Opera and Song

“There were others there that night who had not heard her name, [Roxanne Coss], who would have said, if asked, that opera was a collection of nonsensical cat screechings, that they would much rather pass three hours in a dentist's chair. These were the ones who wept openly now, the ones who had been so mistaken.”—Bel Canto by Ann Patchett

Roxanne Coss: “It's easier to love a woman when you can't understand a word she's saying”—Bel Canto by Ann Patchett

Here's a bit of opera with no words at all (coloratura)—the end of the Bell Song from Lakmé, sung by coloratura soprano Sabine Devieilhe, imitating the orchestra's own bells.

Like film music, opera still tells you what it's about, but within the machinery of directors and sets and costumes and words, it is the music, not the words, that lives front and center.^💬

A simplified interpretation chain might look like: Libretto (words) > Composer > Performers > Audience

We'll listen to opera—and later, art song—with the lens of humanity: a libretto full of people and relationships. In particular, we'll look at how melody and counterpoint carry that life into the music. Then we'll step away from the libretto—the staging, the words, the singers—and see how far the lens holds, or where it doesn't.

Character and Counterpoint

I'm a bit embarrassed to say that for all my love and advocacy of classical music, there are two pillars of the repertoire that eluded me for most of my life: opera, and Mozart. I saw the Magic Flute when I was little, I'm told, but I don't remember it since I was asleep most of the time. I barely played any Mozart growing up—“You'll get Mozart in college,” Natalie told me. I don't know if “get” meant play or understand… either way I didn't.

Only after a concerted effort did I come to the realization—perhaps I always thought Mozart was boring because I didn't understand opera. I still do not claim to be an opera connoisseur, but this chapter isn't about that; it's about using opera to approach all of music.

The first number we'll dive into is the Papageno-Papagena duet from Mozart's Die Zauberflöte (The Magic Flute), when the bumbling, bird-like bird-catcher Papageno finally finds his soul mate, Papagena. They flirt and debate whether their children will be little Papagenos (boys) or Papagenas (girls).

The words and acting make this very concrete music—we know exactly what's going on. The next couple renditions will be more abstract, highlighting how Mozart's music can communicate just as specifically as the words.

There are many delightful recordings of this—this one with Edward Jowle and Sofia Kirwan-Baez at the Royal College of Music is set in a university.

It's hard to imagine how a scene could convey such a strong sense of character so quickly. You may never have met a semi-magical bird-man in your life—and in this university staging, you might not know that's what Papageno even is. Still, you feel like you know him in his somehow relatable, simpleton quest for a partner and mini-papagenos. Even the dramatic pause at the end before they pick up their argument is so clearly jesting rather than fighting.

There are plenty of non-musical cues—the acting, costumes, and the words, just to start. But even just with the way they say the “pa”s of their names, you can tell that the music doesn't just have character; it is inseparable from the character.

A first experience of an opera needs subtitles so you can understand the plot (3 hours without the plot is a long time, no matter how good the music is)—but a subsequent listen will allow listeners to bask in this duet's charm—the effervescence of the strings and the vocal color simultaneously sincere and immature. Let's take one step away and hear a recording session—un-staged and un-subtitled, the libretto now just in our memory.

The character is so latent just in the opening “pa” and all the ways they sing it, even without any other words. And despite all the similarities between Papageno and Papagena—and there are quite a few—we also don't have any difficulty telling them apart in the music's texture, even when there's more than one voice at a time (polyphony). There are always two distinct voices.

This may sound ridiculously obvious to point out. After all, one's sung by a man, one by a woman. They have different ranges, different timbres. In this interpretation, Papageno's “pa” is more spoken; Papagena's is more sung. In short, they're different characters.

So let's take the texture—both singers, and the entire orchestra—and abstract all the layers onto one instrument, no words, any timbral differences limited to what a piano can do (which is not nothing, but far short of what the human voice and the instruments of an orchestra can do).

Listen to the interplay between the various voices—called counterpoint—and see if it can still retain its individual character. ^📖

This was Anderson and Roe's own arrangement of The Magic Flute, adding an additional link in the interpretation chain that injects some of their own personalities. The duet takes on two layers—one between Papageno and Papagena, a second between Greg Anderson and Elizabeth Joy Roe, who are not always just playing one character or the other. But throughout, their control of the independent voices (“voicing”) makes the texture crystal clear. There's no issue picking out the characters.

But many of Mozart's own instrumental works—pieces titled Symphony and String Quartet and Piano Sonata which mean nothing except for identifying which instruments play them—are really operas in disguise. Papageno's main theme, which Mozart supposedly heard his bird sing, even made its debut in a Piano Concerto. The characters in these pieces can be as vivid and accessible to listeners as they are when they're in costume.

So here's a sonata: absolute music, no story attached. But once you're aware of it, it's impossible not to hear the dialogue between voices, and imagine the characters of the opera that Mozart has written here without actual voices. And as in the Papageno-Papagena duet, there are also instances where Mozart writes the orchestra part into the piece as well.

