Temporal Structure in Speech Can Affect Semantic Prediction: Evidence from the N400

No Thumbnail Available
Journal Title
Journal ISSN
Volume Title
Listeners can use the content of the preceding speech to predict the semantics of the following word. At the same time, temporal structure in speech provides listeners the information about when the following words will occur. However, few studies examine the interaction between temporal information and the semantic information coming from the top-down prediction. This study aims to fill the research gap between temporal structure and semantic prediction. We used isochronous speech and changed the onset time of the final word (target) to control the temporal structure of the speech. High and low constraint sentences were designed to control the semantic prediction of the target. This study used EEG to record brain activity when listening to the experimental stimuli. In line with previous studies using natural speech, we demonstrated that the N400 component can be used as the index of semantic prediction in isochronous speech. More importantly, we found that the N400 component was significantly decreased when the onset of the target was earlier, but only in the high-constraint condition. In the low-constraint condition, activation in the N400 time-window was not significantly affected by timing. Our results show that timing of speech and the semantic predictability closely interact: the effect of semantic predictability is largest when the word appears at the expected time.
Faculteit der Letteren