Theories of Aggression

Core Idea

Aggression is behavior intended to harm another person who does not wish to be harmed. The frustration-aggression hypothesis (Dollard et al.) proposed that frustration always produces an urge to aggress, though Berkowitz later revised this to argue that frustration produces negative affect, which increases aggression only in the presence of aggression-related cues. Bandura's social learning theory demonstrated through the Bobo doll studies that children acquire aggressive behaviors through observation and imitation. Biological factors including testosterone, amygdala reactivity, and MAOA genetics interact with social learning; purely biological or social accounts are inadequate alone.

How It's Best Learned

Compare the frustration-aggression hypothesis, the revised negative affect model, and social learning theory across the same scenario (e.g., road rage). The cue-activation component and the observational learning account are often combined in modern research.

Common Misconceptions

• Catharsis — the idea that venting aggression reduces future aggression — is not empirically supported and may increase aggression.
• Media violence research shows modest, consistent effects on aggressive cognition, but its relationship to serious violence is weak and contested.

Explainer

Aggression — behavior intended to harm a person who does not wish to be harmed — is one of the most studied topics in social psychology because it matters so much practically and because early theoretical accounts proved seductively simple and empirically wrong. The history of aggression theory is largely a story of oversimplified causal chains getting progressively refined, and understanding that history builds genuine explanatory depth.

The frustration-aggression hypothesis (Dollard and colleagues, 1939) proposed a clean, powerful claim: frustration — the blocking of goal-directed behavior — always produces an aggressive drive, and aggression always stems from frustration. It was an elegant hydraulic model: frustration builds pressure, aggression releases it. Your social psychology prerequisite gave you the vocabulary to see why this appealed — it offered a single mechanism with broad explanatory scope. But the original formulation was too strong: people experience frustration constantly without becoming aggressive, and aggression occurs in the absence of any obvious prior frustration. Berkowitz's cognitive neoassociation model (1989) salvaged the core insight while abandoning its universality. Frustration matters not because it creates a drive but because it produces negative affect — aversive arousal that primes aggression-related thoughts and behavioral tendencies. Crucially, this priming converts into overt aggression only in the presence of aggression-related cues: stimuli (weapons, aggressive symbols, angry faces) that activate aggressive cognition and lower inhibition thresholds. The "weapons effect" — showing that the presence of a gun in a room increases aggressive responding even without explicit threat — is direct evidence for cue-activation.

Bandura's social learning theory challenged the internal drive accounts entirely by demonstrating that children acquire specific aggressive behaviors through observation, without frustration, arousal, or any prior direct experience of aggression. In the famous Bobo doll studies, children who watched an adult model punch, kick, and verbally abuse an inflatable doll replicated those specific behaviors in detail — including novel verbal labels the model had used. This shows that aggression is learned like any complex behavior: through observation, vicarious reinforcement, and cognitive representations of behavioral scripts. Crucially, learning to perform aggression is separable from actually performing it — children who learned aggressive behaviors in all conditions only performed them consistently when the model had been rewarded (or at least not punished) for aggression. Incentive conditions regulate performance; observational exposure determines acquisition.

Biological factors — from your prerequisite work on hormones and the limbic system — do not replace these social learning accounts but constrain and interact with them. Testosterone is consistently associated with dominance-seeking and reduced inhibition of aggressive impulses, but its effect on actual behavior is modest and moderated by social context, cultural norms, and provocation. The amygdala is central to threat detection and the generation of defensive aggression; hyper-reactive amygdala responding characterizes some individuals with chronic aggression problems. The MAOA gene influences the breakdown of monoamine neurotransmitters (dopamine, serotonin) and has been linked to heightened aggression — but only among those with childhood maltreatment histories, a finding that exemplifies gene-environment interaction rather than genetic determinism. Biology sets parameters; learning and context determine where within those parameters behavior falls.

The practical implication is that single-mechanism accounts of aggression — whether purely hydraulic, purely social learning, or purely biological — are always incomplete. A comprehensive account of a particular aggressive act asks: what frustration or negative affect was present? What cues activated aggressive scripts? What behavioral models had this person observed, and with what consequences? What biological predispositions altered thresholds? Aggression is multiply determined, which is why effective interventions typically require targeting multiple levels simultaneously — changing environments that generate chronic frustration, altering cue exposure, providing non-aggressive behavioral models, and in some cases addressing biological vulnerabilities.