A neurocomputational account of motivated seeing

Do goals, beliefs, and desires affect visual experience? This question has long been controversial in cognitive science. There exists extensive literature documenting motivational effects on perceptual reports, but these findings could reflect biases in what people report seeing rather than what they see. Here, we propose that examining the underlying neurocomputational processes can provide new perspectives on this longstanding debate. We review evidence suggesting that motivation biases both perception and action, but does so via distinct neural systems: amygdala and locus coeruleus (LC)-norepinephrine (NE) activity enhances sensory representations for desirable stimuli, while striatal dopamine biases action selection toward goal-congruent actions. The neurocomputational approach provides a framework to advance a mechanistic understanding of motivated seeing and how these biases are shaped by context.