[CLIP: Theme music]
Rachel Feltman: What is it that turns speech into song? And why did humans start carrying tunes in the first place? Those are questions that scientists are still puzzling out. But some recent studies offer a few clues.
For Scientific American’s Science Quickly, I’m Rachel Feltman. I’m joined today by associate news editor Allison Parshall to learn more about the latest research on folk music.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
Allison, thanks for coming back on the pod. Always a pleasure to have you.
Allison Parshall: Thanks for having me.
Feltman: So I hear we’re going to talk about music today.
Parshall: We are going to talk about music, my favorite topic; I think your favorite topic, too—I mean, I don’t want to put words in your mouth.
Feltman: Yeah, I’m a fan, yeah.
Parshall: Yeah, yeah. Well, I guess I would love to know if you have a favorite folk song.
Feltman: That is a really tough question because I love, you know, folk music and all of its weird modern subgenres. But if I had to pick one that jumps out that I’m like, “I know this is genuinely at least a version of an old folk song and not, like, something Bob Dylan wrote,” would be “In the Pines,” which I probably love mostly because I grew up kind of in the pines, in the [New Jersey] Pine Barrens, so feels, you know, appropriate.
Parshall: Will you sing it for me?
Feltman: Oh, don’t make me sing, don’t make me sing. Okay, yes.
Parshall: Yay, okay! I’m set.
Feltman (singing): “In the pines, in the pines, where the sun don’t even shine / I’d shiver the whole night through / My girl, my girl, don’t lie to me / Tell me, ‘Where did you sleep last night?’”
That’s it; that’s the song.
Parshall: Clapping, yay! Oh, that was lovely. Honestly, I didn’t know if I expected you to sing it.
Feltman: If you ask me to sing, I’m gonna sing.
Parshall: I’m very happy. Well, I will not be singing my favorite folk song—I don’t even know if it qualifies as a folk song—but my grandma used to sing us a lullaby, and that lullaby was “The Battle Hymn of the Republic,” like, “Mine eyes have seen the glory,” or whatever. Yeah, so I think that’s my favorite one, but I don’t know if it qualifies.
[CLIP: “Handwriting,” by Frank Jonsson]
Parshall: But I’m definitely not the only person, like, asking this question; I’m asking it to you for a reason. There’s this group of musicologists from around the world that have been basically going around to each other and asking each other the same thing: “Can you sing me a traditional song from your culture?”
And they’re in search of the answer to this really fundamental question about music, which is: “Why do humans across the whole world, in every culture, sing?” This is something that musicologists and evolutionary biologists have been asking for centuries, like, at least as far back as Darwin. And this year we had two cool new cross-cultural studies that have helped us get a little bit closer to an answer. And actually they’ve really changed how I think about the way that we humans communicate with one another, so I’m really happy to tell you about them.
Feltman: Yeah, why do we sing? What theories are we working with?
Parshall: Well, okay, so there’s generally two schools of thought. One is that singing is kind of an evolutionary accident—like, we evolved to speak, which is genuinely evolutionarily helpful, and then singing kind of just came along as a bonus.
Feltman: That is a pretty sweet bonus.
Parshall: I agree. It’s like we get the vocal apparatus to do the speaking, and then the singing comes along. And the people who buy into this theory like to say that music is nothing more than, quote, “auditory cheesecake,” which is a turn of phrase that has long irked Patrick Savage. He’s a comparative musicologist at the University of Auckland in New Zealand.
Patrick Savage: It’s just like a drug or a cheesecake: It’s nice to have, but you don’t really need it. It could vanish from existence, and no one would care, you know?
So that kind of pisses off a lot of us who care deeply about music and think it has deep value. But it’s kind of a challenge—like, can we show that there are any real, consistent differences between music and language?
Parshall: Savage took this challenge very seriously because, if you couldn’t tell, he belongs to the other school of thought about music’s origins: that singing served some sort of evolutionary purpose in its own right, that it wasn’t just a bonus. And if that were true, if music weren’t just a by-product of language but played, like, an actual role in how we evolved, you’d expect to see similarities across human societies in what singing is and how it functions in a way that is different from speech.
Feltman: Yeah, that makes sense and also sounds like an extremely massive research project.
