Of all the conquests of music’s domain by Kentridge’s conquistadors of the eye, the treatment of the Interludes was the most rapacious. As Kerman long ago pointed out, since in a work constructed strictly as a sung play, it is only in the Interludes that the music can assert its wonted operatic dominance and sum up the emotional content of what’s happened for us. “So musical expansion, which the final force categorically demands,” he notes, “could take place only in the interludes between the scenes . . . A forceful emotional place is provided without breaking the naturalistic illusion. Or more exactly, this illusion is broken by the curtain, a theatrical convention Berg was careful not to override.” And Kentridge assaults us with his eye-candy not only during the interludes between the scenes, but in the spaces between the acts. (The opera is presented without intermission, in flagrant contradiction of its formal framework.) Most unforgivably, he previews onscreen the dancing of the “low tavern” scene, thus ruining the sudden break from the tension of the famous full-orchestra crescendo on B (the climax of the “Invention on a Note” that is the design of the preceding scene as a whole) into the off-key barroom piano polka, one of the score’s great coups. There’s such an artform as one that relates the linear throughline of a story without a break in its visual rhythm, with music in constant supportive attendance. It’s called the movies.

“You can always tell when a director hasn’t worked with his singers,” I read in a recent New York Times article. I’m not sure I can always tell (some singers are just good actors who don’t need much from a director, while others can receive intense directorial attention to no effect whatever), but in general, I agree. Funny thing is, the speaker was Nikolaus Habjan, who’d just finished staging Salome at the Theater an der Wien with puppet doubles, manipulated by the singers, for the roles of Salome and Jokanaan, and who uses puppets a lot in his opera productions. He seems to prefer them to singers. (“They can get away with things that singers can’t,” he says—perhaps those early Maeterlinck plays will finally get their proper due.)  And there’s a puppet in Wozzeck. Now, I’d have no objection to a puppet-show opera. (The Salzburg Marionettes Present Wozzeck?) With my dutifully open mind, I allow that possibly I’d like Habjan’s big puppets, which sound sort of like dark versions of Red Grooms’ kooky-charming figures. But Kentridge’s puppet was a miserable little doll supposedly representing Marie’s child, hand-manipulated by a guest puppeteer while Marie sat around and watched. This was all-too-reminiscent of the one that takes Trouble’s place in Anthony Minghella’s production of Madama Butterfly. There, the excuse was a tradition in one style of Japanese theatre; here, it was Kentridge’s worry that a child actor/singer might make a mistake. In both cases, the result was creepy in the wrong sense, and distracting.

Since we have arrived, by however convoluted a route, at the subject of performance, I’ll enter a few comments on it here, and if I seem to give them short shrift, it’s because in a real sense performance didn’t matter much, having been swallowed and digested by installation-production. And while I assume that Kentridge must have worked with his singers with whatever time and energy remained from co-ordinating everything else, I sincerely doubt that he knows how to be of much help to them. Nevertheless, for the record: