By 
A few months back, I caught up with a childhood friend. We were reminiscing on the days of our youth and he told the story of my first day at school. Here's the short version: I was not happy to be there and knocked over a table of student projects, or something to that effect. Then, my mother had to come to calm me down.
That would be a slightly embarrassing story except for the fact that I remember none of it. You'd think something like that would have stuck in my mind, so I asked my mother, who I thought would confirm the story.
She didn't. She couldn't recall it happening either.
I still don't know what happened on my first day of school, but until I asked my mother, I'd taken my friend's word for it. I've done the same with my mother's stories for years.
It got me thinking about memories. I have 'memories' but are they actually mine? Or have they been constructed from the stories I was told?
There's quite a bit of research on this and it suggests that the stories we’re told can indeed end up feeling like our own memories.
In one well-known 1995 study, Elizabeth Loftus and Jacqueline Pickrell had relatives describe a few events from people’s childhoods, one of which never happened, and some of those people came to remember it anyway.
So other people can put memories in our heads. That’s strange in itself. But what surprised me more is that it seems we don’t need anyone else. We do a pretty good job of rewriting our memories on our own.
In January 1986, the space shuttle Challenger broke apart just over a minute after launch. The morning after, Ulric Neisser, a cognitive psychologist, and his collaborator Nicole Harsch asked 106 undergraduates questions about the event.
Almost three years later, Neisser went back and found 44 of those students. He asked them the same questions, and this time he also asked how confident they were in their answers.
Around a quarter of the students gave completely different accounts from the ones they’d written. Neisser was surprised. “I never thought they’d be this wrong,” he told the Associated Press.
A lot of the wrong memories had something in common. Many students incorrectly remembered hearing the news on TV. Most of them probably did watch the coverage later that day, again and again. For many, the TV part was probably real; it had just moved to the wrong moment.
It seems that none of the students were lying. They fully believed what they were saying, and they were sure about it.
So the researchers did what you might expect. A few months later, in spring 1989, they brought some of the participants back for interviews, gave them hints, and finally put their reports in front of them. You’d think that would be the end of it. Surely, they would recognise what they’d written and correct their account of the event.
They didn't. 12 received a score of zero out of seven for accuracy.
I’d believed my friend’s version of my first day at school without giving it much thought. These students didn’t even believe themselves.
And all that in under three years. What happens to these memories over a longer stretch of time?
After the attacks on 11 September 2001, a group of memory researchers led by William Hirst saw a chance to do what Neisser did, but on a much bigger scale. Within a week of the attacks, they got more than 2,000 people in the US to fill in a survey about how they heard the news. Then they asked them again about a year later, three years later and finally ten years later.
Of course, people dropped off over the years but around 200 people answered all four surveys; almost five times as many people as in Neisser's study and over a much longer period.
So what did Hirst find?
The first year was when people's memories changed the most. More than a third of the details people gave were different from what they’d said that first week.
After that, things settled down. Between year three and year ten, their memories barely changed at all. When a detail changed, it usually stayed changed. They repeated the new version in the next survey, and the one after that.
One person’s answers illustrate this well. The week of the attacks, they said they were in the kitchen making breakfast. A year later, they were in their dorm room folding laundry. Three years later, they were ironing in their dorm room. Ten years later? Still ironing.
Breakfast had turned into ironing, and it stayed that way for the best part of a decade!
Not all mistakes stuck, though. The researchers also asked about the attacks themselves, like how many planes there were and where they crashed. When people got those central facts wrong, they usually corrected them later.
So why would one kind of memory get fixed and the other not?
The researchers’ answer is pretty simple. The facts of 9/11 were repeated over and over in the news, in films and in conversations, so mistakes got corrected. As they put it:
"External reminders are present for event memories, and hence, in the case of the media coverage of 9/11 in the United States, corrections prevailed."
But nobody broadcasts where you were when you heard. That part of the story mostly exists in your head, so there's little to check it against. Once people settled on a version, as the researchers put it, “this memory became their story, regardless of any errors it might include.”
There may well be something going on in the brain here. Hirst and his colleagues thought that their results lined up with two processes that neuroscientists study.
The first is consolidation. This is where a memory gradually settles into long-term storage and becomes harder to disturb. This could explain why people’s stories stopped changing after the first few years.
The second is reconsolidation: the idea that a memory can be reshaped when we bring it back to mind. Perhaps that's how the news and films managed to fix people’s mistakes about the facts, even years later.
It’s worth saying this is the researchers’ interpretation. They didn’t scan anyone’s brain, and the evidence for reconsolidation in people is still debated. But it’s a neat way of explaining how a memory can be rock solid and still open to change.
This brings me back to my first day at school. There was no survey that morning, and it certainly didn’t make the news. All I’ve got is a friend’s story and a mother who can’t remember.
So what does that mean for the stories I’ve been told over the years?
Honestly, I’m not sure. Some of them are probably true. My mother was there for most of my childhood, so her version is likely better than mine. But some of them might have become ‘memories’ simply because I heard them enough times.
From the inside, I can’t tell which is which.
That’s really what both of these studies point to. We only know Neisser’s students' stories changed because someone asked them to write it down the morning after. Most of our memories don’t have a page like that. There’s nothing to check them against, so we just go with the version we’ve got.
And if the studies are anything to go by, we probably feel sure about it. The people in Hirst’s study stayed confident for ten years in stories that had quietly changed in the first year. It seems that feeling certain about a memory and actually being right are two very different things.
As for my first day at school, I’ll probably never know what happened. Maybe I did knock over a table of projects. Maybe my friend has it mixed up with someone else’s first day. I’ve still got my ‘memories’. I’m just a lot less sure where they came from.