[CLIP: “None of My Business,” by Arthur Benson]
Parshall: Yeah, I don’t envy them the job of having to go around and try to perfectly represent the globe, but they made a solid attempt. They got to work recruiting colleagues to submit samples of them singing a traditional tune of their choice. And through what I can only describe as a truly heroic act of coordination—I can only imagine the e-mail threads—he and a small team of collaborators received data from 75 total participants from 55 language backgrounds and all six populated continents.
Feltman: Wow.
Parshall: So each participant submitted four recordings: one of them singing the traditional tune, another one where they play it on an instrument, another one where they speak the lyrics and another one where they speak naturally—just basically giving a natural language sample of them describing the song that they picked. And Savage himself picked the tune that you might recognize called “Scarborough Fair.” Let me play that for you.
[CLIP: Patrick Savage sings “Scarborough Fair”]
Feltman: It’s a classic choice—can’t knock it.
Parshall: Yeah, and I’m not immune to a little “Scarborough Fair.” There were also more upbeat tunes that some of the English-speaking contributors submitted.
[CLIP: Tecumseh Fitch sings “Rovin’ Gambler”]
Parshall: It makes me want to slap my knee and, like, play a fiddle. But that one was from Tecumseh Fitch. He’s an American biologist currently at the University of Vienna.
And this next one that I picked to show you comes from Marin Naruse of the Amami Islands off southern Japan. She’s actually a professional singer and cultural ambassador for the region.
[CLIP: Marin Naruse sings “Asabanabushi”]
Parshall: That vocal-flipping technique I just thought was so cool. And I was also totally taken by this next one from Neddiel Elcie Muñoz Millalonco. She’s an Indigenous researcher and traditional singer from Chiloé Island in Chile, and here she is singing a traditional Huilliche song.
[CLIP: Neddiel Elcie Muñoz Millalonco sings “Ñaumen pu llauken” (“Joy for the Gifts”)]
Parshall: So that’s just a little taste of what this data is like. There’s way more where that came from, and it’s all publicly available too, so you can check it out yourself. But the researchers after this, when they got the samples, got to work analyzing it. So hats off to Yuto Ozaki of Keio University in Japan. He’s the lead author of the study, and to hear Pat Savage tell it, he spent, like, months just processing these audio files full time.
So by comparing the singing samples to the speech samples and then comparing those differences with each other, the researchers found that songs tended to be different than speech in a few key ways: they were slower, they were higher-pitched, and they had more stable pitches than speech.
[CLIP: “The Farmhouse,” by Silver Maple]
Feltman: Yeah, I guess that makes sense.
Parshall: Yeah, like, if you think about the way that maybe a lot of us think about the differences between singing and speech—which, again, we can’t fully trust because there’s so many different ways to sing and speak around the world—but it generally takes more time to sing a lyric than to speak it because we’re lingering on each note for longer. And because we’re lingering that means we’re able to settle on specific pitches, like, instead of—where I’m speaking, I have this kind of low rumble that settles for less time on any specific pitch. I could also go dooo, and that is, for the most part, like, one specific pitch. It’s less upsy and downsy. And then, also, we generally sing with higher pitches than we speak.
Feltman: Yeah, why is that?
Parshall: Maybe because when we speak we’re kind of in this narrow, comfortable window toward the bottom of our vocal range. Like, right now, the way I’m speaking, I could go a little bit lower, but I couldn’t go very much lower, whereas if I’m singing, I can go, like, octaves higher, probably, than the way I’m speaking right now.
I think it’s partly just the way that we’re built, but singing opens up that upper range to us—like, you know, the mi mi mi mi mi mi mi of it all. So these differences where we’re hearing, you know, slower speeds, higher pitches, those are all interesting, but they feel kind of intuitive, and I didn’t have a great way to understand what they were telling me kind of as a whole until I learned about this next study that I’m going to tell you about.
Feltman: Ooh, so what did they find?
Parshall: So this one actually had more of a neuroscience focus, whereas the other one was a little bit more anthropological. This one was conducted by Robert Zatorre of McGill University in Montréal and his colleagues. His team has been asking basically the same question as Savage’s team but in a different way. So that’s: Can we find commonalities in how cultures around the world speak versus how they sing?
