The problem with dreams
28:11 Share

You think you know what working on your wellness sounds like, but there's one thing that truly sounds like the best thing you can do for your overall wellness.

Every great performance starts with a great night's sleep. And every great night's sleep starts with Natrol, the number one drug-free sleep aid brand in America. 100% drug-free and non-habit forming. Get yours now at Target or wherever you shop. These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. Ta-da! Pow!

Check, check. Hello. Hey, Noah. Hello, Meredith. So, as you know, I have the cutest dog of all time. Houdini. Houdini. We've met. He's adorable. He is the cutest little boy. And he's a tripod. He's missing his back left leg. He...

Lost his leg, I think, in a car accident. And we call him Houdini because we say that it's because he made his leg disappear. He really is like a magic dog. Yeah. He just hops around and has this big swinging tail. And it just looks like he's always leading a parade.

Like, we bring him to airports and just, like, every child in the terminal is just magnetically attracted to this adorable dog. Incredible. But there are some times when I wonder, like, you know, does he ever miss having his fourth leg? And I think the place where I see this the most is when he falls asleep and falls into a little puppy dream. He's yipping, all his paws are twitching. Oh, my God.

And so I've always thought that in his dreams, he must have four legs because his little stump is twitching along with the rest of his legs. Oh. So it's like, is he dreaming being a four-legged dog or is he three-legged in his dream? The question of exactly what Houdini is dreaming about might be impossible to answer. But Meredith is in very good company in connecting these twitches with dreams.

During the deepest stage of sleep, when we've got these twitches in our limbs, we've also got twitches in our eyes. Rapid eye movement.

"REM" for short. Scientists are pretty sure that REM sleep is when most dreams happen, at least in people, because they've done experiments where they've literally woken people up and asked them. Just sit up and try to remember what was going on before I came in. I was dreaming about school and I was flunking the grades, that's all I know.

But for animals, these twitches are some of the best evidence we've got. There is some scientific theory that when we see our cats and dogs dreaming or what looks like dreaming, almost barking in their sleep or moving their legs, that they probably are. The idea is that these twitches are bursts of activity that kind of break through the paralysis that happens during REM sleep, which gives us a glimpse of animal dreams.

And this has been accepted for a pretty long time. Darwin's protege actually wrote, quote, "'Ferrets dream, as I've frequently seen them when fast asleep, moving their noses and twitching their claws, as if in pursuit of rabbits.'"

The thing about Twitches is it looks like obviously it's related to dreams, right? Right. Mark Blumberg, neuroscientist, University of Iowa. I mean, we know we have dreams. We know that we are moving around. So it just makes sense to think, oh, movements, movements. Why wouldn't they be connected?

That's what Mark thought at first. But then he started seeing twitches in really young animals, like newborns. And he was like, a newborn animal has had very little waking experiences. What the hell are they dreaming about? If twitching was really related to dreaming, you'd expect that the older you get and the more experiences you have, the more you dream, the more you twitch. But that's not what happens. We move the most when we're young. So what we're dreaming the most when we're young?

In case it's not totally obvious by this point, Mark's kind of over dreams. I mean, can we please just not talk about dreams all the time? You know, that's sort of, sorry, I'm a little bitter. I mean, look, dreams are fascinating, but the focus on dreams is kind of a distraction from what really matters.

So Mark decided to take dreams fully out of the picture by experimenting on newborn rats. We literally surgically disconnected the cortex from the other half of the brain. He cut off the part of the brain responsible for creating dreams. And the animals twitched completely normally. If these twitches were caused by dreams, they should have stopped. But we found no effect at all on twitches. And so I was like, okay, what is this about?

You know, this is happening, as we counted it up, hundreds of thousands of times per day. And pups are in the business of growing, not wasting energy. And, you know, hundreds of thousands of movements, that takes up a lot of energy. Why would you do it unless it had some intrinsic value that had not yet been explored? I'm Noam Hassenfeld, and this week on Unexplainable, how this seemingly small question, why do we twitch in our sleep, has fundamentally shifted how we understand the relationship between the brain and the body.

