Can Illusory Contours be Formed from a Coincidental Alignment of Moving Inducers

It is believed that illusory contours help us distinguish object boundaries as they are sometimes difficult to discern. An example of this is the difficulty with local ambiguity. The generic viewpoint assumption states that we can typically only see illusory contours if the occlude is viewed from a generic viewpoint. The occluding object has to be plausibly occluding the other objects. The illusory contour typically cannot be seen if the viewpoint is an accidental one.

A question that I am attempting to answer is whether illusory contours can be formed from a coincidental alignment of moving inducers. To do this, I took parts of traditional illusory contours and set them far apart. From there, I have them move around and eventually create the original illusory contour. The question is whether the illusory contour is still visible, or if it is harder or easier to detect.

Triangle Illusory Contour


Square Illusory Contour


Albert, M. K., & Hoffman, D. D. (2000). The generic-viewpoint assumption and illusory contours. Perception, 29(3), 303–312.

Comments


Michael Anoke

c) The responses of the participants do help answer my question. It seems as though the contours were still perceivable even though the inducers were moving in both cases.

d) I believe that the participants provided excellent feedback for how the experiment could be improved to better answer the question. Perhaps if the movements were more random than they currently are so it would seem more "coincidental", it would more accurately fit the question I initially posed.

Emily Huang

a) For both gifs, I saw the illusory contour when the objects lined up perfectly, then afterwards they split into separate objects. The square one was a little harder for me to see, maybe because it isn't as clean as the triangle one.
b) I think this experiment answers the question sufficiently - it shows that the illusory contours can be seen when aligned exactly. However, I think it would be a more robust effect if the trajectories were less fixed and more random, because as Ben mentioned, aligning them perfectly might not be considered "coincidental"

Maddie C

I see a very strong illusory contour in both of these, just for a brief moment!

Maria Fernanda De La Torre

a) In both examples, I see the illusory contours in the exact moment all component parts align. After that, I see other non-typical geometric shapes emerge as the parts move away from each other. When they reach their "apart" final location, I cognitively recognize them as their own parts. It is very interesting because for me, the presented moving behavior gives me the impression that each of the parts are agents of their own, or just whole objects and not parts of a composed object (specially in the first example). I would suspect that this interpretation would make slow down the appearance of the illusory contours and they do take slightly longer for me than in the static case. Once these illusory contours appear and I treat it as a composed object instead of three or four different objects, the opposite motion (away from each other) leads me to wanting to find other illusory contours from their non-aligned interaction.
b) This experiment answers the posed question, it investigates what happens when moving inducers fall into positions that create illusory controls in a seemingly coincidental manner. It shows that although the observer might find the resulting contours (triangle or square) less likely, they still appear but for some observers, they take slightly longer to appear.

Ben Radovitzky

a) I noticed that although it definitely works for the split second that the shapes line up, I feel like if it was a little bit longer I would be able to see the effect with a lot more ease. Still, it certainly works as intended.
b) This demo definitely answered the question to some degree. I don't know if I agree with using the term "coincidental" when describing the alignment, since it clearly isn't in this case. However, it is interesting that at the exact moment of the alignment, the effect is especially strong.
Now, pretend you are the researcher and assess whether your experience of the demo was helpful in answering the stated question.