area of vision in the RVF. Method Eight different letters were presented: A, B, E, F, H, K, P, and R. The letters were uppercase, printed in Geneva font, plain, 18 points. Each letter covered a horizontal visual angle of 0.4 degrees. Letters appeared in 13 different locations, ranging from 2.6 degrees to the left of the fixation item to 2.6 degrees to the right of the fixation item, with a distance of 0.4 degrees between the centres of two adjacent locations. Only M.B. took part in this experiment. He sat at a distance of 60 cm from the screen. Single letters appeared 500 ms after a fixation item (plus sign). Their exposure duration was 67 ms, followed by a patterned mask (a string of ampersands), which remained present until the next trial. The task was to identify and read aloud the letter. The experimenter triggered each trial after the oral response. Two blocks of 104 trials each (eight letters in each location) were presented in two COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) 245 SIÉROFF AND LAVIDOR different sessions, with several days in between. Trials were randomized in each block. Results and discussion The results showed first that single letters were correctly identified in the LVF. In addition, there was a clearly spared area for letter identification in a small portion of the RVF. However, letter identification was nearly impossible beyond 2 degrees of visual angle in the RVF, undoubtedly because of the right hemianopia (Figure 2). EXPERIMENT 2: IDENTIFICATION OF OBJECT DRAWINGS IN EACH VISUAL HEMIFIELD In this experiment, objects were presented in the spared area of the RVF and an equivalent location in the LVF. Method The stimuli were 50 object drawings, adapted from Snodgrass and Vanderwart (1980). The different drawings covered a visual angle varying between 1 and 1.5 degrees (vertically and horizontally). They were centred at 1 degree to the left (LVF) or right (RVF) of fixation, so that the maximum distance between the fixation item and the farthest extremity of the object was 1.8 degrees. Each trial began with the presentation of a fixation item, a plus sign, for 500 ms in the centre of the screen. Immediately afterwards, one (unilateral presentation) or two (bilateral presentation) object drawings were presented for 50 ms. No mask was used. The task was to identify and name the object. Unilateral and bilateral presentations were blocked. Each object drawing was repeated four times: twice in each block and twice in each hemifield. Results and discussion We calculated the number of correctly identified object drawings (Table 1). M.B.’s performance was clearly inferior (40.5% correct) to the control J.S.’s performance (90% correct), x2(1) ¼ 54.0; p , .01, showing that M.B. has a global perceptual impairment. We were mainly interested in the difference between the visual hemifields. J.S. did not show any such difference. This absence of asymmetry is in agreement with results in the literature (Biederman & Cooper, 1991; Kanwisher, Woods, Iacoboni, & Mazziotta, 1997). Similarly, M.B.’s capacity to identify object drawings did not differ between his right and left visual fields, for either the unilateral, x2(1) ¼ 1.96; ns, or the bilateral condition, x2(1) ¼ 1.65; ns. To sum up, M.B. demonstrated a general difficulty in processing visual stimuli, but no asymmetry or visual extinction for object drawings. Table 1. Recognition of object drawings in the left and the right visual field Figure 2. Number of correctly identified single letters presented in different locations, in patient M.B. (Experiment 1). 246 COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) Presentation n Participant LVF RVF Unilateral 50 Bilateral 50 J.S. M.B. J.S. M.B. 44 28 47 19 45 21 44 13 Note: LVF ¼ left visual field; RVF ¼ right visual field. PURE LEFT HEMIALEXIA EXPERIMENT 3: READING WORDS AND PSEUDOWORDS AT DIFFERENT ECCENTRICITIES In this experiment, words and pseudowords were presented at different eccentricities to compare M.B.’s reading performance in the LVF and the spared area of the RVF and to evaluate his performance with centred letter strings. Method A total of 140 four-letter words and 140 fourletter pseudowords were presented in seven different locations (20 words and 20 pseudowords per location). Each letter string was presented once. The real words were very common words. The seven lists of 20 words, according to different locations, had similar means of log-frequencies (minimum ¼ 348; maximum ¼ 381), according to Brulex (Content, Mousty, & Radeau, 1990). There were two or three disyllabic words per list, and the remainder were monosyllabic. Pseudowords were constructed with the same letters as words and were pronounceable. Letter strings were in lowercase, printed in Geneva font, 18 points, plain. Each four-letter string subtended a visual angle of 2 degrees. The leftmost and rightmost locations were centred at 3 degrees of visual angle from fixation (L3 and R3). Other locations were at an eccentricity of 1 and 2 degrees (L1 and R1, L2 and R2, respectively), or centred (C). Note that the L1 and R1 locations were equivalent to those employed in Experiment 2. The procedure was similar to that used for single letters (fixation item, patterned mask), except that a 116 ms-exposure duration was used for M.B., and 67 ms for J.S. Different exposure durations were employed for each participant in order to reduce the global difference in performance between M.B. and J.S. The task was to identify and read aloud the letter string. If participants were unable to correctly identify the target, they were encouraged to respond anyhow, by pronouncing a word or a pseudoword that resembled the target, or by spelling some of the letters. Trials were randomized and were presented in two separate sessions for J.S. and three for M.B., with an interval of one week at most between sessions. Several practice trials using different letter strings from those used in the experiment were run before the experiment. Results Two analyses were conducted. The first one considers the number of correct items (words, pseudowords), independently of the number of letters correctly identified. The second one considers the correct identification of each letter for each letter position in the string. Letter string identification and eccentricity The results of the first analysis of the correct items are shown in Figure 3. J.S.’s performance declined as the eccentricity of items increased and was slightly better in the RVF than in the