“I Wanna Dance With Somebody (Who Loves Me)” by Whitney Houston

When we’re listening to a song, how much does the tune’s structure really matter?

How much does it really matter which notes are in the verse or in the chorus?

How much does it really matter when the chorus happens or when the verse happens?

The music psychologist Carol Krumhansl once argued that when we listen to a song, “how the parts are perceived depends on their functions in the whole”. So we’re going to hear a song differently depending on whether we hear the verse or the chorus first. And if we jump into a song halfway, we’ll get a completely different experience.

To really understand why, let’s look at one song where the verse and chorus really do act differently, and how they interact with each other.


Welcome to Song Appeal, where we dive into your favorite songs to answer the question “Why do we like the music we like?”. I’m your host, Hunter Farris, and on today’s episode, let’s take a look at one reason we like “I Wanna Dance With Somebody” by Whitney Houston.

You can find this episode’s transcript, its references, and a link to hear the song at SongAppealOfficial.com/IWannaDanceWithSomebody.

Thanks so much to my patrons for making this show possible. If you want to become a patron, visit Patreon.com/SongAppeal.

This episode will talk a lot about how major scales work and what effect different notes have on a major scale. If you want to learn more about major scales and why we use them in the first place, check out Coursera’s program “Music as Biology: What We Like to Hear and Why” at bit.ly/MusicAsBiology.

[Music: I wanna dance with somebody, with somebody who loves me]


Narada Michael Walden, the producer of “I Wanna Dance With Somebody”, once said that recording this song “was about trying to create the best possible record that will last for hundreds of years”, and that’s exactly what they got. It was a #1 hit when it was released in 1987, and over 30 years later, people still dance to this at weddings and sing their hearts out to it.

Now, when I hear a song get that popular and last that long, I have to ask: How did that happen?

Whitney Houston is a definitely a great singer, but that’s not the only ingredient. The first time I heard this song, it was Fall Out Boy’s cover, and I loved that version enough to go look for the original version. So Whitney Houston’s vocals can’t be the only thing making this song work. The tune had to be enjoyable, too. So how did the melody contribute to the song to make it last this long?

To answer that question, I interviewed two experts:

[Bryn Hughes quote:] “My name is Bryn Hughes, and I am an assistant professor of music at the University of Lethbridge. … I did a PhD in Music Theory at Florida State University, and I’ve been at this position for 4 years.”

[Anja Cui quote:] “My name is Anja. I am a PhD student at Queens University in Canada.”

[BH:] “My research is generally in the music cognition area, so I sort of do interdisciplinary research that sort of draws from music theory… and also behavioral science and cognitive psychology.”

[AC:] “My research is in music cognition, so basically how people listen to music, what happens when they’re listening to music.”

A lot of the research Dr. Hughes does tends to [BH:] “ask general big questions about how harmony and melody and rhythm and meter and all of those sorts of things that we talk about in music theory, how those things work in sort of a more controlled, experimental context”.

Let’s take a major scale for example. “I Wanna Dance With Somebody” uses the first, third, and fifth notes on that scale more often in the chorus than in the verses and pre-choruses. Since Anja and Dr. Hughes have each worked on different versions of an experiment about how notes 1, 3, and 5 on a scale affect us psychologically, they can tell us a lot about why that difference matters through the lens of music theory, music psychology, and cognitive psychology.

When “I Wanna Dance With Somebody” focuses on notes 1, 3, and 5, it’s focusing on the notes that rank highest in what music theory calls a “tonal hierarchy”. [BH:] “The tonal hierarchy is this idea that certain notes are sort of more important than others.”

In a song in a major key, like “I Wanna Dance With Somebody”, you get 4 levels of how important a note can be. Level one: [BH:] “So if you’re in the key of C major, on the top of the hierarchy is the note C. Because that’s the most important note in the key of C major.” After all, that’s the root of the scale, the base of the scale, and name of the scale. It’s the most important note, just like Hamlet is the most important character in Hamlet. Level 2: [BH:] “And the next 2 notes on the hierarchy are E and G.” That’s because when you take C – which (remember) is the most important note – and you make a chord based off of it, the other two notes are E and G. Level 3: [BH:] “And the remaining notes the next rung down are all the notes left in the key of C major that are not C, E, or G, so that would be D, and F, and A, and B.” And lastly, you get Level 4: all the notes that aren’t in C major, like Db, Eb, Ab, and Bb. [BH:] “so you’ve got sort of a 4-tiered hierarchy.”

