Abstract
Goals: Performance in tasks mediated by basic visual dimensions, such as contrast and spatial resolution, is not homogenous across the visual field. Specifically, performance is better at isoeccentric locations along the horizontal meridian than along the vertical meridian (Horizontal-Vertical Anisotropy, HVA), and along the lower than the upper vertical meridian (Vertical Meridian Asymmetry, VMA). Given the lower density of ganglion cells and faster decline of cone density with increasing eccentricity along the vertical than the horizontal meridian, and the greater cone and ganglion cell densities in the lower versus the upper visual field, it has been assumed that performance fields are retinotopically defined. However, ecological hypothesis stress the importance of the allocentric location of visual information. Here we investigated whether performance fields for tasks mediated by contrast sensitivity are retinotopic (eye-centered) or allocentric (display-centered) in origin.
Methods: Observers viewed briefly presented radial arrays of eight suprathreshold Gabors at four cardinal and four 45° intercardinal locations. They determined whether a target tilted relative to the homogeneously oriented distractor Gabors was present. To dissociate retinotopic coordinates from allocentric coordinates, observers performed the task both with their heads upright and tilted −45° (left). In addition, the orientation of the distractor Gabors was varied over experimental sessions, such that each participant viewed vertical (0°) distractor Gabors with tilted targets (±60°) and −45° tilted distractor Gabors with tilted targets (−105°, +15°) in both the upright and tilted head positions.
Results: Observers showed both the HVA and VMA in the upright head condition, replicating previous findings (Carrasco et. al., 2001). Critically, there was a −45° shift in both the HVA and VMA when observers' heads were tilted −45°, with performance fields corresponding to a horizontal and vertical meridian defined by retinotopic coordinates.
Grant: NIH R01 EY016200-01A2. Thanks: Stuart Fuller.