The memory colour effect—the bias to perceive objects in their typical colours—can modulate colour appearance and has been proposed to operate via Bayesian mechanisms, combining noisy sensory input with prior knowledge. As colour perception is substantially degraded in peripheral vision, Bayesian modelling predicts that the influence of priors should increase when sensory reliability decreases, suggesting that memory colour effects may be enhanced in peripheral vision. Across three experiments, we tested this prediction. In Experiment 1, participants adjusted the colour of familiar objects to appear grey or natural, either foveally or at 10° eccentricity. Grey settings for peripheral stimuli were systematically biased away from the object’s typical colour more than for central stimuli, indicating a stronger memory colour effect. Experiment 2 replicated these findings and confirmed that the effect magnitude scales with sensory uncertainty, as predicted by Bayesian modelling. In Experiment 3, using a forced-choice task, participants compared central and peripheral versions of the same object, presumably both affected by memory colour. The peripheral stimulus was more likely to be judged as grey then the central stimulus when shifted away from its typical colour, further confirming an enhanced memory colour effect in the periphery. These results demonstrate that memory colour biases are amplified in peripheral vision, supporting a Bayesian framework and highlighting a functional role for object knowledge in compensating for peripheral chromatic loss. Memory colour may help maintain the vividness of colour perception across the visual field.
Metzger et al. (Fri,) studied this question.