A danger when talking about writing for the stage is the pervasive idea that if a composer does something one way, it’s a manifesto on the way music for the stage should be written: a philosophical stance rather than a combination of skills, techniques, and decisions which, ideally, come from the idea of theatrical storytelling. Although writing for the stage harnesses the power of instrumental writing and non-dramatic vocal writing, it is, in many ways, a very different artistic and practical ecosystem. One kind of story might suggest a small, amplified ensemble and two singers, whereas another might require a grand opera house with a full chorus, orchestra, and live horses clopping around. Each project has its own rules, and telling one story one way should not preclude another story being told an entirely different way. Just because I have written three operas with linear narratives does not mean I won’t write an abstract one at some point; some of my favourite operas are ones where the plot is secondary or tertiary to the experience; the Philip Glass trilogy operas – Einstein on the Beach (1975–6), Satyagraha (1978–9), and Akhnaten (1983) – have plots only in the biggest, bird’s-eye sense, but that doesn’t mean that Glass hasn’t written purely narrative operas, like Appomattox (2007).
You will find that people who write about music want to generalise about composers’ intentions and processes; it’s understandable because of course detecting patterns is a way to predict trends and analyse the ‘state of things’. This is their job, but it is not composers’ jobs. If you open up the newspaper or, god forbid, the Internet, you will find a lot of chaotic shorthand: ‘Why write Grand Opera?’ Or ‘The future of opera is scrappy black-box productions’ or ‘operas which are workshopped feel corporate’ or ‘we’re seeing a resurgence of operas based on myths’ or ‘there are all these operas based on movies’ followed by the writers’ opinions about the same observations. Ignore all of this. Write the piece that makes sense for the company who commissioned it and, most importantly, for yourself as an artist at that time in your life. Writing a grand opera with six million people on stage doesn’t mean that you can’t then write a one-woman site-specific opera in a supermarket lit entirely by smartphones. This is the most crucial task: negotiating your own interests and obsessions as an artist with the restrictions and specific possibilities of the ‘situation’ of the commission (which includes everything from the size of the house to the types of voices to the budget and rehearsal time and so on). Composing for stage is a constant dance, and each situation is different; the process, in turn, will change, and you will change as an artist. Whilst you do not – and should not – have to worry constantly about each restriction, having a bird’s-eye view from the start will help enormously.
Any collaborative process requires a constant negotiation between its collaborators as to the hierarchy of creation: ‘does the dance come first and then the music or the other way around?’ is something one often hears when writing a dance. ‘Does the music come first and then the words?’ With an opera, for instance, the primary document for hundreds of people is the piano-vocal score. It can be a few hundred pages long, give or take, but it is the common language and origin story of all of the practical documents which then emerge: the costume-mistresses’ notebooks, the stage managers’ giant binder, the spotlight operator’s cue list. It is a document memorised by however many singers and covers there are, and a document whose shards are on the individual stands of the flute players in the pit, or the on-stage marching band. The score is so important, and so scrutinised, and yet very rarely seen as a whole. The flute player, for instance, won’t necessarily have studied your intricately worked-out plot, and so the storytelling has to be imbued, in some way, into the instrumental writing. While some directors work off the score, others work only from a libretto and others in combination. No matter what comes first, the score is, for better or for worse, the Bible.
Big Shapes, Tight Collaborations
The three larger operas I’ve written – Dark Sisters (2011), Two Boys (2011), and Marnie (2017) – were created in tight collaboration with their librettists (Stephen Karam, Craig Lucas, and Nicholas Wright) and directors (Rebecca Taichman, Bart Sher, and Michael Mayer respectively). The basic process was one of starting broad and zooming in together as a team. This is only a fraction of how it worked, but, in essence, once we knew the basic restrictions of the piece (fewer than a dozen people, a chamber orchestra), we came up with a single-sentence description of the plot of the piece: A woman born and raised in a polygamist environment decides to leave after the family’s children are taken away by the government. A young boy impersonates a series of different people online to seduce an older boy, who eventually stabs him. A woman steals from her bosses, changes identities, and gets caught, forcing her to confront the horrors of her childhood. From there, we zoomed in: what happens in each act? How tightly linked to a real story should this be? How do we meet all the other characters and when?
