Neural Correlates of Familiar Face Versus Person Recognition: Evidence from self-, associative and repetition priming studies

Abstract

Recognizing familiar faces is a highly efficient process that involves a hierarchy of representations, consisting of structural, long-term visual and semantic (person-related) information. These are reflected in event-related potential (ERP) components: the N250 is associated with access to domain-specific, long-term visual representations of faces, while the subsequent Sustained Familiarity Effect (SFE) is hypothesised to reflect domain-general stages of familiarity processing. However, the properties of these representations remain poorly understood.

This dissertation used self-, associative- and repetition- priming to investigate whether the SFE reflects perceptual familiarity with a face, access to post-perceptual representations of familiar people, or the prediction of a response to the target (familiar/unfamiliar). Consequently, if the SFE reflects perceptual familiarity, it should increase after repeated face exposure. If it indexes access to domain-general person knowledge, it should be enhanced by relevant associative cues (e.g. names). To test this, personally familiar and unfamiliar faces were preceded by: (1) the name of the same or a different person (self-priming), (2) an associated or unrelated name (associative priming), (3) the same or a different face (repetition priming).

Results showed that the N250 is sensitive to within- and cross-domain priming, indicating an earlier shift towards modality-independent processing than previously assumed. Crucially, the SFE was enhanced by self- and associative priming, suggesting that activating person-specific knowledge can facilitate access to domain-general representations. The detected familiarity effect was additionally influenced by the prediction of an upcoming familiar or unfamiliar target. Together, these findings confirm that the SFE reflects post-perceptual processing of contextual and identity-specific associative cues rather than visual recognition of a face. Importantly, priming revealed that context-based expectations and person knowledge influence familiar face recognition, offering new insight into how we dynamically integrate information in real-world contexts – a process essential for successful interactions that has rarely been addressed by face recognition research.