Abstract
It has consistently been shown that prolonged exposure to a face can affect subsequent perception of other faces. For example, prolonged exposure to a Chinese face can result in a face being rated as Caucasian that was previously rated as neither Chinese nor Caucasian but somewhere in between. This and similar face after-effects have been taken as evidence that individual faces are encoded as a deviation from the norm of the population of faces to which the perceiver has previously been exposed. It is thought that this adaptability ensures that we are most sensitive to variations within the range of faces by which we are currently surrounded. If this is the case, adapting to an average Chinese face should allow for better discrimination between other Chinese faces. Participants learnt to discriminate between either four Chinese or four Caucasian individuals. They were then adapted to either an average Chinese or Caucasian face for 5mins. Following adaptation, participants completed a forced-choice task asking them to identify the individuals previously learnt from test faces with identity strength ranging between 5 and 50% of the individual identity morphed with the corresponding racial average. Discrimination thresholds within a group of faces were lower after adapting to an average face from the same ethnicity than after adapting to an average face from the other ethnicity. A functional benefit can be gained by encoding faces in an adaptive norm-based manner whereby an observer's current diet of faces will be the ones to which they are most sensitive.