Memorie, the greatest gift of nature,
and most necessary of all others for this life.
—Pliny the Elder, 1st century AD
I construct my memories with my present.
I am lost, abandoned in the present.
I try in vain to rejoin the past: I cannot escape.
—Jean-Paul Sartre, Nausea (1938)
I’m more than my brain but my memories are what makes me me, so if I don’t remember then who am I? . . . I don’t know when to say goodbye.
—Nicola Wilson, Plaques and Tangles (2015)
Ireneo Funes was bucked off a horse and knocked out. When he regained consciousness, his body was broken but his memory had become perfect—superhuman, you might say. He could recall every dream and every daydream he’d ever had. “He knew the forms of the clouds in the southern sky on the morning of April 30, 1882, and could compare them in his memory with the veins in the marbled binding of a book he had seen only once, or with the feathers of spray lifted by an oar on the Rio Negro on the eve of the Battle of Quebracho.” Funes, his memory now limitless, constructed a madcap numbering system, whereby numbers were assigned words. So the number 7,013 was “Maximo Perez,” and 7,014 was “the railroad.” The system went beyond 24,000 in this way.
Funes, of course, is fictional, a product of the extraordinary mind of the Argentinian writer Jorge Luis Borges. In his short story “Funes, His Memory,” Borges observes that Funes’s method of numbering, and his other similarly confusing mental catalogs, are pointless and intelligible to no one but himself, but they at least show us what’s going on in his mind. “They allow us to glimpse, or to infer, the dizzying world that Funes lives in.”
This is the kind of world we’re going to explore in this chapter. We’ll meet people who approach Ireneo Funes in the magnitude and accuracy of their memory, but are all the more startling for being real. And we’ll go the other way, and see how for all of us, the faculty of memory is delicate, malleable, and utterly unreliable: and that this is a good thing. Memory is just as strange as, and arguably more mysterious than, the richest and weirdest Borges story. Bear in mind as we go that there are two meanings to the phrase “she has an amazing memory.” The first is that she has the ability to store a lot of information; the second is that the content is remarkable. For the most part we’ll be focusing on the cognitive ability.1
Let’s start with a number. It’s one we all know at least slightly from school: pi, the ratio of a circle’s circumference to its diameter. It begins 3.14159 . . . but the numbers carry on forever. It is infinite and irrational, never ending and never repeating, and many fall into its orbit. To some the attraction may be spiritual, to others the pull may be more like the “because it’s there” reasoning of mountaineers. Memory athletes—so called because of their intensive training—in particular are drawn to the endlessness of pi.
Akira Haraguchi, of Kisarazu, near Tokyo, recited pi to 100,000 digits in 2006 at the age of sixty, a feat that lasted more than sixteen hours. To him pi represents a religious quest for meaning. “Reciting pi’s digits has the same meaning as chanting the Buddhist mantra and meditating,” he says. “Everything that circles around carries the spirit of the Buddha. I think pi is the ultimate example of that.” He is the world champion, although Guinness World Records has not recognized his recitation.
The official Guinness record holder is twenty-three-year-old Rajveer Meena from the Sawai Madhopur district in Rajasthan, India. On March 21, 2015, at Vellore Institute of Technology in Tamil Nadu, Meena recited pi to 70,000 decimal places. He was blindfolded. The feat took him nine hours, seven minutes. One of the factors motivating him, he told me, was his upbringing. He wanted to show that despite his humble background, he could win the world’s toughest memory challenge.
These memory wizards have different motivations, and use different techniques, but they all essentially convert the numbers into a story. When they recite the numbers, they are telling the story in their head and translating it back into digits. Haraguchi uses a system based on the Japanese kana alphabet. The first fifty digits of his system read (translated into English): “Well, I, that fragile being who left my hometown to find a peace of mind, is going to die in the dark corners; it’s easy to die, but I stay positive.” One hopes that in the rest of the 100,000 digits the story picks up a bit.
Meena assigns groups of numbers to words, like Funes in the Borges story. When I chatted with him, he gave me an example. “I leave my house and meet Roger Federer, go to the park, grab a pair of jeans, get a cab for fifty dollars to the office, where I earn one hundred dollars.” This translates into the number 749,099,950,100: I (74) leave my house and meet Roger Federer (90), go to the park, grab a pair of jeans (999), get a cab for fifty dollars to the office, where I earn one hundred dollars.”
Memorizing the story for a sequence of 70,000 digits took him over six years. As well as a way to show he was the best in the world, “it was a good way to increase patience and confidence,” Meena says, deadpan. Those qualities were tested in the seven-month wait he endured before his attempt was officially verified: “When I got an email from Guinness World Records saying my record claim was successful, that night I was unable to sleep. Many times I checked that mail.”
Wondering if his memory was like that of Funes, I asked him if he recalled everything that had happened to him every day. Meena said no. He has a good memory for faces and incidents, but he doesn’t inadvertently recall everything he was wearing or eating over every day of his life. Later, however, we will meet someone like this.
• • •
If I’m going to understand the people who memorize pi, I feel I ought to get a bit closer to it, so I spend some time scrolling through a webpage listing pi to one billion digits.2 I’ve been scrolling down steadily for some time, and the scroll marker in the margin is still only about 5 percent of the way down the page. I feel like my mind could unravel if I do this for too long. The tumbling numbers recall The Matrix, but obviously I can’t see anything in them, because there isn’t anything in them. Pi is infinite, and so far has been calculated to 22 trillion digits. No one has (yet?) posted this number online. I go back to the beginning of the sequence, 3.14159 . . . and cut out the first 22,514 digits. It’s a mere snippet (though of course even a trillion digits from an infinite number is still a mere snippet), and look through the numbers at greater leisure.
