Repeats?

Oh gawd - you're infected, Chris! LOL

That's what I did with Firkusny - so a full-on Bazza. :D

To the original question, depends, depends. In Mozart, e.g. the quintet first movement expositions, they seem to me an essential component of the overall structure, throwing what happens subsequently into high relief. In Schubert 9 they turn an already-prolix work into a bore. Surely in many cases they were designed to familiarize audiences with the basic material and their composers themselves would omit them in recordings. With exceptions. It always seems to me that the first movement exposition repeat in Mahler 6 serves an essential purpose. The sense of being on a treadmill, of drudgery and routine is entrenched by the repeat and lends the development a feeling of catharsis, almost iconclasm.

If a commercial recording does it, I suppose it is. But can anyone tell the difference? If a conductor/performer is going to pay it exactly the same, just patch in the first take. Recordings are jigsaw puzzles anyway so not really something to worry about.

Does anyone think a totally different interpretation of a repeat would affect the continuity of the piece? I have seen reviewers moaning because a theme was recapitulated faster so I doubt if it would go down too well.

Naupilus - your metaphysical assertion is correct. And hearing an exposition repeat is just like rereading a chapter to re-familarise yourself with the characters.

I need to hear Sir Adrian's Brahms. I have heard good things about it.

@Bliss, Sir Adrian's remarks sound about right to me: don't be totalitarian about it, have it depend on the occasion, and - if you choose to do the repeats, make sure you do it in a way that it attributes something to the composition apart from adding a couple of minutes. It's important that he says repeats should be observed as a rule - but observing rules aren't always what makes great music.

It's also important to note that Boult says repeats that contribute to the piece's structure and unity should never be skipped. Of course that's a rather subjective view - which pieces and composers are subject to this "rule" and which are not? Are Mozart's repeats really "random" - and Brahms' not?

Personally I think that the more complex the exposition is, the more the need is there for a repeat. Omitting the repeat in a complex sonata form will destroy the sense of formal unity and make the piece sound like a random succession of themes. For instance, in many cases, when the exposition is fairly complex and has a lot of development of the thematic material already, going directly into the development section just makes that sound like it's a progression of the exposition, which can be very confusing to the listener.

I think it's interesting that Schubert is mentioned a couple of times in this thread - a composer who kind of falls between the classical "repeats because sonata forms have repeats, period" and romantic "repeats because the composer choses them" eras. In the case of D960, I must say Brendel's statement shocked me a bit. I always regarded him as a great, very intelligent and theoretically well-informed Schubert player but dismissing the repeat in the first movement of D960 means cutting out a couple of original music (very effective and inspired music, too!), and to me that's close to a mortal sin.

I guess Brendel does it to give the sonata as a whole more classical proportions, instead of a HUGE first movement, followed by 3 shorter ones in most other recordings that observe the repeat. Another great Schubertian like Richter doesn't appear to care - and he even enlarges the disproportion by not only playing the repeats but also chosing a very relaxed tempo for the first movement. But does it really matter? Richter's Schubert is the most magical, mystical and "deep" of them all - and he never ever bores me, even if his first movements make it feel like time itself stopped.

Take his recording of the Fantasy-sonata in G, with that first movement that clocks in little under half an hour. Mahlerian expansion, but both tempo and the observance of the repeats feel absolutely right. I think Brendel made a huge mistake by wanting to streamline his Schubert and give these revolutionary pieces more classical proportions.

@Eyeresist, of course the copypaste option didn't originate in our modern Finale/Sibelius era (I expect many of the written-out repeats in older printed music just being there because the composer's payment was based on the number of pages...) but yes, it's become very easy to make a piece look "bigger" nowadays when you print your own music.

As for my own music, as a composer I never write repeats in my pieces that have a classical sonata form - but I don't copypaste either. In the second exposition I usually cut out some material to make it shorter and make it sound differently by, for instance, transposing the second theme to another key. One could argue that it isn't a true sonata form anymore that way, but some kind of rondo-sonata hybrid, but I'll let the theorists fight about that ;)

Joke went over my head, Bazza... :) Care to explain?

50m, regarding your penultimate post in this thread, I feel compelled to commend you on such a thorough and well-written post with which I cannot agree more. I guess I'm one of the few who finds too Richter's Schubert mesmerizing and magical at the same time. I was not surprised with Brendel's view. He had some strange ones on various composers, but, fortunately, he performs them a bit better.

Parla