Abstract
Coordinate reference systems have been essential conceptual tools for understanding how space is represented in the brain. For instance, the representation of space on the retina is commonly construed as a polar coordinate system with the origin centered on the fovea and with eccentricity (distance) and polar angle (direction) representing two orthogonal, separable dimensions of space. However, although the concept of a coordinate reference system has been essential to understanding how encoded space is defined, the concept leads to an important theoretical prediction that has generally gone untested: specifically, that location is represented in terms of at least two perceptually separable and potentially independent dimensions. The present study addressed the extent to which luminance onsets are encoded as separable vs. integral spatial dimensions using general recognition theory (Ashby & Townsend, 1986). A total of 15 observers performed a complete identification task, in which distance and direction were factorially manipulated at three levels of each dimension (see Figure 1 in the supplemental materials for a description of the stimulus displays), and which required a nine-alternative response task that allowed for direct simultaneous assessment of each observer's perception of both stimulus dimensions on each trial. A sufficient number of trials were run to allow for distributional analyses of response frequencies and latencies at the individual subject level. Response frequencies and latencies were analyzed with respect to marginal response invariance and report independence (Townsend, Houpt, & Silbert, 2012), and the results were used to fit multivariate Gaussian discrimination models. The results indicated that direction and distance were not encoded as separable sources of information, either perceptually or decisionally, with there being evidence for a lack of independence across dimensions. These findings suggest that coordinate reference systems are not appropriate conceptual tools for understanding the composition of spatial representations of the visual world.
Meeting abstract presented at VSS 2018