It’s a reliable movie trope: Our heroes are lost in the woods, and in their valiant effort to make a beeline out of the forest or back to camp or civilization, they inevitably get turned around and wind up back at the same spot where they began.
When a science television show approached Jan Souman, a research scientist at the Max Planck Institute for Biological Cybernetics, with a viewer’s question about the phenomenon, Souman wasn't sure if people actually did accidentally circle back. When lost, he thought, people would probably veer left or right randomly, but not actually circle back.
To find out, he and his research group gathered nine volunteers and stuck six of them in a German forest and the other three in the Tunisian desert. All of them were instructed to walk as straight as possible in one direction for several hours while wearing GPS receivers so that the researchers could analyze their routes.
Souman found that all of them eventually veered off course, and more than half did end up circling back to cross their own paths without realizing it. There was an interesting twist, though. The circling only happened with the four of the forest walkers who made their journeys in overcast conditions and the one desert walker who walked on a night with no visible moon. Those who could see the sun or moon managed to travel in straighter lines and, when they did lose their way, moved as Souman had predicted, veering left and right while generally going in the same direction and not crossing back on their route.
In a second experiment, the researchers had 15 volunteers try to walk in a straight line for an hour while blindfolded. When they couldn't see at all, the walkers circled back sooner, more often, and in tighter arcs, sometimes making a circle about the size of a basketball court.
The two experiments cast doubt on an older idea that this kind of disorientation comes from biomechanical asymmetries—like a differences in length or strength between the left and right legs—that create small but consistent directional bias. That would cause a person to consistently veer off in the same direction, especially when the person is blindfolded and without visual cues to compete with the bias. But only three of the walkers had a tendency to veer in one direction, while the others varied wildly in their circling, with their paths looking like a child had scribbled on a piece of paper. Walking in circles, Souman and his team think, isn’t caused by some physical bias, but an uncertainty about where straight ahead lies that increases over time.
For the walkers in the first experiment, visual cues appeared to be very important. Those who could see some external reference point—the sun, the moon, a hill in the far distance—could use it to recalibrate their sense of direction and maintain a relatively straight path. (Interestingly, Souman notes that the volunteers in the first experiment walked for several hours, during which the sun moved about 50 to 60 degrees; rather than following a correspondingly bent path, they were able to correct for this, even if subconsciously.)
The volunteers who walked when it was cloudy or dark or while they were blindfolded didn’t have this luxury and walked in circles. Without a reference point to maintain their course, these subjects had to rely on other cues, like sounds and information from the vestibular system, which aids in movement, balance and spatial orientation. Small random mistakes in the processing of these cues, Souman and team think, add up over time, especially when the senses are limited. Eventually, the internal compass fails and “random changes in the subjective sense of straight ahead” lead a person off the straight and narrow path and right back where they started from.