We Must Not Do Without It

I first came to know Rienzi in 1983, on record, in the edition conducted by Heinrich Hollreiser with the Staatskapelle Dresden — the orchestra of the world premiere. The box carried on its cover a watercolour of around 1843, the work of Ernst Polycarp von Leyser after a drawing by Ernst Laddey, portraying Josef Tichatschek, the role's first interpreter, in the finale of the third act, and on the back the costume designs made for the 1842 production. Anyone who came to this opera in that way retains not merely an aural memory: he retains a stage image, received as a gift from the cardboard box along with the record. Silver, gold, irradiated marble. A conventional Rome, of course, inherited from French grand opéra rather than from archaeology — but an image, and for forty-three years it has held in my mind and above all in my imagination.

I say so at once, because whoever writes about theatre ought to declare where he is watching from, and because in the case of Rienzi the declaration counts for more than usual. I followed the premiere of 26 July from home, in 4K and Dolby Atmos, not from the Festspielhaus. Everything I shall say about the balance between voices and orchestra applies to a television relay built with microphones in the pit and on the stage, not to the acoustics of that hall. It is a difference that cannot be passed over in an account of a Bayreuth evening.

The facts

On Sunday 26 July 2026, at four in the afternoon, the curtain of the Festspielhaus opened for the first time in a hundred and fifty years upon a work that does not belong to the canon.

That Wagner did not want the early works on the hill is documented; what does not exist is any testamentary or statutory foundation. The set of ten running from the Dutchman to Parsifal rests upon passages from letters addressed to Ludwig II in 1882, and as to how definitive those lines were the Festival itself now dissents — though it is an interested party. That early work its author treated for decades as a noisy monster and a screamer, with a self-critical coquetry that posterity has taken more literally than he did. To bring it onto the hill, a resolution of the Foundation's board was required.

It was not entirely unknown in the town. In July 2013, for the bicentenary, Rienzi had already been staged in Bayreuth, but down in the town, at the Oberfrankenhalle: a multi-purpose hall fitted out with a flat stage and a pit at ground level for a figure said to be around two hundred and fifty thousand euros, directed by Matthias von Stegmann, conducted by Christian Thielemann, in co-production with Oper Leipzig. The operation was avowedly conceived not to breach the precept but to supplement it, and even then the programme book confronted head-on the most uncomfortable chapter, the one devoted to the Rienzi Hitler loved. Thirteen years later, the same question has climbed the hill.

Direction, sets and costumes are signed by Alexandra Szemerédy and Magdolna Parditka, the first women outside the Wagner family to sign a production in the Festspielhaus: before them, only Cosima and Katharina. On the podium the women had arrived five years ago, with Oksana Lyniv. Here Nathalie Stutzmann conducts, with Thomas Eitler-de Lint as chorus master. That same morning, a few hours before the curtain, the Festspielhaus had hosted the commemoration «Verstummte Stimmen» for the Jewish artists driven out and murdered, with an address by Michel Friedman. There are nine performances in all this summer, and then Rienzi disappears from the hill's programmes.

There is no single Rienzi

Before discussing what was seen, it is worth making clear what was heard, because in the case of this work the question is not rhetorical. The autograph score is considered lost. There exists no final authorial version. At the Dresden premiere, on 20 October 1842, the performance ran to more than six hours including intervals, and from that day Rienzi has been cut everywhere, house by house, according to the taste and the stamina of the theatres.

The two major recordings measure the problem exactly. The Hollreiser edition, ample as it is, shortens; the only truly complete one is a British radio recording: in June 1976, in the BBC studios in Manchester, Edward Downes obtained leave to perform and broadcast Rienzi without cuts, reconstructed from the manuscript and printed sources then available, with John Mitchinson in the title role, Lois McDonall and Lorna Haywood, within a project that in May of the same year had already tackled Die Feen and Das Liebesverbot. The most complete Rienzi in existence, therefore, has never been a staged production.

To this history is now added the version prepared for 2026, built upon the surviving material around an explicit thesis: had a definitive redaction been arrived at towards 1871, Rienzi would have found its place in the Bayreuth canon as well. On that premise the working group decided the cuts and the sequence of numbers by following the musical dramaturgy and the composer's own late pronouncements, and claims to have recovered portions believed lost.

