Piano World Home Page Forums Our Most Popular Forums Pianist Corner famous pianists on taking (or not taking) the repeats

This seems to be a very controversial question: Do we take the repeats or no?

Does anybody know if Brendel, Gould, Horowitz, or others wrote about / discussed in interviews the subject of repeats? I'm having trouble finding anything online, except for an excerpt from an essay by Angela Hewitt, on the Goldberg Variations:

http://www.thegoldbergvariations.com/index.php?page=hewitt

Schiff in his Beethoven sonata lectures said, 'you have to take Beethoven's repeat signs seriously.'

Sometimes repetitions fulfil the temporal proportion aims of a composer. Also, they are important for the memory to perceive the formal structure of the music. But the more we know a music (and we know a lot of past music) the less we need repetitions to comprehend it. So i think that a contemporary interpreter should be attentive to our modern impatience and try to limit the obviousness of some repetitions.

Yes, absolutely. And, at least in the sonata, it's also the last breathing before taking the plunge into themes interplay. However, such function does not exist in, say, the dance pieces from Bach's suites and I really don't care for the repeats in this case.

What if among his audience there are those who never listened to the work before?

Horowitz said he wanted to avoid monotony... That meant either playing the repeats differently, or ignoring them.

By the way, does anybody know where all those Busoni references lead to? Is there a book somwhere that collects Busoni's writings on music?

Yes; "The Essence of Music, and other papers" published by Dover (in English translation by Rosamond Ley).

But, upon skimming the table of contents and a few of the essays, I didn't see any titles that would suggest a discussion of repeats in the Goldberg Variations.

Wasn't it Dvorak who wrote "once and for all, without the repeat" into a copy of the score for his 8th Symphony?

Thanks, PJ, though the book seems to be out of print.

I suppose Busoni talks about the repeats in one of his letters, probably referenced in some Busoni biography...

I can remember Brendel somewhere specifically justifying his not taking the exposition repeat in Schubert D.960, but I don't have a reference. Andras Schiff, OTOH, insists that it should be taken , especially because of the interesting material in the first time bar (I think that's in the notes that he wrote for his cycle on Decca.)

I can think of the practice of a lot of pianists (eg. Arrau always took EVERY repeat by ANYONE and his exposition repeats are like carbon copies) but I can think of few of their comments about it off hand.

Sure. May be that’s because music originated in dances have “non progressive” forms, but “accumulative” ones. We need to remember the functional origins of each music, I mean it’s “social ur-ethos”.

Quite a point. I think there’s a cultural learning curve and actual audiences can assimilate older forms easily. We are used to quick comprehension of simultaneous and complex facts even though perceptive deepness and profundity is not necessarily included.

An interesting issue about repetition is the strong denial of it in some contemporary music (integral serialism) and the further acceptance of it in other music (minimalism). I have to recognize that hearing some non repetitive music like Anton Webern microforms, is difficult, to say the least.

I'd say that trying to listen to any serialist work, repeatedly or not, won't make you understand it any better.

I've just remembered an interview Schiff gave when his recent ECM recording of Goldbergs came out. He said that his approach to Bach repeats had changed a lot since his Decca recording when he used to ornament the repeats heavily. He now uses fewer ornaments and employs other kinds of inflection instead. (Incidentally, I have heard him ornament the exposition repeats in both Mozart and early Beethoven.)

Anyway, here is Rosalyn Tureck in the notes to her Partita recording:

"Repeat marks for each section are always present and they form as much a part of the musical form of the dance movements as does the harmonic scheme. In modern time "repeat" means simple, duplicated repetition. Bach's musical style grows out of a time when variation of ornament or embelllishment formed the style, and even form, of a piece. Repetition in that era never meant duplication; it meant embellishment or variation of material already presented. Proof of this can most quickly be seen in Bach's use of the term "double" which serves as titles to his very highly embellished repetitions of Sarabandes, but to us, "double" signifies exact duplication ...

Performance in repeats is a great art. The test and stamp of the level of performance is in the varied treatment - a subtle amalgamation of historical knowledge, musical taste, and understanding of instrumental sonorities ... I think it will be clear to my listeners that in my performances all the repeated sections are treated with variation of nuance and meaning in the repeats which are, however, always related to the whole. The core of Bach's style in the performance of works such as the Partitas, "Goldberg" Variations etc., lies in the informed and imaginative treatment of the repeated sections."

A bit of dogma and self-promotion in there (neither of which Tureck was averse to), but interesting nonetheless.

great post, yok.

I'm learning Mozart's Sonata K457 1st mov., in c minor, I've got two CDs with two pianists played this piece, they played in a normal time, about 6 minutes, and repeat the first part.

But one day, I've seen a DVD, performed by Alfred Brendel, he played again and again, firstly, I thought he's getting older, maybe forgot the end part, so repeat the middle part, but then I checked the notes, I found he's right, not just has to repeat the first part, the middle section also has to repeat.

But do you think it'll be too long for a piece? I remember one of Chopin's piece, the end section repeated about 6 times! What's for?

Yes -- I found his essay, "Schubert's Late Sonatas." He set down a whole list of 8 questions to ask when deciding if a particular repeat should be taken, make specific comments on 3 Schubert Sonatas, and also quoted Dvorak and Brahms (actually, he quoted Fischer quoting Brahms). Too much for me to type -- 4 pages. It's in several of his books (Alfred Brendel on Music: Collected Essays and Music Sounded Out)

Richter's comments on repeats, particularly with reference to the Goldberg Variations are worth quoting:

Talking with Monsaingeon about his film of the Goldberg Variations with Glenn Gould:

Taken from Bruno Monsaingeon's book "Sviatoslav Richter: Notebooks and Conversations".

