“There’s no good speed indication?…We’re in…we’re in a climb?” Bonin responded.
The pilots kept asking each other questions as the plane’s crisis deepened because they didn’t have mental models to help them process new information as it arrived. The more they learned, the more confused they became. This explains why Bonin was so prone to cognitive tunneling. He hadn’t been telling himself a story as the plane flew along, and so when the unexpected occurred, he wasn’t sure which details to pay attention to. “I have the impression we’re going crazily fast,” he said as the plane began to slow and fall. “What do you think?”
And when Bonin did finally latch on to a mental model—“I’m in TO/GA, right?”—he didn’t look for any facts that challenged that model. “I’m climbing, okay, so we’re going down,” he said two minutes before the plane crashed, seemingly oblivious to the contradiction of his words. “Okay, we’re in TO/GA,” he added. “How come we’re continuing to go right down?”
“This can’t be true,” he said seconds before the plane hit the water. Then there are his last words, which make all the sense in the world once you realize Bonin was still grasping for useful mental models as the plane hurtled toward the waves:
“But what’s happening?”
This problem isn’t unique to the aviators of Flight 447, of course. It happens all the time, within offices and on freeways, as we’re working on our smartphones and multitasking from the couch. “This mess of a situation is one hundred percent our own fault,” said Stephen Casner, a research psychologist at NASA who has studied dozens of accidents like Air France Flight 447. “We started with a creative, flexible, problem-solving human and a mostly dumb computer that’s good at rote, repetitive tasks like monitoring. So we let the dumb computer fly and the novel-writing, scientific-theorizing, jet-flying humans sit in front of the computer like potted plants watching for blinking lights. It’s always been difficult to learn how to focus. It’s even harder now.”
A decade after Beth Crandall interviewed the NICU nurses, two economists and a sociologist from MIT decided to study how, exactly, the most productive people build mental models. To do that, they convinced a midsized recruiting firm to give them access to their profit-and-loss data, employees’ appointment calendars, and the 125,000 email messages the firm’s executives had sent over the previous ten months.
The first thing the researchers noticed, as they began crawling through all that data, was that the firm’s most productive workers, its superstars, shared a number of traits. The first was they tended to work on only five projects at once—a healthy load, but not extraordinary. There were other employees who handled ten or twelve projects at a time. But those employees had a lower profit rate than the superstars, who were more careful about how they invested their time.
The economists figured the superstars were pickier because they were seeking out assignments that were similar to previous work they had done. Conventional wisdom holds that productivity rises when people do the same kind of tasks over and over. Repetition makes us faster and more efficient because we don’t have to learn fresh skills with each new assignment. But as the economists looked more closely, they found the opposite: The superstars weren’t choosing tasks that leveraged existing skills. Instead, they were signing up for projects that required them to seek out new colleagues and demanded new abilities. That’s why the superstars worked on only five projects at a time: Meeting new people and learning new skills takes a lot of additional hours.
Something else the superstars had in common is they were disproportionately drawn to assignments that were in their early stages. This was surprising, because joining a project in its infancy is risky. New ideas often fail, no matter how smart or well executed. The safest bet is signing on to a project that is well under way.
However, the beginning of a project is also more information rich. By joining fledgling initiatives, the superstars were cc’d on emails they wouldn’t have otherwise seen. They learned which junior executives were smart and picked up new ideas from their younger colleagues. They were exposed to emerging markets and the lessons of the digital economy earlier than other executives. What’s more, the superstars could later claim ownership of an innovation simply by being in the room when it was born, rather than fighting paternity battles once it was deemed a success.
Finally, the superstars also shared a particular behavior, almost an intellectual and conversational tic: They loved to generate theories—lots and lots of theories, about all kinds of topics, such as why certain accounts were succeeding or failing, or why some clients were happy or disgruntled, or how different management styles influenced various employees. They were somewhat obsessive, in fact, about trying to explain the world to themselves and their colleagues as they went about their days.
The superstars were constantly telling stories about what they had seen and heard. They were, in other words, much more prone to generate mental models. They were more likely to throw out ideas during meetings, or ask colleagues to help them imagine how future conversations might unfold, or envision how a pitch should go. They came up with concepts for new products and practiced how they would sell them. They told anecdotes about past conversations and dreamed up far-fetched expansion plans. They were building mental models at a near constant rate.
“A lot of these people will come up with explanation after explanation about what they just saw,” said Marshall Van Alstyne, one of the MIT researchers. “They’ll reconstruct a conversation right in front of you, analyzing it piece by piece. And then they’ll ask you to challenge them on their take. They’re constantly trying to figure out how information fits together.”
The MIT researchers eventually calculated that getting cc’d on those early information-rich emails and hashing out those mental models earned the superstars an extra $10,000 a year, on average, in bonuses. The superstars took on only five projects at once—but they outperformed their colleagues because they had more productive methods of thinking.
Researchers have found similar results in dozens of other studies. People who know how to manage their attention and who habitually build robust mental models tend to earn more money and get better grades. Moreover, experiments show that anyone can learn to habitually construct mental models. By developing a habit of telling ourselves stories about what’s going on around us, we learn to sharpen where our attention goes. These storytelling moments can be as small as trying to envision a coming meeting while driving to work—forcing yourself to imagine how the meeting will start, what points you will raise if the boss asks for comments, what objections your coworkers are likely to bring up—or they can be as big as a nurse telling herself stories about what infants ought to look like as she walks through a NICU.