What the Dog Saw and Other Adventures

by Malcolm Gladwell

  deeply entrenched individual mind-sets that prize the production of evidence-supported narratives of defendant wrongdoing over the drawing of probabilistic inferences based on incomplete and fragmentary information in order to support decision-making. … Law enforcement organizations handle information, reach conclusions, and ultimately just think differently than intelligence organizations. Intelligence analysts would doubtless make poor policemen, and it has become very clear that policemen make poor intelligence analysts.

  In his 2003 State of the Union message, President George W. Bush did what Shelby wanted, and announced the formation of the Terrorist Threat Integration Center — a special unit combining the antiterrorist activities of the FBI and the CIA. The cultural and organizational diversity of the intelligence business, once prized, is now despised.

  The truth is, though, that it is just as easy, in the wake of September 11, to make the case for the old system. Isn’t it an advantage that the FBI doesn’t think like the CIA? It was the FBI, after all, that produced two of the most prescient pieces of analysis — the request by the Minneapolis office for a warrant to secretly search Zacarias Moussaoui’s belongings, and the now famous Phoenix memo. In both cases, what was valuable about the FBI’s analysis was precisely the way in which it differed from the traditional “big picture,” probabilistic inference making of the analyst. The FBI agents in the field focused on a single case, dug deep, and came up with an “evidence-supported narrative of defendant wrongdoing” that spoke volumes about a possible Al Qaeda threat.

  The same can be said for the alleged problem of rivalry. The Cell describes what happened after police in the Philippines searched the apartment that Ramzi Yousef shared with his coconspirator, Abdul Hakim Murad. Agents from the FBI’s counterterrorism unit immediately flew to Manila and “bumped up against the CIA.” As the old adage about the Bureau and the Agency has it, the FBI wanted to string Murad up, and the CIA wanted to string him along. The two groups eventually worked together, but only because they had to. It was a relationship “marred by rivalry and mistrust.” But what’s wrong with this kind of rivalry? As Miller, Stone, and Mitchell tell us, the real objection of Neil Herman — the FBI’s former domestic counterterrorism chief — to “working with the CIA had nothing to do with procedure. He just didn’t think the Agency was going to be of any help in finding Ramzi Yousef. ‘Back then, I don’t think the CIA could have found a person in a bathroom,’ ” Herman says. “ ‘Hell, I don’t think they could have found the bathroom.’ ” The assumption of the reformers is always that the rivalry between the FBI and the CIA is essentially marital, that it is the dysfunction of people who ought to work together but can’t. But it could equally be seen as a version of the marketplace rivalry that leads to companies working harder and making better products.

  There is no such thing as a perfect intelligence system, and every seeming improvement involves a trade-off. A couple of months ago, for example, a suspect in custody in Canada, who was wanted in New York on forgery charges, gave police the names and photographs of five Arab immigrants, who he said had crossed the border into the United States. The FBI put out an alert on December 29, posting the names and photographs on its website, in the “war on terrorism” section. Even President Bush joined in, saying, “We need to know why they have been smuggled into the country, what they’re doing in the country.” As it turned out, the suspect in Canada had made the story up. Afterward, an FBI official said that the agency circulated the photographs in order to “err on the side of caution.” Our intelligence services today are highly sensitive. But this kind of sensitivity is not without its costs. As the political scientist Richard K. Betts wrote in his essay “Analysis, War, and Decision: Why Intelligence Failures Are Inevitable,” “Making warning systems more sensitive reduces the risk of surprise, but increases the number of false alarms, which in turn reduces sensitivity.” When we run out and buy duct tape to seal our windows against chemical attack, and nothing happens, and when the government’s warning light is orange for weeks on end, and nothing happens, we soon begin to doubt every warning that comes our way. Why was the Pacific fleet at Pearl Harbor so unresponsive to signs of an impending Japanese attack? Because, in the week before December 7, 1941, they had checked out seven reports of Japanese submarines in the area — and all seven were false. Rosenhan’s psychiatrists used to miss the sane; then they started to see sane people everywhere. That is a change, but it is not exactly progress.

  5.

  In the wake of the Yom Kippur War, the Israeli government appointed a special investigative commission, and one of the witnesses called was Major General Zeira, the head of Aman. Why, they asked, had he insisted that war was not imminent? His answer was simple:

  The Chief of Staff has to make decisions, and his decisions must be clear. The best support that the head of Aman can give the Chief of Staff is to give a clear and unambiguous estimate, provided that it is done in an objective fashion. To be sure, the clearer and sharper the estimate, the clearer and sharper the mistake — but this is a professional hazard for the head of Aman.

  The historians Eliot A. Cohen and John Gooch, in their book Military Misfortunes, argue that it was Zeira’s certainty that had proved fatal: “The culpable failure of Aman’s leaders in September and October 1973 lay not in their belief that Egypt would not attack but in their supreme confidence, which dazzled decision-makers. . . . Rather than impress upon the prime minister, the chief of staff and the minister of defense the ambiguity of the situation, they insisted — until the last day — that there would be no war, period.”

