Acquired Dyslexia in Japanese: Implications for Reading Theory

Acquired dyslexia research has been conducted mainly on English neurological patients. A limited number of dyslexia studies on non-alphabetic orthographies are available. Classical case studies for acquired dyslexia in Japanese, which has two distinctive scripts (morphographic Kanji and phonographic Kana), reported 'script-dependent' dyslexia patterns. Although recent case studies showed 'script-independent' dyslexia patterns for surface and phonological dyslexia, a 'script-independent' deep dyslexia pattern in Japanese has not yet been reported. This study examined four Japanese aphasic patients, using psycholinguistically well-manipulated reading stimuli for both Kanji and Kana strings. YT, with phonological impairment, demonstrated the same effects of psycholinguistic variables as observed in English deep dyslexia, but semantic errors rarely occurred in Kana word reading. YT's concomitant deep dyslexia for Kanji, and phonological dyslexia for Kana fit the phonological impairment hypothesis, and this can be treated as a unique characteristic of Japanese deep dyslexia. HW, with semantic impairment, demonstrated a 'script-independent' surface dyslexia pattern. SO, with severe semantic impairment, demonstrated a surface dyslexia pattern in Kanji word reading, but showed substantial difficulty with Kanji nonword reading. ME, with phonological impairment and a visuo-spatial deficit, showed both lexicality and length effects on reading aloud Kana strings, thus suggesting phonological dyslexia for Kana. That is, the double dissociation between Kanji and Kana nonword reading was observed in SO and ME. These results suggest that Japanese acquired dyslexia patterns are not dependent on script-type, but are also not totally independent of script-type. These outcomes of this study are discussed in terms of universality and orthographic-specificity in acquired dyslexia. Moreover, possible workings of the Japanese version of the DRC model (Coltheart et al., 2001) and the triangle model Plaut, et al., 1996; Harm & Seidenberg, 2004) are presented in order to explain acquired dyslexia patterns in Japanese