Tosca And Tradition

In an interview last month, David McVicar, the director of the Metropolitan Opera's new production of Tosca, talked about how, when talking about productions of operas, people confuse "traditional" with "period."

Peter Gelb took over as the Met's general manager in 2006, and since then, most of the company's standard repertory has been replaced with new productions, most of which in turn could be said to be "non-traditional." In general, productions can be non-traditional in one of two ways. They can either be set in a time or place apart from what the libretto specifies, or the choices made could be more artistic, with unrealistic sets and stylized costumes and whatnot.

It's easy to spot a non-period production simply by looking for anachronisms. And the Met's had their share, from Michael Mayer's infamous Vegas-set Rigoletto, to ones that more or less flew under the radar, like Deborah Warner's Eugene Onegin, or Richard Eyre's Werther, both moved to take place around the time they were written

Stylized productions are a bit harder to pin down. Some are obvious, such as Julie Taymor's Die Zauberflote, which makes heavy use of unreal puppets and magical stage effects, or William Kentridge's projection-heavy productions of The Nose and Lulu. But all of these are magical or unreal operas to begin with, and so treating them with realism isn't expected. On the other hand, Willy Decker's minimalist La Traviata has gotten much praise and criticism, and several new productions, such as Dmitri Tcherniakov's Prince Igor or Mary Zimmerman's Armida, heavily abstract the situation, not clearly taking place in any particular time period.

And, of course, some productions do both. In McVicar's Cav/Pag, Cavalleria Rusticana is heavily abstracted, while Pagliacci is shifted temporally. Zimmerman's Lucia Di Lammermoor not only moves the action up several centuries to about the late 1800s, but also features ghosts on stage, a mansion suggested only by a large staircase and a balcony, and a monstrously-sized moon looming ominously in the background. (Apparently visible through the wall of the mansion?)

Of course "stylized" is a weird word to use. Every theatrical production is, by necessity, at least a little stylized. Even in McVicar's new Tosca, the angles on the Castel Sant'Angelo in the final act aren't a proper ninety degrees, and the stage is on a weird slope. This makes it look nice to the audience, but could hardly be called "realistic." But let's say, for lack of a clear delineator, that a "traditional" production of an opera must be set in-period, and must not be more stylized than suits the nature of the opera. (Meaning Die Zauberflote more or less gets a pass.)

Since so many of these non-traditional productions have been met with criticism or controversy, surely, five years into Gelb's tenure, he must have realized he needed a traditional production to win back the fans. And so in 2011, Michael Grandage directed a new production of Don Giovanni at the Met, that was in-period, and basically realistic (excepting, of course, the Commendatore scene).

So, of course, it was raked accross the coals in the reviews, described as "dull," "unimaginitive," and, perhaps most scathingly in the New York Times review, "almost makes you yearn for those new stagings where the creative team is booed on opening night." The basic consensus is that it was too safe. Too bland. There wasn't any vibrancy in the production itself, nor the spectacle you expect from Don Giovanni. Fortunately for this production, Don Giovanni is an opera that sells itself on its own merits, and even with the drabbest production, a good cast can make the opera shine.

This was followed by Bartlett Sher's production of L'Elisir D'Amore, constructed with his usual Broadway team of Michael Yeargan and Catherine Zuber. This was a period-accurate and highly realistic production of a much-beloved classic, but unlike Grandage's Don Giovanni, it was well-received from its first premiere. Why did L'Elisir shine while Giovanni fell flat? I think the answer might have less to do with the productions themselves and more to do with the audience's expectations. Don Giovanni is a highly active opera, with a lot of stuff happening at any given moment, all leading up to an iconic, intense, climactic scene at the very end. L'Elisir D'Amore, by contrast, has a fairly subdued plot, mostly just dealing with the personal quibblings of a small handful of people, and there is no great spectacle to be had. Grandage's production of Giovanni failed to deliver on the expected spectacle. Sher's L'Elisir absolutely delivered above and beyond the lack of it. There is also the point that Grandage was replacing a much-beloved Zefirreli production (which we will come to learn is sacriledge), whereas the previous Copley production of L'Elisir was not so widely liked, being basically functional and vaguely cartoonish. (Addendum: It was well-received when it premiered in 1991; it did not age that well.)

So when Susan Stroman (another Broadway director) was brought in for a new production of The Merry Widow, which had no iconic production to replace, and she set it in-period, staged it realistically, and added the same verve that won her five Tonys, it must have been the toast of the season, right? Wrong. Critics described it as "overproduced," and "too Broadway."

I... I just don't know anymore. Do you want spectacle or not, opera-going audinece? Make up your mind!

