What the Dog Saw

farther out into the ocean, and then to the left. He climbed and descended. He sped up and slowed down. Just a few miles from his destination, Kennedy lost control of the plane, and it crashed into the ocean.

Kennedy’s mistake, in technical terms, was that he failed to keep his wings level. That was critical, because when a plane banks to one side it begins to turn and its wings lose some of their vertical lift. Left unchecked, this process accelerates. The angle of the bank increases, the turn gets sharper and sharper, and the plane starts to dive toward the ground in an ever-narrowing corkscrew. Pilots call this the graveyard spiral. And why didn’t Kennedy stop the dive? Because, in times of low visibility and high stress, keeping your wings level—indeed, even knowing whether you are in a graveyard spiral—turns out to be surprisingly difficult. Kennedy failed under pressure.

Had Kennedy been flying during the day or with a clear moon, he would have been fine. If you are the pilot, looking straight ahead from the cockpit, the angle of your wings will be obvious from the straight line of the horizon in front of you. But when it’s dark outside, the horizon disappears. There is no external measure of the plane’s bank. On the ground, we know whether we are level even when it’s dark, because of the motion-sensing mechanisms in the inner ear. In a spiral dive, though, the effect of the plane’s G-force on the inner ear means that the pilot feels perfectly level even if his plane is not. Similarly, when you are in a jetliner that is banking at thirty degrees after takeoff, the book on your neighbor’s lap does not slide into your lap, nor will a pen on the floor roll toward the “down” side of the plane. The physics of flying is such that an airplane in the midst of a turn always feels perfectly level to someone inside the cabin.

This is a difficult notion, and to understand it I went flying with William Langewiesche, the author of a superb book on flying, Inside the Sky. We met at San Jose Airport, in the jet center where the Silicon Valley billionaires keep their private planes. Langewiesche is a rugged man in his forties, deeply tanned, and handsome in the way that pilots (at least since the movie The Right Stuff) are supposed to be. We took off at dusk, heading out toward Monterey Bay, until we had left the lights of the coast behind and night had erased the horizon. Langewiesche let the plane bank gently to the left. He took his hands off the stick. The sky told me nothing now, so I concentrated on the instruments. The nose of the plane was dropping. The gyroscope told me that we were banking, first fifteen, then thirty, then forty-five degrees. “We’re in a spiral dive,” Langewiesche said calmly. Our airspeed was steadily accelerating, from 180 to 190 to 200 knots. The needle on the altimeter was moving down. The plane was dropping like a stone, at three thousand feet per minute. I could hear, faintly, a slight increase in the hum of the engine, and the wind noise as we picked up speed. But if Langewiesche and I had been talking, I would have caught none of that. Had the cabin been unpressurized, my ears might have popped, particularly as we went into the steep part of the dive. But beyond that? Nothing at all. In a spiral dive, the G-load —the force of inertia—is normal. As Langewiesche puts it, the plane likes to spiral-dive. The total time elapsed since we started diving was no more than six or seven seconds. Suddenly, Langewiesche straightened the wings and pulled back on the stick to get the nose of the plane up, breaking out of the dive. Only now did I feel the full force of the G-load, pushing me back in my seat. “You feel no G-load in a bank,” Langewiesche said. “There’s nothing more confusing for the uninitiated.”

I asked Langewiesche how much longer we could have fallen. “Within five seconds, we would have exceeded the limits of the airplane,” he replied, by which he meant that the force of trying to pull out of the dive would have broken the plane into pieces. I looked away from the instruments and asked Langewiesche to spiral-dive again, this time without telling me. I sat and waited. I was about to tell Langewiesche that he could start diving anytime, when, suddenly, I was thrown back in my chair. “We just lost a thousand feet,” he said.