The videography here is also artistic in itself—highlighting the counterpoint between instruments. In future chapters we'll look at much denser instrumental polyphony in the music of J.S. Bach.

Singing, Speaking, and Imitating

As the music becomes more abstract, more and more questions are posed to the listener. A concerto for piano and orchestra doesn't just ask, “Will Papageno and Papagena end up together in the end?”—it also asks what characters are involved in the first place, what they are saying, and whether they are even speaking or singing.^📖

In Mozart's K. 271 Piano Concerto, Mozart pioneered what would become the modern concerto—a pianist like an opera singer, playing the protagonist, and sometimes other characters, by themself. One usually thinks of Mozart's most memorable moments as the soaring arias, but sometimes music actually speaks, pauses, thinks, and breathes more than it sings. Something similar can be seen in the opening of the slow movement, played below by Elisabeth Brauss with Florian Donderer conducting.

In opera, a singing-speaking hybrid, recitative, is usually used to advance the plot when a song would stop the action. In Mozart, the soloist speaks freely in their own time, and then the orchestra comments in response with just a chord or two.

This piece imitates the recit all over the place: in the first movement, right after its climax, the protagonist second-guesses her confidence. In the slow movement, it comes at least four times, including an extended prayer in the pianist's cadenza. ^📖

These recits entreat you like the character is trying to tell you something—like there's a real, specific human with a life and desire, on the other side of the fourth wall speaking right to you.

So we're primed to look for another recitative-like solo in the rollicking finale, but at the last second Mozart turns the opportunity into a gentle minuet instead. Now he tells us don't worry, be happy.

But there are a number of problems with this interpretation. For one, it's still not that specific. In the slow movement, say, what is the pianist actually saying? Asking for mercy, like in Mozart's Great Mass in the same key? If so, for whom?

Second, it took me months of study before I found the connections between these recits, and discovered the similarities to the Mozart C Minor Mass—and after all that it's still very likely that another pianist will review them and call BS. So what are we supposed to make of it on the first listen?

I think these questions are all fascinating when deciding how to perform a piece. But when we're listening to it, I just don't ask, “is she praying for mercy?” I just imagine that she is singing to me in another language and I don't have subtitles. I know I'll never know what words she's saying. But I can still understand her.

What instrumental music can't do

Interpreting piano music as if it were just imitating singers and orchestras works for a surprising majority of the repertoire. As a pianist, I hated to admit this for a long time. Is our instrument just a poor man's orchestra?

Historically, it was often literally this—when a composer wanted to hear or promote a piece before the days of electronic sampling, they often couldn't just get a whole orchestra together, so they would write an arrangement for piano. Even modern, difficult works, like Stravinsky's Rite of Spring, started out with a version for one piano four hands that's actually quite rewarding to play.

And yet for all of the ways that instrumental music can give us a complete character, there has to be a limit. “Music expresses that which cannot be said and on which it is impossible to be silent,” said Victor Hugo. But when we want concrete specificity, there is sometimes no substitute for words, staging, and acting.

Opera can keep secrets from its characters to comedic effect like in Mozart's Marriage of Figaro, where the Count hunts around his wife's room while a young page hides in the closet, hastily dressed in women's clothes, and her maid Susanna tries to figure out what's going on.

Or opera can make three sincere hearts audible at once: in the moving final trio of Strauss's Der Rosenkavalier, three characters sing of their loves in one interweaving moment. This is still more complex counterpoint, with three completely independent voices sung simultaneously.

But most of all there is a power in the connection to the humanity of the voice that is sometimes imitable but ultimately irreplaceable.

Julia Wolfe's Fire In My Mouth is about the 1911 Triangle Shirtwaist Factory fire, which killed 146 garment workers, mostly immigrant women and girls.

This oratorio (a semi-staged opera) creates a visceral feeling of protest, where the audience feels directly entreated by the victims' voices. If effective, it gives the audience, too, a fire in their mouth.

What Music Wants

The Wagner in the pocket concert brings us to the next step beyond character—what the character, or the music, wants. In Wagner, it's to reach its climax and to resolve—both in its harmony and in its melody.

Technically, “melody” is just a sequence of notes one after another, contrasted with “harmony”, which is multiple notes at the same time—the horizontal dimension of a score as opposed to the vertical.

Harmony will come back in future chapters. But first, what really makes a melody, and how does it fulfill the role of a character needing something?

To me, a meaningful melody is a journey to and/or from a climax. That is to say, it has a sense of direction, and if you were to look at the written notes moving up and down, you might see a shape.

Most of the typical melodies give us interesting shapes to catch our ear, leading us up and down their twists and turns, but most importantly they give us a sense of direction that we are moving in. They make us understand their sense of need innately.

Here are a few measures of Una furtiva lagrima (A furtive tear) from Donizetti's L'elisir d'amore (“The Elixir of Love”), a more traditional melodic shape that mimics the patterns of speech. The peasant Nemorino notices a tear in his beloved Adina's eye, giving him hope that she may secretly return his feelings. The mostly descending melody captures the sadness, while the moments of upward reach capture the longing.

