One of the classic findings in memory research is that we’re better at remembering information when we’re in a similar context to that in which we learned it. This was perhaps most famously demonstrated in a 1975 study, which found that people who learned a list of words while scuba diving had better memory for the words when again underwater, compared to when on land (similarly, those who had learned the list on land were better at remembering it on land).
But it’s not just the external environment that matters: our internal states can also provide memory cues. For instance, people who were intoxicated when learning information were better at recalling that information when drunk than when sober, and there’s also evidence that our recall is better when our mood matches how we felt at learning.
Now a new study published in Child Development has found that the same is true even of babies in their first year of life. The findings have implications for understanding infant memory — and could even help to explain why we can’t remember anything from our early years.
To look at state-dependent memory in babies, Sabine Seehagen and colleagues from Ruhr University Bochum in Germany recruited 96 nine-month-old infants and their caregivers. In the learning phase of the experiment, the babies watched a researcher perform a series of actions with an animal puppet: removing the puppet’s mitten, shaking it three times, then replacing it.
After a short break, there was a test phase where the infant had the opportunity to play with the puppet themselves. They had 90 seconds to do any of the three actions — i.e. remove the mitten, shake it, and replace it — though they weren’t prompted to do so in any way.
Crucially, immediately before the learning and testing phases, the babies had a 5 minute play session designed to put them in a specific state of mind. Sometimes caregivers were instructed to play as calmly as possible with their infant and they were given quiet toys like a book and soft toy. Other times they were instructed to play as animatedly as possible, and given more active toys like a ball and swing.
For one group of infants, the play sessions were the same before both learning and testing phases (i.e. calm both times, or animated both times). In another group, the play sessions were different (i.e. calm at learning and animated at testing, or vice-versa). And a third group only did the testing phase, without the learning part (this allowed the researchers to look at how often the babies performed the actions just by chance).
The team found that infants who had the same kind of play before both learning and testing made an average of 1.47 actions during testing. This was significantly more than those whose play sessions were different, who only made an average of 0.59 actions. In fact, the babies in the latter condition didn’t seem to remember the actions at all: they performed no better than the babies who hadn’t even done the learning part of the study.
The findings suggest that internal states do provide important cues for babies when they’re retrieving memories. And that seems to make sense: while older children and adults have a wealth of knowledge and experiences which allow them to use a variety of cues to trigger memories, young infants haven’t yet formed those connections. Relying on internal states to aid recall is therefore “highly adaptive”, the authors say.
This reliance on internal cues might also contribute to “infantile amnesia”, the fact that we’re unable to recall anything from our earliest years. As we grow our internal state changes considerably, and is never again going to match that of our first years of life — so, like the babies in the “different state” condition, we’re not able to access our memories.
It’s worth pointing out that the research didn’t directly measure babies’ internal states, as the team acknowledges. During the animated play sessions, the babies did show significantly more movement and active behaviour than during the calm sessions. But further work will be required to see whether memory recall is truly dependent on infants’ emotions — it could be that the matching types of activity were somehow providing cues to the babies even without influencing their mood.