Signal-driven and expectation-driven processing of accent types

Abstract

This paper investigates neurophysiological correlates of prosodic prominence in German with two EEG experiments. Experiment 1 tested different degrees of prominence (three accent types: L+H*, H*, H+L* and deaccentuation) in the absence of context, making the acoustic signal the only source for attention orienting. Experiment 2 tested L+H* and H+L* accents in relation to contexts such as "Guess what happened today" triggering expectations as to how exciting the following utterance will be. Results reveal that prominence cues that attract attention, such as a signal-driven high level of prosodic prominence or a content-driven expression of excitement, engender positivities of varying latency. Furthermore, contextual expectations trigger prediction errors, e.g. deviations from an appropriate level of prosodic prominence result in a negative ERP deflection. Hence, the data suggest that the two core processesattentional orientation and predictive processingreflect discrete stages in the construction of a mental representation during real-time comprehension.

Rises on Pitch Accents and Edge Tones Affect Serial Recall Performance at Item and Domain levels

“…The results in Fraundorf et al (2010) and Koch and Spalek (2021) are compatible with findings in inter-disciplinary research showing that rising F0 contours are perceived as more prominent than falling ones (see e.g. Baumann & Röhr, 2015 for German pitch accent types), since intonational rises attract more attentional resources than falls (Hsu, Evans, & Lee, 2015;Röhr, Brilmayer, Baumann, Grice, & Schumacher, 2020 for German; see also early experimental evidence on infants 'attention in Sullivan & Horowitz, 1983). This appears to be relevant in assigning a special contribution to intonational rises in memory performance, also given the role played by attention in fixing stimuli in working memory (Oberauer, 2019).…”

Referential Chains Reveal Predictive Processes and Form-to-Function Mapping: An Electroencephalographic Study Using Naturalistic Story Stimuli

“…Therefore, we have to keep in mind that our results still contain overlapping brain responses to adjacent words, especially in the baseline interval and at latencies of the late components of the rERP (<400 ms). Different variants of the rERP method have already been successfully applied to linguistic experiments in the visual (e.g., Hauk et al, 2006 ) and auditory domain (e.g., Brennan and Pylkkänen, 2012 ; Sassenhagen, 2018 ; Alday, 2019 ; Röhr et al, 2020 ; Ventura et al, 2020 ).…”