What I know and what I see about you: the impact of social information on the spatial localization of others

In everyday life, we rarely perceive others as neutral entities; rather, we attribute social meaning to them. In this study, virtual humans, to whom we assigned a relational category (friend, stranger, enemy), could appear at various distances in the spatial context. In such situations, we simultaneously process relational categories and spatial cues (e.g., location and interpersonal distance). While spatial cognition relies on egocentric (observer-to-stimulus) and allocentric (stimulus-to-stimulus) reference frames, social cognition provides knowledge that guides expectations and interpretations of others’ behavior. However, it remains unclear how these sources of information jointly shape the egocentric and allocentric spatial encoding of others. To address this issue, participants performed a visuospatial memory task in a virtual environment, encoding the positions of two virtual humans and an object. Social relationships (strangers, friends, enemies) and interpersonal distances (personal vs. social) between virtual humans were systematically manipulated. Thirty-eight participants subsequently provided either egocentric or allocentric spatial judgments. Eye-tracking measures were collected during encoding, and gaze transitions were analysed as indicators of encoding-related attention. Overall, our findings indicate that social and spatial cues jointly shape how others are encoded in space. When these cues are clear or suggest a positive context, such as with friends, an egocentric encoding is favored. When these cues are unclear or suggest a negative context, such as with strangers or enemies, an allocentric encoding is favored.