From there, Stephen started providing me little pieces of text in isolation, which worked because we all knew more or less the overall arch of the plot. That way, starting in the penultimate scene (as I did) didn’t feel inappropriate. In the case of Dark Sisters, the music director and conductor, Neal Goren, was, from the beginning, helping figure out precisely what voice type would belong to each character, which in turn helped me think about the kind of text that would set best for a coloratura versus a dramatic soprano, and so on. With Two Boys and Marnie, the Metropolitan Opera’s dramaturg, Paul Cremo, was always available as an extra set of eyes and ears and a guiding hand through the process in all its facets. All of this occurred before I wrote a note.
I’ve found that starting with a large shape helps keep me focused. If you’ve come to an agreement with your collaborators about the general thrust of the thing, smaller amendments and additions feel much easier when they’re happening on the fly. I always liken large-scale structure to an in-flight map. In its most zoomed-out form, you can all more or less agree that the flight is going from London to Singapore. Then, you can zoom in and see which countries you’ll fly over, and then zoom more and see major cities, and then zoom yet again and be above undreamt-of towns in Uzbekistan. If you and the librettist are both in agreement about the large flightpath, and keep that, in some way, in view at all times, it will be easier for you to each zoom in on more obscure areas of concern.
It is crucial to make sure your personal relationships with the artistic team remain under control. History is littered with fights between librettists and composers, or composers and directors, and trying to keep a constant sense of empathy and understanding is crucial in this process. I’ve scored a fair amount of film and am completely accustomed to random people giving me notes and suggestions. Whenever I write articles, I am used to getting them back covered in red ink. Other people are not accustomed to this or are indeed offended by it: if you find yourself setting up defences against what can seem like a fusillade of comments from singers, the director, people you have never met from the music department, and what can seem like people off the street, see if you can coordinate with your collaborators to calm the seas.
A good example of this early in the process could be if somebody points to a scene and says, ‘I don’t get what’s happening here.’ Any number of things could be causing this hang-up: is the text telling rather than showing? Are the text and the music too entwined, almost cartoonishly making a point? Is the text too abstract? Is the music cluttering the text? Or is it something which will only make sense when perceived visually, with blocking or a prop or a scene change? When you get a note like this, it’s easy to think it’s some existential problem with the libretto or the music. Often, though, these seemingly large issues (presuming you agree, in some way, with the note) can be resolved by delegation. ‘Let’s try a different word’ or ‘Let me see if I can have them deliver the line in the clear’ or ‘Maybe instead of the scene changing right after, we let the moment sit before the transition.’
An issue that arises later in the process is a famous one: ‘I can’t hear the singers over the orchestra.’ There are countless ways to solve this issue, and sometimes ‘play quieter’ and ‘sing louder’ are not the right solutions, either practically or artistically. Look into orchestrational details: the odd ways in which sometimes doubling a vocal line makes it pop out, and other times can have a muffling effect. See if there’s a way the director can move a singer farther downstage, or closer to a surface which will reflect the sound. You can also investigate diplomatically to see if an individual singer is saving their voice (known as ‘marking’) or has the habit of singing a bit quieter when getting used to blocking and staging.
It is not the composer’s job to create an easy production for a director. Some operas are director’s dreams, such as Glass’s Satyagraha, which, for instance, was a perfect playground for Phelim McDermott’s production at ENO and the Met. Other operas have very difficult staging issues built into the plot: what do you do about a dragon in Richard Wagner’s Siegfried (1857), or the ocean in Benjamin Britten’s Billy Budd (1951)? How do you depict real people saying invented things, like in Jake Heggie’s Dead Man Walking (2000)? How do you get a giant choral number to work if they’re all facing the side of the stage because of how the set works but need to see the conductor? I feel like these puzzles can draw the best out of a director: Claude Debussy’s Pelléas et Mélisande (1902) is such a tricky opera to get right because it takes place in a variety of different settings (a castle, a forest, a cave, etc.) but the music is so shimmering and quiet that having a giant cave-mouth wheeled onto stage would never work. I saw a production directed by Katie Mitchell that was so brilliant because she had screens framing different parts of the stage, and in the transitions, closed-off areas would open for the audience to realise that a set change had happened silently in secret. The solution itself became part of the structure of the production, and the whole thing held together because of it. Whilst it is fun to imagine what something could look like during the compositional process, it is good to keep in mind that directors and designers have spent as many years in the trenches of creative collaboration as you have and can often find a fiendishly clever and direct way to communicate the score through theatre.
