Adrian Daub | Longreads | September 2021 | 21 minutes (5,894 words)

When a new trailer for the Marvel film Black Widow dropped in April of this year — after the movie had been repeatedly moved back due to the pandemic — the producers seemed intent on reminding people about why they’d been excited about the movie before the lockdowns started. They did so by closing the promo with a new version of the theme from The Avengers, probably to call back viewers to a different, less socially distanced time. How could you know this was a new version of the motif? It was choral, but that was a well Marvel had gone to before. This time it had lyrics. As best I can tell, for the first time.

As fans welcomed the callback in online comments, I was brought back to a question that I’d had when Game of Thrones did something similar at the end of its fourth season and again at the very end of the show. It’s something of a trend these days to take a highly recognizable instrumental theme and make it choral. And I get why: The gesture is big and bold and epic. But my question concerned something comparatively pedestrian: Who decides what the lyrics are? What language are they even in? And who writes them? I decided to find out.

Those of us who listen to soundtracks obsessively do so knowing that that’s not how soundtracks are intended to work on us. Whoever mixed in a chorus for a few seconds of the Black Widow trailer was going for an emotional reaction, not some new layer of meaning to be disentangled. “When I do a film score,” the late James Horner said in a TED talk in 2005, “I am nothing more than a fancy pencil” executing the vision of a filmmaker. You’re not meant to listen to a soundtrack in isolation from the image. It is music in service of the moment.

You’re not meant to listen to a soundtrack in isolation from the image. It is music in service of the moment.

But one place where this fancy pencil has more autonomy is when it comes to the text that a chorus sings. Perhaps it’s better to say that the pencil is condemned to freedom. When the composer John Ottman was hired to score the 2008 Tom Cruise film Valkyrie, he realized that he needed a break in the texture of the soundtrack at the very end of the film. That’s because in the final scenes of the movie basically all of the even remotely redeemable characters get executed. After they had all died and the credits rolled, Ottman decided he wanted a “sense of release, because there had to be a different feeling as the audience walks out of the theater.” So he hit upon the idea of a self-contained choral piece. “The problem was though, what on earth would they be saying?”

What on earth indeed? It’s a moment where blockbuster filmmaking — always so anxiously in control of its meanings — seems to be at a bit of a loss. And it’s a moment where we as an audience suddenly get a sense for how films make meaning, and how it isn’t always the meaning they intend to make.

So who decided what the lyrics to the theme from The Avengers were? The short answer is that I still don’t know. But the long answer to my pedestrian question leads into the high-pressure, highly collaborative world of film scoring. A world in which composers often have just a few weeks to write music that pleases the studio and the director, and potentially even test audiences. And in which they toil with assistants, orchestrators, sound editors, and many, many session musicians to find a sound for a film that is still in the process of evolving. I wanted to find out who among this massive group would be the one to say “hey, let’s add a chorus and have it sung in Sanskrit” or something along those lines.

The answer turns out to be: Pretty much any of them can and sometimes do. What film choruses offer us is a perfect synecdoche for the collective, frenzied, and deeply mercenary magic that creates movies in the first place. It’s as likely that a director had the screenwriter invent specific lyrics early in post-production as that a subcontractor, assistant composer, or orchestrator jotted down some words or went on a Wikipedia deep-dive eight weeks out from release in a desperate late-night quest for a non-copyrighted text to use with a cue that might please a bunch of suits half a world away.

What film choruses offer us is a perfect synecdoche for the collective, frenzied, and deeply mercenary magic that creates movies in the first place.

***

Choruses have been part of film scoring for over a century. People have been singing on screen since the earliest silent reels, and with increasing technical wizardry we could even hear them doing it. But something like the Black Widow trailer is what we call an non-diegetic chorus: These are voices that viewers aren’t supposed to somehow locate within the screen action. In early cinema you had to have musicians physically present, first in the cinema with a viewer, eventually in the scene with the actors. Both of which pretty much ruled out the use of a choir. And, as film music historian Mervyn Cooke points out, once technologies existed that allowed films to have at least a partial soundtrack, filmmakers initially avoided non-diegetic music — precisely because they needed to sell the illusion that the sound was coming “from” the scene.

