Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Support PBS Shop PBS Search PBS

Interview with Daniel Levitin
Part One

What is musical memory? What does it have to do with the way that we perceive and use music in our everyday lives?

Daniel Levitin

Daniel Levitin: One of the things about musical memory is that, in some respects songs stick in our head, and maybe that’s because they’re supposed to. It’s difficult to talk about these things without talking about evolution. When you can remember a song so well, maybe it suggests that evolution wants us to. Maybe songs played an important role in our evolutionary history. A lover out on a hunt for a long period of time wants to be remembered while he’s away, she wants him to remember her, they have their song that they sang to each other, and, you know, that sticks in the head, and it keeps them faithful, and, you know, there’s some evolutionary advantages in that, in terms of raising the kids and, and so on. I’m interested in what attributes of music stay stuck in the head. Is it rhythm, it is pitch, what is it? It turns out to be all of it.  The average person has an extraordinary memory for the components of music. Even when there’s no theoretical reason why they should.  So, take the song “Happy Birthday”. Every time you sing it, you sing it in a different key. It’s still the same song. Whoever it is that’s in the room that starts, they just start any way they feel like, they may not even think ahead. And then you all join in, and some of you are synchronized in the right pitch, and some of you aren’t, and it doesn’t really matter. It’s still the same song.

In fact, it was the Gestalt psychologists, who noticed in the 1890s, Christian Von Ehrenfels and Max Wertheimer and others, that there’s this funny property to songs. You can sing them with any group of notes, and they’re still recognizable as the same song. Even when you change every single note, it’s still the same song.

It’s because songs are defined by the relationship between pitches, not the absolute pitches. Nevertheless, if you ask the average person in the street to just sing their favorite song, they tend to sing it with the right pitches. Their memory has encoded this information that isn’t necessary for maintaining the identity of the song, but it’s there.  Why would evolution create a brain mechanism that holds onto the stuff that it doesn’t need? It must’ve been important, throughout evolutionary time, or it must be that memory is more efficient, if it can hold onto all this detail. People don’t just remember the absolute pitches, but they tend to remember the actual tempo, and a lot of the little nuances of the singer’s voice. When Michael Jackson goes, “Eeh, eeh!” or Madonna has a particular growl in her voice, people remember all of that, and they replicate it when they sing.

So describe how you do this experiment to study this.

One of the ways that, that we study this is we just bring people into the laboratory, or stop them on the street. We ask them to sing their favorite song. And then we analyze their production, and compare it to the CD. Now, in order for this to work, they can’t be singing a song like “Happy Birthday”, or the National Anthem, or “Deck The Halls”, where there is no right key. But if they sing a pop song, a song by U2 or by Backstreet Boys, that song exists in the world in only one version, and it’s the version that people have heard thousands and thousands of times. There is a correct answer to the question, “What is the tempo of that song?” or, “What is the, the right pitch, starting pitch?” You just record them, you compare it to the CD, and you, you look at the pitch and the tempo and you see how close they got.

So, coming back to emotion in music- what are the possible theories about why music affects us emotionally?

There are a lot of different factors that go into our emotional appreciation of music.  Some of it is the memories we have of a particular song, which we heard at a particular time in our lives, or it reminds us of a song that had those qualities. Some of it has to do with just the beat, the pulse. Music like James Brown or March music, for that matter, can be invigorating. It makes you want to move your body. Other music can make you, uh, just sort of melt and relax.  It’s, it’s either composed to have that affect, or it’s performed to have that affect.

We do know that listening to music releases certain neurochemicals. If you listen to music that you enjoy, it releases dopamine, a so-called “feel-good hormone”. It can also release prolactin, the comforting hormone that’s associated with mothers lactating and feeding their infants.

There’s another hormone called oxytocin that’s the so-called “trust hormone.” This is the hormone that’s released when two people- well, if a person has an orgasm, oxytocin is released, and it makes them bond to the person that they’re with. If two people have an orgasm at the same time, they bond to each other. There’s an obvious evolutionary advantage for this. Oxytocin causes feelings of trust with the person. For reasons that we don’t fully understand, when people sing together, oxytocin is released. People trust more, people that they’ve sung and played music with.  So there’s all this neurochemical change that occurs, in response to playing and listening to music.  And we’re just at the beginnings of trying to sort it all out.

Do you think in a very broad sense, that it’s more because we associate music with something that creates an emotion, or because there’s something structural about the nature of music itself?

