Differentiating Horror and fear via elicitor qualities

Harmful events elicit strong emotions, such as fear, horror, anger and pity, even when they are encountered third-hand via news or online posts. But how media audiences respond emotionally varies, depending on how the event is described. For example, to understand how the English-language emotion concepts of “horror” and “fear” represent distinct responses to different types of real-world harm, Taylor & Uchida (2022) investigated how the salience of these terms differed depending on the type of harm described. They observed that when harm was extreme or abnormal, people reported more horror than fear, but when the harm was mild, common or a pre-harm threat, “fear” was the more salient concept. Based on these findings, the authors defined horror as an emotional response to extreme and abnormal harm. While that was a useful first step to clarifying how horror is distinct from fear, key gaps remain in their definition. Specifically, “abnormal” is a vague concept, and the qualities that constitute horrific abnormality have not been defined.
To address this gap, two pre-registered experiments manipulated two aspects of harm hypothesized to affect perceptions of abnormality and, consequently, feelings of horror: severity of harm and graphicness of description. Participants read tabloid headlines describing interpersonal harm that were manipulated in a 2 (Severity of Harm: Extreme, Moderate) x 2 (Graphicness of Description: Concrete, Abstract) design, with fully-crossed repeated-measures. After reading each headline, participants rated their affective response (horror, fear and physiological arousal) and appraisals of the harm (abnormality, graphicness, severity, harmfulness).
To investigate 1) the appropriateness of defining horror as a response to abnormal harm, (which differentiates it from fear) and 2) what aspects contribute to perceptions of abnormal harm, three main hypotheses were tested: relative to Moderate Harm and Abstract Descriptions, Extreme Harm and Concrete Descriptions will increase perceived abnormality (H1), as well as ratings of horror (controlling for fear) but not fear (controlling for horror; H2), and the effects of severity and graphicness on horror will be mediated by abnormality (H3). As additional test of the assumption that horror and fear are different emotions, we tested the alternative hypothesis that horror is merely extreme fear by testing if physiological arousal differentiates horror from fear, rather than abnormality (H4) by including arousal as a covariate in the model testing H2.