Have you ever played around with point-light displays? If not, take a few moments to explore the amazing site I’ve linked. Through these simple animated displays, we can detect gender, emotion, even species. Point-light displays have been studied for decades as a way to understand how we perceive biological motion. Even pigeons, quail, and cats appear to recognize animals when they are shown point-light displays.
You might think the simple fact that other animals can also recognize these displays suggests that perception of biological motion is “hard-wired” into our brains—and it is indeed a good indicator that this might be so, but other animals can also learn. It’s possible that we learn to perceive biological motion, or that we rely on higher cognitive processes to understand it. This is the “top-down” versus “bottom-up” question: does the mind first simply look at the image we see and then later try to determine what it is? Or do we determine what we are looking at early on in the visual process and then later figure out what to do with that knowledge?
Ian Thornton of the Max Planck Institute for Biological Cybernetics and Quoc C. Vuong of Brown University devised a set of experiments to try to answer that question (“Incidental Processing of Biological Motion,” Current Biology, 2004).
The test Thornton and Vuong used to help answer this question is called the “flanker-interference paradigm.” In this test, viewers are shown a display with a central image that is sometimes accompanied by two or more “flanker” images. In every case, they are asked to quickly make a judgment about the central figure (Is it a letter or a number? What color is it?, etc.). If the flankers are present, participants are supposed to simply ignore them. When the central figure has no flanker, users can do the task quite readily. When the flankers are the same as the central image, the results are similar. But when the flankers contrast with the central image, (number flankers with a letter image, or red flankers with a blue image), then both speed and accuracy of response suffer.
For Thornton and Vuong’s experiments, in every case, they showed viewers both animated and static point-light displays and asked them which direction the “person” in the central image was facing. In the first task, the central figure was surrounded by similar figures that were facing either the same direction or the opposite direction (example). The results were the same as in other flanker tasks. Next they tried moving the flankers farther away, with similar results.
In the next two experiments, they tried something new. First they created a “scrambled” figure from each of the right- and left-facing walkers. They preserved the motion of each individual dot, but moved it to a different location, so that the figure was no longer recognizable as human. In this task, they found that there was no difference in whether the “scrambled” figure was “walking” to the right or left. Then they created a “chimeric” figure (example) by combining the left- and right-facing walker into a single display. Again, there was no difference in reaction time between the display with the chimeric flankers and the display where all the flankers faced the same direction as the central figure.
So how does this all relate to the original top-down versus bottom-up question? Thornton and Vuong argue that if we used a top-down process to recognize point-light displays, then the chimeric and scrambled flankers would be just as distracting as figures facing the opposite direction: but they aren’t. So we must be recognizing the biological motion very early in the perceptual process: before we even consider the orientation of the figure.
Thornton and Vuong suggest that this finding may explain a common herding behavior where animals tend to move toward each other rather than all moving in sync: this behavior makes it more difficult for a predator to hone in on a single animal. So, paradoxically, while the early recognition of biological motion might make some things easier for predators, their prey didn’t take long to adapt!