In a typical live concert, audience members experience a piece of music as a linear event. There’s no going back to listen to an earlier spot, no lingering over a particularly catchy chord. My experience as a composer is quite different. Over the course of my career, I’ve perhaps composed two or three pieces in which I sat down at a piano and wrote the work from one end to the other without stopping. It is far more common for me to work on multiple sections simultaneously, skipping back and forth throughout the piece as I work.
This approach works well for me for several reasons. First, it allows for musical ideas that occur throughout an entire piece to be developed at more or less the same time. I can introduce a brief version of a motive early in the piece; I can grow it throughout the main body; I can use it triumphantly at the high point, and then bring it back as a whimper in the finals notes of the piece. For a work with a longer duration, I am able to develop multiple movements at once that share some aspect of a motive. For instance, in the large-scale oratorio that I’m currently writing, I am finding it useful to work on ideas for multiple movements of the piece simultaneously. This allows me to devise a few main motives that will be used in a variety of ways throughout the entire oratorio. This approach is also helpful if I get stuck on a particular spot – I can jump to another part of the piece and keep working.
The key to a non-linear approach is staying flexible. I tend to initially work with pencil and paper and eventually switch over to a computer once my sketches are detailed enough. With pencil and paper, it is easy to not get too attached to anything I put down, as the score at this stage usually looks like a hot mess, with musical ideas scribbled all over the place. But with a good computer notation program, you can be seduced into thinking you're much closer to a final product than you really are, as the music looks so neatly engraved on the screen and sounds like music when you hit the playback button. I avoid the siren call of the computer for as long as I can; when I finally switch over to notating within Sibelius, I don’t get too committed to any music I put into the program until much further into the process.
Another key is to think organically. When I write a musical idea that constitutes the opening of a piece, I try to simplify it. Does the piece really begin four to eight measures earlier than I think it does? Perhaps even sixteen measures before my current point of entry? The same organic approach applies to the climax of a piece. Sometimes I’ve worked on building material into a high point for so long that I think the section lasts much longer than it actually does (it is hard to listen with fresh, objective ears). I also find that once I reach a high point, I can experiment to see if the music can momentarily stabilize and withstand one more organic push to reach an even higher point than I first imagined. J.K. Rowling, author of the Harry Potter series, has this approach down extremely well. When she gets to a structural high point of her story, instead of peaking and letting the tension fall away, she turns the high point into a plateau to which she adds more material that climbs to yet another high point.
Sometimes, after working out of order for most of a piece’s construction, I find it surprising to finally hear it in order at the piece’s premiere. After a while, my ears adjust, and then I can’t hear the piece any other way.