Del "mal" a la agencia dañina: reconstrucción multiescalar del Modelo MADI para la ciencia psicológica, clínica y social From "Evil" to Harmful Agency: A Multilevel Reconstruction of the Integrated Harmful Agency Model

The term "evil" retains considerable moral, historical and narrative force, but it has limited scientific value when used as a primary explanation for harmful conduct. It typically conflates normative judgment, outcome description, dispositional attribution and causal inference into a single label, thereby producing circularity, person essentialism and low mechanistic resolution. This article reconstructs the field through the construct of Harmful Agency (HA), defined as the contextually activatable propensity and capacity to produce, facilitate, delegate, sustain or legitimize avoidable harm against persons or groups who seek to avoid it, through action or omission, when conduct is supported by proximal motives, situational opportunities and mechanisms such as moral disinhibition, dehumanization, dominance, reward, obedience or justification. Drawing on a structured integrative review of foundational literature and high-impact reviews, the manuscript reformulates the Integrated Harmful Agency Model (IHAM/MADI) across five planes: dispositional and developmental vulnerabilities, situational triggers, cognitive-affective gates, biological modulators and institutional embedding. The paper distinguishes HA from aggression, violence, psychopathy, sadism, dehumanization and moral disengagement; synthesizes evidence on hostile attribution bias, callous-unemotional traits, triarchic psychopathy, obedience, social identity, structural violence and neurobiological networks; and derives falsifiable predictions, empirical designs and clinical, forensic, preventive and policy implications. The central claim is that behavioral science gains little from asking whether a person is "evil"; it gains precision by specifying which form of harmful agency is activated, under which conditions, through which mechanisms and with what prospects for inhibition, accountability and repair.