Emotions as a driver of aggressive behaviour

Aggression is often studied as a stable individual tendency. However, aggressive behaviour occurs in moments, between people, and within changing emotional dynamics. To understand why the same person may aggress in one moment but not another, we must shift from asking who is aggressive to when and why aggression emerges. This thesis examines aggression as a dynamic, dyadic, and emotionally driven behaviour. I argue that aggression cannot be fully understood by studying the aggressor alone. To develop this argument, I introduce the Emotional CEEP Model, which outlines multiple emotional pathways to aggression. I show that heart rate variability, a commonly proposed physiological biomarker of aggression, does not reliably predict reactive aggression. Using a newly developed face-to-face dyadic aggression paradigm, I examine how emotion predicts aggression in real time through continuous measures of facial affect and behaviour. Negative affect in both potential aggressor and target predicted more intense aggression, while brief delays between provocation and retaliation reduced aggression. Relationship context shaped aggression: familiarity reduced overall aggression but strengthened the link between negative affect and aggressive behaviour, whereas proximity increased aggression. Finally, both anger- and disgust-related facial affect predicted direct reactive aggression shortly before aggressive behaviour. Together, these findings support a state-based and dyadic account of aggression. Aggression is not simply something that aggressive people do; it emerges from emotional escalation within interpersonal contexts. This thesis contributes a theoretical model and novel analytic framework for studying how emotion drives aggression in real time. More broadly, it suggests that prevention may be improved by identifying and interrupting moments of high negative emotional arousal before escalation.