I Remember Mama and Dada

What do small children remember? And why do memories stick into adulthood?

Illustration by Robert Neubecker.
Children remember far more and at earlier ages than previously thought

Illustration by Robert Neubecker.

Last August, I moved across the country with a child who was a few months shy of his third birthday. I assumed he’d forget his old life—his old friends, his old routine—within a couple of months. Instead, over a half-year later, he remembers it in unnerving detail: the Laundromat below our apartment, the friends he ran around naked with, my wife’s co-workers. I just got done with a stint pretending to be his long-abandoned friend Iris—at his direction.

We assume children don’t remember much, because we don’t remember much about being children. As far as I can tell, I didn’t exist before the age of 5 or so—which is how old I am in my earliest memory, wandering around the Madison, Wis. farmers market in search of cream puffs. But developmental research now tells us that Isaiah’s memory isn’t extraordinary. It’s ordinary. Children remember.

Up until the 1980s, almost no one would have believed that Isaiah still remembers Iris. It was thought that babies and young toddlers lived in a perpetual present: All that existed was the world in front of them at that moment. When Jean Piaget conducted his famous experiments on object permanence—in which once an object was covered up, the baby seemed to forget about it—Piaget concluded that the baby had been unable to store the memory of the object: out of sight, out of mind.

The paradigm of the perpetual present has now itself been forgotten. Even infants are aware of the past, as many remarkable experiments have shown. Babies can’t speak but they can imitate, and if shown a series of actions with props, even 6-month-old infants will repeat a three-step sequence a day later. Nine-month-old infants will repeat it a month later.

The conventional wisdom for older children has been overturned, too. Once, children Isaiah’s age were believed to have memories of the past but nearly no way to organize those memories. According to Patricia Bauer, a professor of psychology at Emory who studies early memory, the general consensus was that a 3-year-old child’s memory was a jumble of disorganized information, like your email inbox without any sorting function: “You can’t sort them by name, you can’t sort them by date, it’s just all your email messages.”

By those standards, Isaiah is a wizard of memory—the Joshua Foer of the preschool set. But it turns out that all children are Joshua Foer: Even very young children have bewilderingly good memories. Twenty years ago, a study on memories of Walt Disney World—the ne plus ultra memorable experience—surprised everyone involved: Children who’d been at Disney when they were only 3 years old could recount detailed memories of it 18 months later. Evidence has piled up ever since. A just-published paper on long-term recall found that a 27-month-old child who’d seen a “magic shrinking machine” remembered the experience some six years later.

Far from having no memories at all, very young children remember a lot like adults. In early infancy, the neural structures crucial for memory are coming online: the hippocampus, which is, very roughly, in charge of storing new memories; and the prefrontal cortex, which is, very roughly, in charge of retrieving those memories.

But these neural regions and their connecting pathways are still developing. And they capture only part of the present as it flows by.

Think of memory as like orzo, Bauer says. “It’s not like one big piece of lasagna noodle. Memories are made up of these little tiny bits of information that are coming in literally across the entire cortex. Parts of the brain are taking those little bits of information and knitting them together into something that’s going to endure and be a memory.” Adults have a fine-mesh net to catch the orzo. Babies have a big-holed colander: The orzo slips through. “What’s happening with the baby is that a lot of the information is escaping even as the baby is trying to get it organized and stabilized.” In early infancy, a lot of experiences never become memories—they slip away before they can be preserved.

Babies remember far more than anyone thought, in other words, but far less than any adult. It’s only around 24 months that children seem to get better colanders: They get better at catching the orzo—at organizing and processing information in a way that makes a memory out of an experience.

The past gets stickier, too: Memories no longer slip away after a couple of months. Children a few months under 2 retain memories of experiences a year earlier—half their lifetime ago. But they won’t retain those memories into adulthood: No one remembers their second birthday party. For a few reasons—nascent neural structures, the lack of knowledge to make sense of early experiences, the lack of language to represent those experiences—it may be impossible for any part of our lives before, say, 24 months to stick around into adulthood. The average earliest memory—fragmented and lonely, but real—doesn’t date until around 3½ years of age.

What makes that first memory stick into adulthood? This is where the new science of early memory takes an unexpected turn: Once memories start to stick, how long they stick around for may be less of a neural question than a social question. It may have less to do with the child than with the adults.

Psychologists have spent a lot of time listening to how parents talk to their children, specifically how parents negotiate the very stubborn truth of parenthood that children aren’t any good at talking back. Kids can’t keep up their end of the conversation. When discussing the past, parents get around this problem in a couple of different ways. They might ask specific, repetitive questions about past events. Or they might narrate the past in a detailed, elaborate way, asking the child questions and then incorporating their answers into the narrative, a style that researchers call “highly elaborative.”

It turns out that children of highly elaborative mothers tend to have earlier and richer memories. A study of adolescents whose mothers were highly elaborative during their preschool years found they had far earlier first memories than those whose mothers weren’t. Conversational style matters, because when children remember and talk about the past, they effectively relive the event—they fire the same neurons and reinforce the same connections. They are buttressing their memory of the event. And when parents scaffold their children’s stories—when they essentially tell the stories for their children, as a highly elaborative parent of a very young child would—they are reinforcing those same connections.

The word story is important here. Children are learning how to organize memories in a narrative, and in doing so, they are learning the genre of memory. “As children learn those forms, their memories become more organized,” says Robyn Fivush, a psychology professor at Emory who studies memory and narrative. “And more organized memories are better retained over time.”

Conversational style may also explain why women tend to have earlier first memories than men. Girls typically have different and more elaborative interactions in early childhood than boys do. “Mothers are more likely to be highly elaborative when talking about the past, and particularly when talking about highly emotional events in the past, and they’re more likely to do it with their girls than with their boys,” Fivush says.

As an intervention, in at least in the short-term, training parents to talk about the past in a highly elaborative way seems to be highly successful: Children begin to tell stories—to process their experience—in richer, more detailed ways. (There’s also good evidence that that these skills correlate with literacy.) The Maori in New Zealand have the earliest average first memory of any culture—2½ years of age—and talk to their children in a highly elaborative way about their shared past. I’d thought of memory as essentially neural. But at a certain point, it may be as much cultural.

By pretending to be Iris, by acting out stories from Isaiah’s past, I was, without knowing it, teaching my son how, and why, we remember. In the long term, this is pretty fantastic. In the short term, it is pretty profoundly stupid: I was sentencing myself to spend more time pretending to be Iris. “Children learn the skills that are both practiced and valued in their environment,” Fivush says. “Kids who grow up in homes where you talk about the past all the time, in these more elaborative ways, grow up with better memories.”

But despite myself, I’m prematurely nostalgic for childhood amnesia; in your most unhappy hour of parenting an infant, its absoluteness is a get-out-of-jail-free card. You can console yourself with the thought, He will not remember this. (At 3 a.m., though, your thinking is often more like, He won’t even remember this! The ingrate!) Isaiah is now at the age when he might remember, for the rest of his life, something that happens—something that I do—right now.

It’s wonderful and terrifying in equal measure.

What’s your earliest memory? Slate staffers reveal theirs here.