(Within the same season, Paul Curran's production of La Donna Del Lago met a similar fate to Grandage's Giovanni, though slightly less scathing, as Curran was not replacing any production at all, and the opera itself wasn't nearly as iconic. Still, it was considered by some to be too safe and bland.)

Obviously it's more complicated than that. Each critic has different tastes, and with the internet, everyone is a critic. But in general, this is how the opera world reacts to new productions. We all want traditional productions, but can't agree on what "traditional" means. When there is an unambiguously traditional production, we hate it. Unless we don't. There's really no pleasing us, to the point where it's practically tradition at the Met to boo the creative team of a new production on opening night.

This brings us to Tosca. In 2009, the Met replaced its old production, a highly acclaimed one by Franco Zeffirelli, with a new production by Luc Bondy. The Guardian review does a good job explaining where they went wrong. It was when they decided to replace the Zeffirelli production. "[Zeffirelli's] Tosca... combines a literalist take on Puccini's score with the composition of a Renaissance painting. He dealt with the central problem any director faces at the Met – the theatre's vast Proscenium arch – by filling it up with detail. So the inside of the Church of Sant'Andrea della Valle in act one could have been imported brick by brick from Rome, so faithfully was it reproduced."

The problem with the new production, therefore, is that it couldn't hope to meet the ridiculously high bar set by the production that preceded it. Robert Carsen's new Der Rosenkavalier last year faced a similar issue. It's not that Carsen's production was bad per se (although there were some seriously questionably choices made in the third act), it's just that Nathaniel Merril's production was so much better (and made no such questionable choices in any of the acts.)

Perhaps knowing that he could never beat the spectacle of Zeffirelli's iconic production, Bondy took the opposite tack. Where Zeffirelli had large-scale paintings of full buildings, Bondy had close-ups of walls in small rooms. It was, by both of our earlier measures, traditional. Both in-period, and basically realistic. But it was, in the words of some critics, austere.

But many productions could be described as austere, and many of them do quite well. David McVicar's aforementioned Cavalleria, for instance, also replaced a Zeffirelli production, and was also highly austere. Now, this new production did attract a lot of criticism, but unlike the more generalized criticism of Bondy's Tosca, which pointed out things the reviewers didn't like, but generally failed to articulate why they didn't like them, aside from taste, the reviews of this Cavalleria tended to better identify what went wrong. The most commonly called-out culprit was McVicar's use of the turntable, which many reviewers agreed was too much, often being distracting, causing the opera to lose focus and motivation. That is a specific criticism of a specific thing in the production that didn't work, and an explanation of why that was the case. David McVicar was lucky to get such criticism. Luc Bondy did not get such a courtesy.

So let's dig into it, one element at a time, starting with the most frequently complained-about element, the set. If there's one thing Franco Zeffirelli's productions are famous for, its the spectacular and intricately detailed scenery, which Bondy tossed out in favor of comparative minimalism. Starting with Act I, here are the sets of Zeffirelli's production, Bondy's, and McVicar's. The location is the Sant'Andrea Della Valle, in Rome.

Zeffirelli's production is obviously visually stunning, and a thrill to look at in a museum. If it has a fault though, it is a fault shared by a number of Zeffirelli stagings, and that is that it is flat and stagnant. The background is simply a drop, and though a beautiful one, it is obviously so. The set is clearly a set, and will enthrall an audience, but not necessarily immerse them. Of course, like Don Giovanni, Tosca is an opera that will do most of the work for you, provided you supply an adequate space for it to do so.

Bondy's production is a clear contrast. Like Zeffirelli's, it is obviously a set (designed by Richard Peduzzi). It is non-immersive, but unlike Zeffirelli's, it doesn't really invite the audience to look on either. It has big flat walls with little to draw attention, and it looks awkward next to the surprisingly detailed costumes. (The costumes in Bondy's production I think are actually more colorful than the ones in McVicar's, although McVicar's production wins out by virtue of giving Tosca a particularly sparkly dress to wear in the second act.) Bondy's production is also dimly lit, which is an effect that can be pulled off, but when overused, especially in a house the size of the Met, can easily put an audience to sleep. (The dim lighting was another common complaint with McVicar's Cavalleria.)

Now we come to McVicar's produciton, with sets by John Macfarlane. And if there's one thing John Macfarlane's resume (which includes the Met productions of Hansel Und Gretel and Maria Stuarda) will demonstrate, it's that he loves perspective. (And also vaguely unsettling drops to be used instead of curtains.) And sure enough, Macfarlane's set here makes fantastic use of perspective, giving a great sense of depth to the stage, which, among other things, really serves to pull the audience in. Theater is a three-dimensional art form, and Macfarlane and McVicar know how to use all three.

