Can you trust your memory?

If you were asked what you were doing during a 45 minute period on Saturday afternoon 3 weeks ago, how much would you remember? Unless there was something special happening, for example your birthday, you probably won't remember much.  Cambridge University's Jon Simons explains to Chris Smith why our memories are not like taking a video of our experiences that we can replay and how this can affect criminal investigations.

Chris -  How does memory actually work? How does my brain store a memory of what I've been doing, where I've been, what I've done?

Jon - Everything that we experience, everything that we witness is registered in the brain to some degree. But it's only really if we're aware that something is important that we'll process it further and pay attention to it, and then that information can then be processed in further parts of the brain.

Chris - What actually physically is a memory?

Jon - A memory is a set of neural processes that occur that contain the features of an experienced event so for example, the things that we've seen, the things that we've heard, smells, perhaps our thoughts and reactions when we experienced the event as well. And generally, this is represented in various different parts of the brain. So, there are bits of the brain towards the back for example that process visual information, parts of the brain around the ears that process auditory information for example. And we store those features there. And then when we're trying to remember an event, we use a bit called the hippocampus which is towards the middle of the brain. That is involved in bringing together all these different features and kind of binding them together into a single memory so we're able to relive a previous experience as it unfolded.

Chris - Is it a bit like then if I go to the library and I want to know - this is in the good old days, not so much now. You go to the library and you say, "I want to know where book X is" and they've got a sort of card file system and you flick through the card file and it tells you which part of the shelf you will find the book you're looking for is a part of your brain like the card file and it routes your request for the memory to the part of the brain that stores that particular recollection.

Jon - That's what people think the hippocampus is probably doing. So, this is a region that's very much a kind of card file system. It's something that enables you to know where these different aspects of a memory might be stored and to be able to hopefully retrieve them reliably later on.

Chris - Now, the crux of this is, how accurate are those memories?

Jon - Memories vary in terms of accuracy. For one thing, if somebody is aware that something they're experiencing is important, then they might attend to it more closely. They might think about it more deeply and that's likely to lead to a more reliable memory later on. But then we know that as you said, memory is not about retrieving a video tape and replaying it. We would to think of it as a reconstructive process at the time of retrieval. So, at the time that we're trying to remember something, we go in and we try and retrieve those various small aspects of memory - a visual image or a sound or something. And we try to reconstruct that into a memory at the time of retrieval and therefore, the way in which we remember can be very much influenced by our expectations or our biases or our thoughts at the time of retrieval.

Chris - Well, they say a story always improves with the telling. So, the more you run over something in your mind, can you almost begin to persuade yourself that something that maybe wasn't true or wasn't there is true and was there?

Jon - Absolutely and that's a real problem of course for eye-witness testimony in that we have to be very careful when people are interviewing witnesses for example, not to introduce other information for example by some kind of a leading question that might influence their memory. And then as that is retrieved at multiple times, that can strengthen that memory and it can become a very, very confident memory and absolute, you know, short recollection that is, "I definitely did see that event in the way that is based on that leading question."

Chris - Actually, could it be that we're biasing our witnesses by the way we ask them questions and ask them to recall information in the environment in which we get them to do that recall in a stressful court. They could actually be remembering or misremembering what's going on.

Jon - Absolutely and this is something that has been a very big concern for quite a long time. now, police officers are trained to be very careful about these sort of things, be very careful about the way in which they question witnesses. And also, in court situations, lawyers and judges have to be very careful in the way that they would cross examine a witness for example to make sure that they're not biasing the way in which the witness remembers a previous event.

Chris - Is there anything practical we could do with this audience to convince them of how easy this is to do?

Jon - Absolutely. So, what we've got is a list of words we're going to try to get our audience to remember. I need a volunteer from the audience who has a very clear handwriting preferably.

Chris - What's your name?

Ciana - Ciana.

Chris - How old are you?

Ciana - 11.

Chris - You have immaculate handwriting we're told.

Ciana - Probably.

Chris - We're going to find out.

Jon - Okay, so what I'm going to ask you to do, I'm going to read out 15 words, a list of words. I want you just to write them on that piece of paper that's in front of you. Audience, I want you to try to remember these 15 words. I'm going to be testing your memory a little bit later on. Door, glass, pane, shade, ledge, sill, house, open, curtain, frame, view, breeze, sash, screen, shutter. Don't show the words to the audience for the moment. What we're going to do is test your memory for those words. So, what I'm going to do is I'm going to read you out some words and if you think that word was present on the list, I want you to shout, yes! Door. Was door present on the list?

Audience - Yes!

Chris - A quite convincing yes.

Jon - What about the word 'banana'?

Audience - No!

Jon - Pretty unanimous no there. What about the word 'window'?

Audience - Yes, .....no.

Jon - Very good. So, we've got some very good potential witnesses here, but I think probably - what do you think? 50% of the audience said yes to 'window'? there was certainly a gentleman at the front here who said no pretty quickly, so he'd make an excellent witness. So, what this shows is that even a word that wasn't on the list, most people can think and often can be very confident it was present. The reason for that is of course because all the words on the list related in some way to the word 'window'. And so, over those repeated instantiations of words that meant something to do with window then you find that you have this expectation that window must've been on that list. And so, most people are going to falsely remember the word 'window'.

Chris - So, we just prove to audience - Ciana, can you just hold up your list of words and you will see the word 'window' isn't there. How many people actually genuinely thought - just hold your hands up - who thought that the word 'window' was on the list you'd heard? More than half the audience thought that they'd heard the word 'window' and it's clearly not one that Ciana wrote down. Thank you very much. Give her a round of applause. What about the whole process of forgetting because there was a bit of research just been published showing that actually, when you recall one memory, this can make you actively forget another one? So, by asking people to dwell on just one set of what the police might think are relevant facts in a case now, in fact, they may be deleting from their memory other salient information that they want to later rely on in court but then it's gone.

Jon - When you remember something you actively supress competing or interfering memories that are similar to that memory. This can be a very useful function because it means that when we really want to remember something, we've got a better chance being able to remember that thing in the future and not sort of have our memory made unreliable by interfering competing kinds of information. But it does mean that if you're repeatedly questioned about the same thing over and over and over again, you're going to be less likely to be able to remember some other facts about that crime that might later actually be really useful bits of information that would've been useful to the police.

Chris - People here, what could they do, what can anyone of us do to make sure that we don't fall prey to misremembering things in the way that you've demonstrated that more than half of us do?

Jon - Well, it's very difficult. As we showed, it's a very difficult effect to - even if you know that this effect is present, it's something that's very difficult for us to avoid falling prey to. So, the only thing you can really do are to pay close attention to things you want to remember. If there's something you're really interested in remembering, thinking about it deeply, thinking about it again, with associated information, things that you know that for example, the person that you saw doing a crime might look like someone else you know for example. It might be useful strategy. Those sorts of things, relating it to other things you know are going to be useful. But also, write it down as soon as you can. If you think you've witnessed a crime and you might not be questioned for a long time, write down all the information that you can remember straightway because that's more likely to be accurate and reliable than something you might tell someone in a few hours' time.