Abstract
The present study investigated whether Japanese 3-, 4- and 5-year-olds were better at tracking a speaker's intentions and her/his false-beliefs. At question was whether children were better at learning a novel word on a word-learning task, compared with their predictions of another person's action on a false-belief test. The results revealed that on a word-learning task children chose the correct toy based on a strong connection between the speaker and her/his novel toy, without using a representation of the speaker's false-belief, because they were able to infer her/his communicative intention as their common goal in the context of language communication. As 3- and 4-year-olds were significantly better at tracking a speaker's intentions than a speaker's false-beliefs and an actor's intentions (an actor's false-beliefs), their understanding of speaker's intentions in the domain of language communication was distinguished from their understanding of an actor's intentions in the domain of action prediction. However, children's ways of understanding a speaker's false-beliefs were not different from their ways of understanding an actor's false-beliefs.
