TCHAIKOVSKY Swan Lake • Neeme Järvi, cond; James Ehnes (vn); Bergen PO • CHANDOS 5124(2) (2 SACDs: 154:41)
As has been noted in previous reviews of recordings of Tchaikovsky’s “complete” Swan Lake, there may well be as many different versions of the score as there have been productions of it. The problem is that Swan Lake is both the earliest (1875–1876) and the longest of the composer’s three great ballets, andRead more it has had so many cooks adding their own ingredients, removing others, and generally revising the recipe that no one can say for sure what made up the original soufflé.
The generally known and accepted facts are these: The ballet, with original choreography by Julius Reisinger, was staged for the first time in February, 1877 by the Bolshoi Ballet at Moscow’s Bolshoi Theater. It was not well received; audience and critics alike felt it was too long and convoluted, its music too heavy, and its libretto, adapted from a story by a German author, an affront to Russian sensibilities. And thus began the tinkering and tampering. By the time the work was revived in 1895 by the Imperial Ballet at St. Petersburg’s Mariinsky Theater there was new choreography by Marius Petipa and Lev Ivanov, along with major musical revisions to the score by the Imperial Theater’s conductor and composer, Riccardo Drigo. It should be noted that by the time of the 1895 revival, Tchaikovsky was dead and had no hand in the new performing version. Tchaikovsky and Drigo had worked together previously, but according to accounts, they didn’t agree on much of anything and their relationship was strained.
The upshot of all this is that there is no definitive Swan Lake. It was no longer a ballet by one composer, but rather a group effort; and you know the saying about a camel being a horse designed by a committee. It’s important to bear this in mind when considering the various recordings of Swan Lake that claim to be complete, for the drastic differences in timings cannot be explained by mere tempo differences alone. There have to be other factors involved, such as omission of some movements, cuts to others, and/or reliance on differing versions/editions. Let’s look at the timings of several well-known recordings, sorted in order by duration.
St. Petersburg Mariinsky O
St. Louis SO
Russian National O
Russian State SO
Michael Tilson Thomas
Royal Opera House O
Right off the bat, I need to offer a disclaimer: My personal familiarity with the above-listed recordings is limited to only four of them—Gergiev, Pletnev, Yablonsky, and now this new one by Järvi. Of the four, Gergiev’s version is the worst in terms of the hatchet job it does on the score. Movements are reordered—for example, the act I Waltz has been moved to act III and its ending abridged—and it’s full of egregious cuts—some 40 minutes of music are sacrificed. Gergiev’s Swan Lake is presumptively based on the Mariinsky performing version; i.e., the above-mentioned Drigo edition prepared for the 1895 St. Petersburg revival.
Looking at Pletnev’s timing of 142:52 vs. Yablonsky’s 148:58 and Järvi’s 154:41, it seems pretty obvious that that while tempo differences over the course of two and a half hours could account for the difference of approximately six minutes between Pletnev and Yablonsky and, in turn, between Yablonsky and Järvi, they’re unlikely to be the cause of the approximately 12-minute difference between Pletnev and Järvi.
Upon closer examination of all three recordings, what I found was that Yablonsky and Järvi both include two often dropped numbers from act III, the Pas de deux that was written after the fact specifically for Anna Sobeshchanskaya, and the “Danse Russe,” added specifically for Pelageya Karpakova. Pletnev omits these two additions, as do a number of others. Whether they should be included or not is a rather complex question.
Ballerinas of the day were not much different from their opera diva counterparts in terms of their egos. They had no shame when it came to demanding custom cadenzas to show off their voices or, in the case of danseuses, their fancy footwork and frilly tutus. The story surrounding Sobeshchanskaya and her Pas de deux is especially messy and borders on scandal. Originally picked to dance the lead role of Odette (the Swan) for the 1877 premiere, Sobeshchanskaya was ignominiously dropped from the cast at the last minute when a high-placed government official with whom she’d had a dalliance accused her of having taken expensive jewelry from him and then pawned it when she married a fellow danseur. On the spur of the moment, she was replaced by Pelageya Karpakova. Sobeshchanskaya survived the indignity and went on to dance the title role when the ballet was staged again a month later with no greater success than at its premiere.
But the intrigue didn’t end there. The ballerina made no bones about the fact that she hated both the choreography and the music, and so off she went to St. Petersburg, where she engaged Petipa to choreograph a new Pas de deux for her that would replace the third act’s Grand pas. Petipa complied and choreographed the new number to music, not by Tchaikovsky, but by Ludwig Minkus, the Imperial Ballet’s composer in residence.
