Every so often—sometimes in lessons, sometimes in the classroom, sometimes in casual conversations about classical music—somebody will remark that this or that characteristic of a musical work “had to be” the way that it appears in the score: in other words, that internal, immanent musical criteria necessitated certain compositional decisions down the line. This is a cool thing to say because it suggests that a piece of music behaves like an integral system, a proposition that satisfies our old epistemological obligation to conceive of the “work” as a thing, an object that seems to contain a subjectivity. (To complement Adorno with Bourdieu: The claim that a musical event “has to be” is also an assertion about the claimant, specifically about his or her discrimination and faith in the doxa of Western musical thought.)
It’s a cool thing to say, but it’s not a particularly true thing to say. Even though its premise is so widely accepted that it represented bona fide news when Maynard Solomon identified a “noninevitability” in Beethoven’s late catalog, we all know that at any time you (and I do mean you) could write whatever note you want whenever you want, however you want to write it. Certainly some practices are normative and therefore privileged—we can usually assume that a piece of 18th-century music will end in the same key it started in, and no doubt there are comparable normativities in effect today, however abstract and whether or not we have the critical distance to recognize them—but that limitation is really coterminous with the boundary of our historically specific imaginations.
The only sense in which something “has to be” in a piece of music is the sense in which to do otherwise would have been unimaginable, which allows for a great deal of latitude. Could the triad at the end of the Ode to Napoleon Buonaparte been something besides an E-flat major triad? Maybe the “rules” of the piece—those explicitly constructed by Schoenberg or those brought into relief by later analysis—demand it (and it would be reassuring to think so), but Schoenberg was a person; the piece, really, is paper. Is it fair to say that he couldn’t have written an E-flat minor triad, or a major triad in a different voicing or at a different dynamic level? To me, it’s a bracing reminder that the “unimaginable otherwise” has always been the only real frontier of new music.