LVF. M.B.’s identification of letter strings was highly dependent on the eccentricity and hemifield of presentation: Location effects differed between M.B. and J.S., x2(6) ¼ 115.8; p , .001, and this was true for pseudowords as well as words. M.B.’s number of correct items was particularly poor in the LVF, for all locations. It was also very poor for the extreme eccentricities in the RVF, which is compatible with an incomplete Figure 3. Number of correctly identified words and pseudowords presented at different locations, in control J.S. and patient M.B. (Experiment 3). COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) 247 SIÉROFF AND LAVIDOR right hemianopia. However, M.B.’s performance was much better in the spared area of the RVF—that is, when the letter string was centred on 1 degree of visual angle in the RVF and covered 2 degrees to the right of fixation. For letter strings (words and pseudowords) in the R1 location, there was no significant difference between M.B. and J.S. (M.B. ¼ 30/40; J.S. ¼ 34/40; x2 ¼ 1.32, ns). However, J.S.’s performance was significantly better than M.B.’s in location L1 (M.B. ¼ 6/40; J.S. ¼ 34/40; x2 ¼ 76.86, p , .001). Most interestingly, M.B.’s identification of words and pseudowords centred on the fixation item and covering an overall visual angle of 2 degrees (1 degree in each visual field) was poor compared to his identification of letter strings located in R1 (entirely in the RVF). Letter position and eccentricity Data were also analysed for each letter position in the letter strings. Letter position analysis was straightforward when responses were correct or when they were spelled. However, most incorrect responses involved a pronounced nonword or word. In this case, the identification of each target letter was evaluated by examining the response. In the first possible scoring method, all correctly identified letters, independent of their correct locations, were recorded. In the second scoring method, only letters that were correctly identified and in the correct position in the response (with no transposition) were recorded. Global performance was found to be lower with the second scoring method than with the first. However, because the patterns of effects were similar with both scoring methods, or were even greater with the second (in accordance with previous studies, Siéroff & Posner, 1988), we report here only the results using the first scoring method (Table 2). Considering the number of correct letters per item, there was a global superiority effect for words over pseudowords, x2(1) ¼ 7.71; p , .01, and this superiority effect was not significantly different between M.B. (57% vs. 55%) and J.S. (94% vs. 86%), x2(1) ¼ 0.3; ns, even in the LVF, x2(1) ¼ 0.01; ns. 248 COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) There was also an eccentricity effect in each visual field. In the LVF, the eccentricity effect was not significantly different between M.B. (100 correct letters in L1, 82 in L2, 65 in L3) and J.S. (158, 123, and 117, respectively), x2(2) ¼ 0.8; ns. In the RVF, there was a significantly stronger eccentricity effect in M.B. (152 correct letters in R1, 94 in R2, and 18 in R3) than in J.S. (156, 157, and 143, respectively), x2(2) ¼ 125.2; p , .001. M.B.’s letter identification declined beyond 2 degrees of visual angle in the RVF, even when the letters were embedded in words or pseudowords, confirming the results of Experiment 1. However, his identification of letters in letter strings presented immediately to the right of fixation (location R1) was high, showing a ceiling effect, without any discrepancy of letter identification from left to right: M.B. (like J.S.) showed no decline in the identification of individual letters between the initial and final letters, and this was true for pseudowords as well as words. In the LVF, M.B.’s letter identification deteriorated according to the letter position from left to right. The decrease in identification between the initial and the fourth and final letters was stronger for M.B. than for J.S. for all locations in the LVF: L3, L2, and L1, x2(1) ¼ 9.96; p , .01. The ratio of correct identification between the final and the initial letter (positions L3, L2, and L1) was .82 for J.S. and .41 for M.B. This ratio was slightly stronger for pseudowords than for words, in M.B. (.31 and .49, respectively) and J.S. (.74 and .90). Note that this position effect in M.B. was strong even in the L1 location, although the third and fourth letters were located within 1 degree of visual angle from the fixation item (Table 2). M.B.’s ratio of correct identification between the final and initial letters is .54 in L3, .31 in L2, .41 in L1. Discussion The results of Experiment 3 indicate that M.B. has some general difficulties in identifying words, because we had to use a longer presentation duration than that for J.S. to obtain comparable PURE LEFT HEMIALEXIA Table 2. Number of correct individual letters for each position in words and pseudowords presented at different locations in patient M.B. and control J.S. in Experiment 3 Note: Max. individual letters ¼ 20. L ¼ left; R ¼ right. performance. M.B. has a global perceptual problem, already noted in Experiment 2. Also, Experiments 1 and 3 clearly show that M.B. found it nearly impossible to read any material in the RVF, when eccentricity exceeded 2 degrees of visual angle. This deficit is certainly due to the visual sensory deficit—namely, right hemianopia—which spares 2 degrees of visual angle. Most importantly, Experiment 3 shows a strong asymmetry between the LVF and the spared area of the RVF in M.B. While recognition of object drawings did not differ between the R1 and L1 locations (Experiment 2), M.B. showed better reading performance in R1 than in L1 (Experiment 3). This asymmetry in reading letter strings was not found for the control participant, J.S. In other words, M.B. shows signs of left hemialexia without left hemiagnosia. His hemialexia could be caused by the left occipital lesion, which most probably involved splenial fibres or disconnected the splenium from left hemispheric areas (Figure 1). Most probably, M.B. had problems transferring letter information from the visual areas of the RH to the language areas of the LH (cf., Cohen et al., 2000). The superiority of words over pseudowords in identifying letters was equivalent for M.B. and J.S., suggesting that the VWFA (Cohen et al., 2000) was intact in M.B. However, even though words were better identified than