Researchers discover unexpected limit on human creativity

Humans can juggle as many as 10 balls at a time. But how many can they navigate through the air using their imagination?

The answer, released last month in Nature Communications, surprised even the investigators examining the topic. The cognitive psychologists discovered that individuals could easily visualize the path of one ball after it vanished. Yet, the imagination struggled to concurrently track two objects that fell out of sight.

“We aimed to probe the limits of imaginative capacity, and we discovered that it was one,” stated co-author Tomer D. Ullman, associate professor in the Department of Psychology. “I found this unexpected, so I understand if others do, as well.”

Ullman, who directs Harvard’s Computation, Cognition, and Development lab, has a longstanding fascination with what is referred to as intuitive physics. Consider the brain imagining a ball rolling down a slope, or alerting to two objects on a collision course.

“How do we engage with the physical reality surrounding us?” pondered Ullman, who is also associated with the Kempner Institute for the Study of Natural and Artificial Intelligence. “I adhere to the notion that the brain could be running mental simulations, similar to a video game.”

These could not be flawless representations of physical settings, down to the atomic and molecular scale. Thus, Ullman’s laboratory has endeavored to comprehend what methods and shortcuts enable mental simulations to occur.

“The human imagination is truly fascinating, and we find that numerous individuals are quite intrigued by its workings,” he remarked.

A large collection of studies has examined the capacity limits of human perception or the number of objects the brain can monitor in a visual scene. “Perhaps you’re a parent observing several children, or you’re a lifeguard on duty,” Ullman mentioned. “Clearly, you cannot track everything.”

Neuroscientists, psychologists, and computational modelers have determined that visual tracking is restricted to a few moving objects. However, few have investigated the limitations of imagination.

In the recent investigation, online participants were shown an animation of a bouncing ball, as if within a racquetball court, before it disappeared. Others viewed two balls bouncing at entirely different rhythms before both vanished. Leading the experiment alongside Ullman was Halely Balaban, an assistant professor of cognitive psychology at the Open University of Israel.

Two computational models were also created to explain how the imagination might follow these unseen balls until the moment of impact. The first model suggested that multiple objects could be tracked simultaneously, while the second proposed engaging with each ball in a more serial manner.

Ullman and Balaban discovered that their online participants were quite adept at predicting when a single unseen ball would land on the ground. However, individuals struggled to keep track of two.

“It was more challenging than any of us anticipated,” noted Ullman, observing how reliably the task elicited laughter.

Based on previous research, the co-authors initially theorized that the mind could likely track around three or four items. There were intuitive reasons to believe that the mind’s eye could maneuver several items simultaneously.

“If I close my eyes right now, I can visualize a tower of blocks tumbling down,” Ullman remarked. “It doesn’t feel restricted. People believe they should be able to manipulate more than one.”

In fact, a subsequent experiment indicated that individuals performed somewhat better tracking two balls that moved in sync before disappearing. Yet their results still lagged compared to a follow-up, where study participants monitored two balls that remained visible until impact.

When addressing the task of following objects that have vanished, the researchers discovered that the human imagination primarily relies on a serial model, processing each piece sequentially.

A different follow-up investigated whether participants might conserve cognitive resources by adopting a serial model. After all, executing a simulation through the parallel model would necessitate greater effort. Picture a computer managing multiple simulations simultaneously.

“We offered participants a significant monetary incentive if they could succeed in this task,” Ullman elucidated. “However, this didn’t seem to affect the results.”

For Ullman, the findings unveil an exhilarating frontier. “There has been extensive research on how the mind employs clever strategies to monitor what’s in front of you,” he stated. “However, there’s been relatively little work on the tricks and constraints of the mind’s eye. I envision a great deal more exploration to be undertaken in this area.”