I’m reminded of another Jorge Luis Borges story, “The Library of Babel,” which describes a library containing a monstrous, as good as infinite, number of books, a vast space holding every possible book that can be made, with every possible combination of letters. Overwhelmingly, the books are total nonsense, but once in a while a word or even a coherent sentence appears. People in the library spend lifetimes searching for books that mean something. Looking through my snippet of pi, I glimpse odd islands of numbers: clustered repetitions of nine, or short strings of zeros and ones that look like binary, or regions where spiky sevens seem to make a snakes-and-ladders zigzag slide through the digits. But that’s all. I know there’s no meaning here, and I don’t really understand those who sink into it. Imagine trying to memorize this number.
To Daniel Tammet the numbers do have real meaning. Numbers to him have an aura. They have color, texture, shape, and even, weirdly, emotion. The number four, for example, to Tammet is blue, but it’s also a timid number, one he feels close to because of his own shyness. It’s his nickname for himself. Numbers on their own can glow and wink and perhaps even snarl, but strings of numbers form sentences of emotions and feelings.
Tammet, thirty-eight, is a bestselling author and translator. British-born, he lives in Paris, where I meet him on a sweltering June day in a cool, green restaurant in Saint-Germain-des-Prés. I’m doubly testing my memory before I meet him, as I used to know this area really well, so I am wandering the old streets in the heat looking for familiar places (my favorite oyster shop, Huîtrerie Régis, is nearby). I’m also dusting off my French, and finding, happily, that some is still there. I remember fondly a long lunch where I consumed twenty-one oysters at Régis’s shop.
Tammet is a polyglot, speaking some ten languages, and also has synesthesia, a neurological condition which for him means that when he sees words or numbers, he also sees colors. “Three is green, five is yellow, and nine is blue—very blue, a different blue to four,” he tells me. “Tammet” is orange. “Rowan” is red, “Hooper” is white. (It’s a pleasing combination, I am relieved to hear.) Tammet also has autistic savant syndrome, and his IQ is 150 or 180, “depending on the scale” they use.
It strikes me as odd that the same person’s IQ test results can vary by such a huge amount, but I let it go—I’m more startled by his insistence, like John Nunn, that IQ measures only how good you are at doing an IQ test, and that it “reduces intelligence to a number.” It’s meaningless, Tammet says. We’re back with another billionaire who says money isn’t everything.
But I’m not here to talk to him about IQ. I want to hear how—and why—he set a European record for the recitation of pi. It took him just over five hours. He recounted it to 22,514 places.
Tammet is the oldest of nine children. He says that the necessity of having to interact with so many siblings as a child tempered the antisocial effects of his autism to some extent. And his autism is relatively mild. He makes as much eye contact as a nonautistic person, but says he has to work at remembering to do so. He had problems making friends and communicating as a child, and his first language was numbers. At school he felt attracted to pi, but was overwhelmed by the idea of an infinite number. In his twenties, still feeling its pull, he printed out twenty pages of pi, one thousand digits per page, and dove in. He basically communed with the number.
“I would look at the numbers and find emotions and shapes—it’s like a poem for me, like Baudelaire in French or Shakespeare in English. Pi is like a poem written in numbers. And the further I went into the numbers, the more sense it made. Because the more material I had to make sense of it, to give color to it.” It reminds me of how Haraguchi speaks of finding Buddhist meaning as he goes deeper into pi.
Tammet constructed an emotion-led poem out of pi, and recited it in public, in Oxford. To him it was very much like giving a public reading of an actual poem, or an actor giving a performance. “It was in the language of numbers, and people in the audience were very touched because although it wasn’t their first language,” he says, “although it wasn’t a language they could understand, I was able to embody the number in my body, in my breath, in the way that I spoke. I used numbers as language and this seemed to make an impact on people.”
If Haraguchi recites pi as Buddhist mantra, and Meena for reasons of village pride and record-breaking bragging rights, for Tammet the motivation was more simple: it was about communicating.
During his pi period, did he dream in pi? “In the moment just before falling asleep I would see the numbers flashing before my eyes, the shapes, the colors, the emotions, and the meanings, the particular meaning of feeling alone or feeling scared,” he says. It makes me wonder if bravery is necessary for something like this. “There are moments in pi when you feel completely on your own. At the first thousand digits, you feel like you’re the only person in the universe, and it’s a terrifying emotion. And then the story continues, you enter a new stage.”
There is physical exhaustion during recitation. Meena says that during the nine hours of his recital he had diarrhea and fever, and could hardly get the numbers out. Tammet says he doesn’t know how actors in long plays do it, but the courage is for the psychological ups and downs that come of a number soaked in emotions.
“There were people who said afterward they had tears in their eyes, and I must’ve had tears in my voice,” Tammet says. “It was strange because I was counting but also recounting . . . it was something that kept me apart from people that finally helped me to communicate directly.” His account reminds me of what Mark Rylance said he had to do after playing Rooster Byron in Jerusalem. The performance was so intense and overwhelming, he’d have to sit curled up in a ball after the show and collect himself—draw himself back. “I needed courage because there were moments when it was unsettling,” says Tammet, “but I knew there was something beautiful or encouraging coming along.”
