"Memory color influences conscious object perception" finally accepted

Our paper on how memory colour affects visual consciousness has finally been accepted by “Neuroscience of Consciousness”. Led by our PhD student Vincent Plikat, the study investigated how prior knowledge about which colours are typically associated with which objects affects visual awareness. To answer this question, he examined how colour pairs, red versus green or blue versus yellow, compete for visual awareness in binocular rivalry. Crucially, these colours were super-imposed on photos of objects that were typical of one of the colours. Observers may thus have been presented a strawberry in red, its typical colour, to one eye and in green to the other eye. From the observer’s perspective, their perception keeps alternating between seeing the red strawberry and the green strawberry as is common in binocular rivalry. Now, would observers be more likely or less likely to perceive the object in its typical colour, i.e., its memory colour? It turns out that, by and large, numerous behavioural metrics showed a shift in conscious report in the direction of the congruent object-colour pair, i.e., the red strawberry, rather than the incongruent, green one.

These findings are important for at least two reasons. First, previous psychophysical studies of memory colour effects on perception relied on self-reported judgements, which made them prone to decision biases, i.e., people’s reports may not reflect their perception but rather just how they judge things to look. The new study, in contrast, uses binocular rivalry, which minimises attention and judgement confounds. At a more general level, the paper thus addresses the fundamental question of cognitive penetrability, i.e., if cognition can influence perception. Second, a large body of literature suggests that the brain is a prediction machine that uses prior knowledge to interpret sensory signals perceptually. It is less clear, what this view entails for perception: On the one hand a violation of visual expectations may carry more information than expectations that are confirmed by sensory input and may thus prioritise their perception. Simply put: if something is already expected, there is nothing “newsworthy” when that expectation bears out. On the other hand, a match between expectations and visual input may strengthen the plausibility of that interpretation and thereby strengthen its perception. It turns out that the latter is the case for memory colour.

The final publication can be accessed on the webpage of Neuroscience of Consciousness. A previous, pre-print version of the paper is available for download on OSF.