The Own-Face Effect: The neurophysiological basis of self-identity
Studies using evoked response potentials (ERPs) have demonstrated that faces are distinguished from non-face objects at a relatively early point in processing (approximately 170 ms post-stimulus onset). Familiar faces, on the other hand, are not differentiated from unfamiliar faces until a much later point in time (~400 ms). While these findings argue that the early stage of face perception can be distinguished from the subsequent stage of face recognition, one limitation of the previous studies is that familiarity effects were tested with "publicly" familiar faces, such as politicians and celebrities, rather than faces that are personally familiar to the participant. In our experiments, we provided a stronger test of potential familiarity effects by recording scalp potentials of participants to two kinds of "personally" familiar faces; the participant's own face and the faces of the participant's classmates or dorm mates. Our main finding was that participants demonstrated a greater N170 to their own face compared to other personally familiar faces (e.g., classmates and dorm mates) and other unfamiliar control faces. The "own-face" effect challenges current models of face recognition by demonstrating that at an early stage of visual analysis (~170 ms), the face processing system is sensitive to aspects of facial identity.