♪

All right, Mark, just to make sure before we dive in here, when I think of sleep twitches, I think of those twitches I get like right when I'm falling asleep. Hypnic jerks, yeah. Is that part of this? Is that different? It's a separate phenomenon. It's more akin to what's called a startle than a twitch. You're not in REM sleep when that happens. And there are a lot of theories about it, but the fact is it's an extremely hard thing to study. So we don't really know what's happening there? No. Dang.

Okay, well, if we're just talking about these REM sleep twitches then...

How common are they? Do all kinds of animals and people twitch? Yeah. I mean, I've got a website that collects all these different videos, you know, and what you see across different animals is that the parts of the body that the animals really, really rely on for bringing sensory information into their brain are the parts that twitch the most. So for us, you know, rapid eye movements are twitches of the eyes. We also twitch our fingers a lot when we're adults.

With cats, you see their paws moving a lot. Ferrets, you see whisker twitches. Rats, you see lots of whisker twitches. They use their whiskers to learn about the world just as well as we use our eyes. And if all of these twitches aren't just, you know, enacting dreams, how do you start figuring out what they actually are? Well, you know, the first thing you have to do is try to figure out

What parts of the brain are producing this? I mean, how is this all happening? And what we started to see when we were recording brain activity is that the brains of neonates, baby rats, were much more active during sleep and much more active when animals were twitching than when they were awake.

Okay. It's one thing to think that sleep has brain activity associated with it. That was a huge finding 80 years ago. It's another thing entirely to see that the brain activity is greater, and I mean much greater, during REM sleep than during wake. And I assume it's reasonable to think all that brain activity is connected with these twitches, right? Is there a way to actually...

test it? Yeah. I mean, the biggest problem was methodological. How do you record brain activity in a very, very small baby rat, which was the best animal for doing this sort of work? You have to figure out how to get them in a stable situation so you can drop these very fine electrodes into the brain. And so it took years to get the methods going. But what we started to see is that every time the animal twitches, you know, 10 milliseconds later,

the part of the brain that's responsive to sensory input for that limb shows a huge burst of activity. So twitch, activity, twitch, activity, not the other way around. This is a sensory signal. So this timing here matters. If you have a twitch and then you get a burst of activity in the brain after that twitch, then you have a pretty good idea that that's a sensory signal that you're picking up on.

Yeah, so like a signal the brain is getting from a nerve or a muscle or something? Yeah, the sensory input. So every time you move a limb, you have sensors in your muscle, you have sensors in your skin and your joints. And those sensors, when you have movement, they produce neural signals that flow up into the brain. That's how we know when our arms are moving or when you touch something. So we have sensors all throughout our limbs. And so when the limbs were moving, that's when we were seeing the brain activity in parts of the brain that are responsive to

to those types of sensory signals. And just to be totally clear here, the activity that you saw in the brain was happening after the twitches. Yes. I mean, you can't really see it because it's happening so fast. You have to get down to milliseconds. So what my student, his name is Ryan Glanz, what he did was he was recording from the part of the cortex that responds to sensory feedback. And for every neuron that he was recording from, he gave it a different musical note

So that you could easily see that when the limb twitches, there was a burst of activity in this part of the cortex. And you have a video of that rat experiment, right? Yeah. Can we take a look? Sure. Wow. It sounds twitchy. So what I'm looking at here is this kind of little paw. And the twitchiness

twitching in the rat paw is kind of generating these sounds that are mapped onto neurons. Correct. So every time the limb twitches, there's a really, really clear burst of activity. And at these ages, all the neurons are firing together and that's why it just sounds like a chord and not a lot of like little different musical notes happening in sequence. And it's just a lot of activity happening simultaneously.

Wow. Okay. So you're essentially flipping the traditional hypothesis on its head, right? It's not dreams causing twitches. It's twitches causing dreams or twitches. I don't know. It's not necessarily twitches causing dreams, right? It's twitches causing some impact in the brain. Yeah. I mean, obviously, twitches are not going to be the sole source of all things in the dreaming brain, but that it is at least providing sensory input to the brain during sleep that we know for a fact.