Non-diegetic music started to become the norm only in the early ’30s. And even then the limitations of recording technology meant that non-diegetic voices were not usually worth the trouble. By the late ’30s this had changed. Snow White and the Seven Dwarfs (1937) had its choir chime in even when it wasn’t for the explicit musical numbers. (Snow White was also the first soundtrack issued as an album, so choruses were part of how film soundtracks traveled semi-independently from their films from the very beginning.)

Alfred Newman had begun relying on wordless “heavenly choirs” going ooo and aaa in the background, in films like Wuthering Heights (1939), How Green Was My Valley (1941), and The Song of Bernadette (1943). As the music historian Donald Greig, who is also an active session singer on many modern scores, has pointed out, in the beginning choruses had to be at least somewhat motivated by theme or screen action — they were there to speak for ghosts, to intimate religious dimensions to the screen action.

And then there was Dimitri Tiomkin’s score for Frank Capra’s Lost Horizon (1937). The film concerns the discovery of Shangri-La in the Himalayas, and when we finally get to the fabled land the soundtrack accompanies the matte-painted wonderland with a chorus singing in … well, in a language that isn’t English and doesn’t seem to be Tibetan either. And thus another Hollywood tradition was born: film choruses belting out perfectly nonsensical prose with utter conviction.

And thus another Hollywood tradition was born: film choruses belting out perfectly nonsensical prose with utter conviction.

Both types of choral performance have never left the Hollywood lexicon. In thinking through how film choruses make meaning, I became obsessed with what the process of recording a soundtrack looks like today and at what point in that process someone actually writes lyrics in fake Tibetan. In the Golden Age, studios kept their own choirs — professional singers would show up at the lot and ooo and aaa for a Miklós Rósza score today and belt out a ferocious battle hymn for Erich Wolfgang Korngold the next. Studios also had their house orchestrators (usually several), and while laypeople remember the composers of Hollywood’s Golden Age, there are other figures that probably shaped the way films sound just as much if not more, all the while just quietly collecting their paychecks.

Speaking with modern singers about their experiences, I was struck by how little their day-to-day job description had changed since Tiomkin’s day. But the world in which they are performing is altogether different. As part of my research for this article I made a massive choir belt out the most menacing rendition of “Mary Had a Little Lamb” ever, and all it cost me was $199 plus tax. The EastWest Symphonic Choirs software allows you to make a virtual choir sing in just about any style imaginable. Want your ooos and aaas to sound like a whisper? More Broadway or more classical? All of that’s in the package.

But there’s more: Due to a system called WordBuilder, you can have this choir sing pretty much anything — you can type in text in English, in phonetics, or a proprietary alphabet called Votox, and the software will assemble it out of a massive databank of vowels and consonants. This is a commercially available product, but there are even bigger sample libraries kept by individual composers: If you’re wondering who’s dropping by to supply a quick “agnus dei” for a Hans Zimmer score, well that’s almost certainly a proprietary sample owned by Zimmer’s film score workshop, Remote Control.

All the professional singers I spoke to were keenly aware of products like EastWest Symphonic Choirs and the sample libraries — because more likely than not they’re in them. If you’re in the business of singing on film, these days you won’t always be asked to sing for an actual score, but instead you might get booked to record samples. There’s a scary possibility that these artists are slowly eroding the industry’s need for their labor — that the fruits of their one day of paid work will perform for the studios in perpetuity and with no extra residuals. Their disembodied vowels are putting their vocal chords out of business. But that possibility hasn’t been fully realized: Often enough when they arrive in the recording studio, singers will find that there is a vocal track already, but it’s done by computer. And yet, the composer wants a live version. Almost all the singers I spoke to expressed some surprise that Hollywood still bothered.

Their disembodied vowels are putting their vocal chords out of business.

One possibility why they do: Composers simply like working with live humans and consider it part of their job to do so. As Jonathan Beard, who has been composing and orchestrating in Hollywood for over a decade, put it to me, choirs are an easy, effective way to give dimension to a scene — “because you have a human body as one of the instruments, and there’s a power the human voice [has] over us in general.”

