More Than a “two hit” Wonder: English National Opera’s “The Dead City” otherwise known as “Die tote Stadt”

The performance of The Dead City, better known as Die tote Stadt at London’s English National Opera was a revelation and frankly, a sensation for reasons few critics have mentioned. But then again, few London critics have heard or seen this opera as often as I have. There was the Royal Opera House’s borrowed Willi Decker production from Salzburg and Vienna from 2007 that was miscast with a conductor who seemed to believe that slow was what was required in order to add gravitas. The vocal casting was not up to the work and killed off by the tempi. The critics were not kind and held Korngold fully responsible. The general tone following a concert performance of his other grand opera Das Wunder der Heliane along with the Royal Opera House’s Die tote Stadt was Korngold as yesterday’s news.

Brendan Carroll reminded me that the recording with Erich Leinsdorf is complete: Or it was recorded as “complete” at the time, and it is possible that since then, a new edition of the score now includes passages that may not have made their way into the 1975 recording.  I’ve heard the recording countless times and seen the opera in Salzburg, Vienna, Paris, London, Berlin, Longborough (this past summer) and Munich (on DVD) but what I hadn’t realised was how crucial experiencing the opera in its entirety as well as experiencing Korngold’s unreduced orchestration was to the work’s effectiveness as a performance.  I’m pretty sure that every performance until now has been cut – both for dramatic reasons as well as demands made on the singers. And yet…when I’ve attended a performance, I’ve always been puzzled how beautiful motific ideas such as Brigitta’s little aria about love at the opening, seem to vanish into the ether. And how could you write something as unique and unforgettable as the Pierrot Song or the Lute Song and then not keep them flickering, ever-present in the musical background throughout?

The answer to this and other puzzles was answered in this complete performance:  you can’t and Korngold didn’t.  By not excising the inter-act interludes and keeping all of the processional scenes, these motifs give the work its melodic continuity as they are revived and developed throughout.  I finally experienced a performance that offered Korngold’s full thematic coherency. The cuts in the past have only made the work feel longer.

The often cut processional scenes are important for two reasons: musically, they allow Korngold to expand his motific material, so melodic ideas don’t just come and die, but live and thrive. But the second reason is perhaps more obscure: they are important dramatically, despite the doubts of countless well-meaning directors. Not understood today is the significance of Meyerbeer’s opera Robert le Diable (Robert the Devil) – an opera that was one of the most popular at Vienna’s Court Opera until its final performance in 1908. Until then, it had enjoyed some 200 performances since its Viennese premiere in 1869. Meyerbeer’s opera, originally premiered in Paris in 1831, also plays with the idea of the female as a combination of sacred virgin and whore and does so in a scene in which the ghosts of dead nuns attempt to seduce Robert. Die tote Stadt’s  “sacred virgin” is Paul’s dead wife Marie and the whore is Marietta, who in Korngold’s opera is a dancer in Meyerbeer’s Robert le Diable. Marietta informs Paul that she is the dancer playing the role of the arch seductress, the abbess Hélène.

The idea of the female as sexual predator disguised as sacred virgin in order to degrade men was already current at the time of Korngold’s composition. Otto Weininger’s book Sex and Character, published in 1903 was a sensation in the German speaking world and made even more sensational when Weininger committed suicide upon its publication at the age of 23 in the house where Beethoven died.

Those who see, study and hear Die tote Stadt and think “ahh, Freud” are barking up the wrong tree. Freud was hardly on the map at the time and Weininger was all the rage.  Franz Schreker admitted in correspondence to the critic Paul Bekker to Weininger’s influence when he, Schreker, created the female characters in his operas. One can hear Weininger’s influence in Strauss’s Salome. Wilde wrote his play long before Weininger published Sex and Character, but Weininger had already set the tone when Strauss opportunistically took Wilde’s play for his opera. 

And so it was that I realised how important the processional music is to the madness of Die tote Stadt. This is where the Viennese Zeitgeist of fin de siècle cascades in like an avalanche. The febrile eroticism is captured in the orchestration – and to reduce it, is to reduce the insane obsession and humiliation of worshiping the virgin while fantasising about the whore.