  But, of course, Zeira gave an unambiguous answer to the question of war because that is what politicians and the public demanded of him. No one wants ambiguity. Today, the FBI gives us color-coded warnings and speaks of increased chatter among terrorist operatives, and the information is infuriating to us because it is so vague. What does increased chatter mean? We want a prediction. We want to believe that the intentions of our enemies are a puzzle that intelligence services can piece together, so that a clear story emerges. But there rarely is a clear story — at least, not until afterward, when some enterprising journalist or investigative committee decides to write one.

  March 10, 2003

  The Art of Failure

  WHY SOME PEOPLE CHOKE AND OTHERS PANIC

  1.

  There was a moment in the third and deciding set of the 1993 Wimbledon final when Jana Novotna seemed invincible. She was leading 4–1 and serving at 40–30, meaning that she was one point from winning the game, and just five points from the most coveted championship in tennis. She had just hit a backhand to her opponent, Steffi Graf, that skimmed the net and landed so abruptly on the far side of the court that Graf could only watch, in flat-footed frustration. The stands at Center Court were packed. The Duke and Duchess of Kent were in their customary places in the royal box. Novotna was in white, poised and confident, her blond hair held back with a headband — and then something happened. She served the ball straight into the net. She stopped and steadied herself for the second serve — the toss, the arch of the back — but this time it was worse. Her swing seemed halfhearted, all arm and no legs and torso. Double fault. On the next point, she was slow to react to a high shot by Graf and badly missed on a forehand volley. At game point, she hit an overhead straight into the net. Instead of 5–1, it was now 4–2. Graf to serve: an easy victory, 4–3. Novotna to serve. She wasn’t tossing the ball high enough. Her head was down. Her movements had slowed markedly. She double-faulted once, twice, three times. Pulled wide by a Graf forehand, Novotna inexplicably hit a low, flat shot directly at Graf, instead of a high crosscourt forehand that would have given her time to get back into position: 4–4. Did she suddenly realize how terrifyingly close she was to victory? Did she remember that she had never won a major tournament before? Did she look across the net and see Steffi Graf — Steffi Graf! — the greatest player of her generation?

  On the baseline, awaiting Graf’s ser
ve, Novotna was now visibly agitated, rocking back and forth, jumping up and down. She talked to herself under her breath. Her eyes darted around the court. Graf took the game at love; Novotna, moving as if in slow motion, did not win a single point: 5–4 Graf. On the sidelines, Novotna wiped her racquet and her face with a towel, and then each finger individually. It was her turn to serve. She missed a routine volley wide, shook her head, talked to herself. She missed her first serve, made the second, then, in the resulting rally, mis-hit a backhand so badly that it sailed off her racquet as if launched into flight. Novotna was unrecognizable, not an elite tennis player but a beginner again. She was crumbling under pressure, but exactly why was as baffling to her as it was to all those looking on. Isn’t pressure supposed to bring out the best in us? We try harder. We concentrate harder. We get a boost of adrenaline. We care more about how well we perform. So what was happening to her?

  At championship point, Novotna hit a low, cautious, and shallow lob to Graf. Graf answered with an unreturnable overhead smash, and, mercifully, it was over. Stunned, Novotna moved to the net. Graf kissed her twice. At the awards ceremony, the Duchess of Kent handed Novotna the runner-up’s trophy, a small silver plate, and whispered something in her ear, and what Novotna had done finally caught up with her. There she was, sweaty and exhausted, looming over the delicate white-haired Duchess in her pearl necklace. The Duchess reached up and pulled her head down onto her shoulder, and Novotna started to sob.

  2.

  Human beings sometimes falter under pressure. Pilots crash and divers drown. Under the glare of competition, basketball players cannot find the basket and golfers cannot find the pin. When that happens, we say variously that people have panicked or, to use the sports colloquialism, choked. But what do those words mean? Both are pejoratives. To choke or panic is considered to be as bad as to quit. But are all forms of failure equal? And what do the forms in which we fail say about who we are and how we think? We live in an age obsessed with success, with documenting the myriad ways by which talented people overcome challenges and obstacles. There is as much to be learned, though, from documenting the myriad ways in which talented people sometimes fail.

  Choking sounds like a vague and all-encompassing term, yet it describes a very specific kind of failure. For example, psychologists often use a primitive video game to test motor skills. They’ll sit you in front of a computer with a screen that shows four boxes in a row, and a keyboard that has four corresponding buttons in a row. One at a time, x’s start to appear in the boxes on the screen, and you are told that every time this happens you are to push the key corresponding to the box. According to Daniel Willingham, a psychologist at the University of Virginia, if you’re told ahead of time about the pattern in which those x’s will appear, your reaction time in hitting the right key will improve dramatically. You’ll play the game very carefully for a few rounds, until you’ve learned the sequence, and then you’ll get faster and faster. Willingham calls this explicit learning. But suppose you’re not told that the x’s appear in a regular sequence, and even after playing the game for a while, you’re not aware that there is a pattern. You’ll still get faster: you’ll learn the sequence unconsciously. Willingham calls that implicit learning — learning that takes place outside of awareness. These two learning systems are quite separate, based in different parts of the brain. Willingham says that when you are first taught something — say, how to hit a backhand or an overhead forehand — you think it through in a very deliberate, mechanical manner. But as you get better, the implicit system takes over: you start to hit a backhand fluidly, without thinking. The basal ganglia, where implicit learning partially resides, are concerned with force and timing, and when that system kicks in, you begin to develop touch and accuracy, the ability to hit a drop shot or place a serve at a hundred miles per hour. “This is something that is going to happen gradually,” Willingham says. “You hit several thousand forehands, after a while you may still be attending to it. But not very much. In the end, you don’t really notice what your hand is doing at all.”