Daniel Bor and his colleagues at the University of Cambridge’s department of psychology put Tammet through a variety of tests, including a brain scan, and a psychological measure called the Navon task. This involves showing the subject a series of large “global” characters—letters or numbers—composed of many smaller copies of another, confounding character, the “local” character. So the subject might see a large number three, made up of small number sevens, and have to identify the number seven. Tammet was faster than average at finding what the local characters were, and was less distracted by the global character.
His synesthesia is also of a particular kind.
As Bor’s paper puts it: “This appears to generate structured, highly-chunked content that enhances encoding of digits and aids both recall and calculation.”3 “Chunking” is the grouping of smaller elements into more familiar units, and is a technique commonly used by memory athletes. For example, the number 10,271,962 might be remembered as October 27, 1962. Tammet’s autism and synesthesia seem to have aided him in memorizing pi.
“Tammet has a predisposition to focus on the details of what he is trying to learn—for instance, the relations between numbers, or the specific features of foreign words—how they sound, what synesthetic details they may relate to, which he can use to aid his memory,” says Bor. “I would speculate that this localist approach enhances his particular form of synesthesia.”
Tammet, in other words, saw pi in synesthetic chunks that he associated with particular colors and emotions, and he knitted these chunks together into a story. “He used a mnemonic method, but one intimately connected to his synesthesia,” says Bor.
• • •
People who train their memories are known as memory athletes. Think of them not as Rocky running up the steps of the Philadelphia Museum of Art, but as hunched over a table, straining to remember the order of a deck of cards. That’s how they train. You can still play the theme from Rocky in your head if you like.
There are competitions at country and regional levels, and annual World Memory Championships, run by the World Memory Sport Council.4 The 2016 world champion was a twenty-five-year-old American medical student, Alex Mullen. Among other things, Mullen was the first person to memorize in under twenty seconds the order of a deck of shuffled playing cards, and the first to memorize more than 3,000 single-digit numbers in one hour.
He, like Daniel Tammet and all memory athletes, uses methods to encode information into a form that makes it more memorable. Our brains don’t learn well if we just try and stuff them with raw information—we need to provide a framework of some kind that the brain can feel at home with. That’s because the part of the brain involved in processing short- and long-term memory, the hippocampus, is also involved in processing emotion and navigation. For Tammet, the framework was the emotion-driven story he constructed from the numbers. Once he’d remembered the story, he could translate backward and get the numbers. So he “chunked” the digits into groups, then mnemonically linked them into a story, in a similar way to how Haraguchi and Meena also wove a story out of chunks of pi.
You don’t have to have savant abilities or be aided by synesthesia to do this. You just need to practice. Anyone can do it, as science journalist Joshua Foer demonstrated when he wrote about memory championships and decided to learn the skills for himself. He ended up winning the 2006 USA Memory Championship, setting a US record in the “speed cards” event (he memorized a deck of cards in 1 minute 40 seconds).
I talk to Martin Dresler of the Donders Institute for Brain, Cognition and Behaviour at Radboud University Medical Center in the Netherlands. He’s shown that anyone can use the techniques of memory athletes to become masters themselves. First, Dresler put twenty-three of the world’s most successful memory athletes through a functional magnetic resonance image (fMRI) brain scanner. These athletes have spent hundreds, even thousands, of hours practicing their memorizing methods. Most use a technique called method of loci, also known as the memory palace technique. With this method, you imagine a place you are intimately familiar with, typically your house, and you populate a route through the house with items corresponding to the list of things you need to learn. It’s like Meena’s example of walking to the park and seeing Roger Federer. The more unusual, startling, even unsettling the images, the more memorable they are. By tracing the route in your head, you can pick up the items along the way, and translate them back into the list.
When Dresler’s team checked the results from the fMRI, they found no structural difference in the brains of the memory athletes compared to untrained people. They found only differences in brain activity, and then only when the athletes were resting.5
Dresler then put volunteers who were new to memory training through six weeks of instruction on the memory palace technique. After this, they had typically doubled their ability to remember words from a random list, and the activity patterns of their brains had started to converge with that seen in the champion memorizers.
Anyone, it seems, can become a memory superhuman. Our potential memory is vast, but the key is to understand how it evolved, and to play to its strengths. “There has hardly been an evolutionary pressure for our ancestors to store abstract information,” says Dresler, “whereas memory for visuospatial information—finding the way home, or to feeding or mating places—is crucial for most animals.”
Understanding the evolutionary context of memory is vital to recognizing its vulnerability to misinformation, as we’ll see. It seems counterintuitive at first that to remember more information we need to encode data into bulkier shapes—we need to build palaces filled with penguins and space stations and Roger Federer; we need to create more information to remember less—but Dresler explains why this is.
“Our brain is biologically still mainly prepared to encode very concrete visuospatial information, and is much less suited for abstract information. Encoding via an additional step of transforming abstract information into visuospatial representation, and thus the encoding of additional information, is more efficient than encoding abstract information directly.”
We learn by listening to stories, and if we create stories we can learn to remember. All the memory athletes Dresler has worked with say they have no innate skill. Everything they do has been learned. But there is a form of extraordinary memory that you do just seem to be born with.