Composers are highly trained musicians, and a lot of their training has involved singing. The composer brothers Harry and Rupert Gregson-Williams (Harry composed for films like Kingdom of Heaven, the Narnia-films, and most of Denzel Washington’s films of the last 15 years, while Rupert is best known for DC Universe films like Wonder Woman and Aquaman) were both choirboys at St. John’s College in Cambridge — it makes biographical sense that choral textures and their creation would be important to them. And that they might like to think through music with a live chorus rather than a computer. Another surprising preference that speaks to a kind of sweet traditionalism: While sometimes vocal tracks get doubled in recording (meaning what sounds like 16 singers is just eight overlaid onto each other), this seems to be the exception rather than the rule. Clearly someone in the process enjoys working with large groups of people and thinks they give you an aesthetic payoff that engineering wizardry would not.

But there’s a more cynical reason as well, and it’s the reason why automation hasn’t displaced human labor in other fields: The process of booking some freelancers through a fixer, having them record for a day, and then paying them no residuals isn’t actually much of an expense. That’s how London became a preferred place for Hollywood to record: a large population of well-trained musicians, whose union doesn’t insist on residuals. Several London-based singers I spoke with suggested that the reason Hollywood doesn’t record in, say, Germany as often is that singers in continental Europe have steadier income and are less dependent on session work. And once a producer decides that even London-based musicians are too demanding — well, then there’s always Prague or Budapest. The gorgeous voices you heard in a John Ford Western were the sound of unions and full-time employment; in a Hollywood score today they are monuments to the globalizing power of the gig economy.

***

So that is the world from which these vocals emerge. Imagine you are a classically trained singer in, say, London who has done some previous work on soundtracks. You get a call from a fixer, who is assembling a chorus, or soloists, for a production company. You book the gig, and you show up for the recording session knowing which film you’re singing for, probably knowing the composer you’re recording for, but nothing else. Most recording sessions take place in the famous Abbey Road Studios, which are expensive, so you’re usually booked for no more than a certain number of union-approved hours.

Importantly, by the time you show up for the recording session, the film is pretty much “in post post production,” as one session singer put it to me. The film is basically finished, the wrangling over what the score is supposed to sound like is over. By the time you record, whatever orchestral parts you are supposed to accompany are fully assembled — you usually have them in your headphones as you sing. When you get there, you are handed a large stack of notes to sing and, according to all the singers I spoke with, you get through some portion of them in the next few hours — never through all of them. Some cues you sing will never be in the finished film, some cues you might do 10 versions of. And then the studio time the composer booked is over, you hand over your stack of notes, sign statements agreeing not to divulge anything about what you just sang, and you are on your way.

As the soprano Catherine Bott said: “You enter a studio and you open the score and off you go. You sing what you’re told, and it’s all about versatility, just being able to adapt to the right approach, whatever that may be for that conductor or that composer.” And part of that, singers told me, was singing the words — whatever they may be. As Donald Greig pointed out to me, a lot of these singers have training in classics; they certainly know their way around a Requiem or a Stabat Mater. And yet often enough when they step into Abbey Road they’re being asked to sing perfectly nonsensical phrases in pseudo-Latin — but the studio is booked, the clock is ticking, and as Bott put it, “that’s not the time to put up your hand and, you know, correct the Latin.”

Or the English: Bott sang on the soundtrack for the 1986 animated feature An American Tail. For a cue where the little immigrant mouse Fievel first lays eyes on New York harbor, composer James Horner had the choir intone the famous Emma Lazarus poem inscribed at the base of the Statue of Liberty. As she was singing through the cue — “Give me your tired, your poor” — Bott realized that whoever had put together the score had written down “your huddled masses yearning to be free” rather than “breathe free.” She was pretty sure she knew better, as did some colleagues, but out of English reserve, deference to the Americans, or professionalism, no one felt it was their place to say anything. The misquote stayed in the picture and you can buy it on CD today.

