Abstract
We examined enumeration in the periphery both in the presence and in the absence of distracters. 1 – 9 black circles (0.5 deg diameter) were presented on each trial, placed at least 1 degree apart within a 5 × 5 deg region centered at 7.5 deg eccentricity. There were three conditions: no-distracters, White X distracters, and Black Square distracters. In the two distracter conditions a total of 4 distracters, each measuring 0.6 degrees, were placed one on each side of the grid at three possible distances (0.35 – 2.9 degrees from the edge of the grid). Subjects reported the number of targets (black circles) and when reaction times were plotted as a function of number of objects, separate ranges for subitization (rapid and precise quantification, average slope=55ms/item for 1 – 3 objects) and counting (slow, serial enumeration, slope=429ms/item for 5 – 8 objects) were discernable in the no-distracter condition. Both of these slopes are significantly steeper than those over the same ranges with foveal presentation, suggesting some modification of the subitizing and counting strategies in the periphery when spacing may be too close to allow all targets to be individuated. Distracters also had a significant effect: black, but not white, distracters significantly increased both error rates and reaction times of responses especially at their closest spacing. Further analysis showed that black distracters specifically interfered with the enumeration of small numbers (slope=162ms/item), suggesting that subjects were counting and not subitizing even at the lower end of the number scale. However, slopes for enumerating large numbers were unchanged. Consequently, our results suggest that the subitization process can discriminate black targets from white distracters but cannot separate targets and black distracters in the periphery, requiring a switch to the slower counting strategy to avoid errors.