and reading), finding a close correspondence between the two may increase our understanding about reading disability.
I have hypothesized in this chapter that developmental changes in phonological representation may set the stage for acquiring phoneme awareness, and hence, for reading acquisition. I have further suggested that a failure along this same dimension may be responsible for the finding that poor readers are characterized by deficient phonological representations as evident in diminished short-term memory, inability to encode phonological structures under stressed conditions, and underspecified lexical representations.
A great deal of work remains both to track the development of phonological representation over the preschool and early school years, and to assess the impact of these changes on memory, analysis, and other tasks. Nonetheless, we are left with a sense that the phonological representations upon which analysis depends are not a preset immutable part of language. Indeed, as suggested here, it is possible they may be influenced by increased vocabulary, word play, phonological awareness, and literacy. Once phonological representation is taken out of the realm of the invariant, it becomes possible to do productive research on the relationship between, and development within, awareness and representation. It allows us to be both more specific about what a phonological deficit consists of and more optimistic about remediating reading disability.
I conclude this discussion by applauding Treiman for her thorough and systematic chronicling of the emergence of phoneme awareness and for her efforts to relate awareness and underlying representations. Where many have indicated a lack of phoneme awareness, Treiman, like Liberman before her, has pressed on to ask the more positive question of what units of sound are salient for the young child. Treiman's further refinement of this developmental progression within the context of modern phonology sets the stage for a more explicit account of growth and change in phonological representation than has been available to date, enriching our understanding of phonological processing and phonological development.
I am grateful to my many colleagues at Haskins ( Catherine Best, Catherine Browman, Lois Dreyer, Alice Faber, Carol Fowler, Louis Goldstein, Eliza Goodell, Alvin Liberman, Isabelle Liberman, Susan Nittrouer, and Donald Shankweiler) for sharing their expertise and providing interesting discussion.