I understand that Kirill Karabits, the conductor insisted on the complete orchestration along with the opera being performed without cuts.  I applaud the ENO for letting him have his way. Critics who say there is too little drama in the work may have a point, but the opera is actually a tone poem representing sexual obsession and barely requires a chronological narrative in the traditional sense of la forza del destino. If Strauss could take a philosophical idea and turn into a tone poem with Also sprach Zarathustra, Korngold took sexual obsession and did the same. Only, he added voices…and here we come to the ultimate dilemma: to accommodate the orchestra Korngold demanded requires a house the size of Vienna’s State Opera. The ENO’s Coliseum holds 2300 seats and is a vast half rotunda. The vocal writing is not Wagnerian, but Italianate spinto, but with a louder and larger orchestra than Puccini ever required.  The original singers in Hamburg, Cologne, Vienna and Berlin were not heroic Wagnerian singers and included Richard Tauber, Lotte Lehmann, Maria Jeritza and later, Jan Kiepura: not a Siegfried or Isolde between them, though Richard Schubert who premiered the role of Paul in Hamburg was admittedly a highly regarded Wagnerian. So the circle to square, is fitting the enormous orchestra, the unrelenting volume of throbbing insanity around voices designed more for the lyricism of Puccini. 

The ENO didn’t manage to square this circle and the fact that both soprano (Allison Oakes) and tenor (Rolf Romei) craved indulgence due to ill health was probably neither here nor there.  They shrieked and shouted to the point I feared for their vocal health and future careers. Voices cracked and broke while vanishing in the whirlpool of the orchestra. (I sat in the first row of the Dress Circle) Sarah Connolly’s Brigitta and Audun Iversen (double role of Franz and Fritz) were luckier in their roles, in so much as they didn’t have to fight against the full battery of the band. Salzburg, Vienna, Munich and Longborough managed to find sopranos who looked more like the seductress Meyerbeer required in the dancing role of the abbess Hélène, but I suspect the thinking of the ENO casting director was finding a voice that might be heard above the din of Korngold’s full orchestration. No chance – and I’m not sure anyone short of Birgit Nilsson singing with her laser beam and tonal sharpness could have been heard. Perhaps there is an argument for amplification so that singers are not committing professional suicide in front of an audience.  In any case, Karabits, the orchestra and chorus were the heroes of the evening. Frankly, I was surprised to see the generous manner in which the singers greeted Karabits on stage for the applause. I half expected them to punch him in the face – the two lead roles were the sacrificial lambs in this marvellous representation of the score.

There must be a solution, but I’m not sure what it was during the peak of the work’s popularity. Maybe the mere fact that surtitles didn’t exist in earlier times required singers being heard and more importantly, understood, leading to even the most tsunami like orchestrations having to be reined in. But again, I’m not sure how to control an ensemble that clearly emerged in the wake of Schoenberg’s Gurre Lieder. As a singer, you can only do what’s required and hope you live another day. But what an experience the performance was.

I don’t consider myself qualified to write about the production except from my personal perspective, Annilese Miskimmon’s direction, Miriam Beuther’s sets, Nicky Gilligbrand’s costumes and James Farncombe’s lighting were all in perfect accord with the music and brought us closer than comfortable into the obsessive fantasies and fears of Paul’s brain.  I was totally convinced and whatever sniffy comments made about re-locating the action to the 1950s (really? I didn’t notice and thought it more like the 1920s, which was when the work was premiered), to be without merit – at least, without merit if the comments were intended critically. Maybe it was just an observation that all operas these days appear to be updated to post-war Britain.

Probably the Leinsdorf recording remains unsurpassed and is uncut. The singers are placed so far forward that the orchestration hardly emerges in its full force.  Perhaps this more modern edition as performed at the ENO has bits that were not included when Leinsdorf recorded.  The running time at the ENO was two and a half hours and the tempi never dragged. Hearing a performance in the opera house is more generous to both listener and composer. At a performance, you can’t do your taxes, wash the dishes, or iron shirts at the same time as with listening to a recording. You’re forced to follow the musical as well as dramatic narrative. For this reason alone, I would urge everyone who wishes to be swept away by the sheer thrill of the music, to attend the ENO’s performance. It’s utter madness, but I believe that was what Korngold intended.