| 研究実績の概要 |
The year began with a detailed analysis of a crosslinguistic corpus to look at the occurrence of silent and filled pauses with respect to discourse and clausal structure of speech. The results showed that silent pauses were more closely associated than filled pauses with both discourse processing as well as clausal processing of speech.
Following these results, a series of psycholinguistic experiments were performed based on the earlier work of Bailey and Ferreira (2003) which had shown that native English listeners’ in-line inferences about the syntactic structure of English speech is influenced by filled pauses. In the present experiments, both native and nonnative listeners were tested with sentences containing filled, silent, or no pauses. Results replicate Bailey and Ferreira for filled pauses but further show that silent pauses are more important as signals of syntactic structure. Furthermore, in a novel result, for nonnative listeners, pauses facilitate listeners’ perception of syntactic structure. The experiment was also adapted to a self-paced reading paradigm where participants read sentences with filled pauses in them. Results matched the listening paradigm.
Finally, the importance of silent and filled pauses was pilot-tested in a highly controlled speech production paradigm to see which type of pauses were dominant in specific sentential positions. Preliminary results are not conclusive because of a small number of participants. But the trends are in the predicted direction with participants using longer pauses at clause than non-clause boundaries.
|