Robert Zatorre: Do they have some kind of basic mechanism that all humans share? Or is it rather that they’re purely cultural sort of artifacts—each culture has a way of speaking and a way of producing music, and there’s really nothing in common between them? As a neuroscientist, what interests me in particular is whether there are brain mechanisms in common.
Parshall: And Zatorre wasn’t going into this from scratch. His own research and research of others had shown that the left and right hemispheres of the brain might be involved differently in speaking versus singing.
Zatorre: An oversimplified version would be to say that speech depends on mechanisms in the left hemisphere of the brain, and music depends more on mechanisms in the right hemisphere of the brain. But I say that’s oversimplified because it wouldn’t really be correct to say that.
Parshall: So what is correct, though, according to Zatorre, is that there are certain acoustic qualities common in speech that are parsed on the left side of our brain and other acoustic qualities common in singing that are parsed on the right side.
Feltman: So pretty much all I know about left versus right brain is all the debunked stuff about being, like, left-brained or right-brained as a personality type. So could you unpack the actual neuroscience here a little bit?
Parshall: Yeah, the whole, like, “Oh, I’m left-brained. Oh, I’m right-brained,” that’s mostly been debunked. But it’s true that parts of the two sides of the brain do specialize in totally different things sometimes, and here’s what that means for processing sound.
[CLIP: “Let There Be Rain,” by Silver Maple]
Parshall: Speech contains a lot of time-based, or temporal, information, meaning that the signal of what you hear, even as I’m talking now, is changing from, like, millisecond to millisecond and, importantly, that those changes are meaningful. Like, each letter or phoneme that I’m pronouncing goes by super quickly, but if I swapped one for the other—like said “bat” instead of “cat”—that would totally change the meaning, and that happens super quick. So those tiny time frames really matter when we’re talking about speech, and that kind of quick-changing information is processed more on the left side of the brain.
Singing, on the other hand, contains a lot of spectral information, which is processed more on the right side of the brain. So when I say “spectral,” I’m referring to the spectrum of sound waves from super low pitch to, like, super high. Those aren’t at all encompassing of the spectrum.
Feltman: Yeah, that was the whole spectrum of sound.
Parshall: I can go way lower than—yeah, it goes way lower than what you think you’re hearing and way higher than what you think you’re hearing. But that information of that spectrum, it kind of contains the “color,” or the timbre, that allows you to distinguish between, for example, a saxophone and a clarinet or even, you know, your voice and my voice if you were listening.
You can really hear this difference in some audio samples that Zatorre sent over from his studies. So basically, for one of these studies, they hired a soprano to sing some melodies and then used computer algorithms to mess with the quality of her voice.
So here’s the original audio.
[CLIP: Audio of singing from a study by Zatorre and his colleagues: “I think she has a soft and lovely voice.”]
Parshall: Then they digitally altered the recordings to degrade that temporal, or timing, information. That’s kind of like the musical equivalent of slurring your speech or the audio equivalent of making an image blurry. They basically make all of those time cues that are so important for speech blur into each other.
[CLIP: Same audio from the study with temporal degradation]
Feltman: Ooh, freaky.
Parshall: Yeah, it’s, like, delightfully alien, I would say. You’ll notice that you actually can’t hear the lyrics, but you can still kind of hear the melody, right? You could probably distinguish it from another melody, and that’s not the case when you do something different and instead of the temporal information, you degrade the spectral information—that’s the sound’s color.
So here’s what it sounds like when they take out all that spectral information.
[CLIP: Same audio from the study with spectral degradation]
Feltman: Whoa.
Parshall: Yeah, like, the only thing I can compare it to are, like, the Daleks from Doctor Who.
Feltman: Totally, yeah.
Parshall: I love it, and I hate it.
So in this one you can hear the lyrics, but you can’t hear the melody at all. So it’s kind of the inverse. And you can hear that both of these dimensions of sound—the temporal and the spectral—are really important for both song and speech. Like, you would not want to listen to my voice for very long if I sounded like a Dalek. But generally speech relies more on that temporal information, and song relies more on the spectral information.
Feltman: And this is true across different cultures, too?
