Here's a question for you composers. I didn't know whether to put this in Pianist Corner or here, but I think this could be interesting.

Can a theme be too good on its own to be used for variation treatment?

Of course, a good theme has to give you something to work with, but still...

I got to thinking about this as I recalled a performance I heard of Cornelius Cardew's Thalmann Variations for piano. I'd never heard it before, and I found the theme incredibly beautiful. Then the variations...not so much. I kept wanting to hear the theme again!

So today I listened online to the beginnings of a few variation sets by Mozart, Mendelssohn, Brahms (Handel Variations) and of course the Rzewski. Nice themes but not earth-shaking. Beethoven's Diabelli Variations, of course, is the prime example.

Of course, maybe that was Cardew's fault, I don't know. But I'm still captivated by that tune.

Just interested in your thoughts.

Ha! Good question. I've been on both sides of this. To get analytical it matters whether you're doing variations on a chord progression (Goldberg Variations and typical of Baroque music) or a melody (pretty much all variations since). The genius of the Goldbergs is that Bach never quite lost the chord progression, but twisted and stretched it in so many ways.

When doing variations on a melody it's expected that there will be significant rhythmic and harmonic variation. The melody has to provide some recognizable points that will be recognizable when varied significantly. The alternative is the composer creates inspired variations that can be tied to the original theme after some thought but the piece as a whole hangs together. Think about Rachmaninoff's Paganini Variations, the famous 18th variation doesn't sound anything like the theme but because it's so beautiful we don't care. It's only when it's explained that the theme is inverted in that particular variation that it makes any sense.

So variations can work either way, in one case the theme is recognizable and the audience enjoys hearing the theme is different forms. In the other case the music is (hopefully) extraordinary and the relationship to the original theme is irrelevant because the music is engaging.

I feel anyone learning the craft of composition should spend a lot of time doing variations. It's a vital skill that gets used in all types of compositions. The advice I usually give is that one should write a theme and then ten variations and then throw out the first five because they are usually the most obvious treatments of the theme. Then compose ten more and that's when one has to get creative. That's when things get interesting.

To answer your question, if the variations aren't as good as the theme what's the point? Of course it was Cardew's fault.

Joe,

A very interesting question indeed and thanks for posting it!

Now, first of all, I should note that composing variations is probably the most difficult task for a composer, and this is why it's usually included in all final exams for composers and why very few works on variations exist by composers... I don't think any composer has written more than 2 variations on other themes... That and because it also seems a tiny bit like cheating a bit!

Anyhow...

It's very interesting that very specific themes are given for variations ("La Folia", paganinni, etc...). not every theme is handy for variation treatment... For example the moonlight sonata is exactly the negative of a great theme for variations...

______________________

But to get to your question (like Steve, who take a while to reply as well...).

If the theme is too beautiful, or too recognizable, things must get out of hand (throw away the first five variations, etc... ) quickly enough. Otherwise it's an 'ouch' situation. If the theme is not beautiful, then there's no reason to vary it.

Joe,

Your homework for tonight: rondo form.

Ed

Sorry Nikolas; Alkan Joe...I really agree with your thoughts; is it not possible for there to be a *best* version of a composition and, as such, any variation upon it would be merely tarnish? Um...rondo *is* rather a good solution...
Xxx

Thanks for all your comments. It's interesting for me because, in spite of my playing a lot of contemporary music, I'm not a composer myself and have no interest in composing. In fact, I am in awe of people who do compose. I especially like your ideas about writing the first few variations and discarding them! That sounds like a valid way of working.

If you call up the Thalmann theme (the only way, I think, is to reference the entire variations) on Youtube (or wherever) I'd be interested if you find it attractive, as I do.