pseudowords, the effect of the letter’s position within a string was similar for both types of letter strings. This result is in agreement with results from other studies showing that pseudoword processing, particularly with short pseudowords, can be relatively COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) 249 SIÉROFF AND LAVIDOR similar to word processing in neglect patients (Riddoch, Humphreys, Cleton, & Ferry, 1990) and in normal participants (Carr, Posner, Pollatsek, & Snyder, 1979; Gibson, Pick, Osser, & Hammond, 1962). Studies using brain imagery have come to similar conclusions (McCandliss, Cohen, & Dehaene, 2003). There was a significant hemifield asymmetry considering the letter position effect. Identification was similar for all four positions in letter strings presented in the spared area of the RVF in M.B. Identification dropped dramatically from the initial to the final position for LVF letter strings, specifically for M.B. This was true even for letter strings located in L1 (centred at 1 degree left of fixation)—thus, including foveal letters—a result that supports the SFT. This differential effect of letter position between hemifields indicates that M.B. had particular difficulties with reading in the LVF. Moreover, the BPT would predict good performance when words and pseudowords are presented in the fovea, with a level similar to performance for letter strings presented immediately to the right of fixation. The poor result that M.B. obtained for centred letter strings strongly supports the SFT. EXPERIMENT 4: SPATIAL CUEING IN READING LATERALIZED WORDS Experiment 3 showed that M.B.’s reading performance was better in the spared area of the RVF than in the LVF. We interpreted these results as indicating a specific reading deficit in the LVF. Another possibility is that M.B.’s divergent performance in the LVF and RVF is explained by the advantage an initial letter has in an RVF letter string, because it is closer to the fixation point than is the initial letter of an LVF letter string. For example, performance could be worse in the LVF because attention has to be reoriented toward the initial letter following the presentation of the fixation item, whereas a smooth rightward orientation of attention is sufficient to identify RVF stimuli, because the initial 250 COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) letter is located immediately to the right of the fixation item. To explore this possibility, a cue was presented before the presentation of a lateralized word at the beginning or end of the letter string. It has been shown with centred letter strings that such beginning/end cues have a strong effect when letter processing is sequential (Siéroff & Posner, 1988), and that beginning cues improve the identification of words presented to the LVF, but not to the RVF (Lindell & Nicholls, 2003). Stronger cueing effects in the LVF than in the RVF have also been widely reported for words, with different types of cue (Gatheron & Siéroff, 1999; Hardyck, Chiarello, Dronkers, & Simpson, 1985; Mondor & Bryden, 1992; Nicholls & Wood, 1998). If M.B.’s worse performance in the LVF is caused by the delay in reorienting from the fixation item toward the initial letter of the letter string, his performance should improve with a cue located at the beginning of the letter string. Method A total of 120 four-letter pseudowords (taken from the stimuli for Experiment 3) were presented unilaterally in the LVF and RVF (centred at 1 degree of eccentricity). The cue was a digit (from 1 to 9). All materials were presented in plain Geneva font, at 18 points. The cue was presented for 83 ms before the presentation of the letter string, slightly beneath it and in three different locations: immediately to the left of the L1 letter string, centred, or immediately to the right of the R1 letter string. The rest of the procedure was identical to the procedure in Experiment 3. The task was to identify and report the cue (digit) and then report the pseudoword. Auclair and Siéroff (2002) found a similar cueing effect on letter string reading whether the cue was a digit that had to be reported or a single object (triangle) that participants could ignore. However, to ascertain that M.B. was actually processing the cue, we preferred to use a cue (digit) that had to be reported. Both M.B. and J.S. were tested. PURE LEFT HEMIALEXIA Table 3. Number of correctly identified letters in four-letter pseudowords in each cueing condition in Experiment 4 Participant M.B. J.S. LVF RVF Left cue Central cue Right cue Left cue Central cue Right cue 14 24 32 37 40 25 61 60 64 55 59 50 Note: Max. no. pseudowords ¼ 80. LVF ¼ left visual field; RVF ¼ right visual field. Results and discussion First, M.B.’s performance improved compared to Experiment 3, presumably because of intense training (this experiment was done one month later). Both participants’ performance was better in the RVF than in the LVF, and the hemifield asymmetry was even stronger for M.B. (29% in LVF, 77% in RVF) than for J.S. (43% in LVF, 68% in RVF), x2(1) ¼ 6.99; p , .01, as in Experiment 3. J.S. showed a differential cueing effect according to the hemifield of pseudoword presentation (Table 3). Cue location had no effect on RVF pseudowords, x2(2) ¼ 2.34; ns, but had a significant effect on LVF pseudowords, x2(2) ¼ 6.44; p , .05, with performance being worse with a right cue than with centre and left cues. This result is somewhat similar to those reported by Gatheron and Siéroff (1999), who obtained a stronger cueing effect with LVF pseudowords than with RVF pseudowords, using a different type of cue (bars underneath the location of the letter string and no centre cues). M.B.’s results in the RVF were similar to those obtained by J.S. x2(1) ¼ 0.78; ns; in other words, there was no difference between left, centre, and right cues (Table 3). However, in the LVF, the cueing effects were different from those obtained with J.S., x2(1) ¼ 9.61; p , .01. More specifically, a cue located at the beginning of the letter string provided no advantage. Thus, M.B.’s LVF performance did not improve with a left-sided cue located near the initial letter of the letter string.1 This result militates against the hypothesis that he may have an LVF reading deficit caused by the need to reorient attention toward the initial letter in the LVF. Thus, it remains most plausible that M.B. has a true deficit in identifying LVF letter strings— namely, left hemialexia. EXPERIMENT 5: READING FOVEALLY