Now that works in any key, so to show how it works in “I Wanna Dance With Somebody” let’s change those letters into numbers that work on any scale: Note 1 is the most important. Notes 3 and 5 come next. Then notes 2, 4, 6, and 7, and then all the notes that aren’t in the scale. So not only do we have music theory that says that some notes are more important than others, but we know exactly how important each note is.

So when “I Wanna Dance With Somebody” plays notes 1, 3, and 5 more in the chorus than the verse, music theory would say those notes are more important.

Now, “important” can mean anything from from “played more often”, to “emphasized”, to “the center of attention that’s occasionally strayed from”. So what does it mean in “I Wanna Dance With Somebody”? To find out, let’s put music theory to the test to see how people actually hear music.

Music psychologist Carol Krumhansl once wrote that we can find certain patterns by analyzing music, but if we want to know how well a listener can pick up on those patterns, she said that we can only know that through experiments and tests. So she set up what would later be known as “the probe tone experiments”, which established a template that music psychologists have used for decades. Dr. Hughes and Anja have each done their own different variations on the probe tone experiments, so I’ll let them explain the idea: [AC:] “If you imagine you are participating in one of these experiments, you’ll get played a short melodic excerpt, say, a scale, then you get played a tone”, whether it was in the scale or not. The researchers called this a “probe tone”.

[BH:]

And she’d say [Jenny Roberts voice:]“How well did that note fit?” The question was intentionally vague.

[Jimmy Murry voice:] “Well, what does that mean?”

[JR:] “Well, you know how some notes sound good and some notes sound bad?”

[JM:] “Yeah?”

[JR:] “Well, like that.”

[AC:] “And then you have to rate on a scale of 1-7 how well you think the tone fits with the music that was played before.” [BH:] “And basically, you do this enough times that statistically the vagueness just comes out of the wash.” [AC:] “And based on those ratings, you can order basically how well people think various tones fit in with the same context.” [BH:] “What Krumhansl found was that the ratings, you know, “how well each note fit” reflected exactly the tonal hierarchy.” People said that to them, note 1 fit best; notes 3 and 5 tied for fitting second best; notes 2, 4, 6, and 7 fit less than those; and all the notes that weren’t in the scale (understandably) seemed like they fit worst in context of the scale. Exactly the same as music theory said they would. When the tonal hierarchy tells us which notes are important, that can mean “how well those notes fit in the scale”.

Now people didn’t say those notes were important or that they fit because music theory said so. [AC:] “I think the most interesting thing about this is that even participants or listeners who don’t have music training show the similar patterns.” [AC:] “So in your head, I’m sure you’re not really thinking “How important is that tone?” You’re giving it a rating based on your gut feeling.”

So what gives us that gut feeling? Carol Krumhansl, the originator of the probe tone experiments, proposed that we’ve heard enough songs in major to pick up on what songs in major do, just like when you watch enough spy movies and you’ve picked up on what spy movies usually do.

Anja suggested, [AC:] “I’m assuming that most people hearing “I Wanna Dance With Somebody” would know that it’s in major even though they might not know it’s called major.” And you subconsciously recognize that notes 1, 3, and 5 are going to be pretty important in this song. You probably don’t consciously recognize it, but your subconscious mind picks up on it pretty quickly.

Music theory didn’t tell people what to think. Music theory predicted what people would actually think, based on people’s experiences. And science showed that the music theory was right!

Since we can recognize which notes “fit” best, we can enjoy the chorus more than the verse and pre-chorus.

Why? Because of something called the “fluency heuristic”, a psychological shortcut our brains use that says “We like things we can process faster”. And with good reason! With so much going on every second, your brain has to focus on the things it can process quickly just to keep up.

What does that have to do with the probe tone experiments and the tonal hierarchy? Anja told me that [AC:] “tones that fit well often are also easier to process.” So not only do notes 1, 3, and 5 fit best, not only are they the most important, but they’re the tones our brains can process fastest. And remember: “We like things we can process faster.”

So how does that fit into “I Wanna Dance With Somebody”? Anja told me [AC:] “The chorus – purely based on the tones that are used – are easier to process than the tones in the pre-chorus.” When we hear the verse of “I Wanna Dance With Somebody”, we hear notes 1, 3, and 5 57% of the time. And we enjoy that, because those notes fit well in the scale and we can process them quickly. And when we hear the pre-chorus, we hear notes 1, 3, and 5 50% of the time. And we enjoy that well enough. By then, our brains think it’s normal for this song to use notes 1, 3, and 5 somewhere around 50% of the time. But then the chorus comes in with notes 1, 3, and 5 happening 85% of the time, and our brains say “I can process this even faster! This is great!” and we like that part even more. The verse and pre-chorus feel like riding a bicycle where it’s flat. Enjoyable enough. But the chorus feels like riding a bicycle downhill: We don’t need to pedal too hard. We can just enjoy the ride.

