Latencies for Naming Bound Morphemes and Words in Kanji

Abstract

Naming latency for kanji was compared under three conditions: (1) the B-B (bound morpheme to bound morpheme) condition where single bound morpheme kanji were presented for naming, (2) the W-W (word to word) condition where two-kanji words were presented with the stimulus kanji used in the B-B condition appearing in the initial position, and (3) the B-W (bound morpheme to word) condition where the stimulus kanji used in the B-B condition were presented but required word-level responses which began with these kanji morphemes. The results showed that, while the response time was not significantly different between the B-B and W-W conditions, it was significantly shorter in these conditions than in the B-W condition. It is concluded that single bound morpheme kanji are generally named by directly accessing phonology rather than first accessing the lexicon.