Here is the full transcript of professor Dr. Robert Stickgold’s TEDx Talk on Sleep, Memory and Dreams: Fitting the Pieces Together at TEDxRiverCity conference. Dr. Stickgold is the Director of Harvard’s Center for Sleep and Cognition.
TRANSCRIPT:
This talk is basically about what sleep is doing for us. There’s this common belief that when you are asleep, the brain shuts off, the body lies down and nothing happens. And that’s really not what’s going on. It’s a little bit like a fairy tale, and in fact, “The Elves and the Shoemaker” fairy tale is about my story tonight.
So, here’s the story: The shoemaker was a poor shoemaker, had just enough leather left for one pair of shoes. He cut out the leather for two pairs of shoes – excuse me – and then, as it was late, he left the pieces on the bench, ready to sew in the morning. But when morning came, two pairs of shoes lay on the bench, most beautifully made and no sign of any one who had been there.
What I want to tell you is that this is not a fairy tale. This is what happens for us each and every night. And that sleep, in fact, is sewing together the pieces of our memory. I’m going to walk you through four examples of this, just to show you how powerful it is. I’m going to show you how, when you sleep, your brain will extract from experiences the gist of what happened – it will extract the gist and kind of forget all those details that weren’t so important.
It’ll discover the rules of our lives. It will take large amounts of experiences and take them all together and figure out what the rules are that explain how our worlds work.
And dreaming, in fact, is probably part of this entire process, and I’ll show you some new findings that suggest that that’s true. So, let’s start with the idea of sleep extracting the gist. In this study, we had subjects listen to lists of words spoken to them. Imagine you’re in a room – and here are the instructions: you are going to hear a series of words, broken up into lists of about a dozen words each. We want you to remember the words because later on we’ll give you a blank piece of paper and ask you to write down all the words you can remember.
Now, I’ll give you an idea of what that feels like by taking you through one of those lists. Now, there’s going to be a test at the end, okay? And you are going to have to raise your hands. So, I’m going to give you a lesson in medicine. On the back of the arm, there’s a muscle, which is an extensor muscle that makes the arm go out, okay? This is not raising your hand. This is not raising your hand. You have to use that muscle and actually lift your hand because we are doing the experiment now.
Here’s the instructions: Watch these words as they go by. Remember them. You are going to be tested on them. [nurse] [sick] [lawyer] [medicine] [health] [hospital] [dentist] [physician] [ill] [patient] [office] [stethoscope] [cotton] Aaah – that’s not supposed to happen. Okay, end of list.
Normally, subjects would hear about a dozen of these lists and then be tested for recall. They would then come back, maybe twenty minutes later or maybe twelve hours later, and be tested on them. We’re going to do it now. If you saw each of these words, raise your hand, okay? [nurse] If you did see it. Okay? [sick] Lots of yeses. A lot of weak extensor muscles. Okay? Cotton, you see cotton? No, okay, nobody for cotton. [table] I don’t know why that plus is in the middle.
Doctor? Don’t say it, they were going to raise their hands. So, if I had done this competently, about half of you would have raised your hand for doctor, and you do that because the list is a gimmick list – we’ve put together a whole bunch of words related to doctor, but we don’t include the word “doctor” in the list. And it turns out, that about half of you will falsely remember that you saw the word “doctor” because your brain is busy extracting the gist of the list.
So, we asked the question if, instead of testing you right away, and we would have you write down those words, if you had to come back twelve hours later and do it, how much would you forget? And the question is what happens if we train you in the morning and test you in the evening, compared to training you in the evening and testing you the next morning, after a night of sleep?
So, what difference does sleep make to that kind of memory? Well, the first thing you see is that across the day or the night you forget somewhere between 25% and 45% of the words that you would have remembered, based on other subjects, after just 20 minutes, and you actually forget less over the night – so, that’s the important factor. You will remember more of this list if you’ve slept on it than if you’ve gone through your day after you first see it.
But that’s the words you actually saw. What happens to those words that you didn’t see? Those gist words? Well, across the day, you still forget about 20% of them. But over a night of sleep, you actually get a bit better at it. It’s as if when you sleep, your brain is figuring out what it’s about. It’s holding on to that information, and it’s throwing out all the rest.
So, over a series of a days, in fact, what we see is that all that people remember are those words that weren’t on the list because you’ve extracted that gist, and that’s what you hold on to. And as we do that in life, it looks like it’s sleep, not waking, that allows us to extract that gist. So, when we figure out what the take-home message is, we do it in our sleep.
