Colourful predictions? Unraveling the resolution and neural mechanisms of colour expectation

2026-07-22

Whether doing sports or driving through city traffic, our visual system constantly builds expectations to predict and quickly react to upcoming events. However, despite the great importance of expectations, their form and neural implementation remain a mystery. It is completely unknown whether expectations encode a clear picture of the anticipated visual input or just a rough sketch. This gap in knowledge is vital, since only things encoded in our expectations are expected to enter conscious awareness faster. Given the high value of colour signals to guide attention in everyday life (e.g., traffic signs), this project asks where and when in the brain colour expectations are formed and if they are in full colour or just comprised of few basic hues. To track cortical colour expectations with high spatiotemporal precision, electro- and magnetoencephalography will be combined with anatomical brain scans and cutting-edge computational decoding in a new, integrative approach. First, a decoding model will be built that reconstructs colour from brain responses. Applying this new model to brain activity of participants anticipating colours will elucidate when and where in the brain colour expectations are formed. Combined behavioural and electromagnetic measures will then determine the influence of these expectations on our perception. Finally, combining the decoding model with colour similarity levels and individual colour categorizations will clarify the resolution of colour expectations. This project will be the first to unravel the neural mechanism, influence and limits of colour expectations. Together with the decoding model as future tool, this will open new perspectives on predictive processing, perception and imagery, but will also have widespread societal implications: designing of better recognizable colour traffic signals, improving colouring schemes in school books to more effectively guide attention, and even for decoding brain responses of locked-in patients.

Netherlands

2025-10-15