ABSTRACT
In this chapter, the authors illustrate the distinction between the word-level and the subword-level procedures for reading aloud with reference to the example word steam, for which either procedure should yield a correct reading response. In contrast, the subword level procedure guarantees correct reading aloud only when the orthographic stimulus is a regularly spelled word or a non-word. The word-level procedure allows correct reading aloud only when the orthographic stimulus is a word. Most models of reading aloud entail some version of the distinction between two procedures for converting print to phonology, although the models differ in the ways in which the two procedures are characterised. These two procedures are lexical procedure and non-lexical procedure. If instead the input is a non-word like vib or slint, there will be no means by which the word-level procedure can compute an address for the phonology corresponding to the whole orthographic string.