From which follows a conclusion reaching well beyond the evening: there is no single Rienzi, there are decisions. Whoever writes about it is commenting on somebody's choice. It is a rare condition in the Wagnerian repertory, and it is the reason why this performance, before pleasing or displeasing, is a document.

What the pit did

The musical outcome was of the highest order, and this must be said without qualification. The Festival orchestra played with a cleanness that within my own listening memory had not been heard for years on that stage, and the technical reason is fascinating: the covered mystic gulf, devised for the works of maturity, blends the brass, softens its attack, gives back colour instead of blare. Applied to a grand opéra score born for an open pit, that device produces an effect no other theatre can obtain. The work Wagner excluded for being too noisy is precisely the work that hall heals.

Stutzmann worked in the direction opposite to smoothing: instead of stitching together the heterogeneous components of the early work — Spontini, Auber, Meyerbeer, and the fabric that would become Wagnerian — she made them distinct, letting the friction between them remain audible. It is the choice of someone who knows singing from the inside, and in a score alternating closed numbers and symphonic surges it is the only one that avoids a uniform sauce.

The chorus was the evening's real protagonist. Rienzi is a choral opera in the strict sense of the term: the people are not a frame, they are a character, and their unanimity is not declared but manufactured by homorhythm. The Festspielchor traversed those massive blocks with a compactness that never slipped into coarseness, and at the same time sustained an individual acting commitment such as the mature repertory never demands of it. Some German voices judged the result uneven; from the relay the objection is not perceptible.

Andreas Schager sang a part that is, in length and tessitura, the most exposed Wagner ever wrote for tenor before he learned to husband it. The reactions of the German press range from those who found him wobbly to those who judged him triumphant, by way of those who observed that he takes a shade too literally the label of screamer that Wagner pinned on his own work. The truth seems to me to lie in between: he is a centre of vocal force that crosses five acts without giving way, and the prayer — the one place in the opera where the tribune is left alone with himself — asks for an inwardness that a voice built like this inevitably pays for. In any case, I liked him.

Gabriela Scherer has the opposite problem, and it is not of her making: Wagner wrote Irene stingily, holding her in reserve for much of the opera and granting her real room only when all is already lost. When she arrives, she arrives.

Jennifer Holloway, Adriano en travesti — at the 1842 premiere the part was sung by Wilhelmine Schröder-Devrient — was the evening's true winner in terms of applause, with a vocality now pointing towards the more dramatic repertory. These are also the two parts I struggled to hear in the relay, during certain passages: the middle register is the region that a mix with pit microphones eats first. But it is not solely an effect of the relay: there are those who, from the hall, observed that the balance Stutzmann builds covers the voices a little.

One detail nobody expected deserves to be recorded: at the premiere there was no applause during the acts. In a number opera, built to be interrupted, the audience applied the Bayreuth ritual of silence until the fall of the curtain. It is the first friction between this work and this house, and the staging did not produce it.

What the production is about

Then there is the stage, and here the case alters. Four residential towers in grey concrete, a wooden semicircle, a stage portal, a videowall that opens and closes, ceiling neons, glossy floors. A people that communicates by telephone and votes by raising a thumb and then pressing it onto a cardboard ballot box. Rienzi reaches office with forty-one per cent, and the German critics at once recognised in that figure the percentage attributed to the far right in the Saxony-Anhalt polls. On the monitors an invitation to fight for Rome flashes in English, and English returns on the emergency exits, on the wall of proclamations and in one of the two Gospel passages: in a German theatre representing Rome, language is not a scenic detail but an address, and it says whom the production is speaking to.

But the centre of the production is not there. It is in the second-act pantomime, and it is the point at which this staging declares what it is really talking about.

In the score, that ballet is no diversion: the herald announces that the Romans will see the death of Lucretia, the revenge of Brutus and the expulsion of tyranny — that is, the founding myth of the Republic staged by the tribune in order to legitimise himself. Theatre within theatre, propaganda inside the drama — and at the climax of the fiction Orsini's real dagger strikes. It is the only reason that half-hour of music exists.