PRESENTED PSEUDOWORDS OF DIFFERENT LENGTHS Experiments 3 and 4 indicated that M.B. has left hemialexia, when presented with short four-letter strings. In Experiment 5, we evaluated the identification of foveally centred letter strings of different lengths. More specifically, we differentially evaluated performance in the left and right halves of centrally presented letter strings. M.B. should have more difficulties in identifying letters in the left half of centrally presented letter strings (LVF), than in the right half (RVF). In order to minimize the effect of guessing strategies, we chose to present only pseudoword letter strings. Method The stimuli were 80 pseudowords of different lengths (4 to 11 letters long, 10 pseudowords per length), presented in lowercase Geneva font, 18 points. All pseudowords were generated from groups of letters from real words and were pronounceable. The overall visual angle of a letter sequence was 4.4 degrees for the 11-letter pseudowords and 2 degrees for the 5-letter pseudowords. 1 Absence of a cueing effect has also been found in a patient, R.C.G., who has problems reading the beginnings of words (Arduino, Daini, & Silveri, 2005). However, R.C.G.’s deficit is different, because it occurs with words presented in the RVF as well as with words presented in the LVF (stimulus-centred deficit). COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) 251 SIÉROFF AND LAVIDOR The fixation item (plus sign) was presented in the middle of the screen for 500 ms. Pseudowords were presented in a centred position around the fixation item, which disappeared as they appeared. The pseudowords were presented for 150 ms for M.B. and 117 ms for J.S. Stimuli were immediately followed by a patterned mask, which remained present until the next trial. Participants’ task was to identify and read aloud the pseudoword, either by pronouncing it or by spelling some letters. Participants were encouraged to answer even if they were not sure of their response. Results and discussion Correct identification of each individual letter in the pseudoword was calculated (using the first scoring method from Experiment 3), and then the results were summed up for each hemifield, in order to obtain a hemifield score. Performance on the central letter in pseudowords with an odd number of letters was halved between the left and right hemifield scores. So the theoretical maximum score in each hemifield was: 2 for a four-letter string, 2.5 for a five-letter string, 3 for a six-letter string, and so on. In Figure 4, a monotonically increasing line (from left to right) would reflect a length-independent reading mode, while a length effect would be reflected in a straight Figure 4. Effect of length on accuracy in the left (LVF) and right (RVF) visual fields when reading centred pseudowords in patient M.B. and control J.S. (Experiment 5). The theoretical score corresponds to the maximum number of letters that can be identified per pseudoword length. 252 COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) horizontal line with the same number of correct letters whatever the pseudoword’s length. As in Experiments 3 and 4, M.B.’s reading was better in the spared area of the RVF (79% correct) than in the LVF (56% correct); J.S. experienced no such variation (82 and 83%, respectively), and the difference between M.B. and J.S. was significant (x2 ¼ 7.12; p , .01). In the RVF, the number of letters that M.B. could correctly identify improved with the pseudoword’s length, and a correlation was found between M.B.’s RVF score and the theoretical score (r ¼ .98; p , .01). J.S., the healthy control, also showed positive correlation between his RVF score and the theoretical score (r ¼ .89 for the RVF). M.B.’s performance correlated with J.S.’s performance (r ¼ .90). In the LVF, the number of letters M.B. correctly identified did not increase with pseudoword length. No correlation was found between M.B.’s LVF score and the theoretical score (r ¼ – .07). On the other hand, J.S.’s performance did improve with length, and his LVF score did correlate with the theoretical score (r ¼ .78). M.B.’s performance did not correlate with J.S.’s performance (r ¼ .35). M.B.’s LVF reading is clearly abnormal; his performance is poor compared to that of the control, J.S., and there is an abnormal length effect. His LVF deficit cannot be explained by a specific response strategy, because, as was the case with J.S., M.B.’s responses were almost all pronounced and not spelled. Most probably, M.B. has difficulties transferring letter information from the LVF (RH) to the language areas in the LH—a problem that can be explained by the interruption of splenial fibres. Although both M.B.’s and J.S.’s performance improved with pseudoword length in the RVF and were correlated with length, they were slightly different. In Figure 4, we can see a different asymptote in M.B.’s and J.S.’s RVF performance: J.S.’s performance reached a plateau with pseudowords approximately 8 letters long, while M.B.’s performance continued to improve up to 10 letters. One possibility is that M.B.’s limited processing of letters in the LVF left him with more resources PURE LEFT HEMIALEXIA for phonological processing and reporting letters presented in the RVF, compared to J.S. We calculated the overall correct performance for the entire letter strings. J.S.’s performance improved gradually up to a limit of 7 letters in the response (4 correct letters for 4-letter pseudowords, 5 for 5-letter pseudowords, 6 for 6-letter pseudowords, 6.5 for 7-letter pseudowords, 6.9 for 8-letter pseudowords, 7.4 for 9-letter pseudowords, 6.5 for 10-letter pseudowords, and 7.2 for 11-letter pseudowords). In M.B., partly because performance in the LVF was poor, this limit was never attained, or at least was not attained with 8letter pseudowords (3.5, 3.7, 4.7, 4.9, 5.4, 5.7, 6, and 6, respectively). Actually, in M.B., the plateau seemed to be at 6 letters. This discrepancy between J.S.’s and M.B.’s overall performance may be explained by M.B.’s having a slight short-term memory problem, or by the location of the rightmost letter of 11-letter pseudowords in the patient’s hemianopic field. Finally, Table 4 presents J.S.’s and M.B.’s performance on individual letters for each letter string length. Only for M.B. is there a clear breakdown between letters located immediately to the right and to the left of fixation. Actually, the number of correctly identified letters within 1 degree of visual angle (excluding letters