Onset and Rime Awareness and Analogies in Reading

It has been known for a long time that there is a strong relationship between a child's phonological awareness and that child's progress in reading (Bryant & Bradley, 1985; Calfee, Lindamood, & Lindamood, 1973; Cossu, Shankweiler, Liberman, Tola, & Katz, 1987; Liberman, Shankweiler, Fischer, & Carter, 1974; Lundberg, Frost, & Petersen, 1988; Maclean, Bryant, & Bradley, 1987; Rosner & Simon, 1971; Tunmer & Nesdale, 1985). It is also known that measures of rhyming and alliteration are especially strong predictors of later reading progress, as they still account for a significant proportion of the variance after other variables are partialed out of the equation (e.g., Bradley & Bryant, 1983; Bryant, Bradley, Maclean, & Crossland, 1989; Ellis & Large, 1987, Lundberg, Olofsson & Wall, 1980). For example, Bryant et al. (1989) showed that nursery rhyme knowledge at 3 years was related to reading ability at 6 years even after differences in social background and IQ were taken into account. Recently, Goswami & Bryant (1990) suggested that the linguistic units onset and rime may be crucial in explaining this robust link between rhyming and reading. Specifically, a child's ability to categorise words by onset-rime units (the onset of the word beak corresponds with the initial consonant, b-; the rime corresponds with the vowels and final consonant, -eak) may be related to that child's awareness of spelling sequences in words (Goswami, 1990; Goswami & Bryant, in press), and this may facilitate the development of reading. This hypothesis is based on the finding that children can make analogies between spelling patterns in words to help them to read new words. For example, children can use the spelling-sound pattern of a word like beak to make an inference about the sound of new