Now, I’m not saying “You just like this song because your brain is lazy”. I’m saying you have a good reason to like “I Wanna Dance With Somebody”. I’m saying there’s an explanation for why this melody works so well. I’m saying the melody of the verse, the pre-chorus, and the chorus interact to help us like this song. If the whole song were full of notes 1, 3, and 5, we wouldn’t enjoy it nearly as much. We would just get used to it. But because of the fluency heuristic, this song can set itself up so that we can get a spike of enjoyment during the chorus.

Sometimes, I’ll mention “I Wanna Dance With Somebody”, and someone in the room will love it so much that they’ll start singing it. But they don’t sing the verse or the pre-chorus. They sing their favorite part: the chorus. Why? One obvious explanation: They just heard the name of the song, which is in the chorus. But also because the chorus uses notes 1, 3, and 5 more than any other part of the song. And music theory, music psychology, and cognitive psychology can give us one explanation about that subtle reason why people love that part so much.

And obviously, this isn’t the only reason to like this song. There’s more to it than just the melody, and even if you look at the melody alone, there’s more to the tune than just how many times each note shows up. [AC:] “…that’s fundamentally the trouble of trying to apply specific scientific experiments to songs. The whole idea of scientific experiments is to try to control as many things as possible, and sometimes that ends up happening by stripping away some of the things that happen in real life: like lyrics, like meter, and so on. And now you have this song, which has all these extra things that are not part of music cognition experiments. And anything you say that the scientific experiments might predict are confounded by the fact that in real life, there are all these other things that weren’t part of the scientific experiment.” Notes 1, 3, and 5 aren’t the only important part of why we like this song. But it is an important part that we sometimes overlook.

And it’s not like the songwriters were trying to manipulate you into liking this song. When the producer said he wanted a record that would last for hundreds years, I’m sure he knew how much money that would bring in. But I’m also sure he had some idea of how much this song would affect culture if it really did last for hundreds of years. People are still singing and dancing to this song at weddings. A few weeks ago, I watched Isn’t It Romantic? and one of the most important scenes was all about this song. When songwriters put out mega-hits like this one, they aren’t crafting every note to be as lucrative as possible. They’re trying to put out songs that people will love and care about, songs that will make people’s lives better. Songs that will bring joy to their listeners. And music theory, music psychology, and cognitive psychology can all really help with that.

This song is a great example of how music theory and psychology can help the songwriting process. Some songwriter could decide they want people to enjoy the chorus more than the verse or the pre-chorus, and one of the many ways they could do that is by using notes 1, 3, and 5 more often in the chorus than in the verse. And I don’t think that would be manipulative. I think it would be giving the audience a great time. And that’s just one way that songwriters can make music more effective by knowing the music’s effect.

This song is a great example of how music theory and music psychology can support and complement each other. Music theorists have been discussing tonal hierarchies (or which notes are most important) for hundreds of years. And when music theorists reached a general consensus, music psychology was able to confirm and support the theory, because [AC:] “There are, I think, intricate links between music theory, between the actual music, and between what people think.” Music theory didn’t make music psychology useless. Music psychology didn’t nullify music theory. They each fit together like puzzle pieces to show us why we like the music we like.

“I Wanna Dance With Somebody” is one great example of we like the music we like. It’s a great example of the power of melody. And it’s a great example of how much the structure of a melody can really matter.


[music: “Yeah, I wanna dance with somebody/with somebody who loves me.”]


Thanks so much for listening. If you liked this episode, check out SongAppealOfficial.com for more episodes.

Thanks so much to my Patrons for making this show possible. You can help support this show on Patreon at Patreon.com/SongAppeal.

You can find more from Anja and her work at the Queens University Music Cognition Lab at Bit.ly/QueensMusic.

You can find more from Bryn Hughes on Twitter: @BrynMDHughes.

Thanks to Jenny Roberts for providing voice acting in the role of Carol Krumhansl. You can find more from her at Bit.ly/JennyRobertsVoice.

Thanks to Jimmy Murray for providing voice acting in the role of the experiment participant. You can find more from him on Twitter: @TheJimmyMurray.

Again, thanks so much for listening. I’ll talk with you soon. In the meantime, have a great day. And enjoy your music.

Thanks to the following Patrons for their generous support:

Brenda Farris

Jake Lizzio