but I am concerned that when we extend the use of this paradigm to study underlying cognitive processes (e.g., verbal memory) that the dual diversity of age and cognitive ability makes it difficult to appropriately match subjects or to interpret the outcome (see Shankweiler, Crain, Macaruso, and Brady, in press, for further discussion).
All these points bring me to the somewhat tiring conclusion that in the reading field we need to refine our knowledge of the individual constructs of speech perception, speech production, verbal working memory, and lexical access. In the earlier chapters in this volume, we can see that the effort to finetune our understanding of the construct of phonological awareness is already underway. I don't want to imply that we shouldn't at the same time attempt to test hypotheses about how the whole picture fits together, but it will be important to be cautious about interpreting individual studies. As Stanovich ( 1985) discusses, we need to adhere to the principle of converging evidence. The power of this principle has been awesome concerning the evidence that reading problems are associated with language deficits and, in particular, with phonological awareness. When we ask, as we have been here, how the different phonological processes relate, convergence is lacking and we need to ask why.
In closing, in this chapter I have attempted to accomplish three goals. In the first section, the widespread association of verbal short-term memory deficits with reading disability was briefly reviewed. The prevalence of this association underscores the need to examine the factors that contribute to working memory performance. In the second section of the chapter I argued that studies of memory development and of memory deficits in poor readers provide tantalizing indications that efficiency of phonological coding may be an important factor in memory performance. In the third section I raised the question of how working memory processes are related to other phonological processes, phonological awareness and lexical access, that have also been associated with reading deficits. At present, attempts to delineate the connections between them have yielded contradictory indications. Conceptual and methodological advances, which hinge on the continued cooperation of disciplines involved in the study of language, will be needed to advance our grasp of these issues.
I gratefully acknowledge the helpful comments of Anne Fowler and Donald Shankweiler on an earlier draft. My research was supported in part by a Program Project Grant (HD-01994) to Haskins Laboratories from the National Institute of Child Health and Human Development. Some of the material