Art works, ontology of

DOI: 10.4324/9780415249126-M012-1
Version: v1,  Published online: 1998
Retrieved May 22, 2024, from

3. Works and their histories

What if our performer, Jones, was following the score of an existing work – the Hammerklavier sonata, for instance – but was doing so as a result of having hit upon that score himself by an act of composition undertaken in ignorance of Beethoven’s previous efforts? Would he be performing Beethoven’s Hammerklavier sonata? Structuralists say yes, assuming Jones was playing the instrument specified by Beethoven. Some writers have argued that the correct answer is that he is not playing Beethoven’s work, and that consequently the structuralist approach elaborated above is wrong; he is playing a distinct work composed by Jones himself. The argument for this is another version of the distinct properties argument of §1. For the well-informed listener, much of what is interesting and valuable in a musical work derives from the work’s art-historical features. Works are variously describable as stunningly original, fresh, deliberately anachronistic, shamelessly plundered from better composers. We notice and enjoy or deplore their quotation from and commentary on other works. If such features are features of the works themselves, we cannot say that Beethoven’s and Jones’s works are identical just because they are correctly performed on the same instruments in accordance with the same score. Jones’s twentieth-century work may reflect the influence of Brahms, express outrage at the practice of atonality and consciously submit itself to the discipline of an earlier age; Beethoven’s does none of these things. So these works have distinct properties and cannot be the same work. Structuralism’s identity conditions are ahistorical and fail to locate this vital divergence on historical properties, so they falsely identify distinct works.

Three responses to contextualism have been voiced. The first agrees that works do have the kinds of historically determined features just mentioned, but says that their having them is an objection to the structuralist account of work-identity only as long as we think of these historical features as essential features of works. This response is best clarified in terms of possible worlds. The contextualist is interpreted as saying that in the actual world Beethoven’s work, B, has a certain historically determined feature, F, while in some merely possible world Jones composes an identically scored work, J, which lacks F. But it is agreed that objects have properties in one world that they lack in another; such properties are accidental rather than essential. If we think that historical properties of the kind just described are accidental, we can therefore continue to assert the identity of B and J. This response misunderstands the contextualist’s position. The world the contextualist imagines as a counterexample to structuralism is one in which Beethoven’s composition of B is just as it is in the actual world and in which Jones also composes J. If it is agreed that in this (merely possible) world, B possesses F and J does not, then certainly B and J are distinct, since an object cannot possess and lack a property in a single world. Contextualism does not depend on any controversial assumption about what properties of works are essential.

The second response claims that historical features such as those described above are not properties of works at all. Instead they are properties of acts of composition. So there is no barrier to identifying B and J, and we may say that there is in this case one work and two acts of composition. The problem for someone who advocates this response is to find a principled way of distinguishing properties of the work from properties of the act of composition. An appeal to common speech or practice will not do, since we commonly speak of works themselves as original or conventional. We might say that, in the musical case, the properties of the work itself are just those that determine how it sounds; the historical properties just mentioned do not do this, since it is agreed that B and J sound the same (in the sense that one could not tell, by attending to sound alone, that something was a performance of B rather than of J). This would have the uncomfortable consequence of driving us back to the original structuralist position, according to which the correctness of a performance of a work depends only on how it sounds and not on the choice of instruments for performing it.

The third response to contextualism says that the historical features described above are features of the work, but insists that they are incompletely specified by expressions like ‘is original’. Rather, they are all properties that need to be relativized to contexts of composition, much as velocity needs to be relativized to a frame of reference. On this interpretation, what looks like the paradox of saying ‘B is F, while J, identical with B, is not’ emerges as the consistent avowal that B is F-in-the-Beethoven-context while J is not-F-in-the-Jones-context. This position faces the same difficulty as the previous one: it must be supported by a principled distinction between relativized and unrelativized properties if it is not to be judged ad hoc. If such a distinction can be made, a version of structuralism may be defensible.

Citing this article:
Currie, Gregory. Works and their histories. Art works, ontology of, 1998, doi:10.4324/9780415249126-M012-1. Routledge Encyclopedia of Philosophy, Taylor and Francis,
Copyright © 1998-2024 Routledge.

Related Searches


Related Articles