Orthographies vary in the degree of transparency of spelling-sound correspondence. These range from shallow orthographies with transparent grapheme-phoneme relations, to deep orthographies, in which these relations are opaque. Only a few studies have examined whether orthographic depth is reflected in brain activity. In these studies a between-language design was applied, making it difficult to isolate the aspect of orthographic depth. In the present work this question was examined using a within-subject-and-language investigation. The participants were speakers of Hebrew, as they are skilled in reading two forms of script transcribing the same oral language. One form is the shallow pointed script (with diacritics), and the other is the deep unpointed script (without diacritics). Event-related potentials (ERPs) were recorded while skilled readers carried out a lexical decision task in the two forms of script. A visual non-orthographic task controlled for the visual difference between the scripts (resulting from the addition of diacritics to the pointed script only). At an early visual-perceptual stage of processing (∼165 ms after target onset), the pointed script evoked larger amplitudes with longer latencies than the unpointed script at occipital-temporal sites. However, these effects were not restricted to orthographic processing, and may therefore have reflected, at least in part, the visual load imposed by the diacritics. Nevertheless, the results implied that distinct orthographic processing may have also contributed to these effects. At later stages (∼340 ms after target onset) the unpointed script elicited larger amplitudes than the pointed one with earlier latencies. As this latency has been linked to orthographic-linguistic processing and to the classification of stimuli, it is suggested that these differences are associated with distinct lexical processing of a shallow and a deep orthography.

Phonological awareness has been found to be strongly related to spelling. Findings on the relations between rapid‐naming and spelling are less consistent and have been suggested to be shared with speed of processing. This study set out to examine these relations in spelling and reading of Hebrew. Children attending the regular educational system were followed longitudinally (N = 70): phonological awareness, rapid‐naming and speed of processing were tested in kindergarten and in grade 1, and spelling and reading were tested in grade 2. Kindergarten and grade 1 rapid‐naming predicted spelling and word reading, and grade 1 phonological awareness predicted spelling, word reading and decoding. Speed of processing was an insignificant predictor. The findings extend the role of phonological awareness in spelling to an orthography with partial phonological representations and concurrently suggest weak relations. The results further suggest a link between rapid‐naming and orthographic knowledge, which may not be explained by shared variance with speed of processing.

No abstract