Electrophysiological Correlates of Emotional Content and Volume Level in Spoken Word Processing

Abstract

For visual stimuli of emotional content as pictures and written words, stimulus size has been shown to increase emotion effects in the early posterior negativity (EPN), a component of event-related potentials (ERPs) indexing attention allocation during visual sensory encoding. In the present study, we addressed the question whether this enhanced relevance of larger (visual) stimuli might generalize to the auditory domain and whether auditory emotion effects are modulated by volume. Therefore, subjects were listening to spoken words with emotional or neutral content, played at two different volume levels, while ERPs were recorded. Negative emotional content led to an increased frontal positivity and parieto-occipital negativity-a scalp distribution similar to the EPN-between ~370 and 530 ms. Importantly, this emotion-related ERP component was not modulated by differences in volume level, which impacted early auditory processing, as reflected in increased amplitudes of the N1 (80-130 ms) and P2 (130-265 ms) components as hypothesized. However, contrary to effects of stimulus size in the visual domain, volume level did not influence later ERP components. These findings indicate modality-specific and functionally independent processing triggered by emotional content of spoken words and volume level.

Keywords: EPN; ERPs; N1; P2; emotional content; spoken word processing; volume level.

Figure 1

Effects of emotional valence and volume level on electrophysiological parameters. (A) The upper graph shows global field power (GFP) across all participants, contrasted for emotionally positive, negative, and neutral words presented at high and low volume level. The lower graph depicts global map dissimilarity (GMD) averaged across all subjects and experimental conditions. Vertical black lines mark the segment borders, which were defined according to the GMD peaks. Between the peaks at 265 ms and 530 ms, event-related potentials (ERPs) were divided into five equally long time windows and after the last clear peak at 530 ms in consecutive time windows of 50 ms between up to 980 ms. (B) Maps show the global scalp distribution averaged across all conditions during the time windows flanked by the borders depicted in (A).

Figure 2

Effects of emotional content and volume level on event-related potentials (ERPs). (A) Grand mean ERP waveforms elicited by emotionally positive, negative, and neutral words are depicted from frontal and parieto-occipital region of interest (ROI) electrodes. (B) Depicted maps show the scalp distributions of the ERP differences between negative and neutral words within the time intervals of significant emotion effects as well as the distribution of ERPs, separated for negative and neutral words. (C) Grand mean ERP waveforms, contrasted for high and low volume level, are depicted for central ROI electrodes. (D) Maps depict the scalp distributions of ERP differences between high and low volume level words as well as the topographies of ERPs for both volume conditions within the indicated N1 and P2 time intervals.