Now, interestingly, subjects also come up with words that weren’t on the list and weren’t the gist words either. So, they come up with words like “blood,” which is probably from that doctor list. They think of it because of that. Someone else came up with the word “plate,” and you can see one of the lists was about cups. Or “spoon” when one of the lists was about cups. That’s pretty common when you do this study. But with this 12-hour interval, we started to see funny words popping up, like “fuzzy”. I mean, where did that come from? And we thought maybe from “rough” or from the list of words about soft, but it’s really not either of those. And we get the word “swirl”. Where did that come from?
And Jessica Payne, my student, said, “Well, maybe it came from ‘mountain,’ like they were imagining clouds or soft, like some swirly soft fabric.” But then some other people said, “Well, no, maybe it’s cup – like, you know, you swirl your coffee, or chair – you swirl your chair.” And it started to feel like what these people are doing is they are coming up with words that don’t just epitomize one list but are now sort of starting to bring all the lists together.
And when she looked, she finds out you’re not getting those creative intrusions, so much, when you’re awake, you’re getting them when you’re asleep. So, again, it’s when you’re sleeping that the brain is doing this work of pulling everything together and seeing how it fits together and how to summarize it.
There were a series of nine digits: 1, 4, 4, 9, 4, 9, 4 – is what it should have looked like, it’s that 9, 4 up there – and you have to process this list to get an answer. Here’s how you can do it: you can take the first two digits – if they’re the same, there’s a 1 and a 1 – see those are the top ones there? – if they’re the same, you just write that down – that’s your intermediate answer. And then you take the 1 and the 4 – and if they were the same, you’d write it down again, but because they’re different, you’d take the third possibility, which is 9, because there is only 1s, 4s and 9s, and you write down the 9. There we go.
Then you take the 9 and the 4, and the third possibility is anybody? 1. Because they’re 1s, 4s and 9s. Then you take the 1 and the 9 and you put down a 4, and you take the 4 and the 4 – they’re the same – you use the other rule, you put down 4 again. This is tricky because it’s really on the other line, you take the 4 and the 9 – that got put on the next line – and they’re different, so you put the 1 and the 4 and you get a 9. And that’s the answer and then you push “return,” and it tells you right or wrong and you go on. If you don’t quite understand it, neither do any of the subjects. But over a period of about ten trials, they get it.
And then over the next hundred trials, they get really bored. Because they’re just doing the same thing over and over by this rote method. Now it turns out, unbeknownst to them, there’s a trick, okay? The trick is that the last three numbers they generate, the 4, the 1 and the 9, in this case, are always a mirror image of the three answers before, a 9, and a 1 and a 4, right? That’s the same sequence backwards, which means that the second number you got, that 9, is the answer, and you can just hit “return” there, and get out of there about 80% faster, maybe even make it home in time for dinner. That’s the trick you’re supposed to discover. But you are not told that there’s a trick.
You’re not told there is an easier way to do it. The question is can you just figure it out. In the course of this hundred trials that students are practicing this in the first session, about 10% of them actually figure out the rule. They take those guys out back and shoot them, but the rest of them, they send away for 12 hours and bring them back, and then they give them another couple of hundred trials to see who can figure it out. What did they find out? If they trained people in the morning, and then test them that evening, 12 hours later, about a quarter of them, 25%, will figure it out over the course of those 200 trials.
If they train them in the evening and keep them awake all night and test them in the morning, you get about the same thing, about 25% figure it out. But if you let them sleep on it – that’s the only thing you are doing differently, you’re letting them sleep – two and a half times as many people figure it out. So, you gain these insights, when you didn’t even know there was an insight to find. Just by sleeping on it. It’s an amazing phenomenon.
It really is. It’s like, how does it do it? I want to end with a fourth example, which is how dreaming seems to be part of this because it actually could give us a theory of the brain, but I don’t have time for that tonight. This is an example where people are trying to learn to navigate in a three-dimensional maze – one of those computer games we don’t play but our kids play, where they’re moving around it. The trick is you have to learn, during training, how to navigate through this maze, and then five hours later, we test you on it and see how good you are at getting to certain places in the maze. There’s two groups of subjects: they’re all trained at noon, they’re all tested at five pm, and half of them get to take a nap. Okay? That’s the only difference – they get a nap Losers, right?
