Do Birds Sing In Their Dreams?

JOHN DANKOSKY: This is Science Friday. I’m John Dankosky.

MAGGIE KOERTH: And I’m Maggie Koerth. Birds sleep, perchance to dream. But what about? We’ve got more insight into this now thanks to researchers from the University of Buenos Aires.

It turns out the syrinx, a bird’s voicebox, silently moves as it slumbers, kind of like the way your sleeping dog might move its paws to run. And these scientists have figured out a way to translate that movement into song. We can now effectively listen to birds talk in their sleep, and that gives us some clues about what’s going on in their heads.

This is all published in an amazingly titled journal called Chaos. And joining me now to talk more about his fascinating work is my guest, Dr. Gabriel Mindlin, professor of physics at the University of Buenos Aires in Buenos Aires, Argentina.

So animals sleep. Animals dream. Or at least they have brain activity that would mean dreaming if we saw it in humans. And that includes birds. Your team found that birds or moving the syrinx in their sleep. So tell me how you found this and what you saw.

GABRIEL MINDLIN: We have been studying the physics of birds for many years. And the framework of the studies, we had to measure the physiological instructions that are changing the configuration of the vocal organ while they sleep, also the respiratory activity when they are producing their sounds. Typically, we are interested in measuring that during the day when the bird was vocalizing.

But one time, one of those birds stayed connected. And when my colleague, we realized that there were those patterns of activity that were happening at night as well.

MAGGIE KOERTH: What you just said, it sounds like you’re saying that measuring the dreams was actually an accident.

GABRIEL MINDLIN: Yes.

MAGGIE KOERTH: Oh, that’s so cool.

GABRIEL MINDLIN: Yes, for us, it was an accident. So my other colleague, Dan Margoliash, at the University of Chicago, he’s a neuroscientist. Then we became collaborators. But he’s the one who first recorded the neural activity at night.

So in a way, we’re now trying to work together, linking the neural part and the periphery part and designing new cool experiments, like, for example, having a bird sleeping, listening to actually what is dreaming, and seeing what impact it has on the behavior the next day.

MAGGIE KOERTH: What are you using to measure these movements?

GABRIEL MINDLIN: So what we do is we do some delicate surgeries where some electrodes are implanted in those tiny muscles. Those surgeries are very delicate. And they have to be performed well enough so that the bird, when it wakes up, is in the mood of singing to the female. So those are delicate surgeries where tiny electrodes are implanted in the masses.

MAGGIE KOERTH: You’re then taking this data of the muscle movements, and you’re turning it into song. You’re using a computer model to translate those muscle twitches into music. So how do you train that model? Are you starting with a bunch of different measurement data from awake birds?

GABRIEL MINDLIN: Yes. Basically that’s research that we’ve been doing for several years now. It consists in writing the actual physical equations that rule the behavior of the labia. The labia, the avian vocal organ, are very similar to the vocal folds in humans. So when the bird is singing, it’s changing the configuration of the vocal organ.

And they can stretch or release those labia. And that will affect the frequency at which they oscillate, therefore, the frequency of the sound that is produced. But basically, you start with physics, which is very simple, like Newton’s equations for the motion of those labia. And you try to find which are all the forces that are acting there and how the activity of the muscles would affect those forces.

In that way, you can reproduce the motion of the labia and, therefore, the modulations of the airflow. And that is what generates the sound.

MAGGIE KOERTH: That sounds fascinating. So I want to listen to what a bird dream sounds like this. One is from the bird you’ve been using for research, the great kiskadee.

[ELECTRONIC CHIRPING]

So now let’s listen to what that same song sounds like when the bird is awake.

[CHIRPING]

So those sound a little bit different. Can you tell from the synthesized song what this bird was dreaming about? Does it really translate easily between awake and asleep?

GABRIEL MINDLIN: So the first thing is that sound that you played at the beginning is a call from a different bird. So probably if you would be playing the syllable corresponding to the actual bird that you are measuring, you would find it more similar. But still, there are differences in between what the bird dreams and what the bird sings.

In this bird species, what they do is they use two sound sources. And the muscles control how they tune the two sound sources are. They try to sound intimidating by generating these rough sounds. And they do that by changing slightly the frequency of the vocalizations that they produce with the two sides of the CDs.

When they dream, the difference between that frequency is a little bit larger. Therefore, the syllable sounds rougher. That’s the main difference in between the actual song and the dreamt, the replayed song, yeah.

MAGGIE KOERTH: Do you– are you able to figure out what the what the dream is about and correlate that to awake behavior?

GABRIEL MINDLIN: Yeah. In this species in particular, it’s possible because they have basically two sets of vocalizations. One are these calls, these very simple calls. And they have another vocal behavior, which is called the trill when they produce several syllables in a very rapid sequence. And they usually do that when they are having a territorial dispute.

And it’s very interesting because, when they are actually having this dispute on top of singing and performing this intimidating trill, they also have these feathers in the head that come up. It’s very intimidating.

