This story began, like so many in the history of science, with something we weren't looking for.
In 2020, during the COVID-19 pandemic, we were studying a practical question: how do people move in a closed space when they are asked to maintain a distance from each other? Watching the images repeatedly, we noticed something strange: almost all participants were walking in circles that rotated in the same direction, always counterclockwise. Even more strangely, none of them seemed to have noticed that they were doing this.
This observationled us on a five-year journey to discover why humans behave in this way.
The effect itself is easy to describe. Have a crowd circulate within a closed space without a specific destination, and they will gradually begin to rotate, almost always to the left. No one directs this movement. No one decides this.
For a physicist, this seems a classic example of "emergent behavior": a collective pattern that arises from countless small interactions, whether between individuals or between individuals and their surrounding environment. These were our two main hypotheses, which we attempted to validate independently.
It wasn't the room.
Our first suspicion revolved around the interaction between pedestrians and the walls of the room. Aprevious neuroscience studyWe had observed that right-handed people tend to turn towards the left, and left-handed people, towards the right – a subtle bias linked to small differences between the two hemispheres of the brain. Given that most of our original group were right-handed, it was reasonable to assume that the collective spiral movement did not go beyond the sum of individual preferences, transformed into a circle by the surrounding walls.
To test this, we designed a new experiment that began even before the crowd gathered in the arena. One volunteer at a time walked in a straight line towards a wall, and we recorded which side they turned to when they arrived.
Knowing each person's preference, we could then classify people based on this characteristic and observe the result. If the walls and handedness were the key, a group of left-handed people, or even better, people who turned exclusively to the right, would eventually reverse the rotation and move clockwise. This never happened. Whether we filled the room with people who turned to the right or left, the tendency to rotate counterclockwise remained.
Perhaps the presence of the walls was behind this orientational effect. To find out, we took the experiment outdoors, in a large, open school courtyard of 3,000 square meters, without defining walls, and let over a hundred teenagers walk without any guidance. The rotational effect not only persisted, but intensified. The room, of course, was not responsible.
It was not the culture
This leaves us with our second suspect: the way we interact with each other. When two pedestrians meet head-on, each one veers to the side, leading to the well-knownphenomenon of lane formation.
But the side we choose is not random and is usually determined by the side of the road where people drive. In Spain, as in the rest of continental Europe - and in Brazil -, we veer to the right, leaving some space to our left for the approaching person to do the same.
Perhaps, we reasoned, the origin of the movement counter-clockwise was nothing more than a manifestation of this phenomenon. If pedestrians tend to move to the right with each interaction, they will end up walking with the wall to their right, moving counter-clockwise. If this were true, a Japanese crowd would be rotating clockwise because, in Japan, people drive and move to the left, and lane formation occurs on the opposite side of ours.
We contacted a colleague in Tokyo, convinced that he would observe the crowd rotating clockwise. New experiments resulted in another failure. The Japanese crowd rotated counterclockwise, so our hypothesis was wrong.
An explanation remained culturally: a habit acquired at some point in life, such as the counterclockwise direction in which almost all athletic tracks are built. So, we turned to preschool children, who had just begun to absorb adult conventions. Far from disappearing, the tendency of them to rotate counterclockwise was the strongest we had measured anywhere.
The answer is within us
If it is not the group, the environment, or our social habits, there was only one place left: a purely individual characteristic, something biomechanical or ingrained in the brain.
To test this, we asked over 200 volunteers to walk alone, one at a time, in an empty arena. Without anyone to collide with and without a crowd to follow, a clear majority still moved counter-clockwise.
For decades, our field of study has assumed that collective patterns like this emerge from the interactions between pedestrians. In this case, the data show the exact opposite: the pattern is already present in a single person walking alone. Indeed, by simply inserting these individual movement data into a model, we were able to reproduce the swirling motion of the entire crowd. In this case, and unlike emergent collective phenomena, the whole is not greater than the sum of its parts; it is exactly the same.
And here is the honest conclusion: we still don't know the cause of this bias at the individual level. We have looked for a relationship with handedness, dominant foot, dominant eye, sex, etc., and have found nothing. Studies with people walking with their eyes covered, which tend to veer in a spiral to one side, suggest something deeper, perhaps subtle asymmetries in the inner ear that minimally alter our perception of a straight line. But, at the moment, we cannot offer any explanation for this.
In any case, knowing that this tendency is individual and not collective is still valuable and can change the way we model and design the spaces we share: places like airports, train stations, and museums, where the harmonious flow of people is important.
The same tendency to the left also appears in animals, for example, in fish schools and in the spiral colonies of army ants. The force that pushes a single human to the left can, in the end, be much older than humanity itself.
Iñaki Echeverría Huarte receives funding from the PID2023–146422NB-I00 project, funded by MICIU/AEI/10.13039/501100011033 with funds from the EU's FEDER.
Angel Garcimartín Montero receives grants from the Spanish State Research Agency.
Iker Zuriguel Ballaz receives funding from the PID2023–146422NB-I00 project, funded by MICIU/AEI/10.13039/501100011033 with funds from the EU's FEDER.