In the following, we will show that crowding strength can weaken if more flankers are presented, crowding occurs with flankers well beyond Bouma’s window, complex features determine low level feature processing, processing

selleck chemicals is not stereotypically but necessitates a grouping stage, and, finally, information is not lost at early stages. We can uncork the bottleneck of vision simply by adding elements. First, according to pooling models, crowding strength increases if the number of flankers increases because more irrelevant information is pooled. For this reason, almost all experiments on crowding have used only single flankers neighboring the target 37• and 38•. However, already in 1979, Banks and colleagues showed that crowding is weaker when a target letter is flanked by an array of flanking letters compared to a single letter (Figure 2A, [39]). These results were forgotten for more than 25 years. Recently, we have shown when bigger is better ( Figure 2B). We presented a vernier stimulus, which consists of two vertical lines slightly offset either to the left or right. Observers indicated the offset direction. When one shorter line to the left and one to the MK-2206 ic50 right flanked the vernier, performance strongly deteriorated. Performance improved when further lines were added ( Figure 2B, red line). The same pattern of results was found for longer lines

( Figure 2B, blue line) but not for lines with Fossariinae the same length as the vernier ( Figure 2B, green line). In this case, performance stays roughly on the same level independent of the number of lines. Hence, bigger can be worse and bigger can be better 11••, 15, 16 and 41. The latter case clearly shows that vernier information is not irretrievably lost at the early stages. By adding further elements, we can ‘uncork’ the bottleneck of vision, that is, we can undo crowding. We proposed that grouping explains these results. When single shorter lines are presented they group with the vernier. However, arrays of shorter lines group with each other and do not group with the

vernier. For equal length lines, the vernier always groups with the flankers. Hence, crowding is weak when target and flankers do not group with each other. Strong crowding occurs only when target and flankers group. It may be argued that, for example, adding lines in Figure 2C, [40] simplifies the Fourier spectrum, that is, ‘the more the better’ argument does not apply. We could not find any evidence that such an approach can succeed [43]. Second, because crowding was thought to occur only by flankers presented within Bouma’s window, flankers were only presented close to the target. However, crowding extends well beyond Bouma’s window. Orientation discrimination of a letter T only slightly deteriorated when flanking Ts were presented outside Bouma’s window (Figure 3A, a–b). Crowding was also weak when a square within Bouma’s window surrounded the target (Figure 3A, a-c).