hide
Free keywords:
Prosody, Entrainment, Sentence processing
Abstract:
Prosodic Entrainment Influences Sentence Segmentation
Speech comprehension is subserved by cyclic electrophysiological activity, so-called neural oscillations. Oscillations within the delta band (i.e., < 4 Hz) are thought to synchronize or entrain to the pace of intonational phrase boundaries (IPBs). By means of entrainment, oscillations inherit acoustic rhythms that are present in speech to persist after stimulation offset (e.g., Kösem et al., 2018). Therefore, it could be possible that through entrainment, delta-band oscillations maintain the prior intonational rhythm (Breen, 2014; Frazier et al., 2006), thus triggering the perception of an IPB in an upcoming sentence.
To test this, we combined an initial prosody entrainment with a subsequent visual target sentence. Target sentences were ambiguous, that is, they allowed for two different segmentation options (Hoeks et al., 2002; e.g., “Max sees Tom and Karl laughs”). We reasoned that the duration of the prior contour, corresponding to the duration of one of the segmentation options, would affect segmentation, and thus interpretation of the upcoming sentence. We conducted three online experiments. A trial started with one of two prosodic contours, repeated three times to induce entrainment. The contour was followed by word-by-word rapid serial visual presentation of a target sentence; the sentence-final words were self-paced. Segmentation was inferred from reading times, as well as from RTs to comprehension questions.
The results demonstrated a significant speed-up after a longer prosodic contour at the first self-paced word—apparently, participants expected the sentence to continue after this word, thus accelerating. Therefore, long contours may enable the prediction of long upcoming segments, while short contours trigger the prediction of short segments. Prosody entrainment may thus serve a predictive function at the prosody–syntax interface, not only helping the reader to segment the current speech input, but also inducing predictions of the duration of upcoming constituents (Breen, 2014; Frazier et al., 2006; Grosjean, 1983). We are currently conducting an MEG experiment to assess the underlying electrophysiological dynamics: we expect to find delta-band phase differences, as well as power enhancement at stimulation frequency.
