Sitting alone in a dim room in Pittsburgh, Pennsylvania, Earl flung his arm to the left. He slowed his movement down, examining the position of a cursor on the computer screen in front of him. Where his hand went, so did the cursor. Earl gestured the dot closer to a colorful target zone, just as he had done thousands of times before. This time, he expected a big reward, but instead—time’s up. Earl, a rhesus monkey, choked under the pressure. He didn’t move the dot into the target before the timer ran out.
Choking is when high stakes cause you to fail when you otherwise would have succeeded. In basketball, it’s missing free throws late in a game; during a dance recital, or spelling bee, or job interview, it’s the paradox of overanalyzing and amnesia that leaves you feeling like an alien in your own body. “You overthink it, you get too much in your head,” says Steven Chase, a biomedical engineer at Carnegie Mellon who specializes in motor learning.
And yet, while choking is a common experience, its basis in the brain remains a mystery. What are the patterns of electrical signals and brain chemicals that explain choking—and where do they occur? Researchers have proposed theories based on human behavior and brain imaging. But to eventually perform neurological tests, the kind that involve implanted electrodes, they’ve needed to first observe the phenomenon in a lab animal.
For now, they’ve got Earl—plus Nelson and Ford, two other rhesus monkeys—and a simpler test that only involves observing their motion with a camera. Chase’s team of researchers from Carnegie Mellon and the University of Pittsburgh have shown for the first time that people are not the only primates that choke under pressure. The results appear in this week’s issue of Proceedings of the National Academy of Sciences.
The researchers show that what triggers this behavior is the shot at an extraordinary prize—and their analysis offers clues as to why that might be. In the cursor-based game, the monkeys were tested on how quickly and accurately they moved a target into a box. The monkeys performed better as the reward offered to them improved: nothing for failure, and increasingly large sips of water for success. Until the jackpot—a really big swig of water. Monkeys who expected that rare and more valuable prize failed at tasks they’d normally ace.
Demonstrating choking in other species is interesting, and valuable for the field, says Sian Beilock, president of Barnard College and a cognitive scientist not involved in the study, who wrote a 2011 book on choking. “What I think it does is open up another opportunity to study it,” Beilock says. “If you can get a better sense of the underlying systems, you can start thinking about different ways to mitigate it.”
“Until this, it was just a weird thing humans did,” says Aaron Batista, a bioengineer at the University of Pittsburgh who co-led the work with Chase. But now a proposed model of choking could help researchers decode the neural signals of movement in high-stakes scenarios—for athletes using their limbs when the game is on the line or, perhaps one day, for humans using prosthetics they control with their brains.
Historically, researchers have held one of two perspectives on what causes choking. One is that it’s a uniquely human fault, emerging from superpowered minds. But if other animals choke, too, it may be a more fundamental issue in the wiring of the brain. Brains—animal and human—may fire cognitive or motor signals differently while chasing rare rewards. If a weird reward makes the brain do weird things, training and evolution may not have had the chance to “prune out” that weirdness. “So we set out to figure out which one of those it was,” Chase says.
The team designed their cursor game to be challenging for the monkeys but still simple to analyze. Motion-capture cameras tracked the monkeys’ arm motion, which controlled the dot on the screen. The game itself was the same each time. Any differences in speed, position, and accuracy, the researchers figured, could only stem from the one variable they tested: the reward.
The monkeys learned to anticipate particular rewards with visual cues on the computer screen—different colored targets corresponded to each reward. Earl and the others excelled during the training period, when they earned nothing for failing or tiny sips for succeeding. They performed a little better when the reward they thought they would get doubled or tripled. If that trend held, a rare jackpot—a drink 10 times bigger than the average reward—should have motivated even better performance. But the jackpot did the opposite. The monkeys put up far more unsuccessful runs when the huge prize was up for grabs. Earl choked on 11 of his 11 jackpot opportunities.
To find a cause, Adam Smoulder, a graduate student on the team, scrutinized what was going on with the monkeys’ arm motions during thousands of trials. Their reaction times and maximum speeds showed no clear trend. “Really the only consistency we saw was this increase in caution,” Chase says.
Imagine the monkeys’ arm gestures as a composite of two phases—a fast, initial “ballistic reach” motion to send the cursor closer to the target, followed by a slower, more precise “homing” step to land on-target. Earl, Ford, and Nelson repeatedly undershot in jackpot trials. Instead of starting as they normally would, with a fast ballistic reach that covered a lot of ground, their reach would stop short; the homing step dragged on until time ran out.
“The monkeys are choking by being overcautious,” says Batista. In humans, psychologists have linked choking to paying too close attention to your movements, a behavior called explicit monitoring. Thinking about your movements makes them slower. And he thinks that’s what’s going on; the monkeys are psyching themselves out and undershooting. “If that's not metacognition,” he says, “I don't know what it is.”
One hypothesis for why big rewards cause choking is that making precise movements depends on a “neural sweet spot” for rewards. The anticipation of a larger reward may cause neurons to release more dopamine. At the right levels, that dopamine helps keep movements sharp. But if motivation jumps, the flood of the neurotransmitter could overwhelm the brain’s communication networks. “Too little reward, we don't perform super well; too much reward, you don't perform super well,” says Chase.
The new study doesn’t pin down an exact neural cause of choking, but it sets the stage for scientists to study the neuroscience of high-stakes performance with lab animals. In future experiments, having an animal model will make it easier to use electrodes to eavesdrop on the brain’s chatter.
“Have they shown that this is the only way in which humans or animals choke? No—but it's one way,” says Beilock. A picture of the underlying systems is important, she says, because several regions could be involved, depending on the situation. Supposing those details translate to humans, it could explain how distinct brain regions cause distinct types of choking. A failed motor task would be like missing the ball; a failed cognitive task would be like forgetting your answers in a job interview. The brain regions involved in each situation could overlap, but they may also be separate and worth exploring.
Rob Gray, a sports psychologist from Arizona State University who studies how pressure affects human performance, says the monkey data looks a lot like explicit monitoring in athletes who choke. “That kind of nonfluent movement is what you expect when you try to control things consciously from the top,” he says. It’s paralysis by analysis: “You're micromanaging your body.”
To Chase, the neuroscience of choking doesn’t have to always be negative. Yes, humans sometimes choke. But if the promise of an outsized reward sends pesky electrical ripples into the brain’s fundamental motor function, then it’s remarkable that we can filter it out at all.
And now that we know that choking isn’t a human-only habit, he and Batista also would like to know if the information could be harnessed in useful ways, like in creating brain-controlled artificial limbs for people. “Issues like how emotions can affect motor control are things that prosthesis designers are going to have to think about,” he says. A prosthetic that can parse only motor signals from a mess of emotional buzzing would be a long-awaited payoff, indeed.