Tag Archives: improvisation

Playing for ballet class tips #22: With harmony too, less is more

Swanilda's famous waltz. The first harmonic change comes only at the very end of the phrase.

Swanilda’s famous waltz from Coppélia. The first harmonic change comes only at the very end of the phrase.

If you’re improvising or harmonising a melody, there’s a lot to be said for just sticking to simple harmonies, and avoiding chromaticism or excessive modulation for the sake of it. It’s not a competition to see who can fit the most chords in. On the contrary, dance music depends on a certain amount of harmonic simplicity for its dance quality and feeling of lift and lightness.

It was one of my dissertation students who first drew my attention to this: if you look at some of the most famous and well-loved waltzes you can think of, many of them of them follow the pattern of Swanilda’s variation, which is to stay on the tonic for 6 bars, and then move to the dominant only in bars 7 and 8. In the case of Swanilda, what then happens is the reverse, like a harmonic palindrome – 6 bars of dominant 7th, followed by 2 bars of tonic.

Strauss does it, and Tchaikovsky does it. Another variant is to stay four bars in the tonic, and four in the dominant. Whatever happens, you get very simple harmony with a bass line toggling between the 1st and 5th degrees of the scale, little more. I looked at Oskar Nedbal’s Valse Triste for ages, trying to work out how he had achieved such subtle and unusual beauty, only to find that most of it was down to what he doesn’t do – he never moves from a bass line of G and D in the first 8 bars;  and again, the harmony is tonic for 6 bars, dominant(ish) for 2.

Likewise, two of the most famous codas in the ballet repertoire, the one in Don Quixote pas de deux and the one from Black Swan pas de deux, sit on a tonic pedal for ages, and modulate properly only right at the end of the phrase.

Yet the temptation when you’re improvising or composing is to try and throw as many tricks as you can into 16 bars of music, like you’re loading your plate at the salad bar.  I’ve seen 16 bar compositions for tendu exercises that have already modulated to a remote key by bar 4 (with a change of key signature), chromatic inner voices and bass-lines, interrupted cadences, and hardly a simple tonic or dominant chord in sight. I’d like to say that the result is a real dog’s dinner, harmonically, but in fact, dogs’ dinners I’ve seen tend to make more culinary sense.

If there’s a principle to follow, it’s to remember that 16 bars of music in a dance class have to be imagined as being a ‘clip’ of something larger, not a self-contained miniature. In fact, who writes 16 bar miniatures? There isn’t enough time to develop and resolve musical tension, so don’t try.

Playing for ballet class tips #18: take a lesson in 18th century composition


One of Riepel’s examples – reprinted by Eckhart (click to view the article)

A lot of pianists are frightened of ‘improvisation’  because the term is widely (mis)understood to mean that you play a chord with your left hand and wait for God to tell you what to do with your right.  It’s seen as a gift, or – even more damning – as a sign of liberation from the constraints that hold lesser mortals (like you or me) back. But this is just wrong. Improvisation is all about working within conventions and constraints. That’s why I haven’t even mentioned it until now – how can you improvise if you’ve got no models to work with? The appearance of spontaneity is the result of years of enculturation and practice, some of which looks much more like ‘composition’  than improvisation.  For more on that, read Peter Martin’s excellent chapters on improvisation in Music and the sociological gaze

The line between composition and improvisation is fuzzy: composers try stuff out in an improvisational way before working that material into something more studied; improvisers privately use compositional models and techniques to develop material that they’ll use ‘spontaneously’ in a performance.  And contrary to our modern concept of composition as drawing inspiration down from heaven to create something completely original on a blank sheet, 18th century composers learned their craft in a way that looks more like what we’d call improvisation, working with models to create music within the conventions of a recognisable style.

18th century composition manuals offer wonderful lessons in how to ‘improvise’ for class, because they draw attention to the mechanics and processes involved in conventions that are so conventional it’s hard to see how they’re constructed.  My favourite is Riepel’s Anfangsgründe zur musicalischen Setzkunst (Fundamentals of Musical Composition), which is a written-out conversation with a student on how to write a minuet. Even better than the original is Stefan Eckert’s article on it, So, you want to write a Minuet?” – Historical Perspectives in Teaching Theory, free to read over at Music  Theory Online. 

One of the first thing that Riepel teaches you is about the phrase construction of a minuet – ideally, two zweyers and a vierer (two two-bar phrases and a four-bar phrase). Don’t screw this up, he says (well, in so many words), because this is what makes a minuet a minuet. Then Riepel has all kinds of advice about how to construct melodies in this framework, using an imaginary student composition as an example, pulling it apart, improving it according to rules and procedures, until it begins to look like a minuet you’d buy in the shops.

Look at your average plié exercise, and you’ll see that this, too, is built on the same pattern (two demi-pliés and a full, 2+2+4). Most of what goes wrong with music for pliés is when the phrasing can’t breathe with the exercise, and all you need to get it right is to think like Riepel about whether you need zweyers or vierers.  And if you think you can’t improvise, then this surely makes it easier – it’s two things, another two things, and then four things. If you think you can improvise, don’t think it’s all about ‘breaking convention’. Some of those conventions help the dancer to phrase their movement.

