Effects of previous experience and associated knowledge on retrieval processes of faces: an ERP investigation of newly learned faces.

Conspicuously absent from face recognition research is a direct comparison of well-known faces with newly learned faces for which the associated biographical knowledge and the perceptual expertise were experimentally manipulated. Such a comparison can test competing assumptions made by serial and interactive activation and competition (IAC) models about the role of previous experience and biographical knowledge in face recognition. We measured behavioral performance and event-related potentials (ERPs) for four classes of faces: unfamiliar faces, faces of celebrities, and two classes of experimentally familiarized faces learned one week prior to the recognition test either with or without associated biographical knowledge. Newly learned faces associated with biographical knowledge showed distinct priming and old/new effects starting around 420 ms in the ERP when compared to faces without such information, and when compared to famous faces. In addition, faces for which all biographical facts were successfully remembered showed not only faster recognition performance than faces without biographical knowledge, but also a priming effect in the N170 latency. Previous experience including biographical knowledge associated with a face thus influenced retrieval processes of faces both in late, semantic and early, pre-semantic processing stages. These findings provide novel evidence in favor of IAC models. In addition, famous faces had the same ERP effects as faces learned without associated facts. Our results show that newly learned faces are experimentally indistinguishable from famous faces yet provide superior experimental control.