Repeating music

[Phil Leeds]

I was at a concert the other night, the orchestra played B's 5th symphony, it never struck me before, but quite a bit of it is repeated.
Why is this done one wonders, should the repeat be played slighty differently?

[Bonn1827]

I was wondering the same thing myself when I was listening to my Scottish Chamber Orchestra/Charles Mackerras performance of Mozart's "Prague", Symphony No. 38. It is full of repeats, especially the first movement - which is actually prolonged through repeats (da capo). This is an interesting subject which has been debated for some time, as some performers/conductors omit repeats today. I've always thought the use of repeats was structural, that's to say part of the re-iteration of themes, especially after new ones are introduced or worked out in the development section. Cynics might say they are there to make the work longer. I've heard both arguments. I cannot believe Mozart was lacking invention and had to rely on repeats, so would suggest it is structural in his case. Also, the particular symphony I quote has a dance-like rhythm in the repeated sections which gives it a happy, joyful, celebratory tone after a somewhat dark beginning. So, a move towards the positive through repeats in this instance?

[Roehre]

At a time in which reproduction of music of any kind was non-existent, and therefore most music performed in a concerthall was unknown or hardly known to the listener (and IF the concert goer knew the piece, it was most likely to be in the disguise of an piano arrangement for 2 or 4 hands, played by the listener him or herself, therefore missing the orchestral colouring of the score).
Bruno Walter e.g. took the second time an exposition slight differently in tempo, and sometimes even in articulation, as it were to show another aspect of the same music to the public.

The repeats were in most cases (the opening movements almost without exceptions) a gesture to the listener to grasp the themes before entering the development section.
This opens the dicussion whether a repeat without any added bars should be reproduced in a fixed performance as a recording by definition is, as the listener can discover the music easily be simply playing the recording (78, LP, MC, CD, MP3, or whatever) as many times as he likes.

[Hofrat]

Also, the repeats were for the listeners in Beethoven's times. In those times, the audiences were more musically acute, being able to hear the key changes. The repeat just reinforced the key changes.

Also, the repeats can be controversal. Beethoven went to his grave without deciding if there is repeat of the scherzo and trio in the 3rd movement of the 5th symphony. And the repeat in the 4th movement of his 5th symphony comes off very weak since he can not "ramp up" the orchestra as he did with in most ingeneous ways (e.g. the lead from the scherzo to the 4th movement and the retreat to the 3rd movement theme at the end of the 4th movement's developement). Many conductors do not play the 4th movement repeat and others are beginning to play the 3rd movement repeat.

Another controversal repeat is the one at the end of the 9th's symphony scherzo. In the MS, Beethoven does not indicate Del Capo (go to the begining), rather he indicates Del Segno (Del Segno Go to the sign). BUT he forgot to write the sign!!! In modern day scores, Beethoven instructions are copied word for word. Everyone took the easy way out and played it as DC (Del Capo). It is very logical that Beethoven did not want to repeat those jumpy tutti blasts that open the beginning of the scherzo, rather he wanted to skip the introduction and repeat from a point after this introduction.

[Michael]

I have mixed feelings about repeats in general, but I think a lot of B's earlier work, such as the Opus 1 trios, have the typical young composer's "fault" of maybe too much material. Dropping the exposition repeats in those works might speed them up a bit (except the C minor).
In the case of the Eroica, Beethoven originally intended the exposition repeat to be played in the first movement, but decided against it because he felt the symphony would be too long. But if the repeat is omitted, we lose a thrilling sforzando which us brings back to the opening.
If the repeat is omitted in the first movement of the Fifth, the proportions are wrong (the Golden Mean and all that) but I think (and "who are you?" I hear Beethoven asking) that the scherzo repeat should have been kept. Many conductors who omit it, keep including the exposition repeat in the last movement. Result: too much C major, too easy victory.
The first movement of the Ninth is designed (like the first Razumovsky) to give the impression that a repeat of the exposition has started but is then pushed off course into the development section.
I have several recordings of the Appassionata where, in the last movement, the whole development and recapitulation are repeated, and, either the Tokyos or the Lindsays (I forget which), do the same in the first movement of the second Razumovsky.
Wilhelm Kempff can be quite arbitrary about repeats but his most astonishing first movement omission is in the "Hammerklavier" - in his stereo set, anyway.
And to me it doesn't sound right but he was a slightly better pianist than I am, so I'll have to bow to his judgement.

[Hofrat]

Then there is the "double reprise" format in first movements (and other movements): a repeat after the exposition and a repeat after the developement-recapitulation sections. Beethoven used them in many early works (like the opus 2 sonatas), and in later works as well (like the "Ghost" trio and the "Hammerklaver" sonata). Some of these works would be quite short without both rests, as in the F-minor sonata opus 2 (in both the 1st and 4th movements). On the other hand, some works would be quite a handful for performers with both rests, as in the "Hammerklaver". But then again, both rests served the early 18th century music goers well, with the 1st rest giving them the opportunity to hear the contrasting themes and the 2nd rest giving the opportunity to hear how these themes were developed.