Well, it turns out that at retest, the people who didn’t take a nap take about a minute longer to get through the maze than it did at the end of training. And the people who took the nap are about a minute faster. So again here, and now only with a 90 minute nap, you’re seeing that the brain is actually getting better than what it had been able to do before. But this is a study done by a student of mine, Erin Wamsley, who’s really into dreams, so she added another study. She started waking people up from the naps and saying, “What are you dreaming?” And the people in the wake group, who are just sitting there and watching a video, she’d say, “What’s going through your mind?” And she asked, what about the people who are dreaming about it or not dreaming about it, or thinking about it or not thinking about it in the wake group?
In the wake group, it really doesn’t make any difference if they were thinking about it or not – they’re still not doing much better than they were at the end of training. Actually, those in the nap group who didn’t report dreaming about it, weren’t particularly doing better, either. But the ones who reported dreaming about it, were showing ten times more improvement.
So, somehow, if nothing else, these dreams are a marker – a predictor – that the brain is going to do what it needs to do to really figure this maze out. Now, we got so excited; we started looking at these reports. Here’s what the reports from the wake subjects looked like: “Well, I was thinking about the game I used to play in high school, Counter-Strike, because it’s the same layout, and also I was just planning and trying to remember the maze and trying to figure out the route.” Or, another one says, “I was thinking about what we have to do in the second maze test, wondered if it was going to be, like, the same.” Those could help; you could imagine thinking about it that way; it could help.
But it didn’t. Those people in the red bar didn’t get very much better at all. So, what were the dreams that were so helpful? “Well, I was thinking about the maze and kind of having people at checkpoints. Um, there is no checkpoints and no people in the maze I guess”, and then that led me to think about when I went on this trip, a few years ago, and we went down to these bat caves and how they were kind of like mazes. Great. Alright. How about this one? You like this one? “Well, I was kind of looking for something in a maze” Or, how about this one? “I was just hearing the music from the game.” These are not dreams that you would expect to help. But those are the people who showed ten times more improvement.
And what that seems to tell us is that the brain is processing this information as a whole brain. You know we got the Whole Food stores? We got the whole brain store here. And that when you are asleep, the brain is processing this information at multiple levels. Part of the brain is stabilizing and enhancing those memories and integrating all the different information from all your exploring in the maze – and that’s probably happening down in that ancient brain that they were talking about: the hippocampus is the structure that seems to hold that kind of information. It’s extracting rules and the gist, maybe getting a sense of how that maze works.
That sort of work in the brain is what is going to make you better at it after you wake up. But that’s not what you dream about. What you dream about are other things the brain is doing. It’s integrating new and old memories, so that, in the future, when you are in a cave, you’ll say, “You know what? I’ve navigated through a maze once and I’ve learned some tricks about how you do this.” And that information from the maze game is suddenly useful when you’re down in that bat cave.
It imagines possible futures – that’s something the brain does when you are sleeping. And it’s these last two that is what we see in our dreams. Those are the brain processes that we have access to, that we perceive, that we think are happening when we dream. If you take all of these things together, what you’re doing is you’re creating the meaning of your lives. Because when we talk about the meaning of events in our life, what we’re really talking about is how this event fits in with everything else.
Remember after September 11th, we all walked around, saying, “I don’t understand what happened?” Well guys, it was easy – big plane, lots of fuel, high-speed, big building – I mean, what’s to understand? Well, what’s to understand is: “I work in a twelve-story building, should I go to work tomorrow?” “I got plane tickets to Amsterdam in a week Should I cancel my flight?” “I’m getting on the subway and there’s a guy in front of me, who kind of looks Arab – whatever that means – and he’s got a big backpack .on Should I turn and run?”
When we create meaning in our lives, we learn how to take those events in our life and fit them in with all the other events in our life, and use that, as the first clip showed, to predict the future. Of course, after September 11th, none of us could predict the future, and we felt that. But that’s what our dreams do, and that’s what our brain does while we sleep.
So, like a nocturnal elf, our sleep sews those pieces together, leaving us, in the morning, with something better than a pair of shoes – with an understanding of our lives. So, get some sleep. Thank you.
Related Posts
- How to Make 2026 the Best Year: 6 Questions to Ask Yourself – Mel Robbins (Transcript)
- Dr. Karl Pillemer on The Mel Robbins Podcast (Transcript)
- Transcript: Denzel Washington On Legacy, Wife & Purpose – NXT Chapter Podcast
- How to Treat Men Better – Alison Armstrong on Modern Wisdom Podcast (Transcript)
- Why Young Men Are Struggling Right Now with Oprah & Scott Galloway (Transcript)