Understanding how unfamiliar faces become familiar

For most people, visual recognition of familiar faces is excellent and seems effortless, but recognition of unfamiliar faces is often poor. But how does an unfamiliar face become familiar? Seven behavioural and two event-related brain potential (ERP) experiments were carried-out to investigate the perceptual encoding process and subsequent recognition ability of same or other views when single-views or two-views had been learned. By systematically changing the types of views to be learned and tested, results from the behavioural experiments revealed that when two-views were accessed during recognition, integration and summation between these views and the information each view type afforded (i.e., its 'view type utility') directly influenced recognition performance of a novel view. ERP experimental findings further suggested that the FN400 'familiarity' ERP component found during learning represented access to an established representation in memory, and in the recognition phase represented an approaching significant marker of 'familiarity', but only when two-views had been learned. This suggested that the FN400 two-view recognition effect, which was not present for single-views, represented access to a memorial representation that was qualitatively different from that of single-views. Taken together, behavioural and ERP results indicated that face learning occurred through the encoding of all visual information available at the time, and that learning more than one view imparted an advantage when tested on a novel view that was based on 'view type utility'. Furthermore, the FN400 memorial representation for two-views may represent an association in memory that occurs due to within-identity variation between the two-views learned.