I think music contains an enormous amount of information. And I mean information in the technical sense of information theory, the amount of unique content.  That there’s more information than speech.  It’s more complex a signal. And so I think that although music doesn’t convey information like, “Hey, would you open the window over there?”, it conveys emotional information that’s very nuanced, and we’re sensitive to that. I think that music was probably an early form of emotional communication between humans, and the reason it lasted even after the introduction of language, is that it’s much better at some forms of communications, in some feelings that you want to communicate, than language is. It’s much better at communicating the dynamics of human emotion.

(5 votes)
Loading ... Loading ...
3 comments
madame pirio -- June 25th, 2009 at 8:10 am

I have constant “earworm”, that background music generated by the brain. Is that the result of the brain creating pleasurable neurochemicals to enhance mood?
I often ask people “What does your brain do when it is in neutral?”and many report music but others say “playing with math” or “nothing”. Does the current generated by math/ deduction also enhance neurochemicals?
Why do some people seem immune to the power of music? My husband who seems so had mastoiditis as a child and has hearing loss. Music does nothing for him, in fact he chafs at it.
Wonderful presentation. Thanks for sharing your expertise.

Bernard Souw -- July 3rd, 2009 at 9:53 am

As a physicist, researcher and violinist I am highly interested in your enlightening presentation. I fully agree with your statement,

“music contains an enormous amount of information of unique content” and that “there’s more information than speech,” and further, “ it conveys emotional information that’s very nuanced, and we’re sensitive to that,” and especially “it’s much better …. than language is; It’s much better at communicating the dynamics of human emotion.”

I think music was not just an early form of emotional communication between humans, but –especially in its form we know today– a very highly developed form of communication. As a vehicle for emotional communication the music “language” is universal, capable of traversing the boundaries between ethnicities and cultures, well beyond the capability of language.

The experiment you showed in your TV presentation demonstrates that most elementary forms of human emotion (happy, sad and scared) can be unmistakably understood even by isolated tribe in Africa that has never been exposed to western music in their whole live. My scientific instinct tells me, this fact might be scientifically extrapolated to conclude that well-developed western music as played by symphony orchestras and opera performers conveys a lot more emotional information, compared to primitive music or pop music, judged by a higher complexity of structure and a larger variety of nuances, in the same manner as language is evaluated.

This fact is already shown by your fractional CAT scan images of human brain taken while listening to certain types of music (of course it must be presumed that the test person is enjoying the music being played, so his brain would not respond indifferently, or even become annoyed). I am quite impressed to see that your CAT scan images showed how the entire human brain is excited or stimulated upon listening to Richard Strauss’s song cycle “The Four Last Songs” (September?). This agrees with my own personal intuition; I shed my tears especially when listening to the last song “Im Abendrot” (In the Red of Sunset). It also agrees with Franco Zeffirelli, Director of Met Opera in the Movie, who said in his introduction to the 2008 production of Puccini’s La Boheme, it is completely okay for the audience to cry, because “Puccini wants you to cry”.

In this regard it might be quite interesting to expand the experiment to test the validity of the “Mozart effect”, thus providing a scientific foundation to the hitherto controversial hypothesis. It might be even more interesting if the result would further show that different parts of the brain would be dynamically excited or stimulated by different movements of a symphony or concerto, thus giving a solid reason, why a classical symphony or concerto needs a multiple of movements to form a complete, dynamic and self-contained work of art.

If these prove to be objectively true –by evidence of experimental results, as always demanded by science– we may then have strong reasons to challenge the relative concept of cultural value, i.e., that we can not say whether one particular culture is more developed than the other, as presently claimed by some liberal scholars, allegedly for lack of standard. The scientific evidence would then be the standard, a very strong and solid standard.

Bernard S., physicist and violinist, Herndon, VA

[...] My assistant texted me about this. I recorded the last half of it. I read about it on Twitter (where Dan Levitin gave a live Q& A). If you are a music therapist, you NEED to read this. [...]

post a comment
Please note that the WNET editorial staff reserves the right to not post comments it deems to be inappropriate and/or malicious in nature, as well as edit comments for length, clarity and fairness. No solicitations or advertisements will be allowed. Users may link to other Web sites relevant to discussion, but most often links to commercial Web sites will not be permitted.

YouTube iTunes

Produced by THIRTEEN WNET New York    ©2009 Educational Broadcasting Corporation. All rights reserved.