Austrian composer Gustav Mahler tended to be a little superstitious. He knew that Beethoven died after composing his 9th Symphony, and that Anton Bruckner had left his 9th Symphony unfinished at his death. Never mind that there were many exceptions to this "curse of the 9th" (such as Mozart's 41 symphonies and Haydn's 104) but Mahler intended to avoid a similar fate with a little numbering trick.
After his 8th Symphony, Mahler composed Das Lied von der Erde ("Song of the Earth," completed 1909), which ostensibly might be classified as an orchestral song cycle like Lieder eines fahrenden Gesellen ("Songs of a Wayfarer") or Kindertotenlieder ("Songs on the Death of Children"), except that in length, structure, and scope — the final "song" is a half hour in length, for example — it actually seems somewhat like a symphony, and Mahler thought of it as such.
When Mahler followed Das Lied von der Erde with his Symphony No. 9 (also completed in 1909), he believed that he had actually composed ten symphonies, and thus was free of the curse. He set to work on Symphony No. 10, but while in New York City in 1911 during a stint conducting the New York Philharmonic, Mahler became ill with a blood infection. He returned to his home in Vienna and died on May 18 at the age of 50.
At the time of Mahler's death, the first movement of the Symphony No. 10 had been pretty much completed. This is a gorgeous and moving Adagio of about 25 minutes in length, and it was performed as early as the 1920s. The remainder of the score was mostly in unorchestrated drafts and sketches. It didn't actually seem to be missing anything crucial, so several composers and musicologists took a stab at "completing" or "reconstructing" or deriving at least a "performing version" of the Mahler 10th Symphony.
The first out of the gate was musicologist Deryck Cooke, whose initial version of Mahler's Symphony No. 10 was recorded in 1966 by Eugene Ormandy and the Philadelphia Orchestra (a recording I owned back in my LP days). In the decades since, there have been several other versions and recordings. (The Wikipedia entry on Mahler's 10th Symphony provides a description of the various versions; ArkivMusic.com lists 20 recordings of the Mahler 10th currently available, with 35 additional recordings of the Adagio by itself.)
Needless to say, the very concept of attempting to complete an unfinished Mahler symphony is quite controversial. Some conductors simply refused to perform it, including my three favorite Mahler conductors, Georg Solti, Pierre Boulez, and Leonard Bernstein. As a listerner, I have struggled myself with this symphony for many years, trying to integrate it into my overall knowledge of Mahler's works, but never being entirely satisfied that it fits.
I now have five CDs of various versions of Mahler's Symphony No. 10:
James Levine with the Philadelphia Orchestra (1980) — Deryck Cook Version 2
Leonard Slatkin with the St. Louis Symphony Orchestra (1994) — Remo Mazzetti Version 2
Simon Rattle with the Berliner Philharmoniker (1999) — Deryck Cook Version 3
Robert Olson with the Polish National Radio Symphony Orchestra (2000) — Joe Wheeler Version
Andrew Litton with the Dallas Symphony Orchestra (2001) — Clinton A. Carpenter Version
In recent years several other versions have been recorded, and there are likely to be more in the future.
The first rather comforting fact is that all these versions sound roughly the same. There are no huge missing chunks that have to be invented, and no confusion about the ordering of the movements. The differences are mostly in orchestration and detail. Not surprisingly, the reconstructions have been pursued with different philosophies, ranging from a conservative approach of discreetly orchestrating only what Mahler left of the score, to more adventurous attempts to get inside Mahler's 1911 head and enhance the score in a way that Mahler himself might have. This can be challenging considering that Mahler was still growing and evolving as a composer when he died.
The four versions that I'm familiar with can be ordered from conservative to speculative like this:
Joe Wheeler — Deryck Cooke — Remo Mazzetti — Clinton Carpenter
The Carpenter version is a bit too much for my tastes. He frequently throws in little counter-melodies or squiggly filigrees that simply cannot be in Mahler's score, and which I mostly find distracting. (On the other hand, if you're overly familiar with one of the more conservative renditions, Carpenter's shake-up can be quite exhilerating!) My favorites of these recordings are those by Simon Rattle (generally a first-rate Mahler conductor) and Leonard Slatkin. The Slatkin includes a 19-minute bonus CD that illustrates some of the differences among the versions.
