We retrospectively analyzed the acquisition process of speech, phonological awareness, and reading hiragana using cognitive neuropsychological model in two patients with expressive language disorder. One of the subjects（case 1）developed dyslexia, whereas the other（case 2）did not develop the condition. Case1 made numerous non-word errors in her speech although she did not hav any deficit in
speech comprehension. This suggests her specific weakness in creating phonologicaloutput representation, which caused delay in the development of phonological output lexicon. Case1 also experienced difficulty in phonological awareness and acquisition of letters at a later stage, indicating weakness the phonological output assembly buffer that operates on letter-sound（mora）translation. Conversely, the speech errors in case2 were partial errors where the target words are recall with a few phonological errors, whereas her speech comprehension was not as intact as that of case1. This suggested mild weakness in the phonological input and output as well as the semantic
system. However, case2 showed better performance in the phonological awareness task and reading than case1, suggesting the strength of the phonological output buffer and letter-sound（mora）translation rules. In correctly, resulting in improved speech output. This training strategy was not effective in case1. However, the use of the semantic system, the strength of case1, helped her learn letters.
View full abstract