So it does flip it on its head and it completely changes the calculus of what's happening in a dreaming brain. So then why would the twitching be happening to begin with? Like, what's the point of all of this twitching? Well, so this is where you have to start to think about, well, what is it that's special about twitches, right? The first thing that you notice is that the movements are discrete.

And it turns out that discreteness is incredibly important. So imagine that you're standing at a switchboard with hundreds of different switches. Let's just say they're neurons. And then all the wires from all of those switches lead to a whole bunch of lights. So every switch controls a different light. Okay? And let's say that those lights are muscles. If you're sitting at that switchboard and you want to figure out which switches control which lights, you don't just start throwing all the switches simultaneously.

Because if you did, you're gaining no information. All you're seeing is a bunch of lights turn on and you throw out a bunch of switches.

The answer is you throw one switch at a time, you see which light comes on, and then you make that connection. And so that's the difference between wake movements and twitches. You know, I'm sitting here talking to you and I'm gesturing and I'm moving all my limbs simultaneously, my posture, my neck, everything, my eyes, everything's moving simultaneously, right? That's waking. That's one of the characteristics of waking movements is that they're continuous and they're simultaneous and they're highly complex. But when you're twitching,

one twitch at a time, you ping your body and the body pings you back. And then you know that the first thing is related to the second thing. And that's the discreteness of twitching. And that explains why these animals are twitching so much. You never grow and develop more than you do when you are young. So they're literally, your theory is that the power goes out and they're flipping switches in a

fuse box or something to see which switch controls which light because there's no other stimuli coming in, right? They're in a controlled environment. So they're essentially doing sort of an experiment to learn their own boxwork.

Yeah, exactly. It's like they're bootstrapping their system. They're self-organizing their sensory motor system, and it's done from within. It's a big mystery as to how we develop things like our sensory motor system. How do you actually learn about your body? When you're a newborn rat or a human and you're born, you have no idea how your body is formed. You have no idea how it moves.

And it's going to be changing every single day as you grow and figure out new things, right? So how do you figure out how to move that body in real time through the process of development? You can't prescribe this. You can't blueprint this. There's no genetic mechanism that can tell you exactly how you're going to be on day three versus day five.

So you need to have a system that's highly adaptable. But you said something that's really important. You said, turn off the lights. And that's actually a metaphor I've used before because that's sort of what sleep paralysis is. Sleep paralysis is like turning off all that background noise, creating a very low noise situation for your body. And now when you ping it with a twitch,

you get a really, really clear signal back. So you shut down the lights, you turn off all that muscular activity, you paralyze the body, and then you just allow these individual twitches to go through. And then you take that information and you put that into the system for the purpose of maintaining your circuits, calibrating your circuits, so that you end up with a finely tuned sensory motor system so that we can function in the world.

And is that, you know, if twitching is about learning, we would assume younger animals would twitch more. Is that the case? Absolutely. And then...

I mean, older animals also twitch. Yes. Why would they be twitching? Good question. First, we don't twitch as much when we're older. But second, some animals do twitch quite a lot. And the part of the body that twitches matters. And this is just a theory because nobody has really explored it with the level of sophistication that we need. But

we have to calibrate our systems. You know, over the day we get tired, we lose control, you know, our vision gets worse and worse through the day, and then you wake up the next day and you're rejuvenated. I think it's possible that twitches continue throughout life for some parts of the body for that purpose, to calibrate a weary system. And there's some hints out there in the world that this could be happening, including work that was done in humans, but they're mostly hints, and it needs to be done more systematically. ♪

And why do you think the scientific community missed this for so long, missed understanding twitches as a developmental process? Because when you label something as a byproduct of dreams, why would anybody spend their time studying it? Like it's just closing off further inquiry? It's just being like done? Yeah. I mean, look, I don't want to be too flippant about it. Dreams are fascinating, but they're kind of a red herring when it comes to studying sleep. There is, to my mind...

Many, many fascinating things about sleep that have nothing to do with dreams. And the focus on dreams is kind of a distraction from what really matters. We all know what that 1 p.m. meeting sounds like when you didn't get a good night's sleep. And you know that's going to lead to a follow-up meeting.

What do you do for a living? Stream on Twitch, my cooking show.

