ABSTRACT
There are at least three strategies that beginning readers can use to identify unfamiliar words in connected text (Ehri, 1991; Ehri & Robbins, 1992). They can use their developing knowledge of correspondences between subcomponents of written and spoken words to generate the word’s pronunciation, they can analogize to known words that are stored in lexical memory, and they can use the constraints of sentence context to narrow the possibilities of what the word might be. This chapter is concerned with three questions relating to the use of the latter strategy. First, what role, if any, does language prediction skill (i.e., the ability to use the constraints of sentence context) play in the development of word recognition skill? Second, if language prediction skill does indeed make a contribution to the development of word recognition skill, is it a source of individual differences in learning to read? And third, how important is language prediction skill in relation to other skills, such as phonological recoding (i.e., the ability to translate letters and letter patterns into phonological forms)?
