Abstract
Boundary cues are crucial for defining the spatial constraints of an environment. A recent neuroimaging study showed that the parahippocampal place area (PPA) is sensitive to the presence of vertical boundary structure in scenes (Ferrara & Park, 2014). On the other hand, developmental research suggests that not all environmental structures can serve as effective spatial boundaries: four-year-old children can geometrically reorient by a continuous curb that is only 2cm in vertical height, but fail to do so by tall, isolated columns connected together by a string (Lee & Spelke, 2011). Such results call attention to the potential importance of the horizontal continuity of a spatial boundary, yet little is known about whether scene-selective brain regions are sensitive to spatial boundaries that vary in this respect. In the current study, we manipulated the horizontal continuity of boundaries by varying the amount of poles included in artificial scene images. There were 6 conditions (see Supplemental Figure): no boundary (0 poles, open field), 3 poles, 9 poles, 17 poles, 33 poles, and 65 poles (wall). Participants (N=6) viewed these conditions in blocks of 12 seconds while performing a one-back repetition task. We measured univariate and multivoxel pattern activity in the PPA, a region known to respond to scene geometry. Preliminary results suggest that, contrary to its acute sensitivity to the vertical height of boundaries, the PPA is relatively insensitive to horizontal solidity of a boundary. Both univariate and multivoxel activity suggest that the PPA is sensitive to the presence of vertical poles in a scene (e.g., open field vs. a scene with 3 poles (p=.029)), but not to variations in the number of poles (3 poles – 65 poles), despite the striking low-level visual differences across these conditions. These results provide insight to which specific geometric features constitute the PPA's representation of scene geometry.
Meeting abstract presented at VSS 2016