I haven't read this whole thread but I think it was Edwin Fischer who said - "if it went okay the first time, I move on..."

- or words to that effect -

Is it correct that Rachmaninoff omitted a repetition in the first movement in Chopin's 2nd sonata 1st mov., and Rubinstein (who apparently worshipped him) quickly made the same, because it 'made more sense'??

How about Schubert's repetitions?

I have always thought (probably wrongly) that the repeats are usually merely a means of the composer or the publisher avoiding writing out the score in full. That is, just an abbreviation to save effort and paper.

If the score was written out in full, then presumably pianists would not seek to edit out segments in this way.

I find omitting the repeats a bit strange. Perhaps lazy of the pianist who may assume that the audience is equally lazy.

But I have no scholarly background on this so I may be misguided?

Adrian

I have a facsimile of Mozart's manuscript for K.333. At the recapitulation of the 1st movement, instead of rewriting the 1st part of the exposition, he just puts an empty measure, and then continues writing where it modulates to stay in the tonic --- assuming, of course, that everyone would know to bring back the opening for that "blank" measure. He was just being lazy.

I think he was being economical, and rational (not necessarily in that order).

By the way, apparently Busoni made an arrangement or edition of Bach's Goldberg Variations, and there's your source for the comment.

I've seen that. It was recently reprinted -I believe- by Music Masters. It's all rather embarrassing, though it certainly harkens back to a different era.

Whether or not I play the repeats depends on the particular piece. If the repeat works musically for me I play it. On the other hand, if it bores me to repeat a section, I skip it.

Rachmaninoff did skip the repeat, at least in his recording. I'm not certain if Rubinstein's reason for skipping the repeat in Chopin's 2nd Sonata was because Rachmaninoff did it, but I concur with their decision. That repeat contains one of the worst modulations Chopin ever wrote--unusual because Chopin was a master of modulation.

Chopin also wrote a repeat into the first movement of his 3rd Sonata, which almost nobody takes.

Specific to Schubert's Sonata D. 960, there are nine bars of music that are omitted if the first movement repeat is skipped, so I favor playing the repeat.

Brendel on Schubert and repeats in 959 and 960. Here's something on the subject.

http://www.nybooks.com/articles/4104

Tomasino

Great stuff, thanks for posting! Brendel's reply is superb.

This is my rule of thumb. The main reason a repeat would be monotonous is if it is exactly the same. This is like trying to emphasise a conversational point by just repeating it, and equally nonsensical. For starters, the repeated section in a piece of music is made different by what came before (the section itself) and MUST have a different feel to it. This can be achieved with the whole gamut of pianistic niceties, different dynamics, ornaments, you name it. Certainly in the baroque and early to mid classical periods it was expected that a repeat would be improvised on, I simply do not understand where this was lost along the way.

Having said all this, I choose to leave out repeats at the ends of movements almost without fail. Imagine, finishing a piece with tonic, dominant, tonic, dominant, a big perfect cadence to the tonic. Then going back and starting a new phrase in the dominant!!!! Sounds wrong to me, no wonder Beethoven didn't really do it. I think as music got less abstract this repeat was rightly thrown out.

Yes, Brendel's reply is indeed superb.

And yet... I'll still take my D960 with the first movement repeat- as long as it's not taken at Richter's glacial tempo. (I wonder what Brendel thought of that?)

Mozart's b-flat sonata k333 comes to mind. I chose not to take that end repeat because it feels excessive.

My favorite comment on the subject was from Brahms, who, after conducting one of his symphonies without the repeat, said that when it was new and nobody had heard it, you had to make the repeat, but now that everyone knows it, you do not have to any more.

On the other hand, I have just been reading a book by another great pianist, Steve Allen, where he mentions that the mood of the audience helped him decide how his comedy performances would go, and how to approach it. In other words, sometimes you have to judge whether the audience wants it or not. Gottschalk said pretty much the same thing in his memoirs.

The other thing to consider is that what modern audiences want is not necessarily what contemporary (to the composer) audiences wanted. One of the most revealing documents to life in the 1830s that I ever saw was an advertising flyer for the Covent Garden Theater for a Wednesday evening in December, 1837, for the double bill that was playing that night, Macbeth and The Revolt of the Harem. Imagine an audience sitting through all of that in one weekday evening these days! But not only did they do that, other plays played on other days of the week. It was the television of their time.

As does the 1st mov't repeat in the Chopin B minor...

IMHO, of course.

I believe Rachmaninov omitted certain variations on the fly in the Corelli set if he felt the audience were slipping. (Better idea: don't programe it in the first place. ) It would be interesting, though, to know which variations Rachmaninov excised.

If the coughing got loud, he would skip the next variation.

I like the Corelli Variations better than most of his other large works. That may be reason enough not to program it, of course.

The ending is practically identical to the ending of the exposition... just in a different key, and with a slightly different melody. So do you also not take the exposition repeat for the same reason?

But the exposition ends with a cadence at the fifth. It's not really the same thing to start a repeat there, when the piece hasn't yet ended, tonally speaking, as it does later (what's the name of that later section?).

http://www.pianoworld.com/ubb/ubb/ultimatebb.php?/topic/2/14621.html

I agree. Especially in regards to classical sonatas, as a listener I find it useful to hear the :exposition: to help familiarize myself with the themes before they're played with in the development.

Repeating the second section I find optional. If I like the development I'd enjoy hearing it repeated, if I like it and it's not repeated I'm left with a sense of wanting more (also good!).

I like the idea of reading the audience - it's for them we play afterall (by the time I'm performing, I already know the piece, I've heard it repeated much more than just twice ).