My favorite internet comment of all time was one I saw way back in middle school, or possibly my first year of high school. It was in a music discussion forum the name of which I am too lazy to recall, and it read simply “Remember everyone! Music theory is a theory and not a fact!”, as tho someone somewhere out there was taking a valiant stand against the forces of analysis in favor of some kind of sonic creationism.

The parallel is inexact, but they’re not wrong. Music theory, that sprawling assemblage of analytic principles and rules for what to do when, is largely a descriptive art. It takes as input not some abstract, a priori assumptions about Art but instead specific pieces of music as composers have written them, with all the contradictions and idiosyncrasies that you would expect from people making art. Pretty much everyone agrees that Johann Sebastian Bach wrote really good 18th–Century counterpoint, so let’s look at his fugues and try to figure out what makes them tick. What do the works of Amy Beach have to show about how earlier generations of composers in the United States attempted to modify and preserve Germanic methods of putting chords together? As Claude Debussy once quipped, “works of art make rules, but rules do not make works of art”.

Admittedly, there have always been composers who were active theorists and who wrote pieces in part to illustrate their theoretical ideas. (Twentieth–Century serialist composers will probably spring to mind for many of you, but the history of this kind of thing is longer than that, including people like Anton Reicha who had quite a lot to say about form in the early 1800s.) Some of these creations have become dazzling and brilliant staples of the concert canon while others flopped abjectly from the start, but even the best aren’t always the rebuttal to Debussy they might seem. For every theorist-composer who boldly puts forth a new work as an illustration of a new theoretical system, there’s a subsequent conversation about whether that theoretical system is really the most compelling one to analyze the work, as well as how well the work hews to it anyway*. Many 19th–Century theories now seem inadequate as descriptions of the music they inspired, and I wouldn’t be at all surprised if a hundred years from now theorists have similar things to say about theory–driven works of the more recent past.

Still, even given this messy, ever–changing landscape, despite the way musical works resist any overly tidy pinning down, I’m very interested in music theory, and I like reading about as many approaches to it as I can. This isn’t just a background interest or a desire to be a well rounded musician; being fluent in numerous theoretical approaches to music is an important part of my composerly toolkit. I know I snarked about composers who write theory–first in the previous paragraph, but I do the very same thing myself, albeit usually with theories others have devised, not ones I’ve concocted myself.

Writing music is hard. Even generating the basic musical materials for a piece can be a struggle at times, but then comes the task of spinning those materials out into a full piece, balancing enough variety to be interesting with enough similarity to keep it sounding like a unified whole. Sometimes this is a fluid, intuitive process, but sometimes I get stuck. I’m not quite sure what should happen next, or I know where I need to go, but I’m not quite sure how to get there from where I am.

And that’s where theory steps in. Music theory, in some sense, is just a collection of descriptions of how other composers have solved similar compositional problems in the past. Not every tool will be applicable in every situation — I seldom care about 17th–Century proscriptions against parallel fifths and unresolved dissonances, for example — but the crossovers can be surprising sometimes. Ideas from Schoenberg have gotten me out of sticky spots in tonal music, and tricks I’ve picked up from music theatre writing have gone into some bristlingly atonal passages in turn.

I think of it kind of like scaffolding. If you covered a building entirely with scaffolding then took away the building, you’d still be able to describe the erstwhile edifice in a fair bit of detail. Over here there’s probably a minaret, that looks like a long flat wall, and — are those flying buttresses? Similarly, with theory, noodling around with various analytical approaches to what I have already won’t magically produce the final piece out of thin air, but it can offer suggestions as to what the next steps might be. This fanfare reduces to the same set class as the first half of that waltz theme; the little throwaway melody here sounds really cool upside down and backwards; if you get aggressive about neighbor tones and the like, you can eke a tonal chord progression out of this dissonant mess. The trick is to generate a lot of possibilities and see which ones stick, and theory helps immensely with that.

When you’re done building a building, you take the scaffolding down. Similarly, once I’m done writing a piece, I expect it to stand just fine on its own, without a protective shell of theoretical justification. Not everyone feels the same way about this, and that’s fine! The world is wide enough for all kinds of music and approaches to it, but I’m not particularly fussed about being theoretically consistent, and am perfectly happy to sweep the details of my process under the rug once I’m done with the writing process. Someone analyzing my work down the line may well reverse-engineer the steps I took, but they might just as well find some alternate framework that explains what I’ve done much more clearly than even I would have been able to. That’s fine! I’m no Reicha or Schoenberg — at the end of the day, I just want to build things that people enjoy. Music theory helps me do that, and I think that’s pretty neat.

____________________
*The lowest–hanging fruit for this may be Igor Stravinsky: In 1924, he published a manifesto describing his Octet for Winds as a paradigmatic work of formalism and Musical Objectivism, claiming to have “excluded all the nuances between the forte [loud] and the piano [soft]” [Scribd] despite the fact that on the very first page of the Octet [NY Phil digital scores archive], the flute is clearly marked mezzo-forte [medium loud]. So, you know, grains of salt all around.