In its place, at Bayreuth, the whole canon of the house files past: Das Rheingold, Die Walküre, Lohengrin, Tristan, Die Meistersinger — with Stolzing and Tannhäuser coming to blows and the Wartburg crossing the stage in the shape of an East German motor car — and, if the eye does not deceive, Parsifal, the work Wagner wished performed in that theatre alone. Conducting this procession is a golden figure with its back to the auditorium, which many will find reminiscent of the Wagner sculptures set out in recent years on the lawn before the theatre: works by Ottmar Hörl, which vanished almost entirely in 2023 and in 2024, and whose jubilee project was turned down by the Festival. Meanwhile the chorus, which within the pantomime plays the audience, rises and goes off to the bar and the lavatories as during the intervals, while from the loggia comes an imitation of the summoning fanfare.

It is not an isolated gag: it is a simulation of a Bayreuth evening set into the second act. And then the conclusion imposes itself. At the point where Wagner shows Rome contemplating its own foundation, this production shows Bayreuth contemplating its own. What the staging is about is not Rienzi: it is the Festspielhaus admitting Rienzi to itself, complete with an induction ceremony for the eleventh title. Every other element of the evening — the man with the placard prompting the speech, the percentages, the wall of proclamations, the golden figure conducting — sits inside that sentence.

That the most expert audience in Europe, and the only one in the world able to decipher those quotations in real time, should have enjoyed itself is not in the least surprising. That pantomime was written for them and struck its target. The trouble is that it missed its own scene.

Three pieces of evidence

The first. Forty-one per cent declares that fifty-nine are against; the score, at that very moment, produces a compact chorus with no internal dissent. Wagner does not have Rienzi elected: he has him acclaimed, and the first act culminates in the refusal of the crown. Plebiscitary legitimation is precisely the device the plot then dismantles, because what is acclaimed can be withdrawn in an afternoon. To replace it with a run-off is to replace a dramatic mechanism with an image that explains it less well. The eye reads a relative majority, the ear hears a plebiscite.

The second. In the finale of the pantomime the inner curtain closes shortly before the assassination attempt is launched. Wagner makes the two coincide: while the expulsion of the tyrant is being represented on stage, someone tries to kill the real tyrant, before the same crowd. By bringing the closure forward, the scenic apparatus — too large to be struck in the time the music allows — devours the effect: Orsini strikes without witnesses, and the tumult the score writes in the following bar is left without visible cause. It is damage that does not depend upon the reading, but upon the weight of the machine built to sustain it.

The third, and the most interesting, because the staging in part corrects it by itself. In the first two acts the production atomises what the score writes compact: residential cells, individual voting, telephones consulted incessantly, black masks that erase faces. But the chorus of Rienzi is a technology of unanimity, and the people of this work are not a sum of individuals: they are a single body that acclaims, then withdraws, then sets the fire. In the third act that body finally appears — and it appears in uniform: caps, dark jackets, high-visibility strips, a squad in place of a crowd. Exactly where Wagner writes one of his most massive choral pages, stage and pit go for once in the same direction. The price is the implicit diagnosis, which is worth acknowledging because it is harsh and coherent: this people can be compact only through obedience, never through acclamation.

What must be granted

It would be dishonest to stop here, because several choices that at first sight appear arbitrary are nothing of the kind.

The child who dies during the overture and reappears throughout the production is not an invented symbol: he is Cola's younger brother, whose murder historically founds the tribune's political vocation and figures in the libretto, when Adriano admits with shame that the killer belonged to his own house. To make of him a recurring figure — who rises again when peace comes, falls back when the peace goes wrong, and at the end climbs a staircase clutching a teddy bear — means endowing with a visual leitmotif a work composed before leitmotifs. It is the cleverest idea of the evening, and at the same time the clearest sign that Rienzi is being supplied with something Rienzi does not have.