In an interview during the Met's Live In HD broadcast of this production, Macfarlane mentioned how, although he modeled his sets closely after the real buildings in Rome in which Tosca is set, theatrical space works differently from real space, and so sometimes accomodations have to be made. In the first act, Macfarlane takes an odd sort of angle, and one which allows McVicar to do something brilliant and truly spectacular. On the left side of the picture, you can see two columns, with a passage behind them. If you extend the line of columns out, the next one would coincide with the proscenium. The use of this specific angle of the church is used to great effect in the Te Deum that ends Act I.

During the Te Deum, spectacle is usually achieved by having the chorus gradually fill on while Scarpia's singing, so that the stage is completely full of people for the big choral finale. Both Zeffirelli and Bondy had this. But McVicar holds off on filling the stage until the end. Instead, he has people walk down that passageway toward the audience, and disappear behind the proscenium. The illusion is one of an infinite procession of people walking straight toward the audience. It's like a cheesy effect in a 3-D movie, except, you know, actually 3-D. The church seems to extend into the house, even with nothing built beyond the proscenium. Toward the end, later than usual, the chorus does fill onto the stage, almost as though they've filled all the space in the invisible part of the church that overlaps with the audience, and now they're overflowing onto the stage proper. Now that's spectacle.

The problem with Bondy's Act I is therefore that, while it may provide a technically accurate and functional space, it does not provide a space well-suited to the demands of the opera. And that has nothing to do with austerity.

I couldn't find a good picture of the second act of Zeffirelli's production, but all three are basically funcitonally the same. A room in which some things happen. Peduzzi's room is starker, and Macfarlane's is at a steeper angle. Zeffirelli's room was presented head-on, with the back wall flat to the audience, with a good deal of surface area for intricate detail. Of Tosca's three acts, the second is the most action-filled, and does most of the work for a lackluster production. I don't know that McVicar made any choices that stood out from Zeffirelli, the main difference simply being in the angle of the set, which is a matter of preference. I prefer the more askew angle, because it feels more active to me, while Zeffirelli's set feels more stagnant. Peduzzi's set for the Bondy production is obviously the least intricate of the three, but putting the set aside, I will come back to Act II when I talk about directorial choices.

Once again, we can see much the same trends. Zeffirelli's production is detailed, but flat, and from a head-on angle, the levels kind of wash out. Bondy's production is tall, but sparse, and somehow overbearing. (Reviews tended to agree that the third act of this production was the least problematic.) Thirdly, we once again have Macfarlane's trademark perspective, which provides a good deal of dimension. From the raked angle of the platform to the somehow inverted angle of the statue, the audience is somehow both looking down on, and up at, the Castel Sant'Angelo.

Curiously, Zeffirelli's production also supplied a good deal of scale and dimension, although you cannot see it here. Unlike almost every other production, where the tenor playing Cavaradossi is brought up to the platform to sing "E Lucevan Le Stelle," Zeffirelli took advantage of the Met's gigantic stage elevator, to lift the platform, revealing an entire second level of set hidden beneath the stage, where Cavaradossi is being kept, and where he sings his aria, after which the elevator lowers the platform back to stage level. This is the sort of spectacle which, by necessity, you can only get at a large-scale opera house, because a smaller one wouldn't have the technical capability to achieve it.

And so once again, we see that Bondy's production is functional, but lacks spectacle. Which is more or less what we'd already agreed before looking at all the sets side by side. But as we've also seen (and I hope you'll forgive me for repeating myself), austerity alone does not a poor production make. McVicar's Cavalleria and Richard Eyre's Carmen both replaced visually stunning Zeffirelli productions, and both received criticism, but neither got the same amount of scorn as Bondy's Tosca, or dismissive indifference as Grandage's Don Giovanni. Although, notably, reviewers tended to agree that Zeffirelli's Carmen, though pretty, was dramatically stagnant. Richard Eyre's production, in contrast, was generally agreed to be exciting and action-packed, bringing out the darker elements and violence in the opera that Zeffirelli was too tasteful to emphasize. McVicar's Cavalleria too added something new that Zeffirelli's production did not have. A new angle, if you'll allow for the pun on the turntable. Where Zeffirelli's production was realistic, and made the most of the pretty Sicillian setting, McVicar's darker production was more character-focused, putting an almost oppressive scrutiny on Santuzza, who was kept on stage the whole time, surrounded by a ring of chairs.

