Cognitive antecedents of early reading ability: A test of the modularity hypothesis

This study tested the hypothesis that the cognitive antecedents of word recognition are uniquely domain-specific and unrelated to higher-order domain-general cognitive abilities. This hypothesis was evaluated in a longitudinal study of 349 Hebrew-speaking children (mean age: 6.0 years) who were tested on a battery of domain-specific (phonological awareness, phonological memory, visual-orthographic processing, and early literacy) and domain-general tasks (general intelligence, higher-order reasoning, and language) at the end of kindergarten. Word recognition and reading comprehension were assessed at the end of Grade 1. Whereas the kindergarten domain-specific measures accounted for significant and substantial variance in word recognition (33%), the domain-general measures explained only 5% of the variance. Furthermore, the contribution of domain-specific variables to word recognition remained unaltered even after controlling for all domain-general and higher-order language tasks. Reading comprehension, in contrast, was predicted by both print-specific skills (51%,) and domain-general abilities (44%). These findings strongly support the notion of word recognition modularity in a well-encapsulated orthography. (C) 2003 Elsevier Inc. All rights reserved.