• • •
I’m waiting in front of a huge bronze head of Nelson Mandela. The bust is next to the Royal Festival Hall on London’s South Bank. There are a few other people nearby, a woman with red hair and a man in a raincoat, wearing black-rimmed glasses. Other people stop to take photos next to the statue. I’m wearing black jeans, leather boots, a striped long-sleeved cotton shirt, and a khaki coat. I’ll spare the details of my socks and underwear, though I will note that there’s a (nonfatal) rip in the crotch of the jeans that I made when I was climbing through the labyrinth in the Jardin des Plantes in Paris last week. I study Mandela while I’m waiting. He’s young in this statue, more the defiant leader than the elder statesman. There are clouds in the sky threatening rain. Now a fair-haired young man approaches. He’s taller than me, wearing black jeans, with a pale shirt open over the top of a black T-shirt. He has a small silver hoop earring in his left ear. He holds out his hand and says my name.
If I hadn’t written that down the morning after I was there, I would have soon forgotten those details. And it happened only yesterday. Ask me what I was doing on Monday two weeks ago, and I’ll struggle to remember, let alone recall what I was wearing and what the weather was like. Yet some people remember these sorts of details, and more, for every day, going back years, even decades. They have a condition called highly superior autobiographical memory, or HSAM.
It was first officially described in 2006, when a woman called Jill Price got in touch with James McGaugh at the University of California, Irvine. It turned out Price was able to recall, accurately, huge amounts of frankly unremarkable information like what I listed above—for every day of the last thirty years. “Starting on February 5, 1980, I remember everything,” she said. “That was a Tuesday.” You could throw a date at her—and McGaugh and his team threw plenty—and she would tell you the day of the week and what she did. For example, October 3, 1987? “That was a Saturday. Hung out at the apartment all weekend, wearing a sling—hurt my elbow.” The scientists were able to compare details against a set of journals that Price had kept over the years, and check with calendars the day of the week she gave. She indeed has an unfailing memory, one quickly dubbed “total recall” by the media.
Since Price first turned up, more people with HSAM have been verified, including the actress Marilu Henner (she played Elaine in the 1980s sitcom Taxi). HSAM is an innate kind of superior memory; it is unlike learned abilities such as the recounting of pi, which use mnemonics. Somehow these HSAM people just do it.
I want to understand more about it. How do they do it? What does it feel like, to have a head so crammed with memories? “Most call it a gift, but I call it a burden,” Price has said. “I run my entire life through my head every day and it drives me crazy!”
Maybe people with HSAM can help tell us what memories are, and how they are stored. I remember seeing a Harry Potter film in which Dumbledore extracts memories from Harry’s head with his wand. It’s hard to rid myself of the image of a memory as a wispy, tendrilous structure. What form does a memory have, and how do we access them? I ask Julia Shaw, a psychologist at the London School of Economics, and author of The Memory Illusion. “A memory,” she says, “is a network of neurons that are physically connected and ‘hum’ together, that is, fire on the same wavelength.”
A network is a physical structure, so a memory does have actual form. It’s like a spiderweb in the head. A memory is activated by sending out a “probe”—like casting a fishing line with an enquiry on the end instead of a worm. Say the inquiry is about a beach. It activates related memories. “The concepts that are strongest in their relationship to that probe are going to be activated automatically,” says Shaw, “for example, ‘beach I recently visited in Florida.’ And when they are recalled, it is at this point we can check the memory to see whether it is the one we are looking for.” Perhaps I don’t want to think about that Florida beach, so I refine my probe to make it about black sand beaches, and I come back with the right memory about a black sand beach in New Zealand. I’ll bear this fishing line idea in mind.
• • •
The taller-than-me fair-haired young man shaking my hand is Aurélien Hayman, and he has HSAM. He is twenty-five. Since around the age of fourteen, he’s been able to remember every day of his life. We have a beer at a quiet table inside the Royal Festival Hall. The bottle had a sky-blue label, it was a local pale ale, brewery name of . . . no, it’s gone.
With Jill Price, who said her memory is “nonstop, uncontrollable, and automatic,”6 the ability seemed to come about overnight, but Hayman is hazy about when his kicked in. “I don’t remember when it happened. I always thought it was like a party trick, just something I could do. It’s not like I banged my head and it happened.” He saw a documentary saying a handful of people in the world had this ability that they could remember every day of their lives, thought, “this is me,” and googled it. This led, domino-like, to a meeting with memory researcher Giuliana Mazzoni of the University of Hull, scientific studies, media articles, TV appearances, and a Channel 4 documentary.
I get in touch with twenty-seven-year-old Rebecca Sharrock from Brisbane, Australia, who has an extreme, almost Funes-like form of the phenomenon. “Whenever I’m reliving a memory it is vividly detailed,” she says. “All of my emotions attached to the event come back, as well as all of the things that my five senses picked up from the experience.”
Asked to give an example, she says that each year on her birthday she typically flashes back to one of her happiest memories. It is her seventh birthday. “I can smell jasmine in the air, and my mind is full of pictures of a pink and gold sunrise. There is also an echo of the excitement within me from before I opened my presents.” She goes on to list the presents she got that year, 1996: a princess tiara, a toy pony, and a model house. But there are also what she calls intrusive memories—involuntary recollections of moments of physical and emotional pain. She can remember grazing her knee, and with the recollection comes a phantom echo of the pain. “Though as unpleasant as pain is, reliving negative emotions is so much worse.”
