RAN and orthographic processing: What can syllable frequency tell us about this relationship?

Many explanations accounting for rapid automatized naming's (RAN) relationship with reading have been proposed. One of the most debated perspectives argues that RAN measures orthographic processing, defined as the ability to process groups of letters or entire words as single units. Given that reading familiar spelling patterns will rely on orthographic processing more than reading unfamiliar spelling patterns, manipulating orthographic syllable frequency can be a useful tool to examine RAN's relationship with orthographic familiarity. To meet this aim, RAN's concurrent and longitudinal contribution to reading speed of nonwords composed of high and low syllable frequency, as well as real words, was assessed in a sample of 142 Spanish children. RAN, phonological skills, and visual skills were measured in kindergarten and Grade 5, whereas reading speed was measured in Grade 5 only. Both longitudinal and concurrent path analyses revealed that RAN made a comparable contribution to the reading of both types of nonwords as well as to real-word reading. This suggests that the reading-related cognitive ability measured by RAN operates at a grapho-phonemic, grapho-syllabic, and whole-word level. The current results do not support the view of RAN as a measure of orthographic processing.