Unpleasant words can affect the detection of morphosyntactic errors: An ERP study on individual differences

Abstract

In recent years, several ERP studies have investigated whether the early computation of agreement is permeable to the emotional content of words. Some studies have reported interactive effects of grammaticality and emotionality in the left anterior negativity (LAN) component, while others have failed to replicate these results. Furthermore, novel findings suggest that grammatical processing can elicit different neural patterns across individuals. In this study, we aim to investigate whether the interaction between grammaticality and emotionality is restricted to participants with a specific neural profile. Sixty-one female native speakers of Spanish performed an agreement judgment task in noun phrases composed of a determiner, a noun, and an unpleasant or neutral adjective that could agree or disagree in gender with the preceding noun. Our results support the existence of two different brain profiles: negative and positive dominance (individuals showing either larger LAN or larger P600 amplitudes in ungrammatical stimuli than in grammatical ones, respectively). Interestingly, the neural pattern of these two groups diverged at different points along the time course. Thus, the negative dominance group showed grammaticality effects as early as 200 ms, along with parallel and autonomous processing of grammaticality and emotionality at the LAN/N400 time window. Instead, for the positive dominance group an early interaction was found at around 200 ms, evidencing a grammaticality effect that emerged only for unpleasant words. Our findings confirm the role of individual differences in the interplay between grammar and emotion at the neural level and call for the inclusion of this perspective in studies on syntactic processing.

Keywords: LAN/N400; P600; gender agreement; individual ifferences; unpleasant words.

FIGURE 1

Distribution of participants according to their RDI and ERP waveforms for the grammaticality effect in the two dominance groups. (a) Left: Scatterplot showing the distribution of LAN effect amplitudes (match minus mismatch; y axis) and P600 effect amplitudes (mismatch minus match; x  axis) averaged over the mean activity at left anterior electrodes (F7, F3, C7A, and C3A) and centro‐parietal electrodes (CZ, PZA, and PZ) for LAN and P600, respectively. Each dot shows data from a single participant. The dashed red line represents equal LAN and P600 magnitudes. Participants above/to the left of the red dashed line show a negative dominance brain profile, and participants below/to the right of the dashed line show a positive dominance brain profile. Right: Violin plot showing the distribution of RDI values across participants. More negative RDI values represent a negative dominance profile, while more positive values represent a positive dominance profile; values closer to zero show a biphasic profile. (b) Grand mean averaged ERPs for the whole scalp are shown for match (blue line) and mismatch (red line) conditions for each dominance group.

FIGURE 2

Grammaticality and Emotionality effects for the total sample of participants. Left: Topography maps of the difference between match and mismatch conditions, as well as between neutral and emotional conditions, averaged across specific time windows where Grammatical and Emotional main effects were significant. The asterisks in bold represent the significant electrodes that form part of each of the depicted clusters. Right: Grand mean averaged ERPs from 12 representative electrodes for match (solid line) and mismatch (dashed line) conditions as a function of neutral (black) and unpleasant (red) conditions for the total sample (N = 54).

FIGURE 3

Individual differences: Grammaticality and Emotionality effects. Left: Topography maps of the difference between match and mismatch conditions, as well as between neutral and emotional conditions, averaged across specific time windows where Grammatical and Emotional main effects were significant, for the negative dominance group (a) and the positive dominance group (b). The asterisks in bold represent the significant electrodes that form part of each of the depicted clusters. Right: Grand mean averaged ERPs from nine representative electrodes for match (solid line) and mismatch (dashed line) conditions as a function of neutral (black) and unpleasant (red) conditions for the negative dominance group (a) and the positive dominance group (b).

FIGURE 4

Grammaticality × Emotionality interaction in P200. Grand mean averaged ERPs for match (solid line) and mismatch (dashed line) conditions as a function of neutral (up) and unpleasant (down) conditions in the specific electrodes for CA region for the positive dominance group.