For a long time, having HSAM was grueling. “I believed that everybody remembered in the same way I do,” she says. “I just thought that most people were better at handling their emotional flashbacks than me. This made me feel broken and depressed.” But now she knows—through visiting Craig Stark’s lab at the University of California, Irvine—that she is not alone with the condition. “Now I feel so much happier.”
Sharrock is unusual too in that she has autism, and has memorized the entire series of Harry Potter books. They helped buoy her when she was trying to sleep and memories were flooding in. I ask her to tell me a favorite scene. She says that given what we’re talking about, it has to be when Harry persuades Professor Slughorn to hand over his true memory of what he told the teenage Voldemort. “That section means a lot to me because Slughorn accurately describes how I myself feel about shameful memories.”
Stark says Sharrock is a good illustration that HSAM comes in strong and weak forms: “There is also a range of autobiographical ability. Some perform better on the tests than others, but all are far, far beyond the typical individual.”
The way Hayman describes his memory, it isn’t a burden. “It’s often expressed as ‘gift or curse?’ but for me it’s a very small facet of me, of who I am. When people say ‘how does it affect your day-to-day?,’ well, it doesn’t really.” Hayman’s HSAM is entirely under his control. “I don’t wake up in the morning and go, ‘oh, it’s June fifth’ and think of all the other June the fifths. If someone asks me a date going back, sometimes I can’t remember it. There are blank patches.”
This seems like a good time to throw him a random date. Let’s say, May 1, 2005. There’s a long pause. He’s staring out the window, obviously looking into the past. I imagine him rifling through endless filing cabinets. “That’s not something I can immediately recall,” he says.
I’m forgetting that in 2005 he was only thirteen. I don’t know what I expected—computer-like recall of a date and an immediate report of the things that happened on that day?—but he continues to reach toward the details of the date I have given just like we all do when trying to remember something. Then it clicks, and he snaps his fingers. “I’ve got it, completely, solidly,” he says. Suddenly the picture has resolved. “It was a Sunday.” (I check this later and he’s correct.) “Me, my dad, and my mum went out to the Chilterns for the day and had a pub lunch. I can remember loads of stuff about that day—clothes and weather—but it’s probably mundane for you.”
I ask him to describe the moment just then when he was trying to remember the date.
“If the memory is not immediately obvious, there are sort of checkpoints in my mind, and I was thinking, ‘my birthday is April 27 and I knew it was the bank holiday around then’ . . . you sort of sift around in your head and find it.”
This sounds like what Julia Shaw said about probes going out and seeking related memories. When I think about memories, I find they are grouped by my physical location at the time, say my apartment in Tokyo, or my flat in Dublin. I find it much harder to order my memories by time rather than space, but that’s apparently what people with HSAM can do. “For people with HSAM, dates are the optimal cue, but there seems to be a sort of chaining effect, that is when they retrieve something thanks to the date, they are able also to use some of the elements that they have retrieved as cues for other memories,” says Mazzoni. If enough time were given to them, she explains, they might retrieve much more, even if not every instance of their life.
“People think that this is all about having limitless memory, a really expansive capacity to remember all this stuff,” Hayman says. He doesn’t believe, though, that his memory is more extensive than other people’s; it simply seems to be arranged in such a way that he can pick things from it at the right point. “There’s some kind of accessibility which is really fine tuned for me.” Given enough prompting, he reckons, I would be able to recall what happened on a given date, too. Maybe, but given enough prompting, I might also just imagine I could remember the event, and this can have devastating consequences, as we’ll see. “In the right conditions many people should be able to retrieve more,” says Mazzoni. It’s why she finds HSAM so fascinating: it gives us a glimpse of the potential capacity of long-term memory. Based on both HSAM and other phenomena, she believes, we probably have mental representations of many more experiences than those we are able to retrieve at a given moment.
Okay, so I see how Hayman can grasp his way to the memory. But how does he remember what day the memory relates to? I struggle with this at first because when I think about myself I’m aware that I don’t know what day it is every day. So how can I remember years back? Hayman shrugs, and agrees—he doesn’t know how that works. But thinking about it, even if I go through a day without explicitly saying the name of the day of the week (as I write this a memory from The Godfather pops unbidden into my head, Michael Corleone’s Sicilian wife saying “Monday, Tuesday, Thursday, Wednesday . . .”), even if I don’t think I know what day it is, of course I do really, deep down. It’s just taken so much for granted, like the sun and the rain, like breathing, that we don’t think we take note of it.
I throw another date at Hayman—July 9, 2012.
“Let me think about this.” He goes into recall mode, staring out the window, and I’m picturing tendrils of thought snaking out and gripping hold of markers poking out of the mist.
“It’s a Monday.” Yes, I later confirm this. “I remember . . . that was the day I was being filmed for the documentary.” He laughs that I’ve alighted randomly on that day. “I remember what I was wearing but that’s because it was on the documentary. It was Wimbledon the day before, my hayfever was really bad at that time. I can tell you the weather if you want.” It was the Wimbledon final the day before: Roger Federer beat Andy Murray.
I wonder how far the memories go. This is ridiculous, I say, but could you remember how cloudy it was, or even the shape of the clouds in the sky? Hayman gamely smiles. He can’t. It’s not photographic in that way, he says. “Sometimes it’s just a sense of what it was like, a mood.”
Indeed, I should know better. I was getting carried away with thoughts of Ireneo Funes. Even Rebecca Sharrock doesn’t remember that sort of detail. There is no such thing as “photographic memory.” The closest thing to it is eidetic memory, which is the ability, seen in only 5 percent of children and never in adults, to view a memory of something as if it was a photograph in front of them. Say you show such a child a room for a minute or two, then blindfold them completely. For just a few minutes they will be able to view their memory of the room as if it was still in front of them.
