Item

Abstract

Previous studies have identified relatively separated regions of the brain that respond strongly when participants view images of either faces, bodies or objects. The aim of this thesis was to investigate how and where in the brain shared properties of faces, bodies and objects are processed. We selected three properties that are shared by faces and bodies, shared categories (sex and weight), shared identity and shared orientation (i.e. facing direction). We also investigated one property shared by faces and objects, the tendency to process a face or object as a whole rather than by its parts, which is known as holistic processing. We hypothesized that these shared properties might be encoded separately for faces, bodies and objects in the previously defined domain-specific regions, or alternatively that they might be encoded in an overlapping or shared code in those or other regions. In all of studies in this thesis, we used fMRI to record the brain activity of participants viewing images of faces and bodies or objects that showed differences in the shared properties of interest. We then investigated the neural responses these stimuli elicited in a variety of specifically localized brain regions responsive to faces, bodies or objects, as well as across the whole-brain. Our results showed evidence for a mix of overlapping coding, shared coding and domain-specific coding, depending on the particular property and the level of abstraction of its neural coding. We found we could decode face and body categories, identities and orientations from both face- and body-responsive regions showing that these properties are encoded in overlapping brain regions. We also found that non-domain specific brain regions are involved in holistic face processing. We identified shared coding of orientation and weight in the occipital cortex and shared coding of identity in the early visual cortex, right inferior occipital cortex, right parahippocampal cortex and right superior parietal cortex, demonstrating that a variety of brain regions combine face and body information into a common code. In contrast to these findings, we found evidence that high-level visual transformations may be predominantly processed in domain-specific regions, as we could most consistently decode body categories across image-size and body identity across viewpoint from body-responsive regions. In conclusion, this thesis furthers our understanding of the neural coding of face, body and object properties and gives new insights into the functional organisation of occipitotemporal cortex.