The Thing with Feathers (20 page)

Dellis was almost as happy about his memory times as his near-summit of Everest (he turned around 280 feet from the top). In the rarified pantheon of elite “mental athletes,” he had become an instant star a couple of months earlier by winning the 2011 USA Memory Championship, a niche competition for
speed memorization. The contest had been founded some fourteen years earlier by a former IBM executive to demonstrate the capabilities of the human brain, but then people like Dellis started signing up—and turned simple memory games into a sport in some ways as grueling as mountain climbing.

Trust human beings to find a way to compete over something as ordinary as remembering a list of names. One of the events at the championship—which is held every year in a meeting room at the corporate headquarters of Con Edison, the main electricity supplier in New York City—involves quickly memorizing 117 different photographs of unfamiliar faces, each image inscribed with a person’s name. Competitors must also recite a fifty-line unpublished poem after only fifteen minutes of study and absorb as many random digits as possible in five minutes. Then there is Dellis’s specialty, memorizing the order of a shuffled deck of cards.

In practice, Dellis has correctly recalled fifty-two cards in order after only thirty-three seconds to glance at the deck. At the competition, though, his best time is sixty-three seconds. Distractions and nerves play havoc with one’s ability to concentrate—hence Dellis’s card tricks at twenty thousand feet. In an effort to showcase drama at the USA Memory Championship, finalists must perform mental feats while sitting onstage, facing an audience (yes, there are spectators, commentators, and TV cameras), though the event still feels a bit like a room full of students taking the SAT.

Make no mistake, memory competitions are serious. There is a worldwide rankings list. The World Memory Championships, held each December in a rotating list of international cities, offers tens of thousands of dollars in prize money and is usually dominated by European and Asian competitors. In the United States, mental sports traditionally have held less
prestige, but that is beginning to change. In 2012, journalist Joshua Foer popularized memory contests with his bestseller
Moonwalking with Einstein
, describing how he had trained to win the USA Memory Championship in 2006. After Nelson Dellis won it, he was featured in
The New Yorker
, on CNN, and in
Forbes
. Dellis now works full-time as a “memory consultant,” gives inspirational speeches about climbing and memory, and spends hours exercising his brain.

When I last checked the world rankings list, the top slot was taken by Johannes Mallow, a German fellow who had swept first place in a streak of memory competitions in Sweden, Germany, and England, achieving the honored title of Grand Master. Mallow’s official stats are impressive. In five minutes, he has accurately memorized 500 numerical digits and 85 random words. After just one hour of intense study, he has been able to perfectly recall 2,245 digits and 1,144 shuffled playing cards (twenty-two decks!) in correct sequence. When asked to memorize years of fictitious events, Mallow absorbed 132 dates in five minutes.

If this all sounds like something from
Rain Man
, it’s not; these people do not call themselves savants, and they insist they have average memories—if they’re not paying attention, they forget their car keys just like anyone else. Photographic memory, at least in the popular sense of being able to instantaneously absorb pages of phone books, is a myth. Nobody in the world can regurgitate reams of random information with just a glimpse (if anyone could, surely they’d show up to collect their $10,000 prize at the World Championships). Competitors train hard, like other athletes, and have developed complex systems to pull off seemingly inhuman feats of memory.

How do they do it? Do human memory athletes use the same techniques as Clark’s nutcrackers?

—

IN HOPES OF LEARNING
how birds memorize things, a graduate student at Northern Arizona University, Stephen Vander Wall, focused his attention in the 1970s on the remarkable ability of Clark’s nutcrackers to remember where their snacks are buried. Vander Wall figured that nutcrackers might relocate stashed seeds in one of five ways: (1) The birds dig randomly and find buried seeds by chance; (2) They search randomly, but only in certain, heavily cached areas; (3) They are able to smell buried seeds; (4) They mark the ground surface in some way to indicate where caches are hidden; or (5) They remember the exact locations as if plotted on a mental map. With these five possibilities in mind, Vander Wall set out to design an experiment to whittle down the list.