Perhaps part of what made me look for the meaning behind the lyrics on some of my favorite soundtracks was exactly this professionalism. A good singer sells the emotion and the conviction, to the point that a listener sort of has to believe that it all means something. Interestingly enough, early in this long tradition of made-up languages, Hollywood felt the need to pretend that it did mean something. When Lost Horizon was released in 1937, Columbia Pictures claimed in its publicity material that Dimitri Tiomkin’s score “includes authentic folk songs of Tibet.” The same press sheet noted that the Hall Johnson Choir, a popular gospel choir, “will sing the folk song arrangements in the native Tibetan language.”

Film music historians agree that this is hogwash. There is no evidence Tiomkin researched Tibetan folk songs for his score — what the ad men were selling as “authentic folk songs” were almost certainly newly written pieces in a made-up language. Tiomkin had started out as a concert pianist and relied on a small army of orchestrators to turn his melodies into actual playable scores. Someone in that group put a pen to paper and wrote these pieces, and either that same person or someone else seems to have made up some fake Tibetan text to distribute to the singers.

But for whatever reason Columbia Pictures’ publicity department didn’t want to frame the vocals in this manner. Perhaps extradiegetic voices were still sufficiently new that they wanted to tell an audience what these voices were doing on the soundtrack. Or it had nothing to do with the soundtrack itself, and was just another way of selling the broader spectacle of filmmaking: Look at the lengths we went to.

At the same time, lyrics have a pesky way of clarifying the intended audience. After all, it is not altogether difficult to imagine why Tiomkin and company wouldn’t have bothered with actual folk songs and actual language. Lost Horizon is one of those movies that stars noted non-Asian persons H.B. Warner as “Chang” and Sam Jaffe as “the High Lama of Shangri-La.” The broad and bogus claims to authenticity are also making a point of who the movie is for. The fact that the Hall Johnson Choir was an African American group best known for singing spirituals, amplifies the sense that Lost Horizon turns non-white people’s authenticity into charming window-dressing for white audiences. Like Shangri-La for its white visitors, even when its lyrics were incomprehensible film music was still “for” white English speakers.

At other times when Hollywood filmmaking relied on choruses, the point was the opposite of exoticism: hyper-comprehensibility. Decades later Tiomkin wrote a rousing score for John Wayne’s jingoistic epic The Alamo (1960). At the end of the movie, with the siege over and one lone survivor and her little daughter leaving the ruined fort, a chorus drifts faintly onto the soundtrack, almost as though the singers were standing somewhere far away in the field of battle. Over the movie’s final shots, the choir takes over the soundtrack, singing a version of what would eventually spend some weeks on the pop charts as “The Ballad of the Alamo.” The first lines a viewer is able to clearly hear are: “Let the old men tell the story / let the legend grow and grow. / Of the thirteen days of glory / at the siege of Alamo.”

This music explicitly tells us why it needs to turn human voices singing in a language the viewer is supposed to understand. The “Ballad” tells us what to do with the story we have just heard: Pass it on, let the legend “grow and grow.” Also — since this was made by John Wayne in the ’60s — the message is probably also don’t be a communist. But note how the movie has to treat three things as essentially the same: the singing has to be audible for the casual moviegoer, over people getting out of their seats early or finishing off their popcorn; the words have to be comprehensible on a purely linguistic level to an audience that has been taught to tune out the music on some level for the last two hours; and the reason why these words were included in the movie has to be clear.

Also — since this was made by John Wayne in the ’60s — the message is probably also don’t be a communist.

The fact that these three factors are separate can be easy to forget for an English-speaking audience reared on American pop culture. I grew up on Hollywood films in dubbed versions — though those didn’t typically dub the music. Meaning, as a kid who didn’t speak English, I became pretty used to following a plot in German, then the music would swell and I’d sort of tune out for a few minutes as the soundtrack, and the English language, washed over me. I’d get the basic idea of course — the characters were happy, or sad, or patriotic — but I had no idea what they were saying, and I was okay with that.