Some people with HSAM seem to have better-than-average working memories. What’s Hayman’s like? “It’s really bad. I am actually quite forgetful. I’ll forget to do something at work. It’s a completely different thing.”
• • •
HSAM is weird in that people with the condition, or ability, can remember only stuff that they personally experienced, and they get the recall ability only some months after the experience.7 It’s not like they’ve immediately encoded their memories in some crystal protective shell. It’s more that the route back to the memory has to have time to bake in. The experiments that showed the existence of this delayed-baking phenomenon were carried out in Craig Stark’s lab. “One of the take-homes from the study is that many aspects of their memory are entirely typical—anything outside of autobiographical memory—and that the fundamental mechanisms involved appear typical as well,” Stark says. “It’s not that people with HSAM are wholly different from us and using some system entirely unknown to the field, but they’re clearly performing in ways that are orders of magnitude beyond the rest of us—in this limited domain.”
Time for another date for Hayman. What about March 12, 2009? He struggles with this one, then says he can do March 14. “But maybe this is a memory I shouldn’t share.” Oh, go on. “It was a Saturday. It was the first time I ever got drunk. We were on the beach, Penarth Beach.” What were you drinking? “Straight vodka.”
Hayman’s memories are not all of the same strength, just like those of us without HSAM. Some memories are quite diluted, some are really sharp. I imagine he snapped toward the March 14 memory of getting drunk because it was much sharper (at least until he got drunk) than the March 12 memory. And all “firsts” are memorable. But generally he doesn’t know why memories vary in this way. Perhaps it’s to do with how much they are pondered afterward.
Hayman shares another characteristic with many people who have HSAM—not that there are lots of them; only some sixty people in the world have been identified with the condition. “I have a very active imagination. I’m naturally a bit of a daydreamer.” It doesn’t seem to be the guided daydreaming that Hilary Mantel spoke of. This is much more . . . unmoored. “Sometimes I couldn’t even tell you what I’m thinking about,” he says. “My parents might say I’m away with the fairies.”
This could be key to their ability, says Lawrence Patihis—formerly another of the UC Irvine group, but now at the University of Southern Mississippi at Hattiesburg. In a paper on HSAM personalities, Patihis wrote: “The personal events in which HSAM individuals become absorbed in and fantasize about later will likely lead to accurate memory.”8
Hayman repeats that it’s a misunderstanding to think of HSAM as a burden, at least in the form that he has it. “It’s not like having your entire life history that keeps playing in a reel in your mind, it’s not like that at all. It’s just, if prompted, you can recall things. It’s a talent of accessing memory rather than just possessing loads of it.”
His feeling is that he has more memories there somewhere, only even he can’t access them. He thinks the rest of us do, too. Perhaps everything we experience is encoded somewhere, but locked away or neglected, and so faint that we can no longer access it. “Could I,” wondered Dorthe Berntsen of the Department of Psychology and Behavioural Sciences at Aarhus University in Denmark, “as a non-HSAM person, have memory from each day in my life stored, but I just can’t get to it?”9
It’s a compelling and widely held view, but it’s a widely held myth. When I mention it to Patihis he slaps it down, hard. “Absolutely not. That is a long-debunked idea going back to the neuroscientist Wilder Penfield.” Penfield, a Canadian neurosurgeon, was the first to map and deduce the function of various parts of the brain, and he pioneered surgical treatments for epilepsy. “Most things we experience are not encoded,” says Patihis, “and even those things we do encode fade over time.”
Dutch neuroscientist Martin Dresler points out that it’s hard to think how you would even test this idea. And in any case, from an evolutionary point of view, it doesn’t make much sense to equip the brain with a vast capacity to encode everything, then conceal it from conscious access. “I think it would make much more sense to encode only experiences fully and permanently that were somehow tagged,” he says, “for example by strong emotions, as very important, and to keep only the gist of more mundane experiences, getting rid of all redundant, boring details.”
The question, according to Hayman, is how he can access those bits that the rest of us can’t. The information he retrieves isn’t really significant. But you can’t blame the brain for not knowing what’s significant and what’s not. A pub lunch with Mum and Dad in the Chilterns, there’s at least potential for significance in that. But then there’s potential in everything. It’s only if we tattoo the memories in, for weddings and special occasions, or use mnemonics, or practice over and over, like hafiz devotees learning the Koran, like actors learning their lines, that they become more deeply engraved.
Psychologists class memories not by emotion, as we tend to (happy memories, sad memories), but more broadly, on whether the memory is about general knowledge, or about personal experience. The former are known as semantic memories, the latter episodic. It is these personal memories that are highly detailed in HSAM people.
Experiments carried out by Patihis and his colleagues have shown that HSAM people may have amazing chronological recall, but the detail of the memories is only average.10 They are no better than non-HSAM people at remembering, say, the words in a list presented in a laboratory setting. This is consistent with Hayman’s own feelings about his memory. “I can remember things in a patchy way and I can remember things around it. If you said a month in a particular year, I could remember the kinds of things I was listening to, the friends I had, what was on the radio. If you threw a date at me, it might unlock a memory for the first time since it happened. Which is bizarre.”
