In the lively debate on the issue of repeats there remain one or two points
which still have not been made.
Although the tendency for composers to write in ever more and more
detailed instructions to performers began in the late nineteenth century,
the practice of textual literalism is a peculiarly twentieth century one.
Prior to that I believe that the separation between composition and
performance had not been so complete, which is to say that it was expected
that a performance have an improvisational element incorporated into it.
Indeed even Beethoven had been famous as an improviser at the keyboard.
The contrast to the concept put forward by Stravinsky in his 'Poetics of
Music' during the twentieth century in which he argues that performers
should cease 'interpreting' the score and instead becomes mere mechanical
'executants' could not be more marked.
There is a grave problem with applying this sort of musical literalism
to music of the more distant past in that this sort of dogma had been
previously quite unheard of. One therefore risks playing music of former
centuries as though it were Stravinsky, much as Mahler performed Bach as
though it he were a late nineteenth century composer. Indeed this sort
of puritanical musical literalism was so foreign to the likes of Bach and
Mozart that they freely improvised their own additions to the score in the
form of decorations as they played.
In any case why not take this literalism to even greater extremes? Why not
insist that elements such as vibrato, or rubato should not be permitted
unless they are so notated in the score. One might then argue that to do
otherwise is tantamount to playing Bach on a synthesiser.
Clearly then everyone must permit some degree of interpretative freedom,
the question is how much is too much. Now, all one simply has to do is
listen to older mono recordings to realise that in the past greater musical
freedom was the norm unlike today, and this undoubtedly held true for many
of the composers of the past. One commonly hears purists react with horror
when someone misses out an exposition repeat from a Brahms symphony - never
mind that Brahms himself did much the same when he conducted his own music!
I will argue that the way forward is to ask to what degree of, and what
sort of stylistic freedom each particular composer expected for their own
composition. For Stravinsky this meant practically none. Having said that
he himself fails not only to follow his own metronome markings for the
'Rite of Spring' but mocks a tempo taken by Karajan as a 'tempo di couchee
couchee' - and he was just following the printed metronome marking! So
even here there can be no mindless formulae such as 'stop interpreting and
just play the notes of the score'.