Why do wheels appear to rotate backwards?

[Colophon]

Here, courtesy of the New Scientist, is a possible answer. I just went to find it online, as I remember reading it in the magazine a few years back. It relates to railway sleepers seen from a train, but I believe it’s relevant. Basically, tiny movements of your eyes might “follow” the motion of each spoke, making it appear stationary.

http://www.newscientist.com/lastword/article.jsp?id=lw96

Question
If you are travelling in a train at about 80 kilometres per hour and you gaze out of the window at the adjacent railway track about 6 metres to the side, the sleepers of the adjacent track can appear stationary. If both sunlight and eyesight are continuous what causes this stroboscopic effect?

Answer
If the train was stationary and you looked at the railway sleepers, their image on your retina would be virtually stationary and you would then see them as stationary. When you are in a fast-moving train and look at the railway sleepers, their image on the retina is swept swiftly across the retina’s surface. You perceive fast movement in the scene outside the train. In this case, where the eyes are looking is constantly changing as we switch from one scene of interest to another (these shifts are called saccades). These saccades are very swift, and if you happen to make a saccade in the opposite direction to which the train is moving, then the swiftly moving retinal image of the sleepers will be momentarily held stationary. Consequently, the sleepers will also seem to be stationary.

David Smith , Phsyiological Sciences University of Newcastle upon Tyne

[CurtC]

After this thread came up back in Febrary, I was repairing a bike tire and I spun the wheel in the sunlight, and I noticed the strobe effect. I meant to post the experience after I did it, but forgot. The effect happened when I was shifting my eyes from one side to the other, so they briefly followed the path of the spoke. It’s related to the “frito effect.”

It reminds me of an exhibit in the Exploratorium in San Francisco. On the wall is a vertical column of LEDs. While I was there, I gradually became aware that I was seeing the shape of an eye in that general direction, but it took a while to figure out what was going on. The LEDs were changing in such a way that if you quickly moved it sideways, it would trace out the shape of an eye. But it wasn’t moving - it was relying on your eye to scan across and generate the image.

[li_l_Dickie_Dirtz]

My son has a toy that has a cylinder with alternating black and white longitudinal stripes. When the cylinder is spun fast enough, the alternating stripes appear to merge into a solid gray color. As the cylinder slows down and the stripes start to become distinguishable, there is a period when they seem to go backwards, against rotation, from my observations and in a sunlit (not artificially lit) room. Does that count? I assumed it had something to do with the way the brain process visual info, as Cecil said in his column.

[Don_Roberto]

In the paper I referred to above (Purves, et al), they consider the possibility of saccades (quick eye movements) generating the illusion. In at least one instance, however, eye movements are clearly not related–when polarized light is beamed into the eye, a propeller-like image can be perceived. This is an “intraoptic” image. The propeller rotates along with the polarization, and counter-rotation can be experienced through this effect. Since the image only exists inside the eye, this rules out eye movements as the sole cause of the illusion.

[chorpler]

The PNAS article by Purves et al. is available online to non-subscribers. Very interesting article:

3693.full.pdf