ABSTRACT
The instructional value of a set of spelling-to-sound rules can only be determined in the context of the search for meaning from print. Most people assume that the most efficient solution to the problem of dealing with novel letter strings is to have a small set of very general principles or rules for spelling-to-sound correspondence. Spelling-to-sound rules may be useful for beginning readers as organizing principles or mnemonic devices until lexical activation becomes an automatic consequence of letter recognition. The best evidence for the conflict between different kinds of spelling-to-sound structures came from the errors made in pronouncing the exception letter strings. A decade of psychological research on reading and word recognition converged on idea that pronunciation required a separate set of abstract spelling-to-sound rules that had little to do with the mechanisms of normal reading. The size of the interference effects depends on overlap between the entries activated in the visual network and those activated in the phonological network.
