Jocelyn Wang replied to me:

>SInce the composers didn't say, "I'm putting this repeat in to be
>conventional" or not, you have no basis on which to speculate for any
>given repeat.

Of course I have!  If, say, a composer includes an exposition repeat in all
his sonata-form first movements, then there is a strong likelihood that for
him that repeat was a routine matter, rather than a result of consideration
in each particular case.  I may not be absolutely certain, but all other
things being equal it would be a reasonable inference.

>>What I am suggesting is that some repeats appear to be more important
>>to the compositional, structural integrity of the work than others.
>>Now as Stirling Newberry pointed out, that is no hard and fast thing.
>
>No, but what is a hard and fast thing is what the composer wrote, and he is
>the one who decides what structure his piece should have.  Altering that is
>vandalizing art.

Of course the composer decides what structure his piece should have.  My
question was and is:  how important a role does the *repeat* play within
that structure? Let's go over this one more time:  you seem to believe
that because the repeat sign is on its face 'unambiguous' it invariably
follows that it is a vital component of the structure.  But things aren't
that simple.  As I freely admitted, very often it *will* be an important
component.  If the repeat is only used in some instances and not others,
then of course there is a high likelihood that that it will be significant
in the cases where it appears.  Otherwise, why leave it out in the other
cases? And even if it is used routinely, the musical material may in some
or all cases be 'built around' the repeat, so that ignoring the repeat
would cause a musical loss of some kind:  it would upset the work's formal
proportions, say, or cause the listener to miss some special compositional
effects that exploit the presence of the repeat.  But is it not conceivable
that there will be cases where the exposition repeat continues to be used
as a convention, but with no clear ramifications for the musical substance?
Certain sonatas after Beethoven come to mind; I chose Schubert's sonata
D960 in my last posting, to my mind D959 would also fit the bill.  Why?
(First things first:  by my reckoning as outlined above, Schubert's 1st
movt exposition repeats *are* conventional, at least in his piano sonatas,
where as far as I can tell they appear in every single one.) Some
suggestions:  looking at these opening movements in isolation, I can't
see how the impact or interest of the development would be heightened by
hearing the exposition twice; the exposition and recapitulation are almost
identical, so neither does the exposition repeat highlight any new paths
in the recapitulation, nor is its added weight required to 'balance' any
significant alterations in the recapitulation; nor is there any long coda
which would also require that balance.  Turning to the sonatas as a whole,
both are so sprawling and essentially epic that I would find it very hard
to tell whether avoiding the exposition repeat in the 1st movt.  would be
detrimental to their overall proportions.  Now you might counter that those
are 'classical', architectural criteria, and in these works Schubert is
no longer writing in a 'classical' idiom.  I would agree, but then, what
justifies the repeat (other than the fact that Schubert wrote it!)? Can you
suggest some musical or psychological necessity for it within the work's
overall design? If you or someone else is able to, then I would of course
agree that it should be played.

Can you not see that whether ignoring the repeat here results in 'artistic
vandalism' is very much an open question? It's certainly not an instant
given.  Note that I'm not saying one *shouldn't* play the repeat here.  If
a performer can do it convincingly within a performance that has a whole
lot of other merits, I won't mind.  But I would not feel cheated if it were
left out.  You must agree that in this case at least, calling an otherwise
insightful performance inadequate merely because of the absence of the
exposition repeat - however uncompromising the line and two dots may look
on the page!  - is a bad case of mistaken priorities.

>>We have to guess as best we can given our knowledge.  In the Romantic
>>era, when people took a narrative, psychological view of musical
>>structures, repeats seemed unnecessary and unnatural.
>
>What "people?" The only "people" whose count in this area are the
>composers.

It probably wasn't very sensible of me to include this point.  I'm not
quite sure of its relevance now myself.  Probably I was relating it to
Schubert.  The point is that repeats fit within certain musical aesthetics
better than in others, and it makes little sense to include them if you're
not able to integrate them into your overall approach.  Stirling referred
to interpreters who see music in 'romantic', narrative, psychological terms
but nevertheless include the repeats because that is the thing to do.  The
result is not likely to be all that convincing.  My submission was that
there will also be composers who include repeats as a matter of course even
when the substance of the work neither demands nor supports those repeats.
You will say 'how can we know?' Of course we can't.  But we can make
intelligent inferences in many cases.  We may well turn out to be wrong
in a given instance.  But at least we will have given some thought to the
matter, and tried to make sense of the musical substance (isn't *that*
what really counts here? The composer is dead, it's his composition
that justifies his continued relevance) rather than blindly following
instructions.  And the great thing about a musical performance is that it
is never final and definite:  we can always change our minds and do things
differently later on.


>>What happens, for instance, in the case of composers like Schubert who
>>continues to routinely include repeats in the first movements of his late
>>piano sonatas, even though they no longer Clearly follow the classical
>>aesthetic that gave life to those repeats? ...
>
>It is for Schubert, and only Schubert, to decide.  I submit that, given
>Schubert's genius, if the repeats seem tedious, then it is probably a
>shortcoming in the listener's attention span or the performance of the
>piece.

Maybe - but can you convince me that this is so in the particular works
I mentioned? Again:  what specifically is lost if we leave out those
particular repeats?

>But don't remove a column from Schubert's architecture just because
>there is another like it in the design, because it sometimes takes more
>than one column to support the structure.

Sometimes!  Not invariably.  I just can't think of these late works of his
in those architectonic terms.  They seem to me epic, sprawling, lyrical
(sorry about the contradiction between 'epic' and 'lyrical'!  but I hope
you know what I mean) and in that context, the repeat just seems to add
length.  I'd be up for arguments that the repeat adds inwardness, epic
breadth, or hypnotic 'heavenly length' etc, if it were not for the fact
that Schubert uses the same device in all his sonatas' first movements,
however divergent their style, whether they be more classical-Beethovenian
or more 'Romantic' like these late ones (note I don't know about other
works - symphonies, chamber music - if exposition repeats aren't ubiquitous
there please tell me, that would require rethinking on my part - but
Schubert was just meant as an example after all.).

Felix Delbruck
[log in to unmask]