What the f*** is Twitch? So Mark, if we take a step back from Twitches here and just talk about sleep in general for a second, what would you say people get wrong about sleep?

I think one thing that people may get wrong about sleep is they think about it as a single thing. They think about it as just a unitary phenomenon. And I think the jury's still out as to what sleep actually is. It's very highly variable across the lifespan, highly variable across different species. And there's no singular definition of sleep that applies. Plus, you know, I liken it to wake. I mean, there's no singular function for wake. Wake is not a singular phenomenon. We do all kinds of things in it. We walk, we talk, we eat, we sleep.

We watch TV. We do all kinds of things when we're awake. Why do we think that wake would be any different than sleep? To my mind, sleep is a conglomeration of things, all these separate components. And you've got to think about, well, why are all these different parts there? What brings all these things together during sleep? Is it that all these different parts of sleep are like the place settings at a dinner table? You know, so you have your fork, your spoon, and your knife.

They all serve one purpose, you know, to eat. But what if they're also more like all the collection of tools in a Swiss army knife? You know, you've got a fish scaler and a magnifying glass and a toothpick and a nail file. What the hell do those things have to do with each other? Nothing except they're all together in one place. So is sleep more like a dining room setting or is it more like a Swiss army knife?

And the answer is, it's probably a little bit of both. And our task is to figure out how all these different parts of sleep fit together. And that's the big question about sleep. Which parts of sleep do these different animals exhibit? How does it contribute to their survival? How does it contribute to their learning? How does it contribute to their evolution? To me, that's the bigger question about sleep. Can you tell me about some of those tools in the toolbox, aside from twitching? Sure. I mean, you know, the more we look...

the more we discover new things. You know, we've only scratched the surface of sleep. But there's the glymphatic system, which is basically being called like a waste clearance system for the brain to remove debris and to remove dead cells. And you have fundamental brain rhythms, for example. You have a rhythm that's involved in movement and learning and things of that nature. Okay.

And then there's a whole bunch of animal and human work that's been done looking at the role of these very specific brain rhythms for the consolidation of memory. So sleep is playing a major role in memory consolidation. And the sleep theory of memory consolidation is so powerful because the idea is that you just can't consolidate memories when you're in the process of learning them. So you have to stop learning new memories in order to consolidate the ones you already have. You know, I've been getting back into playing piano again, and I've been...

going through some Bach inventions. And I feel like after I sleep sometimes, it's like more deeply in me. Does that...

makes sense? Is that something like memory consolidation? It absolutely is. So I have the same thing that happens to me. I'm a drummer, and when I learn a new pattern... Oh, I'm also a drummer. I'm more of a drummer. That's my training. So tell me if you can relate to this. So you know when you start learning a very complex new pattern, it's all very deliberate, slow, and rote, right? Boom, boom. Every limb doing something. And you can't do anything else. I mean, you're concentrating on every limb and what they're doing. And then all of a sudden...

Six days in, I don't know. It's like, right? Everything's happening. And at the same time, you can hold a conversation with somebody, you know, right? Right. It's like gone to a deeper level in your understanding. Yes, it's called automaticity. And, you know, it's a pet hypothesis. It has never been confirmed or tested seriously yet.

But I just think wouldn't that be amazing if sleep were involved in that kind of skill transfer. You're taking it that's something very conscious, you're highly attentive to it and turning it into something that you can do on its own. And, you know, maybe that's how we learn how to do things like walk.

I mean, they start off as pretty difficult, but eventually we don't think about those anymore either. And we're actually doing an experiment that's in the dancing realm with some colleagues out in California. And we want to know when people develop these automatic behaviors, does that show up somehow in their sleep patterns? Because

The development of automaticity is a very, very natural place to think about how twitching and these sorts of sleep-related phenomena might be playing a role in the consolidation of those motor memories that happen, like drumming and playing Bach and stuff like that. Yeah, I think, tell me if this is a wrong way to think about it, but I think about the New York subway system, which is...

you know, the biggest 24-hour subway system in the world. And rats are very important for the subway system, too. Very important. Rats are important. But the 24-hour operation of the subway is kind of one of the reasons why the subway might suck as much as it does, just because it's harder to fix stuff. Like other cities where even if you could just work from like 2 to 6 a.m.,

You can fix some stuff, but in New York, if they really want to fix stuff, they have to just do major surgery. They have to, like, shut down a line. And so a lot of the subway just sort of falls into disrepair. That's a wonderful metaphor. You know, there are just some things we do in wake that seem just obviously to be incompatible with the things that need to happen during sleep. Yeah.

