Phi phenomenon
The term phi phenomenon is used in a narrow sense for an apparent motion that people see when two nearby lights are shown one after another, quickly. Unlike beta movement, which is seen at slower speeds, the lights do not seem to move. Instead, a soft, shadowy shape seems to jump in front of them and hide them for a moment. This shadow is almost the same color as the background.
Max Wertheimer first described this kind of apparent movement in his habilitation thesis, published in 1912. This work started Gestalt psychology.
In a broader sense, especially when the plural "phi phenomena" is used, the term covers all the apparent movements seen when two nearby lights are shown in turn. This includes beta movement, which has been thought of as the illusion of motion in cinema and animation. However, some people argue that beta movement is a long-range apparent motion, not the short-range kind seen in movies.
When Wertheimer introduced the term in 1912, he used "φ-phenomenon" for all the apparent movements in his thesis. He used "pure φ" for the movement with no object that is seen when lights flash very quickly. Some commentators say that he saved the Greek letter φ only for this pure, objectless movement.
Experimental demonstration
Wertheimer's classic experiments showed two lights or curves, one after the other, using a tachistoscope. When the short breaks between the images and the distance between them were just right, his test subjects, his colleagues Wolfgang Köhler and Kurt Koffka, saw pure "objectless" motion.
The phi phenomenon is hard to show clearly and reliably. To make it easier, psychologists in the 21st century designed a livelier setup with more than two stimuli. In this demonstration, called "Magni-phi," identical disks are arranged in a circle. One disk at a time is hidden in a quick sequence, either clockwise or counter-clockwise. This makes the shadow-like movement Wertheimer discovered easier to see. The Magni-phi demonstration still works when timing, size, brightness, number of disks, and viewing distance are changed.
With only two objects, the effect is easier to see if the interstimulus interval is set so that the times the two objects are visible overlap slightly. The gap on each side must be short, no more than about 150 milliseconds, and no more than 40% of the stimulus period.
History of research
In his 1912 thesis, Wertheimer used the symbol φ (phi) to describe a "pure movement" he saw. When two objects appeared one after another, people seemed to see movement, even though nothing really moved. His test subjects said they did not think an object had moved. They only saw movement between the two positions. Wertheimer believed this showed that movement could be seen directly. This idea helped start Gestalt psychology.
From the mid-20th century on, scientists became confused about what the phi phenomenon really was. Wertheimer's thesis was written in German, his writing style was unusual, and it did not say exactly when "pure movement" was seen. The phenomenon is also hard to repeat. Edwin Boring's history of psychology, first published in 1942, added to the confusion. It grouped the phi phenomenon with cases that had long gaps between the two signals. With such long gaps, people do not see movement at all. They only see two objects appearing one after another. This confusion probably led to the phenomenon being discovered again under other names, such as "omega motion," "afterimage motion," and "shadow motion."
Reverse phi illusion
The reverse phi illusion is a reversed form of the phi phenomenon. Two similar still pictures are shown quickly, one after the other, and they can look like they are moving. With reverse phi, the picture changes in the opposite direction from what you might expect, so the movement people see is the opposite of the real change. This illusion often uses black and white patterns. Scientists believe it is caused by changes in brightness. Our eyes and brain handle forward and reverse phi motion in the same way. It seems to work by comparing brightness from one moment to the next, point by point. Certain motion-detecting cells are needed to make this effect happen.
Phi phenomenon and beta movement
Phi phenomenon is often confused with beta movement. In 1912, Max Wertheimer, the founder of the Gestalt School of Psychology, explained the difference between them. They belong to the same general group, but they are not the same.
The difference begins in the brain. Visual information travels along two pathways. One handles position and motion, and the other handles shape and color. A moving object often affects both pathways, which can create beta movement. If an object changes position very quickly, it may instead create pure movement, like phi phenomenon. With phi phenomenon, you see motion going between two stimuli, A and B, that appear one after the other. With beta movement, you see an object actually move from position A to position B.
Hassenstein–Reichardt detector model
The Hassenstein–Reichardt detector model is thought to be the first math model to suggest that our eyes and brain notice motion by comparing changes in light from two nearby points. It is a theory about how brain cells might work together, and it can explain and predict the phi phenomenon and its reversed version. The model uses two locations and two visual inputs. When one input is seen at one location, a signal is sent to the other location. The two inputs change over time in different ways. The light change at one location is multiplied by the delayed change from the other. Then the result is subtracted to give an output. Two positive signals or two negative signals give a positive output. One positive and one negative signal give a negative output. For the phi phenomenon, the motion detector responds to a change in light at one point on the retina and connects it to a change at a nearby point after a short delay.
The Reichardt model is a more complicated version of the simplest Hassenstein–Reichardt detector model. The Fourier method is considered a linear method. The Reichardt model adds a multiplication step when the eye's responses to light changes at different places are combined. One input is delayed by a filter and then multiplied with the input from a nearby location. The input is filtered twice, once before and once after the multiplication, in a mirror-like way. This gives a second-order motion estimate. The phi phenomenon is often seen as first-order motion. However, according to this model, reversed phi could be both first-order and second-order.

Quiz time!
4 questions about Phi phenomenon
What seems to jump in front of the two lights in the phi phenomenon?
Keep exploring
This article is a child-friendly adaptation of the Wikipedia article on Phi phenomenon, available under CC BY-SA 4.0.
Images from Wikimedia Commons. Tap any image to view credits and license.