That’s sort of how most of us feel when we listen to the theme to the 21st-century version of Battlestar Galactica — unless we happen to be familiar with the mantras of the Rig Veda. Still, it’s a culturally specific experience. These days we can’t watch fantasy or science fiction without being sung at in Sanskrit, Old Norse, Dwarvish, Elvish, Uruk-hai, Klingon, and so on. When composer John Williams returned to the Star Wars universe for 1999’s The Phantom Menace, he composed an amped-up piece for the final duel — and over its churning ostinatos he overlaid a chorus belting out a … Sanskrit translation of a Welsh poem. And apparently the syllables of the Sanskrit text were rearranged to the point of incomprehensibility. Clearly, these shows and movies are not addressing us as potential speakers of Klingon or Sanskrit or even Welsh — they’re interested in the feel and a sound of a language rather than its meaning. At one recording session, Donald Greig told me, “they spent ages telling us how to pronounce the Russian and then we realized, ‘well this doesn’t actually mean anything.’” This turns out to be both a pretty new and pretty old way of listening to music.

When composer John Williams returned to the Star Wars-universe for 1999’s The Phantom Menace, he composed an amped-up piece for the final duel — and over its churning ostinatos he overlaid a chorus belting out a … Sanskrit translation of a Welsh poem.

***

Hollywood scores come in waves. The film industry isn’t known for being particularly fond of risk taking, and film scores in particular often build on previous scores. The director will often cut the film to a temp track consisting of existing pieces, and it’s easy to imagine that the filmmakers would eventually want something that sounds like their temp track to accompany the finished film. Choirs have never really left Hollywood, but there are certainly moments when producers and directors seem to have almost reflexively sought them out and others when they have avoided them. The Omen (1976) with its massive latinate choral opener, “Ave Satani,” kicked off one such wave. Peter Jackson’s The Lord of the Rings trilogy kicked off another.

This new chapter in the way films sounded started in the Town Hall, a storied concert venue in Wellington, New Zealand. That’s where composer Howard Shore recorded the earliest parts of his soundtrack for The Fellowship of the Ring (the rest would be recorded in London). The recording involved a full orchestra on ground level and rotating choirs in the balcony. It wasn’t lost on the composer that the scene was weirdly traditional: “The orchestra,” Shore explained, “was set up very much the way a pit orchestra was set up in an opera.” The collaborative process around the composition, too, felt like something Mozart and his librettist Lorenzo da Ponte might have recognized. The screenwriters wrote the text the choir would be expected to sing, an on-site translator would translate them into Tolkien’s languages, and Shore would then set the Dwarven or Elvish text.

Somewhat counterintuitively it’s not actually choral music with incomprehensible lyrics that is novel and needs explaining, it is choral music with comprehensible ones. For a long time, and for far longer than instrumental music, choral music in the West belonged to the church, to the mass, and that meant to Latin. A language as native to Christian religious life as it was foreign to most Christians. The Lutheran Reformation did a lot to hand church services over to language the congregants could actually understand, but throughout Europe the experience of being talked, and in particular sung, at in Latin persisted. That’s of course not to say that people didn’t sing in their vernacular languages — just that the experience of singing words you don’t, or don’t fully, understand would have been very normal to these people.

For a long time, and for far longer than instrumental music, choral music in the West belonged to the church, to the mass, and that meant to Latin. A language as native to Christian religious life as it was foreign to most Christians.

For the German philosopher Arthur Schopenhauer choral music was meaningful only insofar as the words were not the point. In his The World as Will and Representation, which appeared first in 1819, was republished in 1844, and strongly influenced composers like Richard Wagner, Schopenhauer claimed that music was the purest expression of reality because it didn’t linger with “representations” — words and the things they represent — but tapped automatically into something deeper. Choral music would seem to fall short of that standard — being pretty centrally concerned with words and the things they denote — but Schopenhauer didn’t think so. After all, you shouldn’t listen to sung music primarily for the words, and often you may not even know the words. And Schopenhauer thought this was for the better.

Latin still works that way for most modern audiences: You might argue that there isn’t much of an expectation on the part of an American film composer circa 1989 (or on the part of the filmmakers who hired him) that the audience should be able to follow along with the Latin lyrics — in fact, it might well be distracting if they did. What text is included, both singers and composers confirmed to me, has far more to do with the flow of phonemes and how it interacts with the raw sound of the vocals. The words are simply yet another instrument in the repertoire the composer has at their disposal. But it’s an instrument that comes freighted with all the complications that inevitably arise when our loquacious species uses language.

