How well are human eyes attuned to visual motion?
Quite well actually. We have two good ways to see visual motion and one weak way. After all, what perceptual skill over the last half billion years has more practical use for survival?
We have two visual cortexes, and each of them can see motion quite well. Most people have some understanding of the primary visual cortex in the back of the brain, called V1, even if they don’t know how it handles motion. V1 can see the world in two ways, and modern life mostly revolves around its static usage, reading a book, or looking at a screen or other people standing in front of you. I will describe its less common usage today, called smooth pursuit.
The other part of brain we use to sense motion is the collicular visual cortex. This is a spooky part of the brain that provides blindsight when V1 is damaged and supports our perception of motion all the time.
So we have three ways to sense visual motion. Here is summary of how they are used and the advantages each provides.
Normal vision involves holding the eyes steady in order to place the center of vision, called the fovea, on something we want to see in detail. The fovea can only see about the area of our thumbnail at normal reading distance, but its high resolution is needed to pick out the letters on a page of text. At TV watching range, say 8 feet, the fovea has a diameter of about four inches. I referred above to this type of vision as static, but in reality, while doing either of these activities, the eye is in constant motion.
Eye motion in normal vision is called a saccade. The process of normal vision is an alternation of holding the eye steady for about 100 milliseconds, turning the eye rapidly for 200 milliseconds, then again holding the eye steady for another 100 milliseconds. The turning of the eye in a saccade is the second fastest muscle movement we make (after blinking.)
The bad news for normal vision is that we are effectively blind for the 200 milliseconds of movement, meaning that in a typical second we are blind for about 60% of that second. To emphasize the point, the process involves 100 milliseconds of seeing, 200 milliseconds of blindness, followed by 100 milliseconds of seeing and 200 milliseconds of blindness. This is an oversimplification of the general case, but for the case of considering motion vision, it is a fair approximation. While normal vision can perceive motion, it is weak for that purpose, as I alluded to earlier.
What we often use for watching a moving object is smooth pursuit vision. If a bird or a ball is moving in our field of vision, we can place the fovea on it, often with a saccade, and move the eye slowly and continuously to keep the object within the foveal view. In this case, the eye movement is not rapid, and we aren’t blind to that object while the eye is moving. We can follow the movement of a bird or ball continuously, and we can recognize and appreciate the birds moving wings and head while we are following.
The downside of smooth pursuit vision is that we are effectively blind to everything else in the field of view. Actually, we can watch any number of such moving objects, with the limitations that they must all be moving at the same velocity and they must all fit into the foveal view or be very near the foveal region. (This is the basis for man visual illusions.)
When you gaze into the sky and watch a cloud or bird or airplane, you are usually using smooth pursuit vision. It is easy and fast to switch between normal vision and smooth pursuit, so that is also common. Of course, there is a limit to the speed of object that we can continue to perceive during smooth pursuit. I suspect that athletes learn to enhance that speed limit with practice.
The third type of motion vision we have comes from the collicular visual cortex (CVC). It doesn’t have a good name, so I will call it motion vision. It is often mistakenly called blindsight because we only find out about it in people who have lost normal vision as a result of damage to V1.
We’ve known about blindsight and the CVC for some time, but until last year we thought it was an adjunct to V1, acting like a yet another chained visual area like V2, V3, etc. Now we know it is a primary visual area, receiving its signals directly from the superior colliculus (SC) near the thalamus, just one step from the retina, like V1. Moreover, it is an ancient structure in evolution, and it may even have predated our normal visual processes.
I call the vision resulting from the CVC motion vision because it only responds to motion. Interviews with subjects demonstrating blindsight indicate that it provides no actionable information to the brain unless there is relative motion between the observer and one or more objects in the visual field. That is, it has no ability to recognize the shape or color or size of objects, only that “something” is moving and where the edges of motion are.
We aren’t normally aware of motion vision because there is no path from it to our consciousness. It only feeds into unconscious processes. Subjects with blindsight can’t report anything about what is in their visual field, but they can avoid moving objects, whether the object itself is moving or the subject is moving toward the object. The subject with blindsight will avoid such objects but not be able to say why.
Since the CVC has been denigrated since it was discovered, I will speculate briefly about its importance. I believe it was the first form of visual cortex, predating the large tracts of visual cortex most mammals, reptiles, and birds have. It is separated into two locations, one on each temporal lobe very near the superior colliculus on each side, quite far from one another. Since the CVC doesn’t register in conscious thought, we can presume it doesn’t participate in the predictive regime that most senses do. It provides all the most important information about the outside world that benefits survival to a primitive creature. Only later did primitive animals develop the rest of visual machinery to distinguish between different kinds of things.
What has been greatly understudied about motion vision and the CVC is that we use and depend on it all the time. It is not just a spare vision center in case something happens to V1. It has full access to retinal imagery, and it is always finding motion artifacts in the visual field. Because it is so close to the SC and isn’t involved in predictive processing, it recognizes motion much faster than our normal vision, and it does it in parallel without interfering with normal vision. Since most of our responses to the world are unconscious, we aren’t aware that the CVC is always informing us about what is going on around us, unconsciously providing grist for our cognitive prediction mill.
I’ve described three skills we have for perceiving visual motion. They complement each other and cover a wide range of motion types and speeds. Each has limitations, but we have the ability to quickly switch among them. We can also train ourselves to perceive motion better by relaxing our dependence on normal vision.
