Evidence for dynamic attentional bias toward positive emotion-laden words: A behavioral and electrophysiological study

Abstract

There has been no consensus on the neural dissociation between emotion-label and emotion-laden words, which remains one of the major concerns in affective neurolinguistics. The current study adopted dot-probe tasks to investigate the valence effect on attentional bias toward Chinese emotion-label and emotion-laden words. Behavioral data showed that emotional word type and valence interacted in attentional bias scores with an attentional bias toward positive emotion-laden words rather than positive emotion-label words and that this bias was derived from the disengagement difficulty in positive emotion-laden words. In addition, an attentional bias toward negative emotion-label words relative to positive emotion-label words was observed. The event-related potential (ERP) data demonstrated an interaction between emotional word type, valence, and hemisphere. A significant hemisphere effect was observed during the processing of positive emotion-laden word pairs rather than positive emotion-label, negative emotion-label, and negative emotion-laden word pairs, with positive emotion-laden word pairs eliciting an enhanced P1 in the right hemisphere as compared to the left hemisphere. Our results found a dynamic attentional bias toward positive emotion-laden words; individuals allocated more attention to positive emotion-laden words in the early processing stage and had difficulty disengaging attention from them in the late processing stage.

Keywords: ERP; dot-probe task; emotion-label words; emotion-laden words; valence.

Figure 1

An overview of a trial in the dot-probe task.

Figure 2

Mean grand-average ERPs (A) and the topography of cortical responses (B) to the four types of emotional word pairs over parieto-occipital sites for P1. E-label, emotion-label; E-laden, emotion-laden.

Figure 3

Mean grand-average ERPs (A) and the topography of cortical responses (B) to the dots under congruent and incongruent conditions over fronto-central and central sites for N2.