Abstract
Last year we showed that the orientation tuning function in the presence of a superimposed mask was cue-specific. When the mask varied in orientation with respect to the test, an orientation cue narrowed the tuning function, while a location cue reduced the amplitude of the masking effect over the entire tuning function. In this study we explore the effect of cueing the location or the orientation of a test patch in the presence of a pair of masks that vary in location with respect to the test. The test was a Gabor patch presented in one of two locations, symmetric around fixation, and in one of two orthogonal orientations, vertical or horizontal. Each potential test location was masked by two 1D noise patches (one for each possible orientation), whose distance from the test location was varied to obtain a masking function for location. In the first experiment, the test was 7 degrees away from fixation and the mask components moved away along an iso-eccentric circle. As in case of the tuning function for orientation the masking effect was cue-specific. The effect of the location cue on the tuning function for location was wide and unspecific but the orientation cue improved thresholds over a narrow range of mask locations around the test. We wondered whether the failure to obtain location specificity with the location cue was due to the large distance between the two possible locations rather than to a signature of the location cue. So, in a second experiment we moved the two locations to only 1 degree of eccentricity, reducing the separation between the two locations to 4 lambda. At this reduced separation, the orientation cue continued to be effective over a narrow range of mask locations and the location cue showed the same broad tuning as before. The specific narrowing with the orientation cue clearly indicates that cueing the feature and cueing the location cause two distinct effects on the tuning of the underlying detectors.