Another fascinating insight into the improvisational and artisanal character of composition is in the use of Gebrauchs-formulas, described by Robert Gjerdingen in his article of the same name.  That article is behind a Jstor paywall, but you can view his translations of composition manuals at the wonderful site, Monuments of Partimenti

And on that subject, a big round of applause, please, for Simon Frosi, whose bacherlor’s dissertation (The Improvisation of Structured Keyboard Accompaniments for the Ballet Class) is probably the first and only place that you’ll see a discussion of the relationship between partimenti and ballet class improvisation. It’s free to read – press the ‘download’ button.

Update on 16th Feb 2013: If improvsation and the baroque interests you, you might also like this  article just released from Music Performance Research online: Incorporating long-range planning into the pedagogy of Baroque-style keyboard improvisation (Vol. 5, 59-78)

Playing for ballet class tips #17: Composing and improvising


Today I’m delighted to have a guest blog by Andy Higgs, friend, composer and musician who’s currently company pianist at Ballett am Rhein in Düsseldorf. This all came about because he asked if I was going to do something about composing for class. I said that I was going to, but more as a footnote – because I think that class is all about working within rules, conventions and constraints, and as such has much more to do with improvisation than ‘composition’ in the usual sense of the word. But who better to talk about composing than a composer who plays for class? So I asked Andy if he’d be willing to be a guest expert blogger.  As it happens, what he’s written touches on a whole bunch of  issues that lie at the heart of playing for class, and I believe (as I think he does) that it’s holding those different things in constant tension that make a good accompanist, and that’s about as good a tip as you can get about this job.  The improvisation/composition point he raises at the end is a whole other topic, so I’m going to deal with that tomorrow!

Guest blog by Andy Higgs:


I am writing this as I have just played for class and so the thoughts and feelings are fresh in my mind. I was observing myself reacting to the ballet mistress’s requirements and trying to consider exactly how composing original music can be a help as well as a hinderance for class.

If you are a composer, I think it is an excellent idea to have a few of your own creations to add to your repertoire. Good ballet accompaniment requires some variety, and there is room for completely new music within that. The unspoken rule of good party music is something for everyone – not listening to a Madonna CD in its entirety – and I think it is healthy to approach ballet class in the same way. Playing a class entirely of my own music would be a mistake, and teachers who use CDs tend to like to mix and match and not rely on only one for the whole class. I noticed today that the best decision I made after playing something quite experimental for pliés – a piece I had sketched a few days earlier – was to launch immediately into a jazzy Irving Berlin tune for tendus, because it completely altered the atmosphere in the room.

This evening I was playing a short 45 mins warm-up before the performance, and the  dancers usually appreciate a bunch of fun tunes to help them relax or lift their spirits before the show. For daily training in the studio, however, there is room for experimentation, and if you are lucky enough to work with an appreciative director and dancers who like to be challenged, then it is a great opportunity to try out some of your own creations. Playing for ballet often involves throwing yourself into it with the focus being on momentum, energy and expression and it can be a good way to come up with ideas that you can work into proper pieces later. The way ballet is constructed made me think a lot about how music is put together in the simple ways I’d forgotten about at music college, where I’d learnt to obsess over pitch and trying to sound original.

In my experience this experimentation in the studio is more helpful than sitting at a desk and trying to come up with something suitable using my composing brain. There is always that feeling of, ‘it doesn’t quite fit’, when I do this. It is different if you have written and recorded a CD of your music, as the teacher is likely to prepare exercises to the music at home, but expecting your composition to just work to something in morning class isn’t always realistic, especially with very imaginative teachers. A good ballet pianist won’t just play Gershwin’s Stairway to Paradise  for grands battements, they will play their own version of it, responding to the important rhythmic elements of the exercise. Similarly, I find I have to adapt my compositions slightly to make them suitable. From this perspective it is the ability to improvise, adapt and be spontaneous that remain the most valuable skills in the ballet studio.

Also, ballet has an insatiable appetite for music and devours everything that comes into its path, and anyone who has tried writing for class will know that music – which can take hours to write – disappears in a matter of seconds. It’s gone. You have to compose some more. It could take a fairly industrious composer maybe a month to write an hour and a half of music for one class, and even then, how often are you going to use it (the exception being, of course, if it was written for official syllabus work)? You’d need vast amounts of music as you’d soon tire of playing these pieces. Certainly 64 bars of composed music for a Grand Valse is not going to get you very far playing for company class. Even if you have something composed as a foundation, you will soon find yourself having to launch into something else, or extemporise around the piece to keep it sounding fresh. Sounding stale and uninspiring is possibly a ballet accompanist’s biggest fear and I have noticed that this is more likely to happen if I rely too much on one system of playing (for example if I only improvise, or if I only play show tunes). The more tricks I have up my sleeve to get out of this sticky situation the better.

I think there is a certain confidence and completeness about a nicely worked composition that is difficult to achieve with improvisation, which has a risk factor which can work for or against you. But what is required in the ballet studio is very different. It is changeable from moment to moment, and it would make the job less enjoyable by attempting to solve this problem with only one approach to the music making.

Conference on musical improvisation


Interesting conference coming up in September 10th – 13th this year at Oxford University – Perspectives on Musical Improvisation.  I’m half tempted to submit a proposal for a paper, since music improvisation in ballet classes is one of those mysterious and hidden-away things that rarely gets an airing. Just not cool enough, I suppose. Just a shame that this isn’t really my area of interest as a researcher, so I hope someone else will take up the challenge.