Parshall: Yeah, so in a study published this summer, Zatorre’s team found that this distinction holds true across 21 cultures, and they surveyed urban, rural and smaller-scale societies from around the world. And despite how different some of these languages and singing traditions are from each other, it held true that songs had more spectral information and speech had more temporal information overall.
And so, since we can link these differences to different methods of processing in the brain, there’s actually a potential biological mechanism in humans that separates music from speech.
Zatorre: So the story we’re trying to tell is that we have two communication systems that are kind of parallel: one is speaking; [the] other is music. And our brains have two separate specializations: one for music, one for speech. But it’s not for music or for speech per se; it’s for the acoustics that are most relevant for speech versus the acoustics that are most relevant for music.
Parshall: Yeah, and it kind of makes sense to me that we’d have these two parallel communication systems because they basically allow us two separate channels to convey totally different types of information. And, like, imagine how long this podcast would be if I sang everything instead of speaking it. And then imagine that I couldn’t incorporate language at all, like, via lyrics, and I just had to do it with notes. That’s just impossible—unless we came up with some elaborate code. But then also imagine trying to sit here and explain to me your favorite song in words and all the feelings it brings up for you and why you love it. Like, could you do that?
Feltman: Probably not. It would be really hard.
Parshall: Probably not. It’s conveying—there’s, like, something extra that you’re conveying with song that just resists being conveyed via speech.
So all that to say, “auditory cheesecake,” quote, unquote—music as this little accidental cherry on top of language—that doesn’t seem to be the right way of thinking about why we sing. Here’s Savage again.
Savage: It suggests that it’s not just a by-product—like, there’s something that is causing them to be consistently different in all these different cultures. Like, they’re kind of functionally specialized for something. But what that something is is very speculative.
[CLIP: “Those Rainy Days,” by Elm Lake]
Parshall: That speculative X factor that he’s talking about, that reason why we evolved to sing, if you had to come up with a theory, Rachel, what would it be?
Feltman: I mean, when I think about reasons to sing that I, like, can’t imagine humanity just not doing, I don’t know—I picture people soothing babies; people celebrating with each other; people, like, engaging in spiritual practice; like, standing outside a crush’s window with a boom box. Singing is a thing we do to get each other’s attention and share an emotional experience.
Parshall: Yeah, I think that sharing feels really important, and I feel like I have a similar intuition. And that’s basically what Savage thinks, too: that music has played some sort of social role. So that could be really wholesome, like the boom box or us bonding together, singing songs around a campfire. Or—I mean, it could be less wholesome. It could be, like, us singing war songs before we do battle with our enemies.
This is one of those evolutionary hypotheses, as many of them are, that it’s kind of impossible to fully prove or disprove. It’s really hard to get evidence that would be able to say, “Oh, we sing because it, you know, bonds us closer together.” But it’s very compelling.
Feltman: Yeah. So just to recap: we know that we have these two very different ways, from a neuroscience perspective, of conveying information. We’ve got this, you know, melodic musical, and then we’ve got this, like, very straightforward speech. And sure, we can’t go back in a time machine and ask, you know, our distant ancestors, “Why’re you singing? Why’re you doing that?” So what’s next? How do we move this research forward?
Parshall: It can be a little tricky, obviously, to come up with specific proof, but one of Savage’s co-authors is hoping to find some clues in an upcoming experiment.
So her name is Suzanne Purdy, and she’s a psychologist also at the University of Auckland in New Zealand. And she’s involved with something called the CeleBRation Choir. And this choir is super cool because it’s made up of people [with communication difficulties, including people] who have what’s called aphasia, so their ability to speak has been impacted by events like a stroke or like Parkinson’s. But one of the very interesting things about aphasia is, oftentimes, people’s ability to sing remains intact. So that might be because it is relying on different parts of the brain—you know, more varied parts of the brain—than speech does.
Suzanne Purdy: When being with the CeleBRation Choir, with people struggling to communicate verbally, but then hearing them sing, [it’s] so beautiful and amazing. And our research has shown how it’s therapeutic in terms of feeling connected and valuable and able to be in a room and impress people with your singing, even when something terrible has happened in your life.
Parshall: So I actually have a recording to share with you of the choir because I think it’s super cool.
