Belief Perseverance Against Contradictory Evidence

Why This Is an Axiom

Belief perseverance represents a fundamental asymmetry in human cognition: the threshold for belief formation is lower than the threshold for belief revision. This is not a peripheral bias but a core architectural feature that shapes learning, reasoning, and knowledge change. Beliefs, once established, develop a kind of cognitive inertia—they resist modification even in the face of evidence that would prevent their initial formation. This asymmetry cannot be reduced to simpler principles; it reflects deep features of how cognitive systems balance stability and plasticity.

Evidence Base

Classic studies by Ross, Lepper, and Hubbard (1975) demonstrated that even after participants were explicitly told that feedback used to form initial impressions was completely fabricated, the impressions persisted. Subsequent research has shown this effect across domains: people maintain beliefs after statistical evidence is debunked, causal theories persist after disconfirmation, and impressions endure after misinformation correction.

Computational analyses show that human belief updating systematically deviates from Bayesian norms by underweighting new evidence relative to prior beliefs (conservatism). Eye-tracking studies reveal that people literally spend less time attending to belief-inconsistent information, and neural imaging shows reduced processing of contradictory evidence in regions associated with belief evaluation.

Curricular Implications

This axiom explains why misconceptions are so difficult to correct and why simply presenting correct information is often insufficient. Effective instruction must explicitly address existing beliefs, create conditions for genuine cognitive conflict, and provide alternative explanatory frameworks—not merely contradictory facts. It also suggests that preventing misconception formation is more efficient than later correction, highlighting the importance of careful initial instruction.