Intention

I chose the spinning dancer illusion, which, seen from two different perspectives, can give the illusion that the dancer is spinning in two different directions (not at once). We can see the dancer in different ways as a result of our bistable perception, so this illusion is thus in the family of illusions that includes the Necker cube and the face-vase. We have two optic nerves, one in each eye, which meet up at the optic chiasma. Any information about the left field of your view goes to the right hemisphere of your brain, and any information about the right field of your view goes to the left hemisphere of your brain. 

In the natural world, there are usually enough context clues around the object that the object will convey only one image to each person. However, with the silhouette, those context clues are eliminated, and the brain has to make a huge number of assumptions to produce an image that makes sense to the person.

With the dancer, because there are two valid contradictory views, the brain just picks one at any given time. Sometimes it'll pick the dancer to spin counterclockwise. Sometimes clockwise. As I wrote this paragraph, the dancer flipped 3 times for me.

The counterpart of this illusion in sound is the tritone paradox, where tritones are involved. When two notes related by a half octave are played in succession, some people will hear the notes ascending while others will hear them descending.

Examples in the world

Here are some accidental photographs taken from angles that coincidentally make two perspectives possible. Although these do not involve silhouettes, the angles eliminate some of the necessary context clues to arrive at only one visual image, so they unwittingly create the same effect as the spinning dancer. At first glance, you see something, but then realize that there are actually two ways to see it. 

 However, in this case, there are enough context clues to tell us when we see something wrong. Some people will immediately see the right version, while others will see the little visual trick but then do a double take because the information clashes with what they have learned about the world.

Wow, that dude sure is hugging that girl really tightly with such a bored expressio-- oh, wait, it's the other way around. 

If this picture were taken from the front, it probably wouldn't have made for such an interesting image. 

While it may seem that this young boy is trying to yank off the head of a tiny adult on the beach, he is, in fact, eating Doritos. This is a lucky angle, like how some people like to pose in front of the Eiffel tower and have their image taken so that it seems like they're pinching the top. Or how some people want their picture with the moon but at an angle where they are holding it. While those are on purpose, this image seems a lot more accidental.

Because the woman on the right chose to wear a black shirt and the woman in the middle chose to wear a black skirt, we have the accidental merging of the two women. My first reaction was, "Why is the sash around two bodies? This doesn't make any sense--now there's a pair of legs that doesn't belong to anyone." Because the woman on the right decided to lean out so far, it seems as though her torso has merged with the other woman's legs.

Examples in Media/Art

A classic example of media that used silhouettes to create a game out of perceiving and filling in context clues would be Who's That Pokemon. Although most of them are pretty obvious, I still got a bunch of them wrong when I was younger. It depends on how good you are at recognizing shapes and filling in the rest.

This isn't as easy as it seems. Here's an example of someone who became frustrated.

Ah, yes, who could forget about Jigglypuff? Especially a top down view? This game was frustrating because the silhouettes could represent anything. But they were fun because you had to fill in the rest of the information and make a decision because the game always had a correct answer.

This work cleverly uses literal perspective. Although it isn't one created by the brain but rather by the work, I think it still captures the essence of this illusion. Two perspectives (whether it's created by the eye or the artist) create two different meanings. 

Reflection

I've known about the silhouette illusion for an extremely long time but had no idea that it was really just a matter of perception. I've always believed that it was about whether people are left- or right-brained and the brain will just make an assumption and run with it (definitely doesn't explain why the rotation switches though). Instead, with this project, I've found that given too little information, the brain just does a random picking algorithm...but always. So you see something sometimes and another thing other times. It's pretty neat. I also didn't know that this was in the same family of illusions as the face vase and the Necker cube. I thought that the examples I found would all have to do with silhouettes. I was incorrect.

Now that I know about bistable perception, I can create purposefully ambiguous but recognizable artwork and know that the human brain will automatically fill in the gaps to create an image that can be successfully categorized. 

Sources

http://psychology.about.com/od/sensationandperception/ig/Optical-Illusions/spinning-dancer-illusion.htm#step-heading

http://greengabbro.net/2007/10/20/the-spinning-dancer-and-the-brain/

http://www.worldwideinterweb.com/item/2314-the-50-funniest-double-take-photos-of-all-time.html

http://philomel.com/musical_illusions/tritone.php

http://hfboards.hockeysfuture.com/showthread.php?t=1728283

http://pokemon.wikia.com/wiki/IL013:_Mystery_At_The_Lighthouse

http://www.moillusions.com/yes-or-no-make-up-your-mind/