in the central location) was 128 for the LVF and 129 for the RVF in J.S. (n ¼ 160), and 74 and 124, respectively, in M.B.; the difference between participants is significant, x2(1) ¼ 7.0; p , .01. Specifically, for long letter strings (over eight letters), letter identification in LVF was poor between 0 degrees (C) and 1 degree (L1) from fixation in M.B. Thus, M.B. processed foveally presented letters in the LVF poorly, a result that is clearly in favour of the SFT. GENERAL DISCUSSION M.B. was suffering from a left mesial occipital lesion. We can summarize his deficits as follows: a global perceptual impairment, a right hemianopia, and a left hemialexia. The global perceptual impairment is weak and results in lower global performance in identifying objects or reading letter strings. Experiment 2 shows that M.B.’s performance is poorer than that of the control, J.S., when identifying drawings of objects with the same presentation duration for both participants. When identifying letter strings (Experiment 3), M.B. required a longer presentation duration to perform at a level comparable to J.S. We did not further explore this deficit, which was independent of visual hemifields. Table 4. Number of correct individual letters for each position in foveally centred pseudowords, in patient M.B. and control J.S. in Experiment 5) COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) 253 SIÉROFF AND LAVIDOR Most importantly, M.B. also has a right hemianopia and has complained about it. However, there is a spared area in the RVF, within approximately 2 degrees of visual angle from the midline, as shown by Experiments 1 and 3. M.B. presents a left hemialexia without left hemiagnosia. Although no asymmetry between the visual hemifields (LVF and the spared area of the RVF) was found when M.B. was presented with line drawings in an identification task (Experiment 2), the RVF (spared area) was found to be superior to the LVF at processing letter strings (words or pseudowords); this superiority was significantly stronger than that shown by J.S. When M.B. was presented with a centred letter string, an asymmetry also occurred between the left and right halves of the string, in favour of the right one. Not only was reading performance low in M.B.’s LVF, but there was also a strong and abnormal discrepancy in reading performance between the initial and final letters of LVF letter strings (words or pseudowords), although no such discrepancy was found in the spared area of the RVF. In addition, M.B. showed an abnormally strong length effect for the left half of centred letter strings (Experiment 5). One possibility is that similar reading strategies were applied in both hemifields, with the difference in performance being explained by some delay caused by the reorienting of attention toward the initial letter of LVF letter strings, although orienting could occur smoothly toward the initial letter of RVF letter strings. According to the reorienting delay hypothesis, a cue located immediately to the left of LVF letter strings should improve performance. However, Experiment 4 showed that digit cueing the beginning of LVF letter strings did not improve M.B.’s performance, although it did for J.S., the control. 2 Another possibility is that the difference in locations of initial letters in LVF and RVF fourletter strings affected M.B.’s performance. In Experiment 3, the initial letters of LVF letter strings are represented parafoveally (even for location L1), although the initial letters of RVF letter strings always appeared in the fovea. Since the initial letters of words are the most informative, stimuli presented in the RVF would be favoured. However, if the RVF advantage over the LVF resulted from a better acuity for the initial letters, which carry more lexical information (Pynte, 1996), we should not find, in M.B., a RVF advantage for pseudowords, where lexical information is not relevant. Moreover, Experiment 4 shows that cueing the initial letter of LVF letter strings does not improve M.B.’s performance. In addition, the better acuity for the initial letters would predict an LVF advantage for Hebrew (which is written right to left); however, most Hebrew studies have reported an RVF advantage (Faust & Lavidor, 2003; Lavidor, Ellis, & Pansky, 2002). Thus, M.B. may present a true difference in processing LVF and RVF letter strings. M.B.’s deficit is probably affecting the transfer of letters from the RH to the LH. A possibility is that this deficit results in an abnormally slowed sequential reading in the LVF, including the left hemifovea. Another possibility is that M.B. has difficulties in identifying individual letters in the LVF. This difficulty may not be apparent when he is presented with single letters (as in Experiment 1), but would impair his performance when presented with letter strings (like the case of W.L., who presented a letter-by-letter alexia; Reuter-Lorenz & Brunn, 1990). A final possibility is that M.B. is affected by crowding—that is, that he finds it difficult to identify a letter in the presence of nearby letters (Pelli, Palomares, & Majaj, 2004), more in the LVF than in the spared area of the RVF.2 The results of an experiment conducted on M.B. could help us to differentiate between the hypothesis that the RH has problems identifying and/or transferring letters and the crowding hypothesis. We presented normal four-letter strings and four-letter strings with more space between the letters (e.g., w o r d) to the LVF and RVF. The crowding hypothesis would predict better performance with the spaced letters. However, we found that spacing the letters did not improve M.B.’s performance. Moreover, the LVF recognition of the initial letter of spaced-letter strings was superior to the recognition of letters in other positions, just as it was with non-spaced letter strings. 