Abstract
Face recognition relies on holistic visual processing in the right hemisphere and therefore exhibits a left visual hemifield (LVF) half-face bias. We hypothesized that hemispheric competition also contributes to this bias, independently of holistic processing. To test this hypothesis, we performed several face recognition experiments using chimeric faces comprised of hemifield-split half faces. Observers performed a matching task, in which they viewed a target face and then selected a match from two alternatives, one of which differed from the target in either the right or the left half. The first experiment tested whether or not the LVF half-face bias only occurs in the context of a whole face or also applies to half-faces viewed in isolation, in either visual field. Consistent with our hypothesis, we found that the LVF half-face bias only occurs for whole faces viewed with each half in an opposite hemifield. Additional experiments examined the effect of face inversion, which disrupts holistic face processing. While we failed to observe the LVF half-face bias for inverted faces—as predicted by a right hemisphere holistic face processing account of the bias—we nevertheless observed strong hemifield biases. The direction of these biases (i.e., favoring the LVF or RVF) varied across individuals, and therefore did not produce an overall LVF half-face bias for inverted faces. Surprisingly, however, the magnitude of these biases (for inverted faces) correlated with the magnitude of the LVF half-face bias observed for upright faces. This means that the LVF half-face bias for upright faces is not solely due to right hemisphere dominance in holistic face processing. Instead, we propose that it is due to the combined effects of holistic face processing and hemispheric competition during the visual recognition of centrally viewed faces.
Meeting abstract presented at VSS 2018