In this way, it's easy to see how Grandage's Don Giovanni failed. It got rid of the beautiful sets, but didn't provide anything new to compensate. During the tenure of this production, all the new excitement has been provided by the cast, which has included at various points a dashing, roguish Don played by Mariusz Kweicen, the more suave and effortless Don of Peter Mattei, and the frightening and powerful Don of Simon Keenlyside. The Met has also cast a variety of accomplished Leporellos, Annas, Elviras, and others in this production, each of whom bring something new, because Don Giovanni is an opera that is carried by the cast. They can make up for a bland (though functional) production, but when the option of a great cast and a great production together was on the table, getting rid of the latter is bound to leave a sour taste.

This explains the lashback against Grandage's Don Giovanni, but does not in itself satisfactorally justify the extreme hostility directed at Bondy's Tosca. It must be more than that. Grandage showed us what happens when you present the same opera with different trimmings. Bondy must have made some sort of active misstep. And to see what that was, we have to look back to Act II. (Well, we don't have to, but I choose to because it provides a clear example of what I'm going to be getting at.)

The climactic moment of Tosca's second act (arguably the climactic moment of the entire opera) occurs near the very end, when Tosca stabs Scarpia to death. One of the most effective ways to provide excitement in opera is with musical contrast, and Puccini does that admirably, framing the murder on the one side with the showstoppingly beautiful aria "Vissi D'Arte," and on the other about a minute of orchestral interlude. It's in this orchestral passage that something interesting happens. The review I quoted earlier said that a problem directors face at the Met is how to fill the enormous stage. A problem directors face with a great many operas is how to fill stretches of music which, if not treated right, can seem unmotivated.

During these sixty seconds of orchestral music, here is what Zeffirelli has happen: Tosca washes her hands with a pitcher of water on the table. She then goes over to Scarpia's desk to take the letter of safe passage she had him write. She rummages through his papers, before realizing that Scarpia was holding the letter when he died. She goes over to the dead body and pries it from his hand.

In this same space, Bondy has something quite different happen. Tosca, without looking, wipes her hands on the (red) couch. She then goes over to the window, where she stands and looks out as if to jump. (Which you can see in the picture above.) Finally, she takes the letter from Scarpia's hand, as before.

The text is the same. The music is the same. The major event of the act (Tosca murdering Scarpia) is intact. But these two sets of actions, inconsequential to the plot of the opera, do very different things to the character of Tosca. For Zeffirelli, Tosca is a pious character. The opera has major religious overtones, and the imagery of Tosca washing her hands alludes back to the holy water explicitly mentioned in the libretto in Act I. When Tosca rifles through Scarpia's papers on his desk, it doesn't make much sense to the casual onlooker. Scarpia literally just wrote the letter. If he wasn't holding it, it would be the top item on his desk. Tosca making a production out of looking for it where it isn't shows her reluctance to approach the dead body again. She is horrified at having just killed a man, and doesn't want to acknowledge it by prying the paper from his hand. The fact too that, later, Tosca places candles by Scarpia's body and a crucifix on his chest, shows that Tosca can find it in herself to show more remorse for Scarpia's death than the audience ever could. The audience doesn't have any remorse for Scarpia's death. He is a thoroughly reprehensible character, and we specifically came to the opera house to see him get murdered in a thrilling act of theater. These actions that Tosca takes after the murder in Zeffirelli's production make her a pure and good character, who has been driven by circumstance to horrifying extremes.

Bondy's Tosca is quite different. She does not wash her hands in water, but simply wipes them clean on the sofa. (And, religious imagery aside, this really isn't an effective way to wash your hands.) She then spends a good thirty seconds or so at the window, apparently contemplating suicide. When she takes the paper from Scarpia's hand, she shows no remorse. And when she stands at the window, although I'm sure Bondy thought he was giving a clever bit of foreshadowing, this only serves to weaken the character. In the opera as written by Puccini, Illica, and Giacosa, Tosca does not jump off the parapet until she is completely out of options. After she kills Scarpia, her plan is to save Mario and escape the country. She only kills herself when she discovers that Mario has in fact been killed, and Scarpia's body has been discovered. She's trapped on the roof of the building, and the guards are coming to arrest her. She doesn't have a lot of options, so she decides in a split second to go out on her own terms. But what Bondy does is have Tosca consider abandoning her goal and the entire third act. He has her become indecisive at the very moment she wields the most power of anyone in the opera. This is not just presenting the same opera with new trimmings. This is a theatrical curveball for people who know and love Tosca, and it is one that doesn't work. I don't know how many audience members took active note of what was different and what implications that had, but I'm sure they noticed, even if they didn't know it at the time.

