It was a good weekend.

We had a traditional Thanksgiving dinner on Thursday, turkey, stuffing, gravy, and pie, in the company of my son. I performed my usual sous chef and dishwasher role while Marta cooked. We traded awful puns around the dinner table, and took our time with the afternoon.

I also had a wonderfully nontraditional Thanksgiving, personally, in that I spent my free time this weekend finishing the first movement of my symphony. I’d gotten stuck in the middle of the development section, and once I broke through — on Tuesday evening — the rest of the movement simply poured out. Ninety-eight measures in about twenty-four hours spread over four days.

I’ve wanted to talk a little about the composing process, at least as I experience it. I’m sure each composer has his or her own experience. But I’ve encountered the idea out there that “classical” music is a lifeless kind of mathematics or formal language rendered as sound.

I don’t believe that statement to be even slightly true: it certainly isn’t true for me. That idea is instead, I think, an artifact of the way that many academics try to teach musical composition as a left-brained, formal, analytical process — how they try to systematically create composers with a curriculum and measure their own progress in doing so. I think a lot of “music” has been composed formally according to this academic process in the last century, and much of it is indeed lifeless: at its best, it is animated, like an electrified corpse; at its worst, it is simply grotesque. The best composers, like the best writers, somehow escape the consequences of their education.

I’d like to go through the closing passage of the movement I just completed as an example of how this works for me.

Here’s one place formal theory comes in, in the structure of the movement: beginning, middle, end. They call the beginning the exposition, where you expose your main themes. The middle is the development, where you beat the main themes senseless, sometimes to the point where you can barely recognize them. The end is the reprise, where you bring back the main themes for a final airing, before the coda, which is the “…and they lived happily ever after” formula at the end that tells you the music is done.

So here is the reprise of the main theme that appears at the end, as I’m wrapping up for the final climax:

Here’s where the composing comes in: in my head, these two themes fit together. So far in the movement, I’ve not done this — they’ve always been kept apart. In my head, they form a perfect dance, sometimes called counterpoint — two themes or melodies that have their own completely independent structure, yet they fit together like two lovers in a single bed. Here’s what happens when I put these two lovers in one bed:

This isn’t proper counterpoint at all, using traditional music theory. It’s just two independent themes laid on top of each other, and while they aren’t a complete mess, they have some rather nasty dissonances, places where the two themes clash like a boxcar full of fingernails colliding with a boxcar full of chalkboards. Nonetheless, they continue to profess their undying love in my head, so what I normally do next in such a situation is add a bass line to anchor the harmonies. In this case, I add two: a moving line in the cello, and a ground — a single note, like a bagpipe drone — in the string basses. It comes out like this:

You can see how this changes the whole texture of the dissonance. In particular, the ground provides a continuity that builds tension — even though the harmonies (and disharmonies) are running mad all over the place, the ground does exactly what the word suggests: it anchors everything together by tying it to the earth. By grounding it. It also supplies a slightly ominous tone that speaks of endings. So far, so good.

Next, I threw in some french horns to fill out the inner harmonies, just to make sure that the wild harmonic lovemaking in my head wasn’t merely the result of too much wine with dinner.

Why french horns? That’s a mystery. When I hear it in my head, it’s french horns. It isn’t the only option, and if the horns don’t work, I can always try something else. But here’s what happens when that harmony is added:

This is now much closer to what I hear in my head.

It’s straightforward to make the sound richer. In this case, I didn’t even have to work very hard. I added a bunch of brass playing the main theme in unison — all the same notes, with the trombones an octave below — and added oboe, clarinet, and bassoon in unison on the secondary theme. Notice how different the texture of the sound is:

The next part is a bit tricky. There’s a pattern I’ve been using in the violins throughout the movement, and it needs to be folded in here as well. It sounds like this:

Because this is just a bowing pattern for the violins, I can change the notes to anything that fits the rest of the music, and since it isn’t a melody, it’s very flexible. The only precondition is that it needs to keep climbing in pitch: I’m building to a climax, and I want more and more tension. Soaring spiccato violins always build tension.

I felt that this violin part needed a bit more richness to it, so I added a counterpoint in the second violins. They fit together like this:

I don’t know how other composers manage these fast passages. This is one place where music theory is your dearest friend, particularly if you are writing with nothing but pen and paper and the music in your head, and my admiration for composers like Beethoven, Rachmaninoff, and Stravinsky is boundless. They could test out their ideas using whatever instruments they could play, typically piano, but often, they didn’t seem to need to. That amazes me.

I’m not Beethoven, nor Rachmaninoff, nor Stravinsky. I need to hear the music.

But I also get to cheat. I have a captive orchestra in my computer, and it will rehearse at any tempo I like, as many times as I like. The players never get tired, and never quit and go join some other orchestra because I keep making them play the same thing over and over until I get it right. So I end up doing something like this:

Playing it back at much slower tempos lets me hear the bogus notes clearly, and I can mix and match different instruments to figure out if all the harmonies actually work. Often, they don’t: the music in my head is not always accurate. I can often push a note one way or the other to get it to fit. But sometimes I have to throw out the whole idea and start over.

In this case, it all worked out fine.

Then there is one final trick with the viola. This entire passage is in four, meaning that there are four beats to a measure, subdivided into eighths and sixteenths for the fast notes. Adding triplets to a big passage of this sort adds a kind of syncopation that feeds the excitement. So I put in this:

These are exactly the same chords I’ve been doing already, turned upside down and played as triplet quarter-notes — not a lot of work to put together. When all of the strings play, it comes out like this; you can clearly hear the syncopation the violas are adding:

Finally, bringing it all together with timpani and cymbals for dramatic effect, we get this:

---

Now, all of the preceding is merely to set the stage to discuss four measures at the very end. I’m going to start by playing the these four measures (with a two-measure lead-in and a trailer) very, very slowly:

What a hideous cacophony! Every music school in Europe prior to 1850 would have thrown me out on the street for the first two measures of disharmony, and they’d have tarred and feathered me for the next two.

I spent nearly four hours on these four measures, tweaking the string parts note-by-note, trying to work out this extremely complex progression.

Here is what happens when I bring it up to full tempo:

I think it works rather well.

I suspect that any musicologist worth his pay grade could tell me that, “Oh, that’s merely an augmented Phrygian scale with a modified McPherson modulation in the fourth Plagal Filbert.”

My response to such a statement would be much the same as if a physician told me that I was experiencing an unusual number of premature ventricular contractions — “Golly gee! And I’m not even pregnant.”

Maybe this really is simple to a musicologist with a solid grasp of music theory. I wouldn’t know.

The truth is, I have no idea what I did here. If there is actually enough theory in the world to describe it in analytical terms (and I’m sure there is), I don’t know that much theory. I have never known that much theory.

So unless I’m channeling the spirit of a deceased composer who actually does know that much theory — not impossible, I suppose — I didn’t use any mathematical or linguistic theory of music to write these two measures.

Either way, the final music always surprises and delights me. That’s why I enjoy it as I do.