“School of Cognitive Sciences”
Back to Papers HomeBack to Papers of School of Cognitive Sciences
Paper IPM / Cognitive Sciences / 11234 |
|
Abstract: | |
A partially occluded figure looks narrower than a completely modally visible one. Kanizsa (Kanizsa, 1975 Italian journal of Psychology 2 187-195) found this illusion and maintained that such phenomenal shrinkage occurs because the former figure seems to complete amodally. In other words, Kanizsa ascribed this illusion to occlusion and amodal completion. Current study examine whether Kanizsa?s hypothesis can explain the observed shrinkage.
Studies on time course of perceptual completion have pointed out that amodal completion occurs within 100-200 ms after the onset of stimulus appearance and presentations shorter than this time leave the fragmented figure uncompleted (Ringach and Shapley, 1996 Vision Research 36 3037 - 3050). Based on this finding, if Kanizsa shrinkage was because of amodal completion, this illusion must be perceived stronger in presentations longer than 100 - 200 ms when completion has almost happened compared to presentations shorter than these times. We tested this prediction by estimating the illusion magnitude in different presentation times.
In each trial of the experiment, after fixation cross presentation, the occluded rectangle (the occluder rectangle was narrower than occluded one and located on the middle part of it) and unconcluded rectangle were presented simultaneously on two sides of the fixation cross for variable times (80-280 ms with steps of 40 ms in different trials) and then a mask was showed [it has been shown that masking can interrupt perceptual completion (Rauschenberger and Yantis, 2001 Nature 410 369 ? 372)]. Observers had to report which rectangle seems wider. The width of the occluded rectangle was fixed in all trials but the width of the unconcluded one in each trial was chosen randomly among five possible widths.
The results showed that observers were more likely to report the illusion in shorter presentations. This finding is at odd with the claim that Kanizsa shrinkage is due to the amodal completion of occluded figure. It might be possible to argue that mechanisms faster than amodal completion must be in operation for the perception of this illusion. Significantly weaker illusion in longer presentations might implies that a correcting mechanism developing throughout the time can weaken the illusion magnitude.
Download TeX format |
|
back to top |