There are two types of composers: those that revise, and those that do not. I am not referring to making minor corrections—a wrong note here or there or whatever—but major changes. like ripping apart and re-constructing sections, adding new notes and changing lots of dynamics and articulations.
There are many well-known composers who edited older works: Sibelius, Schoenberg, Stravinsky, Boulez and Ives, to name a few, but others simply do not like to go back or feel they do not need to, or only make changes when there are very minor details that need fixing, like Christopher Rouse and Poul Ruders.
This is not a black and white issue, of course, but one with infinite shades. One composer's revision is another composer's minor tweak. Some revisions are musical, while others are technical.
I have been thinking about this lately, because I recently revised a work I wrote fifteen years ago, a forty-minute piece for piano trio called Sun Trio that was premiered in Los Angeles by an enterprising new trio. It a surreal experience, revising something I wrote so long ago. As I became intimately reacquainted with the piece, it felt good—like reconnecting with an old friend, or a son who has moved away (or in this case, a sun).
I also recently made minor tweaks to my eleven year old Duo for Flute and Marimba. Like Sun Trio, I didn't make any major changes, just small tweaks that make it more playable: slowing down a few of the tempos by a couple of notches, or tweaking a note here and there. I wrote this piece for percussionist Ingrid Gordon and never played it myself, so I never had a chance to really internalize the tempos in a physical way. I spent this past summer practicing the part, and now realize that my tempos are a little too fast—playable, but right on the edge. Too on the edge, in fact, so I bumped the tempos down a bit.
This brings up another interesting issue: that of altering works written long ago. Is it worth it? In this case, I made a few errors that could be fixed without harming my original intention. For example, I wrote a few notes as harmonics that work much better now as regular pitches. Back then, I was even more obsessed with dynamic shading than I am now, but in a few spots, I think I went way over the top, so I simplified some of the dynamics. I also increased a few of the tempos, easily shaving off a good minute or two from the entire work. By changing these details, the piece will still sound the same, but will also be much more playable. None of my revisions were huge, but everything just feels better now.
A final issue I have been obsessed with lately is the concept of having an editor. By default, composers are their own editors. I think this, more than anything, is the most difficult role to teach composition students: the importance of being self-critical, without pandering to the trends and preferences of others.
In some disciplines such as film, collaboration is normal. Classical composers almost never have an outsider edit their music, or even suggest edits, unless they work in the film industry, write operas or musicals or are still students. Can you imagine? Long, boring swaths of music by certain long-winded composers (I won't mention names) could be cut. What a thrill that would be, and imagine the millions of hours saved!
As Ruders once said in an interview, perhaps experience is what makes the difference, but if there is a good reason, revisions are OK. In the end, it depends on the composer and the work, but composers should always feel free to make changes if they will make a piece better.