The Boundary Behind the Bat

The Bat on Paper

Eli: Francis Csedrik is eight. He's on the swing in his backyard in Washington, D.C., and he stops mid-swing. He's remembering the morning he found a bat in the house. A black one. The furry kind. He was four, maybe a little younger, and he remembers a lot from back then. He's written the bat story down. "I just like writing that story because I just don't want to forget it," Francis says.

Mara: And a listener asked us something simple: why can't anyone remember being a baby? Today's angle is the vanishing line of childhood memory.

Eli: Francis is eight, and his memory already has an edge. On one side, a bat at four, sharp enough to write down.

Mara: Eight years old and already keeping his memories on paper. Eli, that edge sits somewhere behind the bat.

The Edge of the Trip

Eli: Francis has more than the bat. At school, a friend said, "I want to carry you down the stairs." Francis says, "I didn't want him to, but he didn't listen. He did it. And I fell headfirst on a marble floor." He got a concussion. His mom, Joanne Csedrik, has a reason that one matters: "Because it reminds you to be careful." And there's the day someone stole the family car. "My dad had to chase it down the block."

Mara: Both come with a jolt. A crack on marble, a chase down the block.

Eli: Then his mom asks about the Philippines. He was three. "It was to celebrate someone's birthday," she tells him. "We took a long plane ride, two boat trips." Francis looks puzzled. He says he doesn't remember. The car thief, he's got. A long plane ride and two boats, nothing.

Mara: That blank has a name, and a boundary. Technically, infantile amnesia covers birth to age two, when we commonly cannot recall autobiographical events. Then comes childhood amnesia, a patchy, thin stretch that runs to about seven. Patricia Bauer, a psychologist at Emory University, says most adults have no memories of roughly their first three and a half years.

Eli: So it's not a light switching off at three. It's a fade: almost nothing, then scraps, then a record.

Mara: And the fade keeps going well past toddlerhood. In one study, researchers talked with three-year-olds about recent events, then checked back as the kids grew. Through about age seven, most of those memories held. Then they started slipping. By age eight or nine, fewer than forty percent were left.

Eli: Francis is eight. Right on that slope. The bat made it across, the Philippines didn't.

Mara: Which leaves one possibility open: that a three-year-old on that trip never stored it at all.

The Mobile and the Mirror

Eli: Paul Frankland, a neuroscientist at SickKids, watched this happen at home. His daughter with Sheena Josselyn, his research partner, was two or three. Ask her about a trip right after, and she could tell them. "It's clear she can make these memories and tell us details about the trips," he said. "But within a couple of months, if we ask about the zoo, it's, 'We didn't go to the zoo. I don't remember that.'"

Mara: Same shape as Francis and the Philippines, only in months. And she clearly made the memory. Even at three months, babies make them. In one experiment, a string tied an infant's foot to a colorful mobile, so kicking made it move. That's memory encoding, the brain laying down a record of something that happened. A week later, the babies still kicked when they saw it. Wait longer, and a short reminder got them going again.

Eli: And copying works the same way. Babies watch an adult do something and repeat it later. With a familiar task, they can do that by six months. A new one takes until closer to their first birthday. And Patricia Bauer found that even in the second year of life, children had very robust memories for specific past events.

Mara: So the toddler is recording. One thing that changes is how she files it. As they near two, children start recognizing themselves in a mirror. Mark Howe and Mary Courage call that the arrival of a cognitive self, a "me" that things happen to, which gives a memory somewhere to belong. Then words arrive. One account says moving from wordless thinking to language builds new mental tools that can't fully absorb the older records. Those early memories, in that view, are "blocked or obscured."

Eli: Words that make newer memories stick.

Mara: Often with help. Carole Peterson says an early memory is more likely to stick when a child makes it a good story, with a time, a place and a coherent sequence. That's the scaffolding: a parent helping a child shape what happened into a story. In one study, kids around three later recalled virtually none of the parts of a new play event their mothers hadn't helped put into words.

Eli: Francis does that scaffolding on paper, for himself, with the bat. But words still can't explain why a two-year-old who plainly had the zoo trip, details and all, loses it a couple of months later, Mara.

Wiring Over the Record

Mara: For years, the problem was seeing inside a baby's brain as it remembered. Nick Turk-Browne at Yale says, "We spent many years figuring out how to do functional MRI experiments in babies." Functional MRI watches which parts of the brain are working, moment to moment. With babies, he says, the trick is parents, keeping the baby comfortable, with "a pacifier, bottle, or blankie." Then he and Tristan Yates showed infants, four months to two years old, a picture, and later paired it with a new one. The test: did they look longer at the one they'd seen?

Eli: Babies can't say "seen it," so their eyes say it for them.

Mara: And a memory region called the hippocampus was more active when babies first saw the pictures they'd later stare at. That study, out March 20, 2025, showed memories can be encoded in the first years of life. Turk-Browne says babies "at least around a year old are starting to form memories and retaining things that they experience."

Eli: Which changes the Philippines. At three, Francis was about two years past the age these babies were already making memories.

Mara: Now the part that turns it. The forgetting isn't only human. Mice and monkeys forget their early childhood too, and young rats lose what they've learned faster than older rats. It seems to show up only in mammals born helpless, completely reliant on their caregivers.