Talking with Hayman has been illuminating, not least because it has demystified what HSAM people do. Quite probably they use the same memory mechanisms as the rest of us—think of the inquiry cast out on the fishing line—and there is evidence of this from a false-memory experiment carried out by Patihis and his colleagues. In these psychological tests, people are shown scenarios and then tested on them in such a way that they are seeded with potentially false information. All of us are susceptible to blithely incorporating the fake news into our memory. “The memory distortion tasks used in that study really do tap into the variety of ways we reconstruct memory,” says Patihis, “suggesting that storage and retrieval of memories in HSAM people is similar to the rest of us.”
He thinks it is not so much new memory mechanisms that explain HSAM, but personality differences such as obsessiveness, a vivid imagination, and the tendency to become absorbed.
People with highly superior autobiographical memory, then, are not so highly superior after all. Certainly, we don’t know how they fish out so many memories from thousands of days stretching into the past. But just like the rest of us, they are susceptible to contamination—to the adoption of and belief in false memories. Indeed, the very thing that might be responsible for their superior memories—their absorption and tendency to daydream—could be what leaves them prone to misinformation. Deep absorption in misleading information can produce errors in memory, says Patihis.
HSAM may be characterized as total recall, but if you remember the movie of the same name, there’s a detail that rather spoils the comparison: Arnie’s memories have been faked. He misremembers vast swathes of his life. The incorporation of false memories into our brains can have tragic consequences, but its widespread nature is poorly recognized. Time, then, to look more deeply.
• • •
Ledell Lee was executed in Arkansas in April 2017 for the murder of his neighbor, Debra Reese. However, according to the Innocence Project, a legal organization that campaigns to exonerate wrongly convicted people, there were numerous problems with his conviction. For example, many unknown fingerprints were found at the crime scene, but none of them were Lee’s. DNA testing of a speck of blood on Lee’s shoes was not carried out, nor on hairs found at the scene which the prosecution argued came from Lee. The prosecution case also relied heavily on eyewitness testimony.
Three people identified Lee as a man they had seen in the area and leaving Reese’s house. Yet according to a report by the Innocence Project, of 349 exonerations of convicted prisoners by DNA evidence in the US to date, 71 percent had rested at least partly on eyewitness (mis)identification. The US Supreme Court did not order a DNA test for Lee, and he was executed by lethal injection.
Psychologists and lawyers have long known that eyewitness identification is unreliable. There have been thousands of scientific papers demonstrating problems, not least something called “own race bias.” This is the tendency to be worse at identifying people from ethnic backgrounds different to our own.
But eyewitness identification continues to be a powerful strand of evidence; the science, outrageously, hasn’t filtered down to the courts of law. In one infamous case in 1984, a young college student was attacked and raped in her apartment in Burlington, North Carolina. The victim, twenty-two-year-old Jennifer Thompson-Cannino, who is white, said she’d specifically tried to memorize her assailant’s face as he was assaulting her, so that she would have a chance of identifying him if she survived. Local restaurant worker Ronald Cotton, who is black, was at home the night of the attack, but fluffed his alibi and Thompson-Cannino picked him from a selection of police photos. When Cotton was brought in for a live lineup, he was picked out again by Thompson-Cannino, who said she was 100 percent certain he was her attacker. It’s hard to argue with an assertion of total certainty, and we are prone, both as jurors and simply as people, to place greater weight on something told to us by someone else than on dry scientific evidence. Her testimony at trial helped sentence Cotton to life plus fifty-four years. The story has a happier ending than Lee’s, however: just over ten years later DNA evidence exonerated Cotton (and implicated another man, who had confessed to the crime in jail). Thompson-Cannino and Cotton became friends, have written a book together (it has the great title Picking Cotton), and regularly speak on the need to reform the law relating to eyewitness identification.
“Even when witnesses say they are 100 percent certain of something, studies show that the testimony is only slightly more accurate than people who say they are unsure,” says Jakke Tamminen, a psychologist at Royal Holloway, University of London. There is a positive correlation between confidence and accuracy under pristine conditions, but not in the messy real world. And no one, least of all prosecuting lawyers, takes this into account.
At Tamminen’s lab I take part in an experiment to test memory reliability. He has me witness a couple of crimes, staged for the camera for purposes of memory research. In the first crime, two men steal a computer monitor from a library; in the second, a woman in snowy Cambridge, Massachusetts, has her purse stolen by a man on the street. There are lots of complicating factors in each scenario—other people interacting with the main characters, and several different locations in each. After I’ve watched the crimes, I read a list of statements summarizing the plot of each scenario, and the actions of the people involved. Tomorrow I will be examined on the facts of the crimes. The idea, says Tamminen, is to test the accuracy of eyewitness testimony.
It’s been shown experimentally many times that even if we have vivid memories which we rate as highly reliable and highly likely to be true, this does not mean that they are necessarily accurate. The morning after the space shuttle Challenger exploded in 1986, Ulric Neisser, a cognitive psychologist at Emory University, had students fill out a questionnaire about the disaster. Three years later the students filled out the same questionnaire and Neisser compared the answers: they were completely different. Similar comparisons, conducted following the 9/11 attacks on the World Trade Center in New York, found complete mismatches between what witnesses were sure they remembered and what really happened. Our minds quite literally play tricks on us.
