appreciating performable works in performance 57
directly referring to their tokens.^8 But perception seems to require a more
intimate connection than reference.
There is, however, a deeper worry. As we have seen, the particular appli-
cations that we make of structural, sonic, timbral, and expressive predicates
to musical works are to be understood analogically relative to the application
of the same predicates to their performances. What we are thereby ascrib-
ing to the work is the property of prescribing the relevant structural, sonic,
timbral, and expressive properties for its correct performances. A work’s
performances are certainly audible, and it is in virtue of this fact that we are
able to appreciate the particular audible properties that they possess. But
what of the predication of audibility of the musical work itself, as when
Berthold says that he heard Sibelius’s Second Symphony? How can some-
thing that possesses particular audible properties only analogically possess
the general property of audibility in the same sense (that is, non -analogically)
as its performances? If the doctrine of analogical predication applies to those
particular audible properties of a work’s tokens that they possess in virtue
of being tokens of a given work – such as containing a G sharp in the sev-
enth measure – why does it not also apply to audibility in general? These
considerations acquire additional force if musical works, as types, lack the
structure and temporal parts of their performances. For, it might seem,
something that lacks temporal parts cannot possess the kinds of properties
that any object of audition must possess – duration in time and temporal
boundaries, not to mention particular acoustic or sonic properties. This sug-
gests that, if (contra fictionalism) performable musical works exist and are
any of the things canvassed in the previous chapter, they are only analogically
perceivable. To say that we can hear Sibelius’s Second Symphony is to say
that we can hear its performances. To say that we have seen King Lear is to
say that we have seen at least one of its performances. To this extent at least,
our consideration in the previous chapter of the nature of the performable
work may lead us to revise our understanding of at least some of the things
we naturally say about such works.^9
3 The Goodman Argument
We have seen that different views about the nature of the performable musi-
cal work go together with different views about what counts as a correct
or incorrect performance of a work. This connection is most obvious when
works are taken to be norm-types, since the work itself consists in a set of
prescriptions for correct performance. But norm-types also play a role in
the “contextualized type” and action-theoretic accounts considered in the
previous chapter. And even the continuant theorist maintains that works