Neural correlates of mapping from phonology to orthography in children performing an auditory spelling task

Abstract

Age-related differences (9- to 15-year-olds) in the neural correlates of mapping from phonology to orthography were examined with functional magnetic resonance imaging (fMRI). Participants were asked to determine if two spoken words had the same spelling for the rime (corresponding letters after the first consonant or consonant cluster). Some of the word pairs had conflicting orthography and phonology (e.g. jazz-has, pint-mint) whereas other pairs had non-conflicting information (e.g. press-list, gate-hate) (see Table 1). There were age-related increases in activation for lexical processing (across conflicting and non-conflicting conditions) in left inferior parietal lobule, suggesting that older children have a more elaborated system for mapping between phonology and orthography that includes connections at different grain sizes (e.g. phonemes, onset-rimes, syllables). In addition, we found that the conflicting conditions had lower accuracy, slower reaction time and greater activation in left inferior frontal gyrus as compared to non-conflicting conditions. Higher accuracy was also correlated with greater activation in left inferior frontal gyrus for the most difficult conflicting condition (e.g. jazz-has). The finding of both a conflict effect and a correlation with accuracy in left inferior frontal gyrus suggests that this region may be involved in resolving the conflict between orthographic and phonological representations.

Figure 1

Dual-route versus connectionist models of reading aloud. Dual-route models assume that there is a developmental shift from an indirect grapheme-phoneme rule route (square dotted line) to a direct route from orthography to phonology (circle dotted line). Connectionist models propose that there is a developmental increase in the reliance on an indirect route (hidden units) from orthography to phonology.

Figure 2

Activation for the lexical versus null conditions (red), for the perceptual versus null conditions (green) and their overlap (blue). Major clusters for the lexical versus null conditions included bilateral inferior frontal gyri, bilateral superior/middle temporal gyri, left fusiform gyrus, left inferior parietal lobule, bilateral medial frontal gyri, and bilateral cuneus.

Figure 3

(A) Increasing age correlated with greater activation for the lexical versus null conditions in left inferior/superior parietal lobule. (B) Scatter plot of the age (in months) correlation for the lexical minus null conditions in the most active voxel in left inferior/superior parietal lobule (IPL & SPL). (C) For the perceptual versus null conditions, greater activation correlated with increasing accuracy (red) in left inferior parietal lobule and increasing age (green) in left superior parietal lobule.

Figure 4

(A) Increasing accuracy (red) correlated with greater activation for the O−P+ (the most difficult conflicting condition) versus null conditions in left inferior frontal gyrus and in left inferior parietal lobule. Accuracy–activation correlations for the O−P+ versus null conditions in left inferior frontal gyrus overlap with the conflicting versus non-conflicting conditions (blue). Other regions more active for the conflicting versus non-conflicting conditions (green) included left inferior parietal lobule and left medial frontal gyrus. (B) Scatter plot of the accuracy–activation correlation for the O−P+ versus null conditions in the most active voxel in left inferior frontal gyrus (IFG).