On Musical Material
Wagner looms large over operatic composition; even people who’ve never heard a note of Wagner will recognise his techniques from even a minimal knowledge of how the music in Star Wars functions. Every opera requires some form of decision-making about whether musical material will ‘belong’ to a character or not, and it certainly isn’t composer-specific so much as work-specific, with a work like Britten’s The Turn of the Screw (1954) having character-specific instruments. I have never worked with leitmotifs, but mainly because I’m not sure I’d be very good at it. However, I try to create ‘space’: specific musical connections between the atmosphere in a room where something is happening. Instead of Marnie having a theme that follows her around, she has an oboe who functions as an interlocutor to her music. Her childhood home has a unique sound world, and her mother is shadowed by a viola, but as a timbre rather than as a vessel for thematic information. Two Boys functioned, in some ways, like science fiction in which there is an initial amount of world-building that has to happen; Craig Lucas set this up brilliantly in the libretto, where there’s a clear ‘offline’ style and a clear ‘online’ style, which eventually come together at the very end.
Visually, it is a simple task, albeit also a complicated one where the designers have to essentially invent what a chatroom looks like on a stage. Musically, what does that look like in terms of harmonic language, in terms of orchestration? The version I avoided was where ‘online’ was all synthesisers, but it took me a while to land on something fruitful. I was thinking about the last page of Britten’s Death in Venice (1973), when we have the protagonist Gustav von Aschenbach looking out to sea, and the music turns into this kind of highly ‘Brittenised’ Balinese gamelan music, reminiscent of Tadzio’s music but somehow romanticised and infinite. For me, imbuing our story with Britten’s sense of how you would treat the world beyond, of his own memories of his trips to Bali, felt like the right way to imagine the drug of talking to strangers online, of projecting yourself from one space (in this case, a teenager’s bedroom) into another much more unknown, alluring, and dangerous world.
Some of the best moments in operas are, for me, when you have an ecstatic, climactic meeting point between text, music, and direction. Of course, these three things should always be working in counterpoint, but occasionally a giant unison can have a huge emotional effect. The moment when Miles dies in The Turn of the Screw and the Governess then sings his song (‘Malo’) is one such moment, where you have the indelible image of her in a kind of Pietà pose, holding a dead child; a musical callback to a previous theme but orchestrated grandly; and a textual reference to something which had been obscure before, but is now revealed and amplified. A more subtle moment is to be found in Die Walküre (1856), where Siegmund pulls Nothung (the mythical sword) from a tree, which of course is visually thrilling no matter who’s directing, and a nut Wagner has made us wait rather a long time to crack: only then do we realise that this music connects us all the way back to Alberich and ‘fulfils’ the operatic prophecy. These are amazing moments, which stick in the mind and the eye and the ear.
You cannot have too many of these things, however, or the work starts feeling overbearing, cluttered, and bombastic. Moments where we’re seeing something we’re hearing in the music and the text all at the same time can be dramatically unhelpful, particularly in plot exposition, when it’s much more interesting if the oboe knows that the tenor is cheating on his wife rather than him taking a moment to tell us the same. Having something firm requires other things to be supple. Sometimes, teamwork is about all doing different things to get one big thing completed: think about a restaurant kitchen in which a single dish requires four people doing four different things; it would not flow if four people each did a little bit of the same thing all at the same time.
One final point on material: in stage music, the physical stage is itself a miracle and an obstacle to crafting the piece. Impressions of depth (an off-stage chorus, or a villain way upstage, muttering to himself in Italian and watching our heroine undress) are also practical and musical challenges for the singers: if you’ve worked out an elaborately detailed, rhythmic vocal line which sounds so fantastic in the rehearsal room, wait until the first rehearsal, when you have the same singer miles away, on a raked stage, in stilettos, walking backwards with a spotlight tracing her every move and a scene change happening with video projections all over the floor. This does not mean ‘don’t write anything complicated’ so much as if you ever have the opportunity, do a walk-through on the stage during a technical cue-by-cue, when you can see how stage-space itself becomes a dramatic partner in the production and bear this in mind in your writing. Many of these concerns are the director’s responsibility, of course, but it doesn’t hurt to bear these things in mind, no matter how subliminally.
