The fundamental attribution error is the tendency to over-attribute other people’s behavior to their character or disposition while underweighting the situation they were in — and, often, to do the reverse for ourselves. Ross (1977) named the effect and argued it is a pervasive default in social judgment; Jones and Harris (1967) had already demonstrated it, finding that observers inferred a writer’s genuine attitude from an essay even when told the writer had been assigned the position by a coin flip. We see the person and only dimly see their circumstances, so behavior reads as a window into who someone “really is” rather than a response to pressures we cannot see.
As a manipulation, this bias is the engine of blame-shifting and scapegoating. An operator isolates an act, strips away the context that explains it, and presents it as proof of the target’s bad character — then generalizes that verdict to the whole person and uses it to redirect responsibility. Because the audience is running the same bias, the character explanation feels natural and the missing situational facts are rarely demanded. The technique underwrites everyday workplace blame, political attack messaging, and, in its sharpest form, DARVO — where an offender deflects, attacks, and reverses victim and offender by recasting the person harmed as the one whose character is truly at fault.
The high-band harm rating reflects how ubiquitous and consequential the tactic is: it is both a mundane feature of ordinary judgment and a reliable weapon for evading accountability and manufacturing villains. It pairs closely with out-group homogeneity (flattening a group and then blaming its “nature”) and the horn effect (one flaw taken to define a person). The defense is deliberate context-seeking: before accepting any character verdict, ask what situation the person actually faced, apply the same standard you would grant the accuser, refuse to let one act become a permanent label, and watch for whether the blame conveniently relieves someone else of responsibility.