The End of the Lie

Two University of Pennsylvania scientists have invented machines that can look into your brain and know when you’re lying. The implications are extraordinary — and pretty scary

There’s a second truth machine, just a block or so away from Britton Chance’s lab. It’s less portable than the infrared headband, but far more advanced.

Its father is Dr. Daniel Langleben, an Israeli immigrant with close-cropped silvering hair. “Sorry I’m behind,” he said, bounding into a waiting room at the university’s hospital. He wore a backpack and sneakers, and appeared to have jogged the dozen or so blocks from his office. He and Chance work in separate buildings, in separate university departments, but they share ideas and expertise from time to time. After my weeks spent with 92-year-old Chance, meeting 44-year-old Langleben felt like stepping into the path of a galloping colt.

As we walked deeper into the hospital, toward his machine, Langleben talked about lying. He sounded more like a philosopher than a doctor as he noted that people don’t even know exactly what it means to lie. “St. Augustine told us everything in his treatise,” he said. “We can’t improve on it.”

That fourth-century treatise is called De Mendacio, or On Lying. In it, St. Augustine defined the idea: Lying, he said, is a concealment of the truth. That’s a simple statement, but scientists like Langleben are only now coming to understand its significance. Since lying is a concealment of the truth, it must rely on the truth in some way. We can’t create deception from the air. For instance, a lie:

Grass is blue.

Of course that’s untrue. Grass is green. And furthermore, there’s no way to tell that particular lie without first thinking of grass’s essential greenness.

So: When telling the truth about grass, the brain does one thing. It thinks of green.

But: When lying, it does two. It thinks of green, and then blue. That means something in the brain must suppress the impulse to say “green.”

Years ago, Langleben worked with children suffering from attention deficit hyperactivity disorder because research had shown that they were terrible liars — they tended to blurt out the truth. Some piece of their brains lacked the ability to suppress it. And it hit him: He could capture the essence of a lie if he could just watch it unfold in the brain. So he developed a technique using functional magnetic resonance imaging, or fMRI, to track the brain’s activity.

“Here it is,” Langleben said. We had arrived outside the hospital’s MRI room. A sign warned that no one with pacemakers or “metallic implants” should venture closer. Inside the room, a patient lay with his head inside the fMRI, which looks like a giant white doughnut. It was a “four tesla” magnet, one of the most powerful magnets in the world. Attendants turned as we walked in: “Shhhhh.”

Every atom in creation contains electrical charges — positive for protons, negative for electrons — and the fMRI is so powerful that when someone climbs inside, the machine essentially magnetizes his atoms and causes them to resonate. By manipulating the magnetic field, Langleben can cause different tissues to resonate in their own ways. That makes blood carrying oxygen, for instance, look different from blood without it. So Langleben watches blood flow in much the same way Chance does with his infrared headband, except on a tinier, far more accurate scale.

During his deception study, Langleben handed volunteers a $20 bill and a playing card, and told them they could keep the cash if they could successfully lie about the card they held. The subjects climbed into the MRI; when their card appeared on a screen, they pressed a button indicating otherwise.

The volunteers’ brain-scan images, when combined and viewed together, proved that when we lie, the “truth” part of our brain activates, but so does an additional “suppression” part. The averaged images were interesting on the theoretical level, and grabbed the scientific world’s attention. But to make the technology practical, Langleben needed to prove that each individual brain reveals a lie.

Not long ago, Langleben and another scientist, Kosha Ruparel, gathered members of Penn’s staff — but not fMRI experts — to “read” the individual brain images by looking at lit-up areas of activity. Of the volunteers’ brains, one of the readers correctly identified 25 of 27 liars, with the other readers doing almost as well. That’s better than 90 percent accuracy, in the technology’s infancy.

Langleben says the fMRI is fully functional — it could be used in basic public applications tomorrow. It’s bulky, but Penn has a history of proving that time shrinks all technology: In 1944, the school produced ENIAC, a monstrous machine that consumed an entire city block but whose descendents now sit on laps and in pockets around the world. It was the world’s first computer.

I asked Langleben why the military hadn’t employed his technology at Guantanamo Bay and Abu Ghraib. It might help stave off criticism that the U.S. uses torture to extract information. “I haven’t told them about it yet,” he said. Instead of working directly for the Pentagon, Langleben planned to publish results of his study in a pair of scientific journals, which is the scientific equivalent of a trumpet herald announcing a new discovery.

Besides: Guantanamo Bay is just the start. The real work awaits in New York, Phoenix, Pottstown, and every other American city and town.

“I expect we’ll see parents who maybe find some pot or condoms coming in and saying, ‘I want to bring my kid in once a month,’” said Arthur Caplan, the ethicist. He said that other thinkers around the world haven’t turned their attention to the truth machines because they simply don’t know about them. He does, because the inventors are his colleagues at the university. “Right now, a lot of people are focused on the ethical implications in abortion or genetic advances, like cloning,” he said. “But our greatest challenge will come from the neurological sciences. The brain. Because it’s where we identify ourselves. It is us.”

Like the computers that followed ENIAC, the truth machines may improve, and even save, many lives. Caplan said most people might accept their use in airports and rail stations, if they improve security. People might even welcome them.

But some implications, he said, are “appropriately scary.”