The words are simply yet another instrument in the repertoire the composer has at their disposal. But it’s an instrument that comes freighted with all the complications that inevitably arise when our loquacious species uses language.

After all, unlike a humming chorus, a Latin chorus does create extra levels of meaning for those who want to listen more carefully. Composer Jerry Goldsmith wrote “Ave Satani” for The Omen as a deliberate transposition of various Catholic masses. While the individual Latin may have been hard to pick up on (and wasn’t entirely correct to boot), listeners who were Catholic likely would have recognized what was being inverted here, given that they’d spent most Sundays around the actual Latin texts. It’s not clear how seriously Goldsmith (or the choirmaster who jotted down the Latin lyrics for the composer) grappled with that dimension of the score — for one thing, the very title of the piece messes up the declension of Satan. But that dimension was there nonetheless —The Omen was part of a kind of religious revival in Hollywood, and though it plays as camp today it was taken far more seriously then.

James Horner’s score for the 1989 film Glory relies heavily on a Latin chorus, and in the film’s climactic moment that chorus sings recognizably in Latin. Glory tells the story of the 54th Massachusetts Infantry regiment, an all-Black unit during the American Civil War, and the film ends with most of the unit being mowed down by Confederate soldiers while assaulting Fort Wagner in South Carolina. The piece in question relies on a text drawn from a Latin mass, frequently incorporated into the classical canon in various requiems from Mozart to Verdi. But, as so often, Horner (or his orchestrator) doesn’t stick to the actual text, but rather seems to create a mashup of snippets from the traditional requiem mass.

So is Horner just using the text of the requiem mass the way layout professionals use the phrase “Lorem ipsum?” Hard to imagine. After all, it makes a lot of sense to have a requiem text being sung as your characters are dying one by one. But more importantly, precisely because the text is so garbled, certain words stick out all the more: “Recordare,” Latin for “recall,” “stricte” (severely), and “judex” (judge). These pieces are largely taken from the Dies Irae, the part of the requiem mass that tells of the end of the world and God’s judgment, albeit with admixtures from just about every other part. The text, though hard to parse, is remarkably consonant-heavy for a Hollywood soundtrack, and a lot of it seems to be due (and I hope I’m hearing that right, as no actual text exists for this piece that I was able to track down) to the text’s overreliance of the future active participle, which ends in “-urus”: just in terms of pure grammar, the threatening hissing in the text is literally about what is to come.

So is Horner just using the text of the requiem mass the way layout professionals use the phrase “Lorem ipsum?” Hard to imagine.

So maybe the text, and the fact that it’s in Latin, isn’t about pretentiousness on the part of the filmmakers at all. It’s a mass for the dead and a tale of divine wrath, and it seems to make — over the heads of most of the film’s audience, admittedly — a point about retribution. It is remarkable how sophistic (white) Americans, who are frequently so proud to deal in moral absolutes, get when it comes to their Civil War. Horner’s grammatically challenged remix of the “Dies Irae,” I think, makes a point that is stark and simple and remarkably rare in American depictions of the country’s most bloody conflict: The Confederacy is evil, those who kill on its behalf are committing a sin, and they are bringing God’s wrath (and future judgment) upon themselves. There is, then, in this particular instance something to be gleaned from a text that otherwise we’re not meant to pick up on.

Which gets at an interesting disconnect — namely, that different constituencies will experience the same song differently. The choir members know what they’re saying, even if they have no clue as to what any of it means. And the composer, director, sound designer, etc., although they live with a soundtrack far longer than either the performers or even the most devoted audience, don’t tend to get to the words that go with the music until fairly late in the game. They often have to rely on orchestrators and assistants, or a helpful choirmaster who claims he really knows Latin. Their budget, and thus their time, is not tailored to their needs, but to the dictates of the director and the studio. The prose simply appears, like a ghost in this immense machine. And — in spite of the fact that most parties involved seem to be content to have it not mean very much — it winds up signifying something.

One example: An “exotic” text can only be understood by very specific listeners. But, very much to the point, they are not therefore the intended listeners. Lost Horizon wasn’t banking on a particular reception in the Tibetan community — rather the opposite: Dimitri Tiomkin and his collaborators seem to have counted on not having any actual speakers of Tibetan in the audience.