And so we take our day and we break it up into two very different states. Here's our moving about the world and being friends and talking and eating and doing things. That's one part of our lives. And the other part of our lives is somehow connected to the first part, but is very different. And they have to be complementary so that the sleep is supporting what we need to do when we're awake.

consolidating memories, building proteins, repairing systems. All of those things need to go on, but they're just incompatible with wakefulness. So then if I were to bring twitches back into the conversation here, are they kind of emblematic of this larger lights off, recalibrating learning process of sleep? Is that what stands out to you? You know,

I'm a behaviorally oriented neuroscientist. I want to understand the role of behavior, how it develops and how it influences the developmental process. And we have a lot of ideas out there right now in the world, what I would call very simplistic ideas. People think about, well, you know, you have a genetic blueprint and the animal just develops and then it's like,

blah, blah, blah, okay? And these are just corrosive ideas for thinking about development because development is a process. It's extremely complicated and it's extremely plastic. You know, brains are not isolated, separate, standalone organs. They're embodied. There's a reason why

developing robots is a lot more difficult than it is just to develop like AI that has no body to control. And it's because controlling a body is hard. And the one thing that robots can't do yet is develop.

And I think that there's something fundamentally important about developing in a body that changes and learning how to adapt to those changes. That is partly why we are so facile in terms of how we use our bodies on a regular basis. So for me, it's an extension of everything I've ever believed about the developmental process and what we need to do as developmentalists to understand that process, as opposed to the more sort of static ways that oftentimes people think about what's happening.

Wait, so if we want to make robots that can develop and learn better, we just have to make robots that can sleep? Yes. There are people who have used twitching in robots to produce more adaptable robots. Oh. They've mimicked twitches in robots and have produced robots that were better able to adapt to changes in their bodies. Okay. Crazy stuff.

Now you're going to ask do robots dream? And I'm going to, my head's going to explode. Yeah, do do androids dream of electric sheep? That was Professor Mark Blumberg. And fun story, this episode came about because Mark listened to the episode we did a while back on the baseball player who had the yips. You know, this kind of performance anxiety that people used to think was only psychological, but is now sometimes being connected to involuntary twitches.

So Mark reached out to me and told me the episode had given him some new research ideas on twitches. We started talking and then we ended up with a whole episode on sleep twitches. So if you're a scientist out there and you're listening and you're getting an idea for some new research, let us know. We're always on the lookout for more interesting scientific questions and we'd love to talk to you.

This episode was produced by me, Noam Hassenfeld. We had editing from Meredith Hodnot, who runs the show. Mixing and sound design from Christian Ayala. Music from me. And fact-checking from Katie Penzemug. Thomas Liu is wondering why things make so much sense. And Bird Pinkerton headed back to the octopus hospital, knowing she had her army to fight the birds. But when she got there, she saw the door blown out. Windows were smashed. Cables were pulled out, hanging everywhere. She was too late.

Thanks as always to Brian Resnick for co-creating the show. And if you have thoughts about the show, send us an email. We're at unexplainable at Vox.com. And you can also leave us a review or a rating wherever you listen. It really helps us find new listeners. You can also support the show and all of Vox's journalism by joining our membership program today. You can go to Vox.com slash members to sign up. And if you signed up because of us, send us a note. We'd really love to hear from you.

Unexplainable is part of the Vox Media Podcast Network, and we'll be back next week. Having a great day starts with having a great night, and that can lead to a great year. Better sleep can not only improve your overall wellness, it can also help your productivity while you're awake. Every great performance starts with a great night's sleep, and every great night's sleep starts with Natchaw, the number one drug-free sleep aid brand in America. 100% drug-free and non-habit forming. Get yours now at Target or wherever you shop.

These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure or prevent any disease.