Kanji words are easier to identify than katakana words

The effects of script (kanji and katakana), lexicality (word and nonword), and string length (1, 2, and 4 characters) on the identification of briefly exposed kanji and katakana strings were examined. It was found that identifiability is far higher for kanji words than for katakana words, even though the former are visually more complex than the latter. However, the superiority of kanji completely disappeared in a nonword context, suggesting that it is not visual perceptibility that enhances identifiability of kanji words, but something else, perhaps inferability. Differing effects of visual complexity and form were also found. A theory that explains these results is presented.