When news of this change reached Tchaikovsky, he was miffed; his ego was probably bigger than Sobeshchanskaya’s. How dare she?! He was the composer, and he alone should take credit (or discredit) for the music. After some smoothing of his ruffled feathers, Tchaikovsky agreed to compose the music himself for Petipa’s new Pas de deux, but there was a problem. Tchaikovsky’s new music didn’t synch up with Petipa’s choreography, and Sobeshchanskaya, now back in Moscow, wasn’t about to travel back to St. Petersburg to go through the whole exercise again. She didn’t seem to care much one way or the other about the music, but she was adamant about keeping Petipa’s choreographed number. How exactly Tchaikovsky was prevailed upon to discard his newly composed music and essentially start over, this time following the outlines and rhythmic steps of Minkus’s music is not explained, but that’s what Tchaikovsky did. So, this particular episode apparently had a satisfactory ending for all involved, except, I suspect, for Minkus who surely must have felt put out. The original Grand pas with music by Tchaikovsky was replaced by Sobeshchanskaya’s Pas de deux with music first by Minkus and then by Tchaikovsky.
Based on the foregoing, it would seem that there is every reason to include this number in complete performances of the ballet, yet many conductors, Pletnev among them, don’t. The situation regarding the “Danse Russe” (Russian Dance) is much simpler and appears to be the reverse; it’s one of deletion rather than addition. It was composed for and included in the original 1877 version of the score danced by Karpakova, the premiere’s last-minute substitute for Sobeshchanskaya. The number was then removed for subsequent performances in which Sobeshchanskaya took over the role, for reasons one can easily guess. If two competing sopranos could bitch-slap each other on stage during a production of a Handel opera, there was no telling what professional jealousy might provoke between two rival ballerinas.
This describes only some of the butchery that turned Tchaikovsky’s finely feathered swan into a plucked chicken. It’s well to remember, however, that Swan Lake was not only the composer’s first completed ballet, it was really his first major stage undertaking to survive the ravages of time, even if not entirely intact. He was working on his opera Eugene Onegin at the same time, his first opera to achieve success; and though there had been earlier operatic efforts—The Voyevoda, Undina, The Oprichnik, and Vakula the Smith—they were either destroyed by the composer, recycled, later revised, or didn’t stir much interest at the time. Thus, at 37, Tchaikovsky’s greatest works still lay ahead of him, and he had yet to achieve the self-confidence that fame would bring him to be able to just say no to those who would mess with his music.
Neeme Järvi’s Swan Lake follows his Sleeping Beauty, reviewed in 36:5. I would expect to see a Nutcracker in the near future, perhaps timed to coincide with Christmas (I’m writing this in November 2013). My only objection to Järvi’s Sleeping Beauty was his somewhat business-like approach, which struck me as missing some of the music’s fairy magic. But the Bergen Philharmonic’s polished playing, James Ehnes’s ravishing violin solos, and Chandos’s thrilling multi-channel SACD recording offered much allure.
On relistening to that release, and in listening to this present one, in which Järvi, Ehnes, the Bergen orchestra, and Chandos repeat their earlier accomplishment, it occurred to me that my criticism of Järvi wasn’t entirely fair. There are two ways to conduct a ballet performance for a strictly audio recording. You can approach it as a concert work, in which case you will tend to emphasize the melodic, harmonic, and structural elements of the score, or you can approach it as a suite of dances, in which case you will emphasize the music’s rhythmic and terpsichorean aspects. Järvi falls into the former camp, and there’s nothing wrong in that, as long as he’s not directing a live production of the actual ballet, in which tempo, pacing, and phrasing need to be molded more flexibly to accommodate the movements of the dancers.
I can’t say absolutely that this is the most authoritatively complete Swan Lake on record, though in taking up the original 1877 score and including additional material supplied by Tchaikovsky himself for subsequent performances, Järvi gives us a version that’s certainly more complete than are a number of others. What I can say is that of the four recordings of the score with which I’m familiar, Järvi’s would now be my first choice, and taking all other factors into account—superb playing by the Bergen Philharmonic, James Ehnes’s beguiling solo violin contributions, and a killer recording—I’d extrapolate from this that Järvi’s Swan Lake is now the one to have.