The morning after I’ve witnessed the “crimes” in Tamminen’s lab, I am questioned about what I saw, and asked to rate my confidence in each answer. For example, the woman who had her purse stolen. When I read the summary statements the day before, I was told that the man bumped into her from the front. I thought this was misinformation, because I remember thinking it looked unnatural when I watched the crime that the man had bumped into her from behind. The next day, I am again told that the man who stole from the woman bumped into her from the front, and am asked to agree or disagree, and rate my certainty. I disagree, and say I am fairly certain (4/5 on the scale). But then I am told, “the man put the purse into his jacket pocket,” and I can’t remember if he did put it there, or if it went in his trousers pocket. The same with other details—was a woman in the library wearing glasses? Tamminen’s experiment is designed to test how certainty about memories changes depending on whether they are tested soon after witnessing an event, or with a period of sleep in between.
We know that when even a long-term memory is retrieved, it becomes liable to rejigging. The demonstration of this caused quite a stir when it was published in Nature back in 2000, as researchers thought long-term memories were locked down.11 And work by Jason Chan, at Iowa State University in Ames, shows we are more susceptible to being contaminated by false information when recounting an event straight after it occurs. The act of remembering itself makes a memory more unstable, and this is when misinformation might sneak its way in.12
• • •
We all seem to be prone to believing things that aren’t true. To believing, faithfully but incorrectly, that things we didn’t witness did in fact happen in front of our eyes. Maryanne Garry, a psychologist at the University of Waikato in Hamilton, New Zealand, has studied the acquisition of false memories for many years. “My own hunch,” she says, “is that very few, if any, people would be immune to this effect given the right package of information, method, and circumstances.”
Jason Chan agrees. “Some people are a little more resistant to false memories than others, and frontal lobe functioning—which is related to source-monitoring ability and working-memory capacity—is typically thought of as an important factor.” However, resistance is one thing; immunity is quite another. “We are not aware of anyone who is not susceptible to false remembering at all.”
So are there any superhuman resistors? If anyone could be trained, or selected to be resistant to the tendency of our brains to soak up false memories like blotting paper, you’d think it would be the military special forces, whose training is particularly intense. Many countries train their elite soldiers to resist interrogation.
Charles Morgan III is a psychiatrist at Yale School of Medicine who has worked extensively with the US and Canadian military to advise on eyewitness memory and psychological performance under conditions of high stress, and to aid in selection for special forces. Subjects are typically serving military personnel undergoing special forces training. In the studies, soldiers take part in an intense role-play exercise modeled on prisoner-of-war conditions, being deprived of food and sleep for forty-eight hours, and then being “robustly” interrogated. A day after their “release” from the POW camp, the soldiers are asked to identify their interrogator in a lineup. Several of Morgan’s studies have found that many soldiers are unable to, some even mistaking the interrogator’s gender. “After they are ‘interrogated’ in what sounds like a brutally realistic way, large numbers of them just totally suck at recounting what happened, and at identifying their interrogator,” says Maryanne Garry.
For special forces soldiers it may not be a good thing to be susceptible to memory distortion, but most of us will never be captured and placed under intense stress. We have not evolved that way, and indeed our brains work differently. We are social apes; we learn from a range of different people, at different times and under different conditions, as well as from various cues in the environment. Our brains have to be malleable. “I think it’s pretty straightforward,” says Garry. “We have evolved to learn from multiple sources. We have not evolved to detect errors in the way we often encounter them in real life . . . from sources that otherwise seem legitimate, or trustworthy, or not worth the brainpower to scrutinize.”
Nevertheless, there are suggestions that some people are more resistant, at least.
Bi Zhu, of the Institute for Brain Research at Beijing Normal University, China, has worked with UC Irvine’s Elizabeth Loftus, a legend in the field of eyewitness memory, on testing whether some people have more reliable memories. Zhu ran experiments similar to the one I took part in at Royal Holloway. She had 205 Chinese college students watch staged “crimes,” and then had them answer questions on the crimes, while seeding them with misinformation. The first set of answers were given an hour after viewing the crimes, but the second test came a year and a half later. MRI scans were made to measure the volume of key structures in the brain. Students with larger hippocampi remembered more questions correctly, and were less susceptible to the false information.13 What we don’t know is how much genetics influenced the hippocampus size of those students.
• • •
The realization that memory is unreliable has profound implications, quite literally, for who we think we are. Your life is like a Russian doll, with other individuals nested inside you. If the you from one stage of your life could speak to the you of now, they wouldn’t agree on what each had experienced. At first this concept troubled me, but now I feel quite grateful to evolution for providing us with a method of tweaking our personal histories. I heard a sixty-three-year-old man the other day say he was finally happy with the person he was. It’s thanks to our malleable memories that most of the time we are able to grow into people we like.
Felipe De Brigard, who, tellingly, is from the department of philosophy and the Center for Cognitive Neuroscience at Duke University in North Carolina, has a startling idea.14 Memory isn’t just for remembering, he says. Misremembering is so common it shouldn’t be seen all the time as a malfunction. In his view, many cases help us construct scenarios of past events that might have happened, so as to better simulate possible events in the future. An unreliable memory may also destabilize your personality. You may think that your personality is something unchangeably intrinsic to you, but a study in 2016 that measured personality traits over a sixty-year period showed they can profoundly alter over a lifetime.15
Examining superhuman traits expands our understanding of the diversity of the human species. Unexpectedly, this exploration of memory has shown that we ourselves are more diverse than we think. It doesn’t seem to make sense to say an individual can be diverse. But the truth is that each one of us contains, within our self, different people.