Do Japanese EFL learners activate phonology in reading English words and Japanese kanji

This article explores whether or not there is further support for ‘the dual access model’ of the word meaning access, and if so, examines what relationship can be assumed between the two processing routes (i.e., phonology-mediated versus visual) of the model. In the two experiments to be reported, Japanese university students were required to make semantic and phonological decisions for English and Japanese kanji word pairs presented simultaneously on the computer monitor. Though the overall results suggest the plausibility of the above dual processing routes, the present study may also postulates ‘a universal activation-dominance hypothesis’ of the model, the verification of which seems to be an important next step for the future research. INTRODUCTION It has been an issue of great theoretical interest whether processing printed words involves the phonological recoding of the visual form or is possible directly from the visual orthographic representation (Taft, 1991). So far three hypotheses have been proposed on this subject-matter (Perfetti, 1999): (1) universal direct access hypothesis, (2) orthographic depth hypothesis, (3) universal phonology principle. The universal direct access hypothesis suggests that the access to word meaning is possible silently without phonological recoding in all the languages of the world. According to the orthographic depth hypothesis, the orthographic depth (i.e., the degree of consistency in grapheme-phoneme correspondences) determines the relevance of phonological activation in the processing of word meanings. In the shallow orthographic systems (e.g., Spanish, German) where there is supposed to be one-to-one correspondences between graphemes (G) and phonemes (P), the phonological representations easily obtained through the GP converting rules is indispensable for the meaning access. On the other hand, the meaning access in languages which have little or no regularity in the GP relationships is achieved without activating the phonological representation which is eventually obtained after lexical access. Finally the universal phonology principle hypothesizes that the lexical phonological representation is the inevitable prerequisite for the meaning access.