As for McVicar, well, he does pretty much the exact same thing in these sixty seconds as Zeffirelli. So does Gobbi. In fact, the vast majority of Tosca productions have Tosca, after stabbing Scarpia, wash her hands with water (but no soap), flipping through his papers, and ultimately prying the paper from his dead hands. Finally, she places candles by his body and a crucifix on him before leaving. In this way, Bondy's Tosca may be the most blatantly "non-traditional" production we've discussed here, insofar as it blatantly and explicitly breaks from the tradition of Tosca doing these things in the final minutes of Act II.

But McVicar wasn't copying Zeffirelli. None of these people were copying each other. The reason so many productions have Tosca take this exact sequence of events at this point is quite simple: That's what the libretto instructs. Those stage directions are spelled out. Puccini wasn't just writing sixty seconds of filler music to let the audience process what just happened. He was writing sixty seconds of music intended to accompany this specific sequence of events.

Another buzzphrase people like to use when complaining about nontraditional productions is "respecting the composers intentions." Usually, this is near "stylized" on the scale of useless criteria. How can we know the composer's intent? Maybe we could take it to mean "in-period," but Baroque operas set in Ancient Rome usually weren't staged with authentic Roman garb and realistic scenery portraying the forum. When Verdi wrote La Traviata, it was contemporary. It could be argued that the best way to honor the original intent of the opera would be to continue to set it in the present day, whatever that happens to be at the time. Of course, Verdi might have balked at that. We have no way of knowing. And I'm sure plenty of opera composers couldn't care less about how their operas would be staged hundreds of years after their deaths. Some of them probably wouldn't have expected their operas to last that long anyway. Especially those from the days when opera was a mass-produced form of popular entertainment, and composers would churn out three or four in a year. Even going so far as to make cuts or alterations to the score might not be complete blasphemy depending on the opera in question.

But Tosca, with its unusually high level of specificity, makes this surprisingly easy. It is set on a specific date of a specific year (and not an arbitrary date picked by the librettist; it's historically significant), and each of the three acts occurs in a real and famous building in Rome. The libretto comes with a lot of stage directions, which Puccini has accomodated for in the music. And this is why so many productions of Tosca look so similar. It works that way. It was written to work that way. That's not to say it couldn't work any other way, but it does make it a more difficult opera to mess with than others. And ultimately, that felt like the biggest problem with Bondy's production. It seems that, early in his time as general manager, Peter Gelb wanted to bring in new productions of the classics just for the sake of doing something new. And so the impetus for Bondy was to create a Tosca that was different from Zeffirelli's for no other reason than to make it different. This is not a theatrical motivation, and it did not play well on stage. Now, David McVicar was given free reign to bring back all the elements of Zeffirelli's production that worked (read: the elements that were in the script) while also giving it his own personal style.

Of course, if McVicar's production had followed Zeffirelli's directly, it might have been met similarly to Grandage's Don Giovanni. Sure, it works, but so did its predecessor. And so the question is why bother? McVicar's Cavalleria at least provided something new that the Zeffirelli didn't. I'm not sure that his Tosca does. After all, Tosca itself does most of the work. And when you have a production that works for an opera, unless you feel you can give it something new, it does seem kind of pointless to go through the effort. At the same time, a new production can give vivacity to a work that's fallen into a routine. And short term, a new production almost certainly gives a boost to profits, even if, in retrospect, audience members might have no particularly strong feelings for one traditional production over another. I'm sure the Met has all sorts of complicated financial models telling them when a production will peak and when to replace it to maximise ticket sales. That's not my area.

Which I guess brings us to an almost stupidly obvious conclusion. When taking on an opera to direct, follow the script. There's a reason these works endure in the form they're already in. That's not to say you can't put your own spin on it, but make sure it's motivated by something other than a desire to be different for difference's sake, and think it through. Some operas provide more creative leeway than others. Tosca is pretty strict, and David McVicar is a director who undoubtedly has utmost respect for the composer's intent. For proof, you need look no further than his Met debut, a generally-liked production of Il Trovatore, an opera he openly dislikes. It is not in-period, but McVicar's goal was not to bend the opera into something he liked, but to present what he felt Verdi wanted in the best possible way. The result is a fast-paced, aesthetically pleasing Il Trovatore, that one would happily call "traditional" at a glance, even though there are a million things separating it from the most literal interpretation possible of the opera. It works because McVicar read the script, and thought, not about how to make the opera fit his vision or taste, but how to make the opera work. Bondy's alterations were motivated by a desire to contrast Zeffirelli. McVicar's were motivated by a desire to make the opera work.

As far as a review goes, McVicar's new (and by that I mean old) Tosca is thrillingly done, and I'm sure will be engaging audiences at the Met for years to come. And, I mean, of course it will. It's Tosca for crying out loud.