This gets a lot more troubling in the case of the phrase “Nants ingonyama bagithi baba,” likely one of the most repeated, parodied, and bowdlerized lines of text in any soundtrack. It’s clear that it isn’t addressing the average viewer with the intention of being understood. The very fact that it is in Zulu, but the story of The Lion King appears to take place in the Serengeti, thousands of miles to the north, suggests that the language is here to signal one thing and one thing only: African-ness.

For contrast, look at the way composer Michael Abels’ score for Jordan Peele’s Get Out features Swahili voices: Outside of the considerable number of Swahili speakers in the world, most people watching Get Out won’t know what the singers are saying. But what they’re saying does matter, in a way: Literally “listen to your ancestors,” but as a saying meaning something kind of like “you’re about to be in danger.” The viewer who doesn’t understand this line is missing an important warning about what is to come in the film. As is, of course, the film’s African American protagonist who cannot listen (or at least understand) his ancestors. Peele and Abels manage to wring from this small decision a whole range of subtle points.

***

But as with all exoticism, there’s a strange tug of war between condescension and appreciation in these kinds of borrowings. When Ottman decided to use a choral piece at the end of the 2008 film Valkyrie, he clearly needed a German text, and I suspect any German text would have sufficed. But he didn’t pick any German text. The film stars Tom Cruise as Claus Graf Schenk von Stauffenberg, a historic figure who led the only attempt by members of the Nazi state to get rid of Adolf Hitler. The text is “Wandrers Nachtlied,” one of Johann Wolfgang von Goethe’s most memorable, well-known texts, and if it’s a little bit treacly by the great poet’s standards, it’s hard to deny it’s a deeply appropriate choice for this moment. Not overtly about politics, it is nevertheless about history, about reflection, about resignation. And about a different use of the German language than one is used to in Hollywood films.

For any German person it’s weird to hear bad guys so consistently speak (and butcher) your language. I’m not complaining, mind you, it makes perfect sense. But what’s remarkable about Valkyrie is that it seems unusually careful for a Hollywood-film in how it deals with the German language. Earlier in the film, Cruise’s character says that “people need to know we were not all like him,” and this final poem seems to do something similar for the German language — the filmmakers close their movie by pointing out that this language is capable of beauty and deep humanity. The poet Paul Celan — himself a Holocaust-survivor — pointed to the strangeness of writing in a language that was both “my mother’s tongue” (Muttersprache) and “the murderer’s tongue” (Mördersprache). Ottman seems to want to recover the former after showing plenty of the murderers.

The strange thing is: I am pretty sure Goethe’s “Nachtlied” is the first utterance in actual German in this film about Germany. Cruise sort of tries a German accent every other scene, the largely British supporting cast doesn’t even bother. And no one speaks any German, the way Sean Connery does with Russian at certain moments in The Hunt for Red October, or Alan Rickman in Die Hard. The film’s supporting cast is stacked with Germans who belt out accented English throughout. It almost feels like the film wants to bend over backwards a little too much: remind us what beauty and thoughtfulness this language is capable of — even though it never shows us the barbarity, which the film renders in English.

I suppose it’s moments like that one that made me obsess over what choirs sing in movies, and who decides what they sing. Because it’s a moment when blockbuster film or TV, which increasingly is created for the greatest possible global audience, which has been focus-grouped and test-audienced within an inch of its life, manages to speak far more directly, more improvisationally to a much smaller audience. All of us are sometimes in that smaller audience, sometimes not. But we’re aware it’s there. When cinema is literally speaking in tongues, how could we not? And to be the person who hears a call the object of fascination never knew it was putting out there — what better definition could there be of what a fan really is?

* * *

Adrian Daub is professor of Comparative Literature and German Studies at Stanford University. He is the author of four books on German thought and culture in the nineteenth century, as well as (with Charles Kronengold) “The James Bond Songs: Pop Anthems of Late Capitalism” (related story here). He tweets @adriandaub.

* * *

Editor: Krista Stevens
Fact checker: Julie Schwietert Collazo