Often I wonder if rests are not dropped do to pressure exerted by the recording companies trying to sqeeze another work on the CD.

[Michael]

Often I wonder if rests are not dropped do to pressure exerted by the recording companies trying to sqeeze another work on the CD.

The original vinyl release of Szell's Beethoven Five omitted the exposition repeat but it was restored on the CD version. I think I mentioned before that on a CD recording of the second Razumovsky by the Guarneri (I think), the exposition repeat was faked. A background sound - probably a train - can be heard in exactly the same position when the exposition is repeated. Obviously, for some reason, they just used the same recorded portion again instead of playing it twice.
Hofrat, the F minor sonata (Opus 2, No. 1) is one of those works that I think would be better without repeats - especially the development and recap sections. With them I think the themes outstay their welcome. The Hammerklavier is another story, although it would make for a very long performance indeed.

[Chris]

Personally, I am not convinced of the arguments for omitting repeats because they were only there because people needed to hear it again in a performance. They may not have had recordings, but they had encores, arrangements, transcriptions, and reductions. They seem to me to be essential parts of the structure. Omitting repeats of the exposition in sonata form always seems wrong to me, because the primary material needs to be strongly (doubly) established before it's development can have its maximum effect. And compare it with the Classical-era concerto - a double exposition, and not just a repeat, but once through without the soloist and once with. If the purpose of the repeat was simply because the audience needed to hear it again, there would be no point in writing two versions of the exposition; rather, the double exposition is an essential part of the structure. And even though the repeats in non-concerto movements are exact repeats, I still think they serve a similar structural purpose by strongly establishing the primary material and making the proportions right.

And even if the "we have recordings now, so we don't need repeats" crowd is right, I object to the whole idea that in making a recording, one should assume repeat listenings in order for the work to have maximum effect. A recording should be complete in itself, on one listening. A lot of times you CAN'T just repeat it as many times as you like - if you hear it in a broadcast, for example, it may be the first time you are exposed to the work, and you won't have the chance to immediately hear it again. The piece should be to maximum effect in this case too.

And I also don't buy the "you have the CD, so you can listen to it again" argument, aside from the fact that that may not be true, as I explained above, but because the same argument could be made for the reverse - "you have the CD, so you can skip the repeats if you want". Yes, we have CDs. We can listen to the music any way we want, one way or the other.

But really, none of this matters. Because the repeats are there, part of the original scores, and they should be taken always, including on recordings. Trying to figure out a composer's intentions and using that to determine whether or not you take the repeats necessarily involves some degree of speculation. Why speculate? They are there, so do them. There is no reason NOT to do them, except laziness and cheapness.

[Peter]

Personally, I am not convinced of the arguments for omitting repeats because they were only there because people needed to hear it again in a performance. They may not have had recordings, but they had encores, arrangements, transcriptions, and reductions. They seem to me to be essential parts of the structure. Omitting repeats of the exposition in sonata form always seems wrong to me, because the primary material needs to be strongly (doubly) established before it's development can have its maximum effect. And compare it with the Classical-era concerto - a double exposition, and not just a repeat, but once through without the soloist and once with. If the purpose of the repeat was simply because the audience needed to hear it again, there would be no point in writing two versions of the exposition; rather, the double exposition is an essential part of the structure. And even though the repeats in non-concerto movements are exact repeats, I still think they serve a similar structural purpose by strongly establishing the primary material and making the proportions right.

And even if the "we have recordings now, so we don't need repeats" crowd is right, I object to the whole idea that in making a recording, one should assume repeat listenings in order for the work to have maximum effect. A recording should be complete in itself, on one listening. A lot of times you CAN'T just repeat it as many times as you like - if you hear it in a broadcast, for example, it may be the first time you are exposed to the work, and you won't have the chance to immediately hear it again. The piece should be to maximum effect in this case too.

And I also don't buy the "you have the CD, so you can listen to it again" argument, aside from the fact that that may not be true, as I explained above, but because the same argument could be made for the reverse - "you have the CD, so you can skip the repeats if you want". Yes, we have CDs. We can listen to the music any way we want, one way or the other.

But really, none of this matters. Because the repeats are there, part of the original scores, and they should be taken always, including on recordings. Trying to figure out a composer's intentions and using that to determine whether or not you take the repeats necessarily involves some degree of speculation. Why speculate? They are there, so do them. There is no reason NOT to do them, except laziness and cheapness.

Spot on Chris! Just because it is a matter of a few dots in the score why on earth should people imagine they should take no notice? These same musicians would shriek in horror if a staccato was ignored! There are examples where Beethoven marked no repeat such as the Razumovsky quartet no.1 making it quite clear the difference between the two.

[Phil Leeds]

Thanks for your replies - very interesting. Have other conductors followed Bruno Walter and changed the repeat slightly I wonder