Babies are Forgetful
With Memories Like Sieves, It’s No Wonder We Can’t Remember Our Infancy


by Nicky Penttila

May 14, 2007

The fact that we remember nearly nothing from our infancy and early childhood does not mean we created no memories then. It’s just that we forgot.

Babies start with at least some method of memory—they need it, says Lisa M. Oakes, principal investigator at the Infant Cognition Lab at the University of California at Davis. Infants must make sense of their brand-new world, its dizzying colors, sounds, shapes and motions. Without some functions of memory, they would not know from eyeblink to eyeblink that nothing has changed in their field of view (short-term memory). Without memory, they would not know that this four-footed creature is not the four-footed creature they saw this morning—and also not the one that lives at their home (longer-term memory).

“All babies are innately extremely intelligent, because these tasks are difficult, and they all can do it," says Oakes, who studies what babies remember of what they see.

So why don't we, as adults, remember what we knew as babies? Several potential explanations have been offered about what causes “childhood amnesia,” from the idea that babies can’t form any memories and thus have none to recall to Freud’s theory that our early memories are so traumatic that the mind represses them.

Until recently, though, none of these theories was tested by directly studying the affected group: infants and young children. They are hard to test, since they don’t yet have language skills and so can’t tell us what they remember. But through visual and spatial experiments that don’t involve language, researchers at Duke, UC-Davis and elsewhere are starting to tease out answers.

"What our work can help us understand is how are those processes developing: What are they remembering, what are the limitations of that memory and what are the processes involved in it," Oakes said.

First findings

So far, researchers are finding, babies' memory abilities improve, growing in stages toward the adult standards, much as the rest of their abilities do. Babies appear to use what they have.

For example, Oakes and her lab have found that babies as young as four months old will recognize that the orange horizontal bar they see on a video monitor is not the same image as the blue horizontal bar or the orange vertical one they saw earlier, but only if they have seen the first images for a full minute. Children 7-8 months old can make that distinction after seeing the first images for one-half a second.

That suggests that infants may not use the fleeting short-term memory systems adults employ to, say, walk quickly through a crowded department store without bumping into anything. Instead, infants may be using a longer-term memory system until the better system comes on-line.

While researchers don't yet know exactly when memory brain structures develop, Oakes says, this pattern matches their current best guess: The medial temporal lobes (part of longer-term retention systems) develop early, and the dorsal and ventral visual streams--particularly the parts of the parietal lobes that combine the information on what and where objects are--develop later in the first year. 

Researchers discover what babies know by discovering what "bores" them, building on earlier studies that show infants prefer to look at new images and complex images, Oakes says. In one test, a baby is seated on a parent's lap (the parent wears blackout glasses so as not to cue the baby) and shown images on side-by-side video monitors. Investigators run a slideshow of dual images—changing, unchanging, changing again—and watch the child's eyes. If an image on one side is "new," meaning the baby doesn't remember seeing it before, the baby's gaze will flick to that side and stay longer.

Different order of memories

Between 6 months and 18 months of age, babies’ ability to form long-term memories improves dramatically—from remembering a new activity or something they have seen for about 24 hours to remembering it for a year or more, says Patricia Bauer of Duke.

"Infants are forming memories and holding them for a long time," she says, but they aren't exactly the sort of memories adults hold. For example, in adults, "episodic memory" is defined as a memory of a single, unique, exact event.

Babies don’t have exactly that—some infants can remember the unique, exact event of pulling off a puppet’s glove and shaking it to hear a bell inside the glove tinkle, but they need to see it done three times in the lab before they can recall it (if they do) 24 hours later. All two-year-olds, though, can be shown only once how to make a hanging gong from four or five building blocks and make it sound and they can remember how to do it a year or more later. “Infants and children are getting closer and closer to the ideal of episodic memory,” Bauer says.

Their recall is not yet as exact as it will be when they are adults, either.  “They won’t remember you were wearing a red sweater, but that you were wearing some kind of covering,” she says.

Everything is new

One current theory of infant amnesia is that if babies are using alternate processes for encoding and storing memories, these memories may not be retrievable using the different processes adults use. It also appears that making and holding memories takes a lot of babies' brainpower and energy. They might need to divert some of that energy to learning to walk or dealing with their immediate environment and a hundred other developmental tasks, Bauer says, and a memory that is not tended will wither and die off.

Another theory is that not a lot of information is retained right at the beginning. Even as adults, if we are presented with something totally new (say, a lecture about the intricacies of brain surgery) we will not be able to recall much of the detail of what we saw and heard. The second lecture, we’ll likely retain a little more.

“Without prior exposure, most of it goes by you,” Bauer said. “You’ll get the gloss, the gist, but not a lot of detail.”

For infants, making memories appears to have the same difficulty, except for them, everything is completely new. They have nothing to compare it to, or even language to help them encode the experience into memory. So a lot of detail falls out at the start, it’s never saved into short-term memory and so can’t be stored in longer-term memory, Bauer says.

And much memory, even in adults, is transitory. We keep it for a day, a week, and then let go of it. Remember what you ate for lunch last Tuesday?

“Adults encode, hold on to it, then it falls out," Bauer says. "The same thing happens with babies, but it falls out faster.”

“Think of when you’re cooking pasta,” she says. “When you’re making fettuccini, you can use a colander with big holes; when you’re making orzo, you use one with tiny holes.” Babies and adults can have the same experiences coming in, the same thousands of millions of orzo-sized details going into the colander, but babies are using the fettuccini colander, and most of the details are rushing out in the stream.

"Are there qualitative changes in those memories?" says Oakes. "Is it just something like, ‘Oh, well, our brain gets too full so you can’t remember everything’?

"What we're trying to understand is how memory changes over those first few years of life until we get the kinds of memories that we actually, as adults, can articulate."