Friday, May 8, 2009

Does music describe things?

The blog Cognitive Daily has conducted an experiment to find out something that musicians and composers have always known: music is ineffectual at describing things in the external world.

Of course, a lot of music is meant to accompany extra-musical images or stories. There are tone poems, movie soundtracks, the Fantasia movies, Peter and the Wolf, and for that matter, song lyrics in general.

But you need to be explicitly shown or told what the music is supposed to evoke -- which means the music on its own doesn't evoke specific things.

Here's how Cognitive Daily did the experiment: they conducted a survey of their blog readers, in which they embedded audio clips from supposedly descriptive pieces of music and asked the readers to say what the music described. For instance, the survey included

a selection from Claude Debussy's La Mer, from the movement intended to represent the wind and the sea. Only 36 of 357 respondents answered correctly. Even when I gave half-credit for mentioning either the wind, or a storm, or waves, or a boat, only an additional 90 got it. Most respondents -- over 200, in fact, got it completely wrong.

I picked seven different clips like this, from seven different works that were all intended by their composers to represent specific things, not just emotions or adjectives. I tried to pick pieces that seemed relatively obvious, based on the composer's initial intentions. I scored each response on a scale of 0 to 2, with 2 being perfect, and 1 meaning some portion of the response was correct. The average score was a mere 0.38, and 72 percent of the time people got the answer completely wrong.
Respondents with musical training did better than those without it, but "not much better."

I'm reminded of an anecdote about a critic who was writing a review of Mendelssohn's Symphony No. 3, the "Scottish" Symphony (sometimes awkwardly called the "Scotch" Symphony) when it first came out. The critic exulted that Mendelssohn had perfectly captured the essence of Scotland. He had inadvertently listened to Mendelssohn's Fourth, the "Italian" Symphony, which, as you might have guessed, was supposed to evoke Italy, not Scotland.

I do think music is meaningful and important, but for other reasons (which I'll go into in a future blog post). It's not important because it "describes" or "depicts" nature, or a city, or a person. Music doesn't "describe" or "depict" anything.

The famous composer Aaron Copland wrote, in his book What to Listen for in Music (1939):
My own belief is that all music has an expressive power, some more and some less, but that all music has a certain meaning behind the notes and that the meaning behind the notes constitutes, after all, what the piece is saying, what the piece is about.
This whole problem can be stated quite simply by asking, "Is there a meaning to music?" My answer to that would be, "Yes." And "Can you state in so many words what the meaning is?" My answer to that would be, "No."
Therein lies the difficulty. Simple-minded souls will never be satisfied with the answer to the second of these questions.
They always want music to have a meaning, and the more concrete it is the better they like it. The more the music reminds them of a train, a storm, a funeral, or any other familiar conception the more expressive it appears to them.
This popular idea of music's meaning -- stimulated and abetted by the usual run of musical commentator -- should be discouraged wherever and whenever it is met.
One timid lady once confessed to me that she suspected something seriously lacking in her appreciation of music because of her inability to connect it with anything definite. That is getting the whole thing backward, of course.*
Copland said more about this question, based on intuition and experience, than any scientific experiment could.



(That's the movement from Debussy's La Mer that the Cognitive Daily survey had readers listen to, "Dialogue du vent et de la mer," conducted by Herbert von Karajan.)

* Line breaks added for readability.

7 comments:

Jason (the commenter) said...

But you need to be explicitly shown or told what the music is supposed to evoke -- which means the music on its own doesn't evoke specific things.

They don't evoke the things the composer always intended, that's for sure, but everyone who listened to the music thought it meant something and wrote it down.

It's like listening to a language you can't understand. It may be a very pretty language that makes you dream of clouds in moonlight, but once you find a translator you can enjoy it on a different level.

I'd bet most people don't know what the music they are listening to is supposed to mean, but they still enjoy it, and that is the power of music.

John Althouse Cohen said...

I think there's some truth to that analogy, but there's also a distinction between programmatic music (music that's supposed to fit a story, like a tone poem or movie soundtrack) and words in a language you don't understand. Music (of any sort) generally expresses its "meaning" directly through its combination of sounds: melodies, rhythms, etc. Programmatic music gives you those things even if you're not aware of what stories or images are supposed to go along with it. But language doesn't work like that. Language doesn't directly mean things through its sounds; you need to first learn the arbitrary meanings the words correspond to.

Jason (the commenter) said...

Music (of any sort) generally expresses its "meaning" directly through its combination of sounds: melodies, rhythms, etc.

As Cognitive Daily shows, the meaning is only emotional, the same sort of thing you'd get from spoken words.

I don't know about you, but I like watching foreign language films with subtitles, not dubbing, because I want to hear the emotions the actors had in their voices.

Language doesn't directly mean things through its sounds; you need to first learn the arbitrary meanings the words correspond to.

Creek, gurgle, honk, katydid, hiccup, zoom, bang, beep, splash...

Most people could also listen to a conversation and tell if someone is being scolded or being cared for. Most dogs could.

Even the emotions we hear in music may not be the types of emotions other people pick up on. Bach might say Debussy is anti-religious, earlier musicians might say that about Bach. Not to say that they wouldn't pick up on the emotions intended, they'd just weight them differently and have different reactions. You could say even the emotions we get from music are in part learned.

Early western musicians thought you could express specific meanings with music. They were able to produce music that expressed things like numbers, running up and down hills, death etc.

But how often do you run up and down hills?

Also, the sound track to the book Anathem has some examples of music that could be decoded by an alien to describe mathematical concepts.

John Althouse Cohen said...

All valid points ... and your observation about Bach/Debussy could be the basis for a whole other blog post about music's connection to religion. On one hand, a lot of music is clearly supposed to express profoundly religious feelings. On the other hand, as someone who's not religious, I feel like I can experience the some profundity in listening to a mass, but without the actual beliefs accompanying it. You could even argue I'm more directly experiencing the music itself than religious people are...

AES said...

Leonard Bernstein makes the same point in the first of his recorded concerts for kids. He moves from the William Tell Overture, which kids in the fifties would recognize immediately from the Lone Ranger and seems to have a "story" to other pieces. Classical music is often taught as if it "tells a story." He wants to tune their ears differently, since that expectation inhibits what they'd intuitively experience.

If you can get your hands on it, you might find it interesting. At the very least, you'll see a great teacher at work. It's also quite cool to see all the boys and girls in their best dresses and bow ties coming out to the concert.

John Althouse Cohen said...

AES: Great citation for this point! I recently got a set of DVDs called The Unanswered Question as a gift -- several hours of Leonard Bernstein lectures. I haven't watched all of them yet, but I'm almost sure he addresses programmatic music, so I'll be on notice for this point.

John Althouse Cohen said...

Oh, I forgot to point out:

Language doesn't directly mean things through its sounds; you need to first learn the arbitrary meanings the words correspond to.

Creek, gurgle, honk, katydid, hiccup, zoom, bang, beep, splash...

These aren't counterexamples. Yes, there's a tiny portion of words whose sounds have some resemblance to their meaning, but you still need to learn that "hiccup" means hiccup. It's a different process than listening to the first prelude in Bach's Well-Tempered Keyboard and hearing it as tranquil (which isn't to deny that the latter process requires some training). Also, the very fact that onomotpoeias stand out from most words shows that most words' sounds have nothing to do with their meanings.