In my reflection on the death of Karlheinz Stockhausen last week, I accused Igor Stravinsky of being envious of the jazz that was being played when he was living in Paris, because, in spite of his many efforts to do so, he just could never compose music like that. Watching the final performance of The Rake's Progress by the San Francisco Opera yesterday, I realized that there was probably another target of Stravinsky's envy, which was probably just as strong, if not more so. That target was Benjamin Britten, and I believe that this hypothesis cannot be ignored in addressing Stravinsky's only attempt at a "grand" opera.
This is not to say that I dislike Stravinsky. What he composed in his attempts to reproduce rags and tangos has a unique enough sound that the works are interesting through their very infidelity to their sources of inspiration. The same can be said of the rather convoluted ways in which Stravinsky set English text, not just in the Rake but in many shorter efforts. The counterintuitive phrasings must be enough to drive any singer crazy, and they ultimately highlight the extent to which English texts would fit within Britten's music like a hand in a glove. The comparison is most explicit in "The Flood," that "musical play" commissioned by CBS Television in 1960, which I find more worthy of Ivan Hewett's "cringe-making" epithet than just about anything Stockhausen ever produced. While it is possible to imagine that Stravinsky was unaware that Britten had composed a similar work, "Noye's Fludde," in 1957, the fact that Stravinsky drew upon exactly the same Chester Miracle Play texts makes that possibility highly unlikely. Nevertheless, I have seen many performances of the Rake that I have had no trouble enjoying for what they were and I treasure my CDs of the recording that Stravinsky made in London in 1964.
Perhaps some of that joy comes from this being the end of the second major "road" that Stravinsky followed in his musical career: the path of neoclassicism that stands between his more Russian period and his explorations of serialism at the end of this life. The Rake's Progress is classical in just about every way you can imagine, from the harpsichord-accompanied recitativo secco to the wisps of Handel and Mozart that are never quite tangible in the score but always seem to be suggested. If W. H. Auden and Chester Kallman were not able to produce a more salutary text for Stravinsky's musical language, then, as is the case with his experiments in jazz, one can still enjoy the results. Trouble only ensues in a production that does not let Stravinsky be Stravinsky.
Unfortunately, Robert Lepage provided the San Francisco Opera with such a production, reminding me of my having invoked the verb "undermining" when I wrote about Robert Carsen's setting of Gluck's Iphigénie en Tauride last June. Much of my discontent probably resides in the fact that Stravinsky was very good at falling back on structure when intuition was not serving him as well as he wished. The Rake's Progress is impeccably structured, so I was immediately on my guard when I picked up my Opera program and saw that Lepage had smashed Stravinsky's structure to bits. To be more specific, Stravinsky composed this opera in three acts, each consisting of three scenes; and this was the perfect complement to the Auden-Kallman libretto, which structures the narrative of Tom Rakewell around three wishes he makes and how they are implemented by Nick Shadow. Thus, when my wife complained that she felt the final scene of the opera went on far too long, my reaction was that it felt long because she had been deprived the benefit of an intermission before the "unit" of the third act got under way. With all of his experience in ballet, Stravinsky knew all about audience fatigue; and Lepage's attempt to restructure the opera in two acts was a serious "sin" against Stravinsky. When we then recognize that this revised structure also distracts from the narrative role of the three wishes, we must further conclude that he has "sinned" against Auden and Kallman. Finally, to honor a tripartite structure that may be the best part of the opera, I would suggest that Lepage also "sinned" against Hogarth by rejecting everything about his original conception of how the narrative should be set and conjuring up an image that sings of London but exhibits Hollywood. To be fair Hogarth is honored by both the music and the libretto only in spirit, but Lepage decided to have none of that spirit. Also to be fair most of Lepage's images were pretty impressive, fascinating enough to be a viable audience magnet. The problem was that they were an "alternative universe" to the one that Stravinsky, Auden, and Kallman had conceived. Those two universes were almost violently incompatible, which is why I can argue that the former undermined the latter.
The good news was that the music was still pure Stravinsky at his best. All of the primary voices, William Burden (Tom), Laura Aikin (Anne Truelove), Kevin Langan (Anne's father), James Morris (Shadow), Catherine Cook (Mother Goose), Denyce Graves (Baba the Turk), and Steven Cole (Sellem), had their parts solidly nailed and knew how to work with Donald Runnicles, the orchestra, and the chorus to make this production "work" as well as it could. Graves was particularly effective in her full-out-no-shame conception of Baba; and I suspect it would have been even more fun to see what she would have done with a staging that had been more true to the libretto!