Have we had a thread about revisions before? I was prompted to start this by the discussion elsewhere about the Wingfield edition of the Glagolitic Mass. While clearly there's something to be learned from hearing a composer's "first thoughts" in the form of a work which was later revised, this applies in different ways to different composers, doesn't it? Speaking for myself and having done my share of revisions and in some cases total rewritings, I would be horrified if the earlier versions of those things were dredged up and performed rather than the later ones. But then one reason for that is that the later versions aren't exactly familiar repertoire either. Mahler's final thoughts, being generally the result of the practical experience of conducting the works in question, always (to my knowledge) exhibit a further degree of refinement relative to previous versions. On the other hand there are examples of composers whose original versions are almost invariably more interesting (to me) than the later revisions - Hindemith in Das Marienleben or Cardillac, for example, where the later versions are the typical stodge Hindemith served up after he'd systematised his musical technique and published a book about it. And as for Bruckner, I am not going to go there in this first post.

So - revised or original, for which music?

I'd rather have all RVW's symphonic revisions - the London Symphony's original version may have some nice extra music, but it's just too episodic for my taste: the version we're more used to is far tighter. The tiny alteration of the final note of the Andante Moderato of the 4th symphony from a tonic F to an E (the original can be heard on the composer's 78s) seems to me infinitely preferable (though I know Ollie disagrees): it also creates a musical cell that links it to the 6th and 7th symphonies, amongst other works. The extra lines added in the Scherzo of the 6th, too, strike me as an improvement; a reinforcement of the texture. Again, the original version survives in early recordings: the 78s by Stokowski and Boult, though HMV later recorded the revised scherzo, and at least two reissues have included both versions for comparison.

Stravinsky - Symphonies of Wind Instruments  The original 1920 version for me please.

RVW - Symphony No 2  I'm a novice RVWer and would be interested to know how opinion divides on this one. I hope I don't end too many beautiful friendships by saying that my preference is for the later, tighter, version.
[Ah, Ron's beaten me to it. We both like it 'tighter'.]

Is being reincarnated as a conductor a reward for a good life?  I would think most conductors are going to be reincarnated as something with too many legs/scales/self-generating slime (delete as appropriate), so I had assumed coming back as a conductor is part of a downwards slope...

I'd certainly agree about the Council Chamber scene - IMO one of the greatest in the Verdi canon.

One of the issues in opera is that sometimes compoisers made changes to accomodate different singers: such as the alternative version of Susanna's aria in Act IV of Figaro and the Dalla sua Pace/Il mio tesoro question in Don Giovanni, where the former was written to replace the latter so a performance containing both is a solecism, but most listeners wouldn't want to be without either, and both fit quite neatly in the dramatic structure.  And it's difficult to imagine Faust without the arias that Gounod wrote for Santley (although I for one am quite glad do do without the recits Gounod wrote for the Paris Opera).  Opera composers do appear to have been affected more than most by the practicalities of performance - budget limitations, whims of singers, desire of tenors to sing higher and louder etc, changes of language.

Wagner of course made extensive alterations to Tannhauser, but Act I of this work is where my Wagnerism is less than perfect and I could easily do without it altogether; but the small changes he made to the orchestration of Der fliegende Hollander do seem to me to be a real improvement - the original scoring does have overtones of the Teutonic town band at times.

Back to RVW, and the piano concerto, a work I'm still not completely decided on. I still prefer the original version to the later one which has a second piano added to beef-up the texture: both are available on the recent EMI box.

Opera-wise, the 2 act Billy Budd works better for me than the 4 act original, though I prefer Walton's original Troilus and Cressida to its revision, with its downgrading of the soprano part to a mezzo for Janet Baker. There don't seem to have been many revisions to Tippett works, apart from the various methods used to achieve the 'breathing' in the 4th Symphony. As for Stravinsky, revisions are a minefield: some for no other purpose than to re-establish copyright, others in attempt to make the orchestral forces more manageable, both, in the long run, to increase revenue. I love all three (well, two-and-a-half) versions of Svadebka/Les Noces pretty equally, though the two earlier attempts, with their highly idiosyncratic scoring, might just have the edge, especially in the one chord for the strings that the composer was unable to reassign to the four pianos, and omitted completely from the final edition...

Oh. Well, in that case... One result of that experience was a very refreshing sense that a piece of work can be an ongoing, non-finite thing. Double barlines? Ptwah. I did a talk a while back about revisionism (tinners was there), and my sense that 'revisions', as we normally understand them (i.e. as 'improvements', 'rethikings' etc.) are, for me, mostly an irrelevance. I like (prefer) the idea that the next piece one writes is a de facto 'revision' of the piece one finished earlier. Which is why I hardly ever revise pieces, just build on the strengths and weaknesses of the previous ones. It's a bit like the Rothko idea of work en serie - a canvas is only so big; its dimensions necessarily curtail work that necessarily continues on the next one.

I wouldn't mind seeing the text of that talk some time.

I personally tend away from the idea of a work as an "ongoing" thing, except in terms of what performers and listeners do with it. It irks me that there are still a few pieces in my catalogue that I wouldn't want to be performed in their current state, and god knows whether I'll ever get round to doing the necessary work on them, because in a couple of cases it would involve basically writing a new composition on the ruins of the old one. The novelist B S Johnson viewed writing as partly a kind of exorcism -  "in order that it may be over there, in a book, and not here in my mind" - and that's closer to my attitude: once something is finished, it has to make its own way in the world and I have other things to get on with. (This also applies in a different way to the electronic music of FURT: once we have what we think of as a good recording of one of our "compositions" we don't perform it any more, and move on to the next one: the "completed" improvisation becomes a fixed-media composition.)

That Dutilleux revision is a very strange one, and strangely related perhaps to the story of the last chord of Ives' Second Symphony.