The Flashed Face Illusion as Conditioned Deployment of Self-Motion Compensation

The Flashed Face Illusion (FFI)---in which neutral faces briefly presented in the periphery appear grotesquely distorted---is typically attributed to feature-motion binding errors arising from poor peripheral resolution. We propose an alternative: the FFI reflects conditioned deployment of the brain's self-motion compensation mechanism, triggered by an ecological heuristic that produces a false positive in the laboratory context. Fixed forward gaze during peripheral transients ordinarily signals forward locomotion; the compensation mechanism engages accordingly, applying a transform appropriate for optic flow to static faces. This account explains three phenomena that resist standard explanations: face specificity (mediated by FFA gating of compensation gain), the peripheral requirement (reflecting the eccentricity of self-motion processing), and sensitivity to optic flow geometry (tangential flow produces more distortion than radial outflow because the mechanism is not pre-engaged). We distinguish our account from configural distortion, attention-based, and generic motion anisotropy alternatives. We further propose that this compensation mechanism is implemented through classical conditioning to the reliable sensory predictors of forward locomotion, which accounts for its robustness to conscious override and predicts cross-cultural variation as a function of habitual optic flow environment. The framework situates the FFI within a broader class of conditioned compensatory phenomena and generates six testable predictions that distinguish it from competitor accounts.