Across alphabetic languages, reading speed of beginning and dyslexic readers is slow, whereas in skilled readers the sight of a word’s written form is often sufficient for the immediate activation of its pronunciation in long-term memory. The general belief is that beginning and dyslexic readers lack the orthographic knowledge that is needed for the quick recognition of a word, and therefore often have to use a slow sublexical process of print to sound translation. Obviously, the development of orthographic knowledge is a key issue in reading research.
According to one prominent theory of reading development, Share’s self-teaching hypothesis, the acquisition of orthographic knowledge is acquired through phonological recoding, the ability to translate an unfamiliar written form of a word into its spoken counterpart. In addition, this theory makes the claim that orthographic knowledge is word specific. Phonological recoding of novel words and word-specific orthographic knowledge are also assumed in the DRC computational model of skilled reading. We will present a number of studies on the use and importance of phonological recoding and the development of (word specific) orthographic knowledge in normal reading and dyslexic children.