Do the multiple (ON and KUN) readings of Japanese kanji induce a more direct visual to meaning association than Chinese kanji with only a single phonetic code? The rate at which Japanese kanji and Chinese kanji access meaning and pronunciation was tested in comparison to the rate at which English words and photographs access verbal and semantic codes. A Stroop-paradigm was employed where the pattern of disruption caused by the simultaneous presentation of incongruent photograph-word (Japanese, Chinese and English) pairs revealed the order of information access by each stimulus. It was found that the categorization of an object (a semantic task) represented in a photograph caused severe interference by the presence of an incongruent word ONLY for the Japanese. In addition, naming photographic images (a phonetic task) in the presence of an incongruent word was also adversely affected in the Japanese case. There is greater disruption in the processing of similar stimuli than stimuli that are very different. Reading Japanese appears to have something in common with the processing of photographs, which Chinese and English words do not. It is suggested that the ON/KUN aspect of kanji make it more "semantically transparent" than Chinese single pronunciation kanji.