Why we can't remember being a baby

Have you ever wondered why you can’t remember being a baby? New research suggests it’s not because we didn’t form memories at that time, but because we can’t access them. The study found that babies do indeed appear to be encoding memories in the brain’s hippocampus from very early on. This agrees with other previous studies done using mice, and it leads to the tantalising hypothesis that all of us may potentially have infant memories locked away in our heads that we can’t directly recall. Here’s Nick Turk-Browne, a professor of neuroscience at Yale University…

Nick - We learn an astounding amount as infants. It's in fact, I think, the greatest period of learning in life. And so it's particularly striking in that context that we're lacking a specific kind of memory from that time, which is memory for specific events in space and time. That's the kind of memory we're studying here. It's called episodic memory. That kind of memory is what's missing from early life. In this case, we tried to hone in on when this ability to store specific memories for events in our life comes online.

Chris - So were you actually brain scanning babies to do this?

Nick - Yeah, we have been working for several years to figure out how to do this kind of research in babies. And this is very challenging. They're the worst possible study participants for this kind of a method because you can't move a millimetre. Imagine taking a photograph in low light with your phone. If the subject moves at all, you get a blurry image. It's very similar in the case of functional MRI. If the baby moves, then we can't figure out what's going on in their brain. They're also challenging because they have very short attention spans. And to perform these kinds of experiments, you typically need a lot of data. But in a baby, we're lucky if we get five or 10 minutes of data. So we had to design really engaging tasks that the baby could do to probe their memory.

Chris - Talk us through what the experimental design was then. What did the babies have to do that enabled you to get at this question of, well, are there memory circuits for episodic memories? What I ate for dinner last night, are they working or not?

Nick - Well, most of the babies we're working with were pre-verbal. They can't speak and possibly can't fully understand language yet. And so you can't tell them what to do or ask them to do things. So we had to come up with indirect ways of probing what a baby remembers. And so in this case, we're going to use a measure called looking time. Where a baby looks sort of reveals what they know. And so we can use patterns of looking time to figure out what it is that they remember. The experiment involves showing them photographs of faces and objects and places. This is while they're lying in the fMRI machine and we're recording their brain activity. And then we test their memory by showing them two photographs at a time. And one of those photographs was something that they had seen earlier in the experiment. And the other photograph was another image from the same category. So it might be the same woman's face paired with another face that they hadn't seen before. If they remember having seen that woman's face before, they'll look longer at that photograph than at the new one. And so with this, we can quantify how strong their memory was for that picture. And based on that measure of where they're looking and their memory, we can go back and say what was happening in the brain when they first saw that picture. And we can compare brain activity during that period of storage with what was happening in their brain when they saw a picture that they later didn't remember. And so that's the basic logic of this experiment is to compare brain activity during encoding. That is the initial storage of a memory for things that are later remembered in this case, looked at more versus later forgotten.

Chris - And where do you, based on these experiments, see the bottleneck then? Is it that they don't encode or make the memory in the first place? Or is it that there is some kind of issue with encoding that into long-term memory? Or is there a third possibility? They can do all those things, but they just don't know where they encoded the long-term memory, so they can't get it back.

Nick - Yeah. What we demonstrate in this particular study is that hippocampal activity is greater when babies are looking at something for the first time that they later remember. And so what this study establishes is that beginning around 12 months of age, the infant hippocampus has the capacity to encode memories. What this means is that our inability to remember early life, or what's called infantile amnesia, is not a deficit of storage. Memories are getting into the system, but how long they last and why they're not accessible are the next questions that we're asking.

Chris - And why do you think that might be the case?

Nick - Yeah. Well, there's some really beautiful animal research in mice where you have a baby mouse, they have some experience using molecular techniques. You can tag which neurons help to store that baby memory. And then when the mouse grows up to maturity, which in the case of a mouse is just a few months, then you can stimulate those neurons that were part of the memory that was formed in the baby mouse, and they'll express the memory. So there's evidence in these experiments that the mice are able to form memories as a baby that persist until when they're adults. And so what this suggests is that at the limit, this may be possible in humans that into some period of childhood or perhaps beyond, that we may still have these memories in the hippocampus, but that they're not accessible. So what's going on is a deep mystery. There's one data point in the mouse experiment that's really, or mice experiments that's really critical, which is if you just put the mouse back in the environment where they formed the memory as a baby, now when they're an adult, they don't show memory. So just being back in the environment wasn't enough. Their hippocampal memory had to be stimulated directly. So what this tells us is that in humans, as we grow up, we may have memories in our brain that are inaccessible because the cues from the environment might not be enough to retrieve the memory. And this is the hypothesis that we're developing now. This is just a speculative theory that, whereas people originally thought that infantile amnesia was because the hippocampus is immature, I actually think what might be going on is that the rest of the brains develop and may be responsible.