Saturday, October 25, 2014

Schubert and Harmony

We have discussed harmony a lot here at the Music Salon, most recently in this post on Harmonic Deficiencies. The so-called "common practice" period of harmony, lasting from around 1600 to around 1900 is well behind us now and its felicities are really only heard at classical music concerts. Nowadays the more old-fashioned composers are still writing atonal music, which is music that defies all the rules of harmony. The more up-to-date composers are writing some kind of modal music or using drones or perhaps some kind of polytonality. The "common practice" of harmony is less-known than it should be. The truth is that common practice harmony is the most developed, subtle, sensitive, finely-calibrated harmony ever developed in music--little wonder that it reigned for three hundred years!

I was listening to several pieces by Stravinsky in the last couple of days and while his harmony is inventive, fresh and stimulating, it is, compared with the best examples of common practice harmony, crude and harsh to the ears.

This is all prompted by reading a post on Luke Dahn's blog. He is an excellent theorist and wrote a fascinating post about the possibility of using one piece of music as material for teaching everything there is to know about chromatic harmony. The piece he chose is the masterpiece by Schubert, the song-cycle Die schöne Müllerin. Luke has prepared a beautiful chart showing just what harmonic devices Schubert used in the songs. Here is the link. I can embed it here as well:

But it is much more legible if you follow the link. This is a beautifully graphic analytic overview--the kind of thing I might do if I were teaching a theory course! You can download the complete score here. And here is a complete performance by Dietrich Fischer-Dieskau and Gerald Moore with the score. It even includes the poet's prologue:

So you can, through the wonders of the internet, give yourself a complete course in chromatic harmony taught by Luke Dahn, with the aid of Franz Schubert. How, exactly? Well, let me walk you through the first song. First, go download that score from IMSLP (what a wonderful resource!). Then print out the two pages of the first song, Das Wandern. Now, let's have a look at Luke's chart. As we can see, Das Wandern is in the key of B flat. That is shown in Luke's chart by the "Bb:" at the beginning. This is a very simple song and the only harmonic device he shows is the little purple "V" in measure 13. The purple indicates that this is a "tonicized key area". Tonicized? Whazzat?

What Luke is teaching is not the most basic level of harmony, i.e. all those chords that are a normal part of the key, but rather chromatic harmonic, that is, all those devices that use accidentals, chromatic alterations, to either strengthen chords that are in the key, as the augmented sixth chords do, or to intensify certain chords by "tonicizing" them, i.e. making them momentary tonics in their own right by using secondary or applied dominants. Let's look at the Schubert song to see how this works. The song is in B flat. In measure 13 (you have numbered your measures, right?), which is the first measure on the second page in the score I downloaded, we see this:

Looks ok--hey, wait a minute, what is that F# doing there in the bass? That's the chromatic part. In order to temporarily make a chord into a tonic that is not the tonic in the key, Bb, we have to use an accidental. In this case the F# is the leading tone in the key of G minor, which is the relative minor (meaning it shares the same key signature of two flats as Bb major) of Bb. This makes the last harmony of m. 13 into a dominant of G minor, what we would analyze as a "V6" of V (V6 because the F#, the third of the chord D, F# A, is in the bass). Now go and listen to the song and see if you can hear this momentary departure from the key of Bb--it adds harmonic richness to the third phrase, creating the climax of the song. Incidentally, this tonicization forms the first part of a sequence, which means that the same idea is repeated at a different pitch. The tonicization, using F#, of G (minor) is followed in m. 15 with the tonicization of the dominant ("V of V"), F. This is done with its leading tone, E natural.

In the second song, Wohin?, we get not only more tonicization, but also, in mm 38-40, use of the Italian augmented sixth chord, shown with a little Italian flag. You can read up on augmented sixth chords here.

That should get you started!! With all the resources available online these days, anyone who actually wants to learn about music, can do so---for a song!



Anonymous said...

At the risk of sounding pedantic, I want to bring up the more recent research on Schubert which shows that this is not the best way to think about his harmony. Essentially, the attempt here is to use the cycle of 5ths as the template and try to fit Schubert voice leading to it. It's like figuring out the solar system with the earth in the middle. It's doable but awkward.

There's a harmonic system called neo-Riemannian which is based on the hexatonic cycles (4 sequences of 6 notes). In the key of C, you have C,D#,E,G,G#,B). Once you replace the cycle of 5ths by the hexatonic cycles, Schubert's music becomes as harmonically "proximal" as, say, the Beatles's music is proximal with respect to the cycle of 5ths (in the sense that it never strays far from the tonal center).

Roughly, the hexatonic system is built by composing three operations: parallelization; leading tone exchange; and relativization.

People like Cohn and Clark have worked out the details, but it's all extremely simple. In the hexatonic systems, Schubert's harmony becomes essentially trivial -- there is no "momentary departure from the key" or surprising chromaticisms --, which is why I think it's the right way to think about it.

Bryan Townsend said...

Anonymous, here at the Music Salon you can be as pedantic as you want! We don't dumb down anything. In fact, a very hearty thanks to you for alerting me to neo-Riemannian theory. I had never heard of it before and after reading the Wikipedia article I am extremely interested. In fact, I just ordered the Oxford Handbook of Neo-Riemannian Music Theories. Why am I so excited?

As a composer, I have been working on re-thinking harmony for a long time now. A while ago on this blog I called myself a "post-modern neo-medievalist" composer and stated as the reason that I used voice-leading principles as a fundamental practice in my music. In private conversation with a violinist friend, I told her that the secret of a cadence I had just worked out to end a movement in one of my symphonies was entirely due to my use of a certain kind of voice-leading. So when I read that neo-Riemannian theory is very interested in the efficiency of voice-leading, it piques my interest.

I have been arriving at these ideas purely through exercises in composition, so it is fascinating to see how theorists have been arriving at, if not the same place, then a similar place.

Neo-Riemannian theory seems to be quite new. The Oxford Handbook was just published on May 1st of this year and an important book by Richard Cohn was published in 2011.

My deepest thanks to you, Anonymous, for alerting me to these ideas! Are you a professional music theorist?

Anonymous said...

I am not a professional music theorist but I know Suzannah Clark, who told me about her wonderful work on Schubert.

Bryan Townsend said...

Great! But did you really mean to say that "Schubert's harmony becomes essentially trivial", or that ANALYZING it becomes trivially easy?

Anonymous said...

I meant analyzing becomes much easier and more natural. Nothing "trivial" about Schubert.

Bryan Townsend said...

You bet!