The incestuous allusion between Rienzi and Irene, which likewise is not in the text, is argued by the directors as an anticipation of the Walküre constellation, and remained on stage in suspension, never becoming assertion — the more so since the third act opens with Irene and Adriano in bed together, that is, with the love story the libretto provides for brought fully into the light. The wall papered with text-only proclamations — Thomas More, paradise lost, the acts of the senate, the law on Vespasian's imperium — restores the authentic Cola far better than any videowall: the papal notary who governed by posting allegorical paintings on the façade of the Capitol and on a wall of Sant'Angelo in Pescheria, and who in the Lateran rediscovered the bronze tablet of the lex de imperio, set it at the centre of a fresco and read it aloud before the assembled magnates.

And then the two Gospel passages projected on the backdrop in the second act: Matthew's eschatological discourse, first in the King James version and then in the Vulgate. Above the scene of the pardon runs the warning about the many who will come in Christ's name and will lead many astray. It is a documentable reading, not an invention: from the end of 1348 to the middle of 1350 the historical Cola lived among the fraticelli of the Maiella, Spiritual Franciscans awaiting the reign of the Holy Spirit; in 1350 he presented to Charles IV of Bohemia the Oraculum Cyrilli, a prophecy of the Antichrist of Joachimite descent, and it cost him two years in prison; and on 1 August 1347, in the Lateran, he had assumed the title of candidatus Spiritus Sancti miles. Which is then the cry with which, in the third act, Wagner sends his tribune into battle: santo spirito cavaliere. The staging therefore imports no alien apocalypse; it makes explicit something the libretto already contains. The choice of Latin, in particular, is as exact as anything could be: it is that man's language.

It must be added, finally, that at the curtain call dissent was negligible. On a stage where directors are booed as a matter of course, this team came out to convinced applause. Whoever writes against this production writes in a minority, and must know it.

The subtraction

The finale remains, and it is there that the fracture becomes total.

In Wagner's fifth act the Capitol burns and drags down with it Rienzi and Irene, with Adriano hurling himself into the fire to reach them: it is the only gesture of love in the work, and it is a physical catastrophe. Here there is no fire. The building remains standing; what collapses is a metropolis of skyscrapers drawn on the videowall, which has nothing Roman about it. Adriano survives. Irene, who had cut her wrists, rises again. Rienzi cuts his palms with a knife and falls with a wound in his side at the top of the staircase — and there are those, in the hall, who saw in that ascent not a death but a dissolve towards an imaginary heaven, reading in it an allusion to the transfiguring effect this work had upon the admirer we all know. The two readings do not coincide.

Whatever the answer, the fact is that the catastrophe is quoted and denied. The emergency exits lit on both sides, the fire escapes, the whole apparatus of signals made ready for a blaze that never comes: the staging builds the expectation of an ending and then subtracts it. In terms of thesis one can understand it — a work whose reception passed that way cannot allow itself a liberating catastrophe, and whoever stages it may legitimately refuse catharsis. In terms of theatre the price is very heavy: the score performs an ending the stage contradicts, and this time the gap is not one of detail but structural.

There is more, and it comes from earlier. In the third act this Rienzi does not win a battle: he kills Orsini and Colonna himself, with a pistol, before running cameras. In Wagner the two nobles fall in combat. And if the final wounds too are self-inflicted — whoever cuts his own palms with a knife inflicts them — then the tribune shoots first and kills himself. In history Cola was lynched by the crowd; in Wagner he dies in the fire the people have set. In both cases others kill him. Here the people leave the stage with clean hands twice over, and it happens in the production that for five acts has been asking what politics makes of men.

Let it come back

All told, this evening demonstrated something no declaration of intent could have demonstrated: Rienzi in that hall sounds as it sounds nowhere else. The canon is not a testament, and the last verdict against it is that of an author by then very far removed from his own early work, who had every artistic reason and no philological one to disown it. Nine performances in one summer, with a redaction reconstructed for the purpose and a period of preparation no repertory theatre can afford, and then silence: it would be a waste hard to justify.

About the stage one may dissent as much as one likes, and I dissent. But the right place for a rarity is the theatre that plays it best, and that theatre we now know. In the second act, on the backdrop, Matthew's apocalypse was running, and inside it there was a line that stands as a prediction: all this must come to pass, but the end is not yet.