Imagine that you stand at a railway switch in horror as a runaway train zooms down the track. On its current trajectory, the train will kill five people on the tracks ahead. If you throw the switch you can divert it to where only one person would be doomed. What would you do?

Now imagine that you could save those five people if you pushed one person on to the tracks under the train’s wheels. Again, what would you do?

These questions are gripping to imagine in the moment but they are also manifestations of a distinctly human nature that has been developing for hundreds of thousands of years. With a curiosity and training that spans philosophy and evolutionary biology, psychologist Fiery Cushman seeks to explain how people have come to grapple with moral dilemmas.

This summer Cushman — who employs a variety of tools ranging from classical behavioral experiments to online surveys, brain scans with functional magnetic resonance imaging, and devices that measure physiological responses to emotional and stressful situations — joined the faculty of the Department of Cognitive, Linguistic, and Psychological Sciences as an assistant professor.

“Broadly what we’ve learned in the last 10 years is that those classic moral and philosophical dilemmas happen because our brain is designed with a toolkit of a lot of different solutions to problems,” said Cushman, who conducted his undergraduate biology, and graduate and postdoctoral psychology studies at Harvard. “Sometimes you’ve got two brain systems that are trying to answer the same question but are doing it in different ways.”

For example, Cushman asks why we punish a drunk driver who falls asleep at the wheel and then hits a person much more severely than another drunk driver who falls asleep at the wheel and then hits a streetlight. Of course one does much more harm, but the results, determined by chance, derived from the equivalent moral failing of driving while intoxicated.

His research has found that there are separate psychological mechanisms at play. One determines how we judge the wrongfulness of actions. The other leads us to decide how much punishment is deserved.

“If you just ask people about wrongness, they’ll base their decisions solely on the choices that the driver made and his or her mental state,” he said. “It’s only when you ask about the punishment that suddenly you get these divergent feelings where it seems like a light punishment is warranted for the person who does no harm, but a very harsh punishment is warranted for the person who does a lot of harm.”

Given this dichotomy, Cushman was curious about whether a goal of punishment — teaching a lesson — is better served by punishing (and rewarding) the intent or the result. To test this, he devised an experiment in which people were asked to play a modified version of darts. Cushman asked them to call their shot and then to try to hit that spot. The game had rules about what spots to hit and which ones to avoid, but rather than simply explaining them with language (a modern invention from an evolutionary standpoint) Cushman wanted subjects to learn from reward and punishment. The subjects didn’t know it, but some were rewarded and punished based on their intended shots, and others on their shot results.

Preliminary results from the experiment indicate that people rated on their results learned much better than people rated on their intentions. Perhaps, then, it’s no surprise that we punish the drunk driver who kills a person more harshly than the drunk driver who smacks a lamppost.

“Is our criminal justice system reflecting moral intuitions inherited from a long evolutionary history?” he asked.

The matter of the train tragedy that can befall five people or just one, known as the “Trolley Problem,” is part of what got Cushman into this business in the first place. When he heard it in a class in high school in Washington, D.C., he was fascinated.

“I couldn’t stop talking about it for months,” he said.

At Harvard he found a like-minded group of researchers working on the Trolley Problem. Together they found that again, people make distinct choices even when that requires looking past some moral equivalences.

“You find that 90 percent of people would flip the switch and only 10 percent of people would push the man,” he said. “But at a mathematical level it’s just the same — it’s just five versus one either way.”

Happily at least one aspect of Cushman’s work is defined by harmony rather than dilemma: his decision to come to Brown. Last year the Department of Psychology and the Department of Cognitive and Linguistic Sciences merged into one.

“That marriage is absolutely ideal from my perspective because morality is a social psychological topic,” he said. “But my training and some of the techniques that I bring to bear are more traditionally cognitive science. It was a really unique opportunity to get to be seeing groups of researchers coming together and thinking about ways to marry those traditions because that’s exactly where my research sits.”