[CLIP: The CeleBRation Choir sings “Celebration,” by Ben Fernandez]
Parshall: So partly inspired by her experiences with the CeleBRation Choir, Purdy, [Savage and their] team are currently developing an experiment where they test whether singing can actually make us feel more connected to each other. So they’re going to bring in students and have them sing together and then compare that to the experiences of students who have just talked together in a group. And then they’ll measure their feelings of connectedness to each other. And they’re planning to actually do this cross-culturally, too. So they’re going to do this for groups of Māori students, Māori being the Indigenous people of New Zealand, and then students of European descent to see if there are any cultural differences in the impact of singing together.
Purdy: It’s the kind of thing that, you know, companies do with team-building exercises. They don’t usually get people to sing, do they? But they do get people to problem-solve or to talk together. So this—part of this next phase is: Can you achieve the same level of social cohesion through just coming together with a shared purpose without singing? Or does the singing add a special quality, and is that more effective?
Parshall: Okay, I can’t tell if the idea of a company team-building choir sounds fun or like the worst idea ever, but I do have a feeling that it would be kind of effective.
Feltman: Yeah, I mean, I guess it’s not so different from a karaoke night. And, you know, what brings people together more than a karaoke night?
Parshall: That’s a good point. Why did I not think of karaoke night? Okay, we’re gonna have to go to our boss with this one. I think it could be really fun.
It is just still a hypothesis whether music really did evolve—or singing, specifically, really did evolve to bond us together. Like, again, this is not something we have necessarily a lot of proof for. And even if this study that Purdy is developing comes up and shows, you know, these groups of students did feel more bonded together when they sang versus when they spoke, that’s still only just, like, a little bit of clues and proof.
Feltman: Right, that could just show that we gained this incredible benefit from singing over time. It doesn’t necessarily tell us that that’s why it evolved.
Parshall: Right. But then I’m always fighting against myself—the instinct to be like, “Oh, but it’s true,” because it feels true, right?
Feltman: It does feel true.
Parshall: Like, based off of my personal experience and a lot of people around me, it feels like, you know, when you’re in a concert and you look around and you feel, like, the oneness of the world when you’re all singing together in this packed stadium, music, regardless of what science shows, it does have these effects on us personally.
Feltman: Yeah, and we can definitely get a better understanding of why it’s so important.
Parshall: Yeah, like, regardless of how we got here, regardless of how we evolved, we can still look at the impact it has on us now.
Feltman: It’s interesting, I’ve been thinking this whole time—my sister does shape-note singing, which is this old musical notation style that was basically created so that people who were not otherwise musically literate could, like, all come and sing together in a group at, like, a moment’s notice. And it has, like, a big following these days, and people just get together and open these giant old books of, like, mostly Shaker songs and stuff. And I find the shape-note stuff very confusing. It’s very confusing until you learn it, and then it’s allegedly easier than reading other music.
[CLIP: Theme music]
Feltman: But yeah, it’s just amazing how connected people feel within, like, five minutes of sitting down together and singing together. We don’t need researchers to tell us that that’s a universal experience, but I think it’s awesome that they’re asking these questions to help us understand, you know, just why music is so important to us.
Allison, thank you so much for coming in to chat about this and for sharing all of these lovely musical snippets. I think that was my favorite part.
Parshall: Thank you so much for having me.
Feltman: That’s all for today’s segment. If this Friday Fascination left you with a song in your heart, we’d love to get your help with a future episode we’re working on. We’ll be taking a look at the science behind earworms, those songs you just can’t get out of your head, and we’d love to feature some of your favorite—or maybe most infuriating—examples. If you’d like to share an earworm with us, make a voice memo on your phone or computer and send it over to ScienceQuickly@sciam.com. And, yes, we do want to hear you singing, or at least humming, the earworm in question.
Science Quickly is produced by me, Rachel Feltman, along with Fonda Mwangi, Kelso Harper, Madison Goldberg and Jeff DelViscio. Today’s episode was reported and co-hosted by Allison Parshall. Marielle Issa, Shayna Posses and Aaron Shattuck fact-check our show. Our theme music was composed by Dominic Smith. Subscribe to Scientific American for more up-to-date and in-depth science news.
For Scientific American, this is Rachel Feltman. Have a great weekend!
