ROBERT STECKER Methodological Questions about the Ontology of Music There are probably more views about musical on- tology in play today than ever before. If we go back far enough, there were few views about this topic because few were interested in it. More re- cently the main debate seems to be about what type or kind a work of music is: an eternal struc- tural type, an indicated type, a norm-kind, and so on. 1 But most recently, there has been an explo- sion of interest in and views about musical ontol- ogy. 2 There are still strong proponents of the views just mentioned, but now many others are on offer, some of which will be discussed below. We are fur- ther from a consensus now than we were a decade ago. This raises the question of how we should adjudicate different views about ontology in gen- eral and musical ontology in particular. One might think that the answer is obvious. An ontological claim must be backed up by an argument, and we know how to evaluate deductive arguments: we see if they are valid and if their premises are true. However, quite a bit of the argumentation for ontological theses is not strictly deductive. It often resembles an inference to the best expla- nation. More specifically, much argumentation in this arena looks like this: desiderata are presented; various views are rejected for failing to meet the desiderata; one view is then defended as meeting them best. So we need some way to evaluate claims that such and such is a desideratum, D, for an ad- equate ontology of music, as well as “best-meets- D” claims that are put forward to defend a favored view. In this article, I argue, first, that desiderata are better derived from pragmatic rather than seman- tic or modal considerations, second, that there will always be more than one ontology that satisfies the desiderata, and finally, third, that appeals to desiderata do not exhaust the grounds for choos- ing among these ontologies. However, even the pragmatic considerations are highly controversial. While I defend them against some alternatives, I leave it open here whether they should ultimately be accepted. i. modal intuitions One method of argument appeals to intuitions about modality, that is, intuitions about necessity and possibility. If we have clear intuitions about properties a musical work could have or must have, the intuitions would clearly constrain our theory about the kind of object a musical work might be. Further, if we have clear intuitions that a musical work could have been different in some specific way (for example, composed in a different era or by a different person), then they would pro- vide an even greater constraint. However, there are few occasions to think about works of art in counterfactual situations, since for the most part our intercourse with artworks involves noticing as rich a set of their properties as we can to properly appreciate them. Where works are physical ob- jects or (not to beg any questions) are closely as- sociated with a single physical object, some coun- terfactual thoughts about them will be inevitable. This is because physical objects undergo visible changes, and so we will ask how a work had been at an earlier period, how it might change in the future, or whether some recent change, such as a restoration, captures some phase of the work’s existence, or instead, destroys the work. Musical works are typically abstract. They cannot change in the way physical objects (or physically embod- ied objects, if one thinks this a distinct category) do, and it is not clear that they ever undergo a The Journal of Aesthetics and Art Criticism 67:4 Fall 2009 c  2009 The American Society for Aesthetics