The experimental results on lexical decision times for Japanese homophones with multiple mates
are inconsistent. Kawakami (2006) and Tamaoka (2007) have reported longer lexical decisions for
homophones than for nonhomophones, which are called homophone effects, while Hino, Kusunose,
Lupker, and Jared (2013) showed shorter latencies, i.e., facilitatory homophone effects. This
study explored the possibility that extremely low frequency words in Hino et al.’s (2013) stimuli
produced this aberrant finding. In Experiment 1, we replicated Hino et al.’s (2013) experiment
with undergraduate participants. The error rates were extremely high, which correlated with their
frequencies as well as their lexical decision time, and produced inconsistency in the results. In
Experiment 2, faculty members, who presumably had a large vocabulary, performed the same
experiment. Facilitatory homophone effects were not found for homophones with multiple mates, and
their latencies did not differ from those for nonhomophones. These findings suggest that extremely
low frequency stimulus words can alter findings on homophone processing. Finally, the possibility
that high phonological familiarities for homophones with multiple mates might shorten their latencies
was discussed.
抄録全体を表示