Eli: So words can't carry this alone. Rat pups don't tell stories.

Mara: Paul Frankland and Sheena Josselyn point to neurogenesis, the brain making brand-new neurons, which runs fast in a young hippocampus. More cells sounds like more memory. In their mice, it went the other way. Adding new neurons after a memory formed promoted losing it. Cutting them back after learning can head off the fast forgetting of infant memories. "As new neurons integrate into the hippocampal brain circuits they modify those circuits," Frankland says, "and this leads to the loss of information stored in those circuits, or forgetting."

Eli: Scientists think that fast neuron production in childhood could add to the higher rate of forgetting when we're young. That's one way to read those kids dropping under forty percent: the record gets made, then the wiring around it keeps changing.

Mara: What nobody has settled is what that loss is. Rewired could mean wiped, or just harder to reach.

Eli: At Trinity College Dublin, a researcher in Tomás Ryan's lab looked at a different player: microglia, immune cells in the brain. In very young mice, when microglia were suppressed, the mice remembered fearful experiences better. Ryan says, "It seems that the brain is filing away the neuronal units that store memory, the engrams, for later use," and that microglia help organize how engrams are stored and expressed across a lifetime.

Mara: Two players now, then: new neurons rewiring the circuit, and immune cells shaping how the engrams get kept.

Eli: Filing away is his phrase. A filed memory could be shredded or just shelved, and that split is what the mouse labs went after next.

The Light and the Lost Key

Mara: Start with an animal that doesn't do this. Rats are born immature, missing many of their adult movement skills. Guinea pigs arrive behaviorally and neurologically advanced. Researchers trained guinea pigs to escape a T-shaped maze, some at five days old, some at a hundred. Seventy-five days later, both groups relearned it equally. No infantile amnesia.

Eli: And if someone gives guinea pig pups a burst of new neurons?

Mara: Their memory gets worse afterward. In the reading Frankland and Josselyn hold, the mouse work goes further: new neurons after learning cut how much the original memory cells fire up again further along the circuit, and the memory fades.

Eli: Is that memory gone, or just unreachable?

Mara: Think of a burned letter versus a lost address. Storage failure means the record itself is destroyed. Retrieval failure means the letter's still there, and the way to it is lost. Mark Howe and Mary Courage argued for storage: failing to recall early life comes from failing to keep information in storage, not from missing the right cue. In their view, early memories are lost.

Eli: Susumu Tonegawa at MIT holds the other reading. His team worked with engram cells, the specific cells holding a memory, and blocked protein synthesis, the process of making proteins to strengthen cell connections, right after learning, so their connections never strengthened. A day later, the mice showed no memory. Then the researchers switched those cells on with optogenetic tools, and the mice showed all the signs of full recall. "The majority of researchers have favored the storage theory," Tonegawa says, "but we have shown in this paper that this majority theory is probably wrong."

Mara: Tomás Ryan draws the line this way: "The strengthening of engram synapses is crucial for the brain's ability to access or retrieve those specific memories, while the connectivity pathways between engram cells allows the encoding and storage of the memory information itself." In his account, the links between those tagged cells hold the memory, and strong synapses are the way in.

Eli: Then the infant version. Researchers tagged cells in the hippocampus while mice learned as infants. As adults, the mice seemed to have forgotten. Stimulating those tagged cells brought back fear memories from up to three months earlier, though only partly. And Paul Frankland, the new-neuron researcher himself, said of that work in 2018, "These results suggest our earliest experiences are not completely forgotten or erased from the brain. Instead, we can bring them back through direct stimulation."

Mara: So the scientist behind the new-neuron idea also found a way back in, at least in mice.

Eli: And traces can work without being remembered. In another mouse study, cells tagged during an infant experience overlapped with cells active when the adults recalled, in part of the prefrontal cortex. Remove those tagged cells, and the adult relearning weakened. The authors say early experiences aren't lost but leave physical traces in a latent state.

Mara: Two readings, then, side by side: Howe and Courage's lost record, and the intact trace Tonegawa and Ryan can switch back on. What would change the picture is evidence from a human brain, and every one of these recoveries happened in a mouse.

A Life in Sequence

Eli: Picture kids at five and a half. Ask about last month, and they recall about eighty percent of it. Ask about events from back when they were toddlers, and it's under forty. So Francis is writing the bat down from inside the stretch where memories are still coming loose. A steadier record of a life doesn't really begin until the early school years, and there's no single age when childhood amnesia ends.

Mara: So nobody remembers being a baby, but not because babies don't record. It appears to come from both: new neurons remodeling the hippocampus's early circuits, and words changing how memories get organized and found.

Eli: Whether his Philippines trip is gone or shelved is still open. Recall starts with a cue at any age, and a smell, a picture or a melody may bring back something long forgotten.

Mara: Two signals to watch: whether the latent traces found in mice can survive intact across decades in a human brain, and whether ordinary sensory cues, nothing invasive, can ever reach a memory from infancy.

Eli: A topic to look into is the infantile amnesia boundary.

Mara: Stay with science for more episodes as research tracks what happens to those lost childhood memories.

Eli: And that's the story. I'm Eli.

Mara: And I'm Mara. Thanks for listening, and until next time.