I have never seen the scores and sketches that Mahler left at his death. The only way I can judge the legitimacy of the 10th Symphony is in the context of my 35-year familiarity with the rest of Mahler's work. Is the 10th structured like the other Mahler symphonies? Does it sound like Mahler? And most importantly, does it give you that typically transcendent and terrifying trip through Mahler's psyche?
Mahler's Symphony No. 10 is in five movements with a symmetrical structure. These movements are titled differently in the various versions, but here's how I like to think of them:
1. Adagio (~ 25 minutes)
2. Scherzo No. 1 (~12 minutes)
3. Purgatorio (~4 minutes)
4. Scherzo No. 2 (~12 minutes)
5. Finale-Adagio (~25 minutes)
This five-movement symmetrical structure is reminiscent of Mahler's 7th Symphony, and the use of slow movements to begin and end the symphony is similar to the 9th Symphony.
Certainly Mahler is very present in the extraordinary first movement, which we know was left in a much more complete state than the other four. Towards the end are several hair-raising dissonant chords, like cries of agony. There is nothing quite like these chords in Mahler's earlier work, and yet they sound perfectly Mahlerian, perhaps giving us a little taste of where Mahler may have gone had he lived just another 10 years, and perhaps been able to assimilate the music of late Debussy, early Schoenberg, and early Stravinsky.
The second movement also sounds like pure Mahler to me, with a bouncy opening, lovely interludes, a shifting metric scheme between 4/4, 6/4, 3/4, and 5/4, and a great exhuberant ending.
The first hint of a problem with the 10th Symphony comes with the third movement. It has a spooky aura and a "night in the graveyard" feel similar to parts of the Mahler 6th, yet it seems to get stuck, and then actually truncated by a sudden ending in a sweeping flurry of harps. At about four minutes in length, it is by far the shortest purely orchestral movement in all of Mahler's symphonies. Only the 4th movement of the 2nd Symphony and the 5th movement of the 3rd Symphony are of comparable length, and these are both song settings.
We are back on firmer ground with the second Scherzo, a bristling can't-sit-still malange of conflicting ideas, tempi, and dynamics that shouldn't make sense but does, and it fades out with a wonderful quiet little pattern on the timpani.
What follows is a transition into the fifth movement that consists of a series of loud naked thuds on the bass drum, with the music slowly arising between these thuds. I remember when I first heard the Ormandy recording in the 1970s thinking how uncharacteristic of Mahler this passage was, and I have never been able to shake that impression. I cringe in embarassment every time I hear it. We know that Mahler derived these thuds from the sounds he heard while living in a hotel above the streets of New York City when a funeral procession of a fire chief passed by. Yet, they just don't sound right to me.
As the fifth movement gets going, there soon follows a little melody for the flute that some commentators have found "beautiful" or "unearthly" or "sweet and ethereal" but to me sounds banal. As the fith movement proceeds, we frequently encounter appealing passages, little bits and pieces of brilliance, occasional recapitulations of themes from the earlier movements (including those dissonant chords from the first movement), but for me nothing holds together. Ultimately I find the entire movement to be incoherent. It just doesn't make sense. Mahler is going through the motions but without connecting to the listener (or at least to this listener). No other movement in no other Mahler symphony is as baffling and emotionally distant to me as this final movement of the 10th Symphony.
This impression is, of course, extremely subjective. (Your mileage may vary.) I can't identify where it goes wrong, and what Mahler might have done to fix it. But I cannot believe that had Mahler recovered and returned to work on the 10th Symphony in the summer of 1911, he would not have performed some serious surgery on this final movement — perhaps even rewritten it entirely.
But if we're going to listen to the Mahler 10th Symphony at all, we need this movement (as flawed as it may be) to bring the symphony to its proper conclusion. Obviously there is nothing we can do about it, so I have come to accept it. Few people's last words are their best ones. There is a certain sadness in the last movement of the 10th Symphony that arises from Mahler's inability to express himself, but I wish the sadness could have come from the music itself.