A face inversion effect without a face

Abstract

Numerous studies have attributed the face inversion effect (FIE) to configural processing of internal facial features in upright but not inverted faces. Recent findings suggest that face mechanisms can be activated by faceless stimuli presented in the context of a body. Here we asked whether faceless stimuli with or without body context may induce an inversion effect as large as the FIE. In Study 1 participants performed a sequential matching task for upright and inverted faces, faceless heads with full, minimal or no body context, headless bodies and bodies viewed from the back. Results show inversion effects as large as the FIE for faceless heads with full or minimal body context, but not for faceless heads without body context, headless bodies or bodies viewed from the back. These findings remarkably show that in contrast to the well-established configural explanation for the FIE, the FIE does not necessarily depend on the processing of internal facial features, but can be also triggered for faceless stimuli presented in body context. In Study 2 participants rated the extent to which they detected a face in stimuli presented with or without faces briefly followed by a mask. We found that faceless stimuli that generated a large inversion effect were rated higher for the existence of a face than faceless stimuli that generated small or no inversion effects. These findings further suggest that the FIE can be generated by a contextually induced face percept at the face detection stage rather than the face identifi-