Identifying intrinsic constituents of focus through ‘‘imitation via restoration.’’

In this study we test the hypothesis that although certain parts of an observed intonation may seem dispensable in perception tests, they nevertheless are consistently produced by speakers. We refer to all consistently produced parts of an intonation as its ‘‘intrinsic constituents.’’ To identify the intrinsic constituents, we developed an experimental paradigm called ‘‘imitation via restoration.’’ In this paradigm, the intonation under scrutiny is first recorded by a native speaker. Then words carrying a potential constituent of the intonation are replaced by a loud noise. During the experiment, the sentence containing the replacement noise is presented to the subjects together with the text. The subjects’ task is to repeat the sentence in exactly the same way as they hear it. The consistency with which subjects restore the missing parts of the target intonation would therefore provide a reasonable indication as to which of them are truly intrinsic to the intonation. Our first such experiment was conduct...