254 COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) PURE LEFT HEMIALEXIA Normal reading of letters displayed in the LVF is based on projections from the right-sided visual cortex such as V4 toward the VWFA, located in the left fusiform gyrus (Beauregard et al., 1997; Cohen et al., 2000; Puce, Allison, Asgari, Gore, & McCarthy, 1996). This pathway goes through the splenium of the corpus callosum, possibly its most ventral segment (Suzuki et al., 1998), and over the lateral ventricles (Molko et al., 2002). Several patients have been described with a left hemialexia, either accompanied by a left visual anomia (see Gazzaniga, Bogen, & Sperry, 1965) or pure, without deficits in naming objects in the LVF (Abe et al., 1986; Binder et al., 1992; Cohen & Dehaene, 1996; Damasio et al., 1980; Degos et al., 1987; Habib et al., 1990; Lausberg et al., 1999; Levine & Calvanio, 1980; Maspes, 1948; Sidtis et al., 1981; Sugishita et al., 1978, 1986; Suzuki et al., 1998; Trescher & Ford, 1937; Wechsler, 1972). Most such patients had a lesion of the splenium of the corpus callosum. More specifically, patients presenting pure left hemialexia seem to suffer from lesions in the inferior part of the splenium. On the other hand, a lesion of the anterior four-fifths of the corpus callosum with the splenium left intact does not produce left hemialexia (Geschwind & Kaplan, 1962). Consequently, the explanation of left hemialexia is a disruption of the callosal transfer of information from the RH to the language areas in the LH. Callosal fibres remain grouped at a certain distance within the occipital lobes, and an occipital lesion may result in a disconnection. In the case of M.B., the disruption of the transfer of visual information may have been located at some distance from the midline of the corpus callosum, in the left occipital lobe. Similarly, disconnection can sometimes occur between the left primary visual cortex and the VWFA, as shown in cases of right hemialexia after a left occipital lesion (Castro-Caldas & Salgado, 1984). In M.B., the VWFA, probably located in the left fusiform gyrus (Cohen et al., 2000), was certainly spared by the lesion, because reading in the spared area of the RVF was good: Even though his performance was poorer than the control’s, M.B.’s letter processing in the RVF was length insensitive. However, the transfer of letter information from the RH lower order retinotopic cortices to the VWFA in the LH was disrupted. Still, reading in the LVF was not impossible. One possibility is that a partial lesion of the splenial fibres slowed the transfer to the left VWFA. A second possibility is that this transfer was completely disrupted, and letter reading in the LVF was accomplished via access to the R-VWFA, a right-hemispheric cortex symmetrical to the VWFA (Cohen et al., 2004), and then through access to higher level cortices through an effortful verbal working-memory process. Because M.B.’s LVF superiority for words over pseudowords was equivalent to the control participant’s, we suggest that the transfer to the VWFA was only partially disrupted, but other patients with a complete callosotomy have been found to have a superiority effect for words over pseudowords in the LVF (Reuter-Lorenz & Baynes, 1992). A better argument in favour of a partial splenial lesion is that M.B.’s LVF reading performance was much better than that of most other hemialexic patients in the literature. It remains difficult to reach a firm conclusion, though, because the efficiency of the right VWFA may vary between patients. A spared area around the fovea is found in two thirds of patients suffering from hemianopia (Leff, 2004). Several arguments have been put forward to account for macular sparing, suggesting that the phenomenon reflects: (a) incomplete damage to the visual pathways (Horton & Hoyt, 1991); (b) dual blood supply to the occipital pole (e.g., Smith & Richardson, 1966); (c) unusual fixation strategies adopted by hemianopic patients (Bischoff, Lang, & Huber, 1995); or (d) bilateral foveal projection. A challenge to the bilateral projection theory, however, is the fact that not all hemianopic patients show foveal sparing: About 36% of them show no sparing at all—that is, they present foveal splitting (Celesia, Meredith, & Pluff, 1983). The bilateral foveal representation theory cannot explain foveal splitting (Leff, 2004), COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) 255 SIÉROFF AND LAVIDOR but foveal sparing can be explained according to the SFT. It seems likely that the reason for the foveal sparing, according to this theory, is that the cortical representation of the fovea is so extensive that a single, random lesion is unlikely to destroy the entire representation (Leff, 2004). M.B.’s lesion may well leave intact the posterior part of the calcarine cortex, explaining the spared area he presented. Strong arguments favouring an explanation of macular sparing in our patient by an incomplete lesion of the primary visual cortex (compatible with the SFT, rather than the BPT) emerge from M.B.’s different reading performance across his two visual fields (see above). Moreover, his global performance for centred four-letter strings in Experiment 3, subtending an overall visual angle of 2 degrees, was similar to the performance obtained with stimuli projected immediately to the left of the fixation item (centred on an eccentricity of 1 degree) and clearly inferior to the equivalent performance in the RVF, a result that can only be explained by the SFT. The split fovea theory also predicts that hemispheric differences will be found in the processing of the left and right segments of centrally presented targets (Lavidor & Walsh, 2003). Experiment 5 showed that M.B. was affected by the length of pseudoword halves in the LVF, but not in the RVF. Although the number of letters that M.B. reported correlated with the number of letters actually present in the RVF half, as was the case with the control J.S., the number of letters reported from the LVF did not vary with length, unlike what J.S. did. This result indicates that even for centred letter strings, M.B.’s reading strongly differed between left and right hemifovea. Moreover, for foveally centred strings of more than eight letters, M.B.’s performance was poor for letters located immediately to the left of fixation, even within 1 degree of visual angle. In conclusion, M.B. presents a pure left hemialexia with a spared macular area of vision in his RVF. The difference in reading efficiency between the LVF and the spared area of the RVF cannot be accounted for by a theory that explains macular sparing hemianopia by a bilateral retinal projection 256 COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) on visual cortices. These results, however, are in agreement with the split fovea theory. Manuscript received 27 October 2005 Revised manuscript received 23 November 2006 Revised manuscript accepted 8 December 2006 First published online 5 March 2007 REFERENCES Abe, T., Nakamura, N., Sugishita, M., Kato, Y., & Iwata, M. (1986). Partial disconnection syndrome following penetrating stab wound of the brain. European Neurology, 