Philosophy of the Performing Arts

the nature of the performable work 25
as we have seen, it is part of the very concept of a performable work that it
can be properly or fully appreciated through, and only through, performances.
A work of art whose proper appreciation was not constrained in this way
would thereby fail to be a performable work. This suggests that the relationship
between performable works and the performances through which they are
appreciated is in some ways a very intimate one. But this should not lead us
to think that such works can somehow be identified with their performances.
On the one hand, it seems obvious that a performable work cannot be identi-
fied with a particular performance, since work and performance may differ in
their properties. Sibelius’s Second Symphony is a stirring work that may have
limply disappointing performances, for example. And if we try to identify the
work with an ideal performance that shares such artistic values with the work,
the work will still differ from the performance in that the former, but not the
latter, might be performed in a limply disappointing manner.
This rests upon a more fundamental difference between performable works
and individual performances. The former, as we have seen, are at least in princi-
ple repeatable in the sense that they can be performed on more than one occasion.
But to say that a performable work can be multiply performed is not to say that
it is in fact multiply performed – indeed, it is not even to say that it is performed
at all. Imagine if we were to discover a score by Sibelius for what was announced
on attached papers to be his Eighth Symphony, something written late in his life
which, as we learn from the papers, would have been performed had not an
untimely demise prevented him from making his intentions known to anyone.
We have surely discovered a hitherto unknown and unperformed performable
work, a work that might never have been discovered. It would sound very strange
to say that the work came into existence only when it was performed, or that,
had we not discovered the score, there would have been no work to discover.
Furthermore, it seems equally clear that a performable work that has been mul-
tiply performed could have had fewer or more performances than it actually has.
For example, the performance attended by Berthold and Magda might have been
cancelled because a misfortune befell the conductor without thereby affecting
the identity of Sibelius’s work. Thus we cannot identify a musical work with the
collection of its actual performances, since it may have none, and may have had a
different set of performances from the ones that actually take place.^2
Thus, if there is an intimate link between being a performable work and
being open to performance, the work must also be somehow independent
of its actual performances. The performances themselves seem unproblem-
atic – they are just events of a particular kind, no different in this respect
from other kinds of events that fit easily into our ordinary conception of
what there is in the world. But the performable work begins to look much
more mysterious. Again we may ask, what sort of entity can have the kinds
of properties just described? What sort of entity can be somehow tied to