hide
Free keywords:
-
Abstract:
In this talk I would like to present results from (on-going) eye-tracking experiments, using the Visual World Paradigm, investigating the processing of disfluencies. Since spontaneous speech is strewn with disfluencies (such as uhm’s, silent pauses, repetitions, repairs, etc.), one may pose the question how listeners cope with these disfluencies. Do disfluencies hinder the processing of the content of the speech signal or can they actually be helpful to listeners in predicting what the speaker will say next?
I will demonstrate that listeners use disfluencies in reference solution: upon hearing the uh in a sentence like “Click on uh the sewingmachine”, our participants showed more anticipatory looks to low-frequent pictures (e.g., the sewingmachine) as compared to high-frequent pictures (e.g., the hand). In our view, listeners took the uh as a sign that the speaker was having trouble naming an object. These troubles are more likely to occur in naming low-frequent pictures than in naming high-frequent pictures, leading to more anticipatory looks to the low-frequent picture. Our data reveal that listeners are sensitive to disfluencies and that they make use of disfluencies, when listening to a native speaker, to anticipate subsequent content. The next step in our research is to investigate the processing of disfluencies in non-native speech. To illustrate this, I will present some recent data from a study into the supposedly beneficial effects of native and non-native disfluencies on subsequent memory.