25, 233– 239. Arduino, L. S., Daini, R., & Silveri, M. C. (2005). A stimulus-centered reading disorder for words and numbers: Is it neglect dyslexia? Neurocase, 11, 405–415. Auclair, L., & Siéroff, E. (2002). Attentional cueing in reading words and pseudowords of different length. Quarterly Journal of Experimental Psychology, 55A, 445– 463. Beauregard, M., Chertkow, H., Bub, D., Murtha, S., Dixon, R., & Evans, A. (1997). The neural substrate for concrete, abstract, and emotional word lexica: A positron emission tomography study. Journal of Cognitive Neuroscience, 9, 441– 461. Biederman, I., & Cooper, E. E. (1991). Object recognition and laterality effects: Null effects. Neuropsychologia, 29, 685– 694. Binder, J. R., Lazar, R. M., Tatemichi, T. K., Mohr, J. P., Desmond, D. W., & Ciercierski, K. A. (1992). Left hemiparalexia. Neurology, 42, 562– 569. Bischoff, P., Lang, J., & Huber, A. (1995). Macular sparing as a perimetric artifact. American Journal of Ophthalmology, 119, 72 – 79. Bouma, H. (1973). Visual interference in the parafoveal recognition of initial and final letters of words. Vision Research, 13, 767– 782. Brysbaert, M. (1994). Interhemispheric transfer and the processing of foveally presented stimuli. Behavioural Brain Research, 64, 151– 161. Brysbaert, M. (2004). The importance of interhemispheric transfer for foveal vision: A factor that has been overlooked in theories of visual word recognition and object perception. Brain and Language, 88, 259– 267. Brysbaert, M., Vitu, F., & Schroyens, W. (1996). The right visual field advantage and the optimal viewing position effect: On the relation between foveal and PURE LEFT HEMIALEXIA parafoveal word recognition. Neuropsychology, 10, 385– 395. Bunt, A. H., & Minckler, D. S. (1977). Foveal sparing: New anatomical evidence for bilateral representation of the central retina. Archives of Ophthalmology, 95, 1445– 1447. Carr, T. H., Posner, M. I., Pollatsek, A., & Snyder, C. R. R. (1979). Orthography and familiarity effects in word processing. Journal of Experimental Psychology: General, 108, 389– 414. Castro-Caldas, A., & Salgado, V. (1984). Right hemifield alexia without hemianopia. Archives of Neurology, 41, 84 – 87. Celesia, G. G., Meredith, J. T., & Pluff, K. (1983). Perimetry, visual evoked potentials and visual evoked spectrum array in homonymous hemianopia. Journal of Electroencephalography and Clinical Neurophysiology, 56, 16 –30. Chiarello, C. (1988). Lateralization of lexical processes in the normal brain: A review of visual half-field research. In H. A. Whitaker (Ed.), Contemporary reviews in neuropsychology. New York: Springer-Verlag. Choplin, N., & Edwards, R. (1998). Visual fields. Thorofare, NJ: SLACK Incorporated. Cohen, L., & Dehaene, S. (1996). Cerebral network for number processing: Evidence from a case of posterior callosal lesion. Neurocase, 2, 155– 174. Cohen, L., Dehaene, S., Naccache, L., Lehéricy, S., Dehaene-Lambertz, G., Henaff, M. A., et al. (2000). The visual word form area: Spatial and temporal characterization of an initial stage of reading in normal subjects and posterior split-brain patients. Brain, 123, 291– 307. Cohen, L., Henry, C., Dehaene, S., Martinaud, O., Lehéricy, S., Lemer, C., et al. (2004). The pathophysiology of letter-by-letter reading. Neuropsychologia, 42, 1768– 1780. Content, A., Mousty, P., & Radeau, M. (1990). Brulex. Une base de données lexicales informatisées pour le français écrit et parlé [A computerized lexical database for the French language]. L’Année Psychologique, 90, 551– 566. Damasio, A. R., Chui, H. C., Corbett, J., & Kassel, N. (1980). Posterior callosal section in a non-epileptic patient. Journal of Neurology, Neurosurgery, and Psychiatry, 43, 351– 356. Degos, J. D., Gray, F., Louarn, F., Ansquer, J. C., Poirier, J., & Barbizet, J. (1987). Posterior callosal infarction: Clinicopathological correlations. Brain, 110, 1155– 1171. Ellis, A. W. (2004). Length, formats, neighbours, hemispheres, and the processing of words presented laterally or at fixation. Brain and Language, 88, 355– 366. Ellis, A. W., Young, A. W., & Anderson, C. (1988). Modes of word recognition in the left and right cerebral hemispheres. Brain and Language, 35, 254– 273. Faust, M., & Lavidor, M. (2003). Semantically convergent and semantically divergent priming in the cerebral hemispheres: Lexical decision and semantic judgment. Brain Research: Cognitive Brain Research, 17, 585– 597. Gatheron, D., & Siéroff, E. (1999). Right hemifield superiority in reading and attentional factors. Brain and Cognition, 40, 122– 125. Gazzaniga, M. S., Bogen, J. E., & Sperry, R. W. (1965). Observations on visual perception after disconnection of the cerebral hemisphere in man. Brain, 88, 221– 236. Geschwind, N., & Kaplan, E. (1962). A human cerebral disconnection syndrome: A preliminary report. Neurology, 12, 675– 685. Gibson, E. J., Pick, A. D., Osser, H., & Hammond, M. (1962). The role of grapheme– phoneme correspondence in the perception of words. American Journal of Psychology, 75, 554– 570. Gill, K. M., & McKeever, W. F. (1974). Word length and exposure time effects on the recognition of bilaterally presented words. Bulletin of the Psychonomic Society, 4, 173– 175. Habib, M., Ceccaldi, M., & Poncet, M. (1990). Syndrome de déconnexion calleuse par infarctus jonctionnel hémisphérique gauche [Cerebral arterial ischaemic complications in young adults: Aetiology and prognosis]. Revue Neurologique, 146, 19 – 24. Hardyck, C., Chiarello, C., Dronkers, N., & Simpson, G. (1985). Orienting of attention within visual fields: How efficient is the interhemispheric transfer? Journal of Experimental Psychology: Human Perception and Performance, 11, 650– 666. Horton, J. C., & Hoyt, W. F. (1991). The representation of the visual field in human striate cortex: A revision of the classic Holmes map. Archives of Ophthalmology, 109, 816– 824. Iacoboni, M., & Zaidel, E. (1996). Hemispheric independence in word recognition: Evidence from unilateral and bilateral presentations. Brain and Language, 53, 121– 140. Kanwisher, N., Woods, R. P., Iacoboni, M., & Mazziotta, J. C. (1997). A locus for human COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) 257 SIÉROFF AND LAVIDOR extrastriate cortex for visual shape analysis. Journal of Cognitive Neuroscience, 9, 133– 142. Lausberg, H., Göttert, R., Münßinger, U., Boegner, F., & Marx, P. (1999). Callosal disconnection syndrome in a left-handed patient due to infarction of the total length of the corpus callosum. Neuropsychologia, 