Abstract
Aggression is a phylogenetically stable behavior, and attacks on conspecifics are observed in most animal species. In this review, we discuss translational models as they relate to pathological forms of offensive aggression and the brain mechanisms that underlie these behaviors. Quantifiable escalations in attack or the development of an atypical sequence of attacks and threats is useful for characterizing abnormal variations in aggression across species. Aggression that serves as a reinforcer can be excessive, and certain schedules of reinforcement that allow aggression rewards also allow for examining brain and behavior during the anticipation of a fight. Ethological attempts to capture and measure offensive aggression point to two prominent hypotheses for the neural basis of violence. First, pathological aggression may be due to an exaggeration of activity in subcortical circuits that mediate adaptive aggressive behaviors as they are triggered by environmental or endogenous cues at vulnerable time points. Indeed, repeated fighting experiences occur with plasticity in brain areas once considered hardwired. Alternatively, a separate "violence network" may converge on aggression circuitry that disinhibits pathological aggression (for example, via disrupted cortical inhibition). Advancing animal models that capture the motivation to commit pathological aggression remains important to fully distinguish the neural architecture of violence as it differs from adaptive competition among conspecifics.