Does logographic script occupy an intermediate position between pictures and alphabetically written words? When confronted with a word, will readers of English and Japanese scripts employ different strategies in reading it? The present experiment tested the hypothesis that in reading, kana and English words have faster access to verbal codes than semantic, and that kanji, like pictures, have faster access to meaning than pronunciation. It was found that words, no matter whether they were written in English, kana or kanji were found to access the phonetic code prior to the semantic in reading.