The appearance of a new object in the visual field captures visual attention. Moreover, detection is faster for a probe presented in a region adjacent to the corner of a stimulus, compared to a probe adjacent to the straight edge. This corner enhancement effect is believed to show that probes near corners receive enhanced processing (Cole et al 2007, Attention, Perception and Psychophysics 69, 400–412). We tested the corner effect for convex and concave corners for surfaces arranged in depth. We used coloured regions with cast shadows to specify foreground and background and a square stimulus that could be perceived as either an object or a hole (a figure-ground reversal). The probe was a small red line that could appear near a corner or a straight edge 100 msec after the stimulus onset. We asked the participants to discriminate the orientation of the probe (horizontal or vertical). The corner effect was found for both convex (Experiment 1) and concave (Experiment 2) vertices but only when the probe was near the corner of the foreground surface (the pattern reversed for objects and holes). In Experiment 3 we tested a situation in which the probe was perceived as a small object not located on any surface—ie, a floating probe. The corner effect disappeared when the probe was not attached to any specific surface. In summary, the corner enhancement effect was present only when the probe was on the surface that owned the corner.