Working with Singers
The first time that singers are likely to encounter and inhabit your work is at a workshop. A key question in the process is whether a workshop can happen, and, if so, is it necessary and what kind of workshop will it be. There are some which are four days of people just learning the music and then singing it from the start to the finish without pause, and some that get to basic blocking and a more detailed sense of physical space – sets, props, rakes, and so on. Some composers find workshops a bizarre process where essentially the creative team’s work stops, and a bunch of people from the opera company are suddenly watching the show in a completely unrealised way. I have found these workshops or sing-throughs or whatever you want to call them useful for the obvious practical reasons of ‘this note sounds crazy on that vowel’, but also for the – absolutely necessary – process of seeing if the piece ‘works’ in real time. When I am writing, it is very difficult for me to check my work solely as a function of time passing, and in a sing-through, the director, composer, librettist, dramaturg, and singers can get a real sense of flow, of architecture, and of breath, with the idea that there is enough time to fix it before the wheels of the final production start churning too vigorously. I would recommend some version of this to any composer, particularly if a goal is to experience the piece in real time, in front of a few other people besides the creative team. Even if it feels like somehow there is a committee judging you, you cannot put a price on experiencing the piece as a single gesture.
Sometimes you might want to start reimagining the performance. I have learned that the use of vibrato is a meta-conversation worth having maybe one time per project. For a variety of reasons, individual singers have different relationships to singing without vibrato on stage. Some are happy to do it provided the space isn’t too big. Others basically will not do it; it’s not been part of their training and they’re not going to stand in the middle of the stage and do something uncomfortable. Some will do it but not on purpose, and others are obsessed with doing it and cannot wait to be asked. Singing without vibrato really is a special effect in most pedagogical contexts, and being asked to do it at length on stage is something so outside of the embrace of most singers’ experience that it’s a giant ask, and not always a welcome one. I find that it’s alright to ask one time, but not to press it too far. For singers who are uncomfortable doing it, their professional and artistic desire to follow the score and the composer’s directions comes in direct conflict with their own instrument, and this is not an energy you want in the rehearsal room. My recommendation, if you find yourself really wanting this sound in your life, is to think about using amplified voices (as in Steve Reich’s Tehillim (1981)) or writing for groups who specialise in early music. The one thing I have had success with regarding vibrato is encouraging singers to treat it as one of many elements of their expressive toolkit. I’ve met many young singers whose dynamic range, subtlety with diction, and rhythmic flexibility goes from 1 to 100, but whose deployment of vibrato of any kind hovers in the 40–60 range.
During the learning process, synthesised demo tracks are helpful ways to support singers. Not only do I make them, but I try to be extra accommodating. If a singer wants music-minus-one (which is to say, the score with their part left out), I’ll make it. If they want the score but with their part banged out on a piano, I’ll make it. Not all singers have the luxury of a repetiteur to hand at all times, so anything I can do to make their lives easier is helpful. However, there is a great danger in artificial versions because everything is always in focus. This is more of a problem in dance, but it’s easy to learn where to get your note from (‘oh okay it’s the same as the bassoon’); on stage, however, you won’t hear that bassoon at all. Another danger of MIDI is the flexibility of tempo that makes live performance so delicious; this is also a bigger problem in dance (a dancer once told me ‘you can’t go ANY faster than this because I can only fall down so quickly – you can’t conduct gravity!’), but if the conductor decides to luxuriate in a tempo when the MIDI has been tick-tock rigid, it can create a dramatic disconnect.
When the singers are off-book (i.e. have their parts memorised), you need to stop making structural changes unless it is vital. In Marnie, we cut a four-minute aria at the end, where it was so clear what was happening from the staging that having her stop and sing about it shut the entire flow of the opera down. I would never start cutting dramatic beats once they’ve been memorised. Then another thing I do (and this is a slightly luxurious thing to be able to do) is that when the singers are off-book, I go off-book. I don’t open a score in rehearsals once the singers have stopped looking at it, because I think it puts them at ease a little bit about the difference between doing it Right with a capital R (i.e. note perfect), and doing it right, as in communicating properly with the audience. Once the singers have learned it, they have internalised it in a far more intimate way than I have. I will almost never tell somebody they sang a wrong note, because they either already know or will get a note from the conductor or repetiteur. As the author of the thing, a statement from me has an inappropriate power in the vulnerable rehearsal process.