This inability to sense, experientially, what your plane is doing is what makes night flying so stressful. And this was the stress that Kennedy must have felt when he turned out across the water at Westerly, leaving the guiding lights of the Connecticut coastline behind him. A pilot who flew into Nantucket that night told the National Transportation Safety Board that when he descended over Martha’s Vineyard, he looked down and there was “nothing to see. There was no horizon and no light…I thought the island might [have] suffered a power failure.” Kennedy was now blind, in every sense, and he must have known the danger he was in. He had very little experience in flying strictly by instruments. Most of the time when he had flown up to the Vineyard, the horizon or lights had still been visible. That strange, final sequence of maneuvers was Kennedy’s frantic search for a clearing in the haze. He was trying to pick up the lights of Martha’s Vineyard, to restore the lost horizon. Between the lines of the National Transportation Safety Board’s report on the crash, you can almost feel his desperation:

About 2138 the target began a right turn in a southerly direction. About 30 seconds later, the target stopped its descent at 2200 feet and began a climb that lasted another 30 seconds. During this period of time, the target stopped the turn, and the airspeed decreased to about 153 KIAS. About 2139, the target leveled off at 2500 feet and flew in a southeasterly direction. About 50 seconds later, the target entered a left turn and climbed to 2600 feet. As the target continued in the left turn, it began a descent that reached a rate of about 900 fpm.

But was he choking or panicking? Here the distinction between those two states is critical. Had he choked, he would have reverted to the mode of explicit learning. His movements in the cockpit would have become markedly slower and less fluid. He would have gone back to the mechanical, self-conscious application of the lessons he had first received as a pilot—and that might have been a good thing. Kennedy needed to think, to concentrate on his instruments, to break away from the instinctive flying that served him when he had a visible horizon.

But instead, from all appearances, he panicked. At the moment when he needed to remember the lessons he had been taught about instrument flying, his mind—like Morphew’s when she was underwater—must have gone blank. Instead of reviewing the instruments, he seems to have been focused on one question: Where are the lights of Martha’s Vineyard? His gyroscope and his other instruments may well have become as invisible as the peripheral lights in the underwater-panic experiments. He had fallen back on his instincts—on the way the plane felt—and in the dark, of course, instinct can tell you nothing. The NTSB report says that the last time the Piper’s wings were level was seven seconds past 9:40, and the plane hit the water at about 9:41, so the critical period here was less than sixty seconds. At twenty-five seconds past the minute, the plane was tilted at an angle greater than forty-five degrees. Inside the cockpit it would have felt normal. At some point, Kennedy must have heard the rising wind outside, or the roar of the engine as it picked up speed. Again, relying on instinct, he might have pulled back on the stick, trying to raise the nose of the plane. But pulling back on the stick without first leveling the wings only makes the spiral tighter and the problem worse. It’s also possible that Kennedy did nothing at all, and that he was frozen at the controls, still frantically searching for the lights of the Vineyard, when his plane hit the water. Sometimes pilots don’t even try to make it out of a spiral dive. Langewiesche calls that “one G all the way down.”

4.

What happened to Kennedy that night illustrates a second major difference between panicking and choking. Panicking is conventional failure, of the sort we tacitly understand. Kennedy panicked because he didn’t know enough about instrument flying. If he’d had another year in the air, he might not have panicked, and that fits with what we believe—that performance ought to improve with experience, and that pressure is an obstacle that the diligent can overcome. But choking makes little intuitive sense. Novotna’s problem wasn’t lack of diligence; she was as superbly conditioned and schooled as anyone on the tennis tour. And what did experience do for her? In 1995, in the third round of the French Open, Novotna choked even more spectacularly than she had against Graf, losing to Chanda Rubin after surrendering a 5-0 lead in the third set. There seems little doubt that part of the reason for her collapse against Rubin was her collapse against Graf—that the second failure built on the first, making it possible for her to be up 5-0 in the third set and yet entertain the thought I can still lose. If panicking is conventional failure, choking is paradoxical failure.

Claude Steele, a psychologist at Stanford University, and his colleagues have done a number of experiments in recent years looking at how certain groups perform under pressure, and their findings go to the heart of what is so strange about choking. Steele and Joshua Aronson found that when they gave a group of Stanford undergraduates a standardized test and told them that it was a measure of their intellectual ability, the white students did much better than their black counterparts. But when the same test was presented simply as an abstract laboratory tool, with no relevance to ability, the scores of blacks and whites were virtually identical. Steele and Aronson attribute this disparity to what they call “stereotype threat”: when black students are put into a situation where they are directly confronted with a stereotype about their group—in this case one having to do with