He could see that following the birds in their natural habitat—across some of the most rugged country in the American West—would amount to a wild nutcracker chase, and that controlled laboratory tests would give much clearer results. So Vander Wall conducted his experiments with captive birds. He covered the floor of a large aviary with a couple inches of loose soil, and then arranged various perches, rocks, logs, and other “landmarks” around the room. After two individual nutcrackers, code-named Orange and Red, had been trained to bury seeds inside the aviary, Vander Wall began his tests.

First, he released Orange and Red inside the aviary and watched in separate trials to see where they made their caches. When both birds had stashed at least 150 seeds in the loose soil, in alternate sessions so they couldn’t watch each other, Vander Wall snuck in and buried 100 seeds himself. He also removed 50 of the birds’ caches. Then he let Orange and Red back into the aviary to see which seeds they would dig up. If
the birds foraged by remembering the locations of their own caches, Vander Wall reasoned, they wouldn’t be able to find each other’s stashes or his additional buried seeds.

And that’s exactly what happened. Orange found 63 of his own seeds, but none of Red’s or Vander Wall’s. Red found 61 of his seeds, three of Orange’s, and none of Vander Wall’s. The birds were not searching randomly, or they would have found an equal share of each group of seeds. The first two hypotheses—two variants on random searching—could thus be neatly ruled out.

Both birds also dug unsuccessfully for their personal caches that had been secretly removed by Vander Wall, which indicated that his third hypothesis—the birds found food by smell—was also false. That left just the fourth and fifth options: Either the birds had somehow marked the soil surface or they were able to remember locations in relation to nearby landmarks.

Vander Wall conducted a second experiment to investigate whether the birds might be marking the soil. He let Orange and Red cache additional seeds inside the aviary, then raked half of its floor smooth to obscure any ground cues. If they were looking for surface disturbances, then Vander Wall predicted that the birds would be able to find their caches only in the unraked half. But when the two nutcrackers were returned to their aviary, they proceeded to dig up seeds in both the raked and unraked areas, indicating that surface texture didn’t impede their ability to remember locations. Hypothesis number four was rejected, too.

Only one possibility remained, that nutcrackers were using spatial memory to locate their stashes. To test it, Vander Wall designed a third, elegant experiment. As in the first two tests, he kept Orange and Red inside the aviary until they had made
numerous seed caches in the loose soil. Then, as before, he removed the birds and made some secret modifications. This time, he rearranged the landmarks in half of the room, moving each of the rocks, logs, and other perches exactly twenty centimeters in the same direction; in the other half, he left the landmarks alone. He predicted that the nutcrackers would be able to locate their caches only where the nearby landmarks had been left in place.

When Orange and Red returned to the aviary after a period of fasting, they searched for their stashed food. This time, they found only half of the seeds—the caches in the undisturbed half, just as Vander Wall had predicted. In the other half of the aviary, the hungry birds dug about twenty centimeters away from the actual locations of their caches, almost exactly corresponding to the distance and direction that the local landmarks had been moved. The only exceptions were a few caches in the middle of the aviary, between the two zones, for which the birds dug at an offset of about ten centimeters.

Vander Wall showed that Clark’s nutcrackers draw on spatial memory to remember locations—a pretty impressive feat, considering how many thousands of points they memorize every year. The basic idea is intuitive. The birds must build a three-dimensional mental map, plotting the locations of temporary caches within their visualization. They can remember where they store food only by tying each individual spot to an already intimate knowledge of landmarks within their home territory. This method orders the information in space, which turns relatively abstract knowledge into a useful foraging map that helps them get through the winter.

It’s as if the birds are telling themselves:
Dinner is on the stove, my car keys are on the bedside table, and I left the car parked in the space by the blue lamppost.

—

ACCORDING TO LEGEND,
the Greek poet Simonides (556–468 B.C.), a charmingly witty, miserly international man of lyricism—dubiously credited, among other things, with inventing four letters of the Greek alphabet—was once invited to a banquet to celebrate the victory of a boxer. At some point during the festivities, Simonides rose from the table and stepped outdoors for a moment of solace; during that brief interval, the building collapsed and killed everyone inside, leaving the poet with a narrow escape. Excavation revealed the bodies of his friends to be unrecognizably mangled, but Simonides realized that he could remember where everyone had been sitting just by closing his eyes and visualizing himself at the banquet in the moments before he stepped out. Thus, he was able to identify the corpses.

