Cameron offers a at-first-sight-bizarre description on musical ontology in this paper, and I worry about what it actually answers. So he develops this idea that music doesn't exist in a sort of "ontologese" language when in fact music does exist, and this is not contradictory because, well, I think the easiest way to think about it is that the word music is used the same in each case, but the words exist and exist have two distinct meanings, the first defined as the independent ontological existence of music as an object not reliant on any other objects, and the second referring to music as merely a specific collection of objects combined in such a way to to create a new ontological object, but rather to be so named only in virtue of the sound structure's relative locations to each other. He is admittedly nihilistic in this regard, refusing to commit to creation in general as opposed to creation, so that statues are created and exist yet are not created and do not exist, since nothing is created and statues are statue-shaped marble, or whatever specific claim he wants to make. My worry is about whether or not any of this is actually interesting; or in other words, whether or not it solves any problems. I'll give him everything: that music doesn't actually exist, that it is music-like conglomerations of sound, et cetera. And then I'll ask the question: What is music? Because many possibilities of what music is--namely, all of the possibilities that involve music existing-- we know are wrong. But that still leaves exactly the same number of options that existed before, right? Goodman's argument for the set of all performances- that seems to work. Kivy's argument for extreme Platonism doesn't quite work, but a slight varient does: instead of music being an eternal abstract object instanced in the real world, it is the particular way in which sounds are organized such that a work sounds like it should (such that it sounds of its particular music-type); and really, that sounds awfully close to exactly what Kivy was arguing for. The entire argument about tracks and works and scores, and so on, is exactly the same argument here, I think, except that instead of instancing an eternal object, it's instancing a earthly non-object.
And so we get to the great message: that music is not an object. And therefore that it has no ontology. So, therefore, it is not the sort of thing that Cameron studies. Well, that's interesting. But it feels like the sort of paper in which the author is saying "I have thus conclusively proven that music is not in my particular field of study, and I thus leave it for others to determine its exact nature". Because in the end, why does it matter if music exists or not, just so long as music exists?