37, 253– 265. Lavidor, M., & Ellis, A. W. (2002a). Orthographic neighborhood effects in the right but not in the left cerebral hemisphere. Brain and Language, 80, 63–76. Lavidor, M., & Ellis, A. W. (2002b). Word length and orthographic neighborhood size effects in the left and right cerebral hemispheres. Brain and Language, 80, 45 – 62. Lavidor, M., Ellis, A. W., & Pansky, A. (2002) Case alternation and length effects in lateralized word recognition: Studies of English and Hebrew. Brain and Cognition, 50, 257– 271. Lavidor, M., Ellis, A. W., Shillcock, R., & Bland, T. (2001). Evaluating a split processing model of visual word recognition: Effects of word length. Cognitive Brain Research, 12, 265– 272. Lavidor, M., & Walsh, V. (2003). Magnetic stimulation examination of orthographic neighbourhood effects in visual word recognition. Journal of Cognitive Neuroscience, 15, 354– 363. Lavidor, M., & Walsh, V. (2004). The nature of foveal representation. Nature Reviews Neuroscience, 5, 729– 735. Leff, A. P. (2004). A historical review of the representation of the visual field in primary visual cortex with special reference to the neural mechanisms underlying macular sparing. Brain and Language, 88, 268–278. Leventhal, A. G., Thompson, K. G., & Liu, D. (1993). Retinal ganglion cells within the foveola of New World (Saimiri sciureus) and Old World (Macaca fascicularis) monkeys. Journal of Comparative Neurology, 338, 242– 254. Levine, D. N., & Calvanio, R. (1980). Visual discrimination after lesion of the posterior corpus callosum. Neurology, 30, 21 – 30. Lindell, A. K., & Nicholls, E. R. (2003). Attentional deployment in visual half-field tasks: The effect of cue position on word naming latency. Brain and Cognition 53, 273– 277. Maspes, P. E. (1948). Le syndrome expérimental chez l’homme de la section du splénium du corps caleux: Alexie visuelle pure hémianopsique [The experimental syndrome of the corpus callosum splenium section in man: Pure hemianopic visual alexia]. Revue Neurologique, 80, 100– 113. 258 COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) McCandliss, B. D., Cohen, L., & Dehaene, S. (2003). The visual word form area: Expertise for reading in the fusiform gyrus. Trends in Cognitive Sciences, 7, 293–299. Mishkin, M., & Forgays, D. G. (1952). Word recognition as a function of retinal locus. Journal of Experimental Psychology, 43, 43 – 48. Molko, N., Cohen, L., Mangin, J. F., Chochon, F., Lehéricy, S., Le Bihan, D., et al. (2002). Visualizing the neural bases of a disconnection syndrome with diffusion tensor imaging. Journal of Cognitive Neuroscience, 14, 629– 636. Mondor, T. A., & Bryden, M. P. (1992). On the relation between visual spatial attention and visual field asymmetries. The Quarterly Journal of Experimental Psychology, 44A, 529– 555. Nicholls, M. E. R., & Wood, A. G. (1998). The contribution of attention to the right visual field advantage for word recognition. Brain and Cognition, 38, 339–357. Pelli, D. G., Palomares, M., & Majaj, N. B. (2004). Crowding is unlike ordinary masking: Distinguishing feature integration from detection. Journal of Vision, 4, 1136–1169. Puce, A., Allison, T., Asgari, M., Gore, J. C., & McCarthy, G. (1996). Differential sensitivity of human visual cortex to faces, letterstrings, and textures: A functional magnetic resonance imaging study. Journal of Neuroscience, 16, 5205– 5215. Pynte, J. (1996). Lexical control of within-word eye movements. Journal of Experimental Psychology: Human Perception and Performance, 22, 958– 969. Reuter-Lorenz, P. A., & Baynes, K. (1992). Modes of lexical access in the callosotomized brain. Journal of Cognitive Neuroscience, 4, 155– 164. Reuter-Lorenz, P. A., & Brunn, J. L. (1990). A prelexical basis for letter-by-letter reading: A case study. Cognitive Neuropsychology, 7, 1 – 20. Riddoch, M. J., Humphreys, G. W., Cleton, P., & Ferry, P. (1990). Interaction of attentional and lexical processes in neglect dyslexia. Cognitive Neuropsychology, 5, 479– 517. Shillcock, R., Ellison, T. M., & Monaghan, P. (2000). Eye-fixation behavior, lexical storage and visual word recognition in a split processing model. Psychological Review, 107, 824 –851. Sidtis, J. J., Volpe, B. T., Holtzman, J. D., Wilson, D. H., & Gazzaniga, M. S. (1981). Cognitive interaction after staged callosal section: Evidence for transfer of semantic activation. Science, 212, 344–346. Siéroff, E., & Posner, M. (1988). Cueing spatial attention during processing of words and letter PURE LEFT HEMIALEXIA strings in normals. Cognitive Neuropsychology, 5, 451– 472. Smith, C. G., & Richardson, W. F. (1966). The course and distribution of the arteries supplying the visual (striate) cortex. American Journal of Ophthalmology, 61, 1391– 1396. Snodgrass, J. G., & Vanderwart, M. A. (1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174– 215. Stone, J. (1966). The naso-temporal division of the cat’s retina. Journal of Comparative Neurology, 136, 585– 600. Stone, J., Leicester, J., & Sherman, S. M. (1973). The naso-temporal division of the monkey’s retina. Journal of Comparative Neurology, 150, 333– 348. Sugishita, M., Iwata, M., Toyokura, Y., Yoshioka, M., & Yamada, R. (1978). Reading of ideograms and phonograms in Japanese patients after partial commissurotomy. Neuropsychologia, 16, 417– 426. Sugishita, M., Yoshioka, M., & Kawamura, M. (1986). Recovery from hemialexia. Brain and Language, 29, 106– 118. Suzuki, K., Yamadori, A., Endo, K., Fujii, T., Ezura, M., & Takahashi, A. (1998). Dissociation of letter and picture naming resulting from callosal disconnection. Neurology, 51, 1390– 1394. Trescher, J. H., & Ford, F. R. (1937). Colloid cyst of the third ventricle. Report of a case; operative removal with section of the posterior half of the corpus callosum. Archives of Neurology and Psychiatry, 37, 959– 973. Wechsler, A. F. (1972). Transient left hemialexia: A clinical and angiographic study. Neurology, 22, 628– 633. Weekes, B. S. (1997). Differential effects of number of letters on word and nonword naming latency. Quarterly Journal of Experimental Psychology: Human Experimental Psychology, 50A, 439– 456. Whitney, C. (2001). How the brain encodes the order of letters in a printed word: The SERIOL model and selective literature review. Psychonomic Bulletin Review, 8, 221– 243. Young, A. W, & Ellis, A. W. (1985). Different methods of lexical access for words presented in the left and right visual hemifields. Brain and Language, 24, 326– 358. COGNITIVE NEUROPSYCHOLOGY, 2007, 24 (3) 259