  Under conditions of stress, however, the explicit system sometimes takes over. That’s what it means to choke. When Jana Novotna faltered at Wimbledon, it was because she began thinking about her shots again. She lost her fluidity, her touch. She double-faulted on her serves and mis-hit her overheads, the shots that demand the greatest sensitivity in force and timing. She seemed like a different person — playing with the slow, cautious deliberation of a beginner — because, in a sense, she was a beginner again: she was relying on a learning system that she hadn’t used to hit serves and overhead forehands and volleys since she was first taught tennis, as a child. The same thing has happened to Chuck Knoblauch, the New York Yankees’ second baseman, who inexplicably has had trouble throwing the ball to first base. Under the stress of playing in front of forty thousand fans at Yankee Stadium, Knoblauch finds himself reverting to explicit mode, throwing like a Little Leaguer again.

  Panic is something else altogether. Consider the following account of a scuba-diving accident, recounted to me by Ephimia Morphew, a human-factors specialist at NASA: “It was an open-water certification dive, Monterey Bay, California, about ten years ago. I was nineteen. I’d been diving for two weeks. This was my first time in the open ocean without the instructor. Just my buddy and I. We had to go about forty feet down, to the bottom of the ocean, and do an exercise where we took our regulators out of our mouth, picked up a spare one that we had on our vest, and practiced breathing out of the spare. My buddy did hers. Then it was my turn. I removed my regulator. I lifted up my secondary regulator. I put it in my mouth, exhaled, to clear the lines, and then I inhaled, and, to my surprise, it was water. I inhaled water. Then the hose that connected that mouthpiece to my tank, my air source, came unlatched and air from the hose came exploding into my face.

  “Right away, my hand reached out for my partner’s air supply, as if I was going to rip it out. It was without thought. It was a physiological response. My eyes are seeing my hand do something irresponsible. I’m fighting with myself. Don’t do it. Then I searched my mind for what I could do. And nothing came to mind. All I could remember was one thing: if you can’t take care of yourself, let your buddy take care of you. I let my hand fall back to my side, and I just stood there.”

  This is a textbook example of panic. In that moment, Morphew stopped thinking. She forgot that she had another source of air, one that worked perfectly well and that, moments before, she had taken out of her mouth. She forgot that her partner had a working air supply as well, which could easily be shared, and she forgot that grabbing her partner’s regulator would imperil both of them. All she had was her most basic instinct: get air. Stress wipes out short-term memory. People with lots of experience tend not to panic, because when the stress suppresses their short-term memory they still have some residue of experience to draw on. But what did a novice like Morphew have? I searched my mind for what I could do. And nothing came to mind.

  Panic also causes what psychologists call perceptual narrowing. In one study, from the early seventies, a group of subjects were asked to perform a visual-acuity task while undergoing what they thought was a sixty-foot dive in a pressure chamber. At the same time, they were asked to push a button whenever they saw a small light flash on and off in their peripheral vision. The subjects in the pressure chamber had much higher heart rates than the control group, indicating that they were under stress. That stress didn’t affect their accuracy at the visual-acuity task, but they were only half as good as the control group at picking up the peripheral light. “You tend to focus or obsess on one thing,” Morphew says. “There’s a famous airplane example, where the landing light went off, and the pilots had no way of knowing if the landing gear was down. The pilots were so focused on that light that no one noticed the autopilot had been disengaged, and they crashed the plane.” Morphew reached for her buddy’s air supply because it was the only air supply she could see.

  Panic, in this se
nse, is the opposite of choking. Choking is about thinking too much. Panic is about thinking too little. Choking is about loss of instinct. Panic is reversion to instinct. They may look the same, but they are worlds apart.

  3.

  Why does this distinction matter? In some instances, it doesn’t much. If you lose a close tennis match, it’s of little moment whether you choked or panicked; either way, you lost. But there are clearly cases when how failure happens is central to understanding why failure happens.

  Take the plane crash in which John F. Kennedy, Jr., was killed. The details of the flight are well known. On a Friday evening in July of 1999, Kennedy took off with his wife and sister-in-law for Martha’s Vineyard. The night was hazy, and Kennedy flew along the Connecticut coastline, using the trail of lights below him as a guide. At Westerly, Rhode Island, he left the shoreline, heading straight out over Rhode Island Sound, and at that point, apparently disoriented by the darkness and haze, he began a series of curious maneuvers: He banked his plane to the right, 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.