So when we recorded for the first time the muscle activity when the bird was producing these kind of vocalizations, the bird was completely asleep, silent with the eyes closed. But still the feathers of the head would come up.

MAGGIE KOERTH: Oh, that’s great.

GABRIEL MINDLIN: Yeah. So you could figure out that, really, this guy was experiencing a nightmare probably, recreating the whole experience of having a fight in his sleep.

MAGGIE KOERTH: That reminds me of my toddler having a nightmare about sharing. So how did you feel getting this window into a bird’s mind?

GABRIEL MINDLIN: It’s really exciting. The first time I did a synthesis for dreams, I was laughing the whole morning. I couldn’t stop laughing. I was very happy. I thought it was very cool. That was my first impression.

Then the second time I synthesized this vocal behavior, which was this nightmare, what I felt actually was a lot of empathy because I thought, well, this guy is belongs to a species that is so distant from us. But probably they’re all alone at night in a tree with his fears. And I felt that there were more similarities between them and us than we usually acknowledged.

MAGGIE KOERTH: Bird ennui. Are other bird species doing the same thing?

GABRIEL MINDLIN: Yes. So we also work with songbirds, which are birds that are a little bit more interesting in a way because they share with us. They need to be exposed to a tutor in order to learn their vocalizations like, for example, a canary, zebra finch, a cardinal, those are songbirds. There are 4,000 species of those birds. And

They need to be exposed to a tutor. They learn to vocalize. The kiskadee in the study you are mentioning is not a songbird. But with songbirds, it’s very interesting because, since they need to learn to vocalize, their behavior is a little bit more complex.

I call them crazy dreams. And what we’re working now is in a precise way to produce those muscle patterns into sounds to listen to more complex dreams.

MAGGIE KOERTH: You’ve been studying the physics of birdsong for 20 years now. What drew you to this field?

GABRIEL MINDLIN: Well, it’s interesting. I was working on the physics of human voice production back then when I started. And I visited some colleagues at the Rockefeller University in New York. And they asked me about birdsong. And I thought it was such an exotic question because I didn’t know it was such an active field of research, particularly in neuroscience.

So at the beginning, it started like a side project. But then what happens is that you have these 10,000 species, and the kind of physics that is involved, the kind of dynamics is fascinating, and very diverse, and changes from species to species. So whenever I thought I had figured it out, there was some weird vocalization that implied new physics, new behavior.

And I just couldn’t stop it. Basically that was it. It was just too interesting.

MAGGIE KOERTH: That’s fascinating. So you ended up a bird guy from physics.

GABRIEL MINDLIN: Right.

MAGGIE KOERTH: So humans, and birds, and all these other animals are having similar brain activity while they’re asleep. Does that all represent dreams the way we’d think of them, playing out these scenarios that help form memories, help us learn? Are dreams the same thing everywhere?

GABRIEL MINDLIN: Well, that’s– actually, it’s a very interesting question. So far, there are many conjectures, but we don’t have any experiment that is supporting one or the other. In general, there are behavioral experiments that say that the activity that the birds display at night plays a role in the process of learning. But obviously, there is more because non-learning species display these activity patterns, too.

MAGGIE KOERTH: The idea of replaying, playing these dreams back to birds is such an interesting idea. Is that the next step for your research? What are you hoping it’s going to teach you?

GABRIEL MINDLIN: One of the things I’m very interested in is in the relationship between the learning capability and the flexibility, the vocal flexibility that the species has with the kind of dreams and replays that it can exhibit because one of the fascinating issues with birds is, as I said before, we have 10,000 species.

4,000 of them learn. But even among the ones that learn, there is a wide variety of behavior. For example, the zebra finch needs to learn the song but will learn only one song. And it will sing it all the time during all his life.

But the canary has a larger variability, has a larger repertoire. And then you have other species, which are stellar vocal performers, like, I don’t know, a starling. Therefore, even in the songbirds, you have birds with different complexity in terms of their repertoires, how they use their vocalizations, how they combine them, how they use it to express what they need to express. And therefore, the comparative study between that complexity and what they dream is, to me, a very fascinating subject.

MAGGIE KOERTH: Thank you, again, for joining us. This has been amazing.

GABRIEL MINDLIN: No, thank you for your interest.

MAGGIE KOERTH: Dr. Gabriel Mindlin, professor of physics at the University of Buenos Aires in Buenos Aires, Argentina. We have to take a quick break. But when we come back, one third of superfund sites are located in 100-year floodplains, how climate change is impacting the safety of superfund sites. Stay with us.

Copyright © 2024 Science Friday Initiative. All rights reserved. Science Friday transcripts are produced on a tight deadline by 3Play Media. Fidelity to the original aired/published audio or video file might vary, and text might be updated or amended in the future. For the authoritative record of Science Friday’s programming, please visit the original aired/published recording. For terms of use and more information, visit our policies pages at http://www.sciencefriday.com/about/policies/