The next research seminar is on Wednesday March 2nd, where Anna Farennikova will be speaking on ‘Unexpected Perception’. Usual time and place of 4.15-5.45pm in Room 0.02, John Percival Building.
According to Bayesian approaches to perception, presence of bias optimizes perception. This raises a question about the cognitive status of perception of the unexpected. Perception of the unexpected occur when we encounter novel or atypical events. Because this form of perception is a result of invalid expectations, it might be treated as suboptimal: it decreases accuracy and amplifies uncertainty. I argue that we need to rethink the notion of optimality for the experiences of unexpected. Focusing on two forms of perception of the unexpected – experiences of change (noticing a new building on the way to work) and of absence (seeing an elephant vanish in a circus trick) – I explain the sense in which both can be understood as involving optimal decisions. I then explain why optimization is harder to achieve for the perception of absence than it is for the perception of change.