Computational Modeling of Word Reading, Acquired Dyslexia, and Remediation

Many researchers assume that the most appropriate way to express the systematic aspects of language is in terms of a set of rules. For instance, there is a systematic relationship between the written and spoken forms of most English words (e.g., GAVE ), and this relationship can be expressed in terms of a fairly concise set of grapheme-phoneme correspondence (GPC) rules (e.g., G , A E , V ). In addition to being able to generate accurate pronunciations of so-called regular words, such rules also provide a straightforward account of how skilled readers apply their knowledge to novel items—for example, in pronouncing word-like nonwords (e.g., MAVE ). Most linguistic domains, however, are only partially systematic. Thus, there are many English words whose pronunciations violate the standard GPC rules (e.g., HAVE ). Given that skilled readers can pronounce such exception words correctly, GPC rules alone are insufficient. More generally, skilled language performance at virtually every level of analysis— phonological, morphological, lexical, syntactic—requires both effective handling of exceptional items and the ability to generalize to novel forms.