This feat supposedly led him to invent a new way of memorizing information, a necessary skill for poets past and present. Simonides imagined himself walking through a familiar setting—say, inside a palace—and, as he mentally traced his path, he would paint vivid images at various points along the way. If he wanted to remember a poem that started with a line about a lion, he’d imagine a lion sitting on the front steps of the palace. If the next line talked about the moon, he’d think of the moon crammed into the entryway. And if the next line dealt with a beautiful woman, Simonides would picture her waiting at the foot of the staircase after he’d squeezed past the moon. He found that by using this strategy, he could keep many images ordered in his mind, and retain the nuggets he needed to remember.

At least, that’s the story. Simonides’ accomplishments are now chronicled only on a few papyrus fragments, so nobody
knows whether he really escaped from a collapsing building 2,500 years ago (although it’s clear that his poetry and ideas did affect the celebrated Classical period of Greece, and by extension all of Western civilization). But the memory-palace method of memorization—often called the method of loci by today’s psychologists—endures, and continues to be used by those who need to absorb large amounts of ordered information, including memory athletes such as Nelson Dellis and Joshua Foer.

You don’t have to imagine a palace. Just pick any familiar space—your childhood home, your commute to work, the interior of your favorite restaurant—and picture moving through it as you normally would. The key is to tag vivid images throughout your journey at points where you’ll remember to look for them when you mentally retrace your steps.

At the USA Memory Championship, all of the top contenders use memory palaces to absorb packs of playing cards and lists of random digits and words. Individual approaches vary, but the strategy is always the same: Hang that information on a familiar spatial frame. If you can convert a pack of cards into, say, a list of celebrities strewn throughout your house, the deck becomes a story—and a narrative of surprises is much easier to remember than a bunch of meaningless numbers and suits.

Nelson Dellis, who broke his own record and defended his championship title in 2012 by memorizing the order of 303 random numbers in five minutes flat, has insisted that he has no natural talent for memory. He hadn’t even heard of mental athletes just three years before he won the event. Anyone, he says, can train to speedily cram information. Memorization stunts are all about forming associations and creating story lines from static data, even if, as in the case of numbers and playing cards, those stories have nothing to do with the content.

To prepare for the random-numbers event in the contest, Dellis painstakingly pre-memorized a list of 999 people, each one associated with a number, action, and object. Number 124, for instance, might be Tiger Woods hitting a golf ball. Number 423 could be George Bush driving a limousine, and 858 could be Britney Spears singing in her underwear. When confronted with a list of random digits, Dellis breaks the string of numbers into chunks and translates them into composite images of people, actions, and objects, in that order, so that 124-423-858 would naturally bring to mind Tiger Woods driving in his underwear. (858-124-423, by contrast, would become Britney Spears hitting a limousine, and 423-858-124 would represent George Bush singing to a golf ball—you get it.) The possible combinations are nearly endless, which keeps the images fresh. As he works through a list of numbers and their resulting pictures, Dellis concentrates on placing the images chronologically in one of several memory palaces from his everyday life, so that when the time comes to regurgitate the list of numbers, he can simply envision walking through a familiar setting with these crazy scenes happening at intervals, and recite the random digits in accurate order. It takes practice to make the associations quickly—Dellis works at it for up to six hours per day—but the technique is basic.

The same method applies to memorizing a pack of cards; each card has been previously associated with a specific person, action, and object so that they can be vividly coded into a memory palace. The more outlandish, bizarre, and surprising an image is, the more likely you will remember it. Don’t just think of the TV set in your living room, in other words—imagine that it’s on fire, and oozing miniature unicorns, with George Bush slapping his underwear on top. This is exactly how Nelson Dellis was able to recite the order of a shuffled deck after
looking at it for only 63 seconds, and win the USA Memory Championship twice in a row. The power of association is strong in the human mind.