Monday, January 17, 2011

Why is Melody a Dirty Word?

I was sitting in a composition seminar one day when a fellow composer presented a piece that was fairly atonal (meaning there was no tonal center, or key- see Natalie's previous post on tonality for a brief explanation). The discussion after the piece was presented focused largely on whether or not there was a melody at one point, which clearly outraged several of my colleagues. Part of the problem was that they felt a melody didn't fit in the piece based on the rest of its contents, which may have been a perfectly valid point. But a few disparaging comments about melody were made, and that bothered me. What's wrong with a good melody? And what have melodies done to deserve our wrath?

I think melody became a dirty word to some classical/new music composers because of its association with tonality.  Think about a melody from one of your favorite songs or pieces, or something you sang or hummed in the shower this morning.  It probably was a tonal melody that used traditional tonal structures--the "tension and release" of moving from dominant to tonic--and had a satisfying ending.  I say that because tonal melodies are a lot easier to remember and sing than a cool part of your favorite new music piece (maybe).  And if it's tonal, it will be shunned by some (or many) new music composers- again see Natalie's previous post that I referenced earlier.

What is a melody?  Is it something you can hum easily?  Is it a recognizable phrase, a string of notes with a distinct rhythm that somehow fit together?  Must melodies necessarily have patterns of tension and release, whether tonal or not?  Can a melody be good if it isn't tonal, if it doesn't have a "satisfying" ending on tonic?  At least to answer that last question, I certainly think so.

To demonstrate my point, below are two examples by one of the least tonal composers of the 20th century, Arnold Schoenberg.  Schoenberg created the 12-tone system for composition, basically the antithesis of tonality.  Both videos show parts of his landmark (and awesome) piece, Pierrot Lunaire, a song cycle made up of 21 poems split into three sections of seven poems each (if you're curious, here's some background).  While Pierrot was written before Schoenberg came up with his 12-tone system, he had already begun to experiment with free atonality (and if you've never heard Schoenberg before, you'll know exactly what I mean by "free atonality" as soon as you listen to this).

The first clip is from the beginning of the piece.

This next clip is the eighth poem, "Nacht," which begins the middle section of Pierrot Lunaire.  I partly wanted to post this just because the video is pretty amazing/creepy.

Are there melodies in this piece?  Absolutely.  Are they tonal?  Absolutely not.  I will let you be the judge of whether or not they're satisfying.  But I think what is interesting to note is that these atonal melodies are just as expressive and evocative of emotion as traditional tonal melodies, albeit in a very different way.  But if expressiveness is one of the criteria that supposedly makes tonal melodies "good," then the melodies in Pierrot Lunaire get an A+ from me.

Of course, it is true that some of my favorite melodies in classical music (even newer classical music), are tonal.  That isn't to say that great atonal melodies don't exist, it just means that we will have to redefine the parameters we use to judge whether a melody is good.  Satisfying doesn't have to mean ending on tonic.  A great melody doesn't have to be tonal.  But if it is tonal, I'm okay with that.  You can even have a tonal melody in a largely atonal piece.  We are living in a new music world where (almost) anything goes- so why do melodies often seem unacceptable?

Many great pieces don't have one single recognizable melody.  And that's fine, too.  I'm not advocating that every piece should use melodies, whether tonal or atonal.  The composers today who avoid writing melodies (and I used to be one of them) favor exploring interesting textures, gestures, timbres, etc.  But if and when a melody does creep into our music, let's not be ashamed of it.

Posted by Sarah


  1. Really interesting, Sarah! I'd always sort of thought of melodies as something which a person might end up humming after hearing it, which the Schoenberg bit is not, at least for me, but it certainly lingers in a different way.

  2. Thanks, Annaka!

    We could probably talk all day about what defines a melody, and it's true that melody is sometimes defined as something you can hum (which I think is why it's often shunned by new music types). But at its most basic (and here I will turn to the Oxford Dictionary of Music for a more elegant definition than I would have written) melody is "a succession of notes, varying in pitch, which have an organized and recognizable shape."

    Most people would probably have difficulty humming Schoenberg's (or any other composer's) atonal melodies, especially after only listening once or twice. But I think that's partly because they're so different from the music most of us listen to on a regular basis- after listening to Schoenberg repeatedly, I bet you could hum or sing some of his melodies (even if you weren't a singer...).

    Thanks for reading (and posting!)!

    - Sarah