Seeing Is Feeling: How Aphantasia Alters Emotional Engagement With Stories

Abstract

Visual imagery is thought to act as an "emotional amplifier," potentially contributing to narrative engagement. To examine this, we conducted two experiments in which participants were presented with emotionally charged audio and video story excerpts. Experiment 1 included 84 online participants from the general population, while Experiment 2 involved 25 individuals with aphantasia (the inability to generate mental images) and 25 controls. In both experiments we assessed narrative engagement behaviorally using the Narrative Engagement Questionnaire (NEQ), while for Experiment 2 we also measured physiological responses. We found a main effect of modality, with video stimuli scoring higher across all NEQ subscales in both experiments. Notably, in experiment 2, a significant group effect on emotional-but not cognitive-engagement emerged, with aphantasics reporting less emotional engagement than controls. Moreover, controls experienced higher heart rate during audio narratives, while aphantasics had a similar heart rate across both modalities. Our results suggest that the enhanced physiological response seen in non-aphantasics during audio narratives is driven by the mental effort required to generate imagery. Furthermore, this capacity for visual imagery appears to enhance emotional engagement with stories. This highlights mental imagery's role in both subjective and physiological responses, emphasizing distinct cognitive processes during narrative engagement.

Keywords: aphantasia; emotional amplifier; narrative engagement; storytelling; visual imagery.

FIGURE 1

Flow diagram illustrating the study's conceptual model and the possible outcomes. Predicted physiological responses of (A) controls and (B) aphantasics assuming physiological responses reflect mental effort. Predicted physiological responses of (C) controls and (D) aphantasics assuming physiological responses reflect emotional arousal. Predicted self‐reported emotional engagement of (E) controls and (F) aphantasics.

FIGURE 2

Experiment 1. (A) Model‐predicted means and 95% confidence intervals of the narrative engagement questionnaire subscales split between story modalities. (B) Correlations of familiarity with the narrative engagement questionnaire subscales split between story modalities. Note that we have marked only the effects driving interactions (presence and attention); we have not highlighted the familiarity main effects. ***p < 0.001. **p < 0.01.

FIGURE 3

Experiment 2. Model‐predicted means and 95% confidence intervals of the different dimensions of the self‐reported engagement split between story modalities (audio vs. video) and groups (controls vs. aphantasics). Note that we have marked only the significant results driven by group effects (presence and character engagement); we have not highlighted the modality effects. ***p < 0.001. **p < 0.01.

FIGURE 4

Experiment 2. Model‐predicted mean and 95% confidence interval of the heart rate z‐scores split between story modalities and groups. **p < 0.01.