A mesial-to-lateral dissociation for orthographic processing in the visual cortex

Abstract

Efficient reading requires a fast conversion of the written word to both phonological and semantic codes. We tested the hypothesis that, within the left occipitotemporal cortical regions involved in visual word recognition, distinct subregions harbor slightly different orthographic codes adapted to those 2 functions. While the lexico-semantic pathway may operate on letter or open-bigram information, the phonological pathway requires the identification of multiletter graphemes such as "ch" or "ou" in order to map them onto phonemes. To evaluate the existence of a specific stage of graphemic encoding, 20 adults performed lexical decision and naming tasks on words and pseudowords during functional MRI. Graphemic encoding was facilitated or disrupted by coloring and spacing the letters either congruently with multiletter graphemes (ch-ai-r) or incongruently with them (c-ha-ir). This manipulation affected behavior, primarily during the naming of pseudowords, and modulated brain activity in the left midfusiform sulcus, at a site medial to the classical visual word form area (VWFA). This putative grapheme-related area (GRA) differed from the VWFA in being preferentially connected functionally to dorsal parietal areas involved in letter-by-letter reading, while the VWFA showed effects of lexicality and spelling-to-sound regularity. Our results suggest a partial dissociation within left occipitotemporal cortex: the midfusiform GRA would encode orthographic information at a sublexical graphemic level, while the lateral occipitotemporal VWFA would contribute primarily to direct lexico-semantic access.

Keywords: complex graphemes; grapheme processing; reading; visual word form area.

Fig. 1.

(A) Example stimuli of the main experiment. Words and pseudowords were split into colored fragments whose coloring either matched the grapheme boundaries (congruent chunking) or was incongruent with them (incongruent chunking). (B) Effects of graphemic congruency on error rates (Left) and RTs (Right). Histograms show the mean (± standard error of the mean, SEM) of the beta estimates of the graphemic congruency effet, after regressing out stimulus-level variables such as word length. *P < 0.05; **P < 0.01. Incongruent trials induced more errors and slower responses only during reading aloud, and especially when reading aloud pseudowords (i.e., the conditions that relied most on phonological processing).

Fig. 2.

(A) Main fMRI effect of task. (B) Main fMRI effect of lexicality across tasks. Black dots indicate the group peaks of the GRA and VWFA for reference, see Two Functionally Distinct Orthographic Areas in Left Occipito-Temporal Cortex. (C) Main fMRI effect of graphemic congruency averaging both tasks. Incongruent stimuli induced stronger activations in parieto-frontal regions involved in attentional control. (D) Activation at the peak of the left IFG cluster. The histogram represents the mean BOLD signal per condition (± SEM after subtraction of each subject’s overall mean). LD, lexical decision; RA, reading aloud; W, word; PW, pseudoword.

Fig. 3.

(A) Comparison of the graphemic congruency effect between conditions involving the most vs. the least phonological demands (i.e., reading aloud pseudowords vs. lexical decision on words). (B) Activation profile at the peak of this midfusiform cluster. Histograms represent the mean BOLD signal across participants (± SEM after subtraction of each subject’s overall mean). LD, lexical decision; RA, reading aloud; W, word; PW, pseudoword.

Fig. 4.

(A) Positive correlation between reading activations vs. rest and individual behavioral sensitivity to congruency, in the VOT cortex (blue), overlaid on overall activations vs. rest (hot), showing a reproducible cluster in the left midfusiform sulcus (arrows). (B) Whole-brain analyses showed an additional area of positive correlation between behavior and average activation vs. rest in the left IFG (Center). Scatterplots illustrate the correlations in the left VOT and IFG peak voxels. LD, lexical decision; RA, reading aloud; W, word; PW, pseudoword.

Fig. 5.

Localization of the proposed GRA (blue), as identified based on its modulation by graphemic congruency (Fig. 3), compared to the VWFA (red). The VWFA was identified based on the localizer experiment by subtracting word reading minus checkerboards. The GRA was located mesially to the VWFA identified both at the group level (surface view, Left) and at the individual level (Right: overlap of individual ROIs across subjects, thresholded at 20% of all participants). Both the GRA and the VWFA were more activated by words than fixation (gold on the surface view).

Fig. 6.

Comparison of the responses of the GRA and VWFA to variables modulating phonological and lexico-semantic mappings. (A) Effects of congruency, lexicality, and task. The GRA showed an interaction of congruency with phonological demands, as defined, while it was insensitive to lexicality. The VWFA showed the opposite pattern. Histograms show the mean BOLD signal ± SEM after subtraction of each subject’s overall mean. (B) Classification accuracy of words vs. pseudowords (50% = chance level; mean ± SEM after subtraction of each subject’s overall mean). Classification was above chance only in the lateral region, and significantly more accurate laterally than mesially. (C) Standard beta estimates of the length effect (mean ± SEM). The effect was present in both regions, with no significant difference between regions. (D) Effects of word regularity and frequency in the localizer experiment (mean ± SEM after subtracting each subject’s overall mean). Only the lateral region showed an effect of regularity, while both regions were sensitive to frequency. ***P < 0.001; **P < 0.01; ^∼P < 0.07.

Fig. 7.

Functional connectivity of the GRA and VWFA. (A) Across the main experiment, the VWFA was overall more connected to most of the brain, while the GRA was more connected only to its surroudings, the precuneus and the ACC. (B) Modulation of the connectivity of the mesial region by phonological demands (RA PW – LD W). Phonological demands notably increased the connectivity of the GRA to the VWFA and left IPS. (C) Modulation of the connectivity of the lateral region by phonological demands (RA PW – LD W). (D) Differences between the GRA and VWFA in the modulation of their functional connectivity by phonological demands. The GRA was more connected to the left IPS in conditions with high phonological demands compared to the VWFA. (E) Meta-analytic maps of functional coactivations of the 2 regions. Beyond the substantial overlap of the 2 maps (purple), the GRA tends to coactivate more with dorsal frontoparietal and supramarginal regions (blue), while the VWFA has preferential coactivations with the IFG, lateral and anterior temporal lobe (hot colors). LD, lexical decision; RA, reading aloud; W, word; PW, pseudoword.