Putting the Work ‘On Its Feet’
To me, the biggest question at any point is ‘whose rehearsal is this?’; and it’s the hardest thing to learn. Whose rehearsal is it and for whom is it? Whom is it helping? You realise that there’s a technical truth, which is ‘this is the conductor’s rehearsal. This is the director’s rehearsal,’ but then there are also many sub-strata of who needs to get what out of the rehearsal. In general, music-only rehearsals belong to the conductor. Staging rehearsals belong to the director. Once you get on stage, it gets a little bit more complicated, and suddenly everybody has different needs. I, for instance, am always most nervous about structure. I am therefore much more interested in running sequences: let’s run three scenes in a row. Of course, the singers think, ‘oh no, I got that wrong,’ and they stop; or the director says, ‘oh no, that’s not going to work’; or the conductor says, ‘no, we need to go faster’. None of that is helpful if you’re worried about structure, but it’s incredibly helpful for the singers and the director. There are a lot of divergent agendas, and that’s something that you have to be mindful of. The composer doesn’t necessarily get a rehearsal. There’s just not the time for that, which is why you have to use every rehearsal as yours, but quietly, and think about exactly when the right moment is to say, ‘excuse me, Maestro, do you mind just running straight through this transition without stopping?’ If you hear something strange and you are thinking, ‘she will never be able to get that note’ or ‘I need to add a note to the bass clarinet’, remind yourself afterwards instead of asking the director or conductor to stop. Little things like that need to happen in the background.
There must be a latticework of tasks that everyone is doing, even if the actual business of the rehearsal is running at variance to your desires. Some, if not most, issues fix themselves. Chances are, the video team knows that they’ve missed a cue, or the flute player knows that it’s a C♯ but just misread it because the lights were lower in the pit than in the rehearsal room. Light in the pit is a sidebar which could take up many pages; I inadvertently found myself in a situation where the musicians’ union was at loggerheads with the projectionists because of the divergent needs of being able to see the notes and being able to see the projections given the bleed from the pit to the stage. We had a half hour in which the pit lights were adjusted by single percentages until everybody was not so much happy as less unhappy.
A practical necessity but a missed opportunity, I feel, is that the creators of a stage work tend to sit in the expensive seats to work on a piece: the ones with the best views and cleanest acoustics. In a grand opera house, it is worth a quick hike up to the gods to see if you can actually hear that vowel on that note from that far away. The singers have spent months on that note, and years preparing, and they can’t tell how things are landing in the hall, which can create a sense of unwelcome uncertainty. Always bear in mind that even the conductor can’t always hear how well things are balanced. As the composer, unless you’re conducting or performing, you don’t have much to do (aside from buying everybody granola bars), so running around the space to reassure everybody that it sounds great is a positive way to engage. Remember that everybody in the house, from the lighting spot operators to the electricians to the stage-managers, is in a constant state of compromise and motion. The wigs-mistress might have to change a beautiful wig last-minute because the clip-on microphone won’t properly adhere. You might need to totally get rid of that bass drum because even though it sounded great in the sitzprobe, it is dominating what should be a subtle moment. The video team might need to completely re-write a cue on the fly, if a projection surface can’t slide into place quickly enough.
One of the scariest moments in my life was during a stage and piano rehearsal which was the first time I’d seen Two Boys on stage. I was worried about everything, of course, but most specifically: ‘does this thing work?’. The point of the rehearsal was for the director to sketch things in on stage, so there was endless stopping and starting, and over the course of a day, we never ran a single transition. For him, this was totally normal and productive and useful, and the singers were delighted to be able to work the same passages over and over; meanwhile, I found myself sitting in the audience, biting my fingernails and debating the merits of a 10 a.m. Bloody Mary. Learning to let go of that tension, and to trust the slow and collaborative nature of this stage of the process felt like a milestone in my development as an artist.
Another oddity is that the only time you will see your piece from start to finish in real time with an audience is opening night, or maybe the dress rehearsal if it’s an open dress. Some scenic and costume elements are added at the very last minute, so sometimes it really is the case that you never really see it until it’s happening right in front of you and everybody you know. It’s terrifying, but there is nothing in the world like the moment 10 minutes before curtain when you realise how many people it’s taken to get to this one single point, and the one single moment when music and theatre come together in a necessarily imperfect but beautiful dance.