Chopin was influenced heavily by opera, and the opening of his F Minor Ballade takes a similar shape to Una furtiva lagrima—a gorgeous melody in the top voice that is equal parts speech and song—and incidentally, the subtle counterpoint of supporting voices underneath.

What does this opening want—or not?

The idea of a melody wanting something can also be built over a longer period of time, and the more the climax is withheld, the more we feel we need it.

Copland says, “the creation of that continuity and flow—that long line—constitutes the be-all and end-all of every composer's existence.” In this sense, melody isn't just a “local” phenomenon that takes its course over a few seconds, but also a global one, in which the entire work has an arc as well.

Tchaikovsky's Nutcracker builds a melody that starts with just a single downward scale. Legend has it he did it to win a bet.

He won the bet, but the scale is actually not the whole melody. It is repeated so that it forms a longer line, and when he withholds it, we need it to come again. Somehow, Tchaikovsky has made that final repetition feel inevitable, so it becomes the climax we long for.

Sometimes it's more ambiguous where the melody is going. The pianist Jerome Lowenthal once told me a story: Someone proudly toasted Debussy at a cocktail party, “to Debussy, the man who killed melody” as if it were an avant-garde badge of honor. Debussy looked miffed, and replied, “everything I write is melody.”

The opening chords of his Cathédrale Engloutie (The Sunken Cathedral) could easily be seen as barely going anywhere—just a reverbating effect that represents still, calm water. Or they can be seen as melody that “wants” to go to the top of the line as ripples “want” to circle away from a disturbance in the water. Different pianists will try to give different impressions.

On the other hand, Rachmaninoff is known for writing many of the most gorgeous Hollywood-esque melodies in the repertoire, constantly reaching for the heavens.

But you can really see his mastery of melody in his cello sonata, which begins with his trademark repeated bells. Bells don't care about the connection or the direction from one to the next—they just toll in isolation. Rachmaninoff takes the tolling of bells, played by a piano's hammers striking the strings, and reveals that there was actually a hidden melody in those repeated notes all along.

This performance switcher shows three different ways different pianists might shape the same repeated bells. Lugansky plays them evenly, like true-to-life bells. Kanneh-Mason plays them with direction, like a melody. Horowitz has a type of magic that only he owns: he sets up the expectation that they'll be even, and then at the last moment, he pulls away and plays one tenderly. And when he resumes, you hold your breath because you can't predict what he's going to do. But it feels like he's putting an entire lifetime into each bell.

What's in a song: Lieder and Art Song

We now return to the human voice to put character and desire, texture and melody all together in a single, standalone song.

The character is Gretchen. What (or, rather, whom) she wants is Faust—the man who sold his soul to the devil.

Rather than composing new music for a custom-written libretto, plot and all, art song (lieder if you're fancy, or German) takes a poem that the composer happens to love, and sets it to music. It is the precursor to the modern pop song. ^📖

The interpretation chain is now Poet > Composer > Performers > Audience.

A composer like Franz Schubert co-opts a poem for his own goal, almost like a film director co-opts a piece of music for their soundtrack. And like in film, the added layer of perspective shifts the audience, the accessibility, and the messaging.

Here is the genre's classic duo: Schubert setting a poem from Goethe. I had never read Goethe before. Maybe it's my own ignorance, but it never even occurred to me that I might enjoy it. But just within a “song,” Schubert and Goethe take us for a ride.

This song is called “Gretchen at the Spinning Wheel” (Gretchen am Spinnrade). The performance is from Renée Fleming with Christoph Eschenbach at the piano. The English subtitles are below the German. (It can take a second to adjust to reading them.)

If we imagine a first listen to this song, the title (Gretchen and the Spinning Wheel) is the obvious place to start. There are two elements here, Gretchen, who is obviously the singer, and the spinning wheel which must be the piano.

Gretchen's melodic line is bursting with desire, cutting out only at the soprano's scream of “his kiss”—the climax marked as much by the cessation of motion as by the energy of its output.

The symbolism of the spinning wheel is an additional layer of interpretation that Schubert adds to Goethe's text. It's possible that in the poem, the spinning wheel is a metaphor for the speaker's inner turmoil, spiraling and aching and trapped in a loop.

But it is the texture that comes from the music's other voice—the piano accompaniment—that converts it into something that is maddening and terrifying.

Even if its omnipresence goes unnoticed, at Gretchen's cry of “his kiss,” it cuts out, its absence loud, and sputters attempting to return.

Meanwhile the pulsing underneath the wheel is barely audible: that pulse is literally the pedal that powers the wheel, yes, but also the singer's heartbeat, and maybe the listener's too.

Pianists love to play Schubert's lieder because the texture is as indispensable as the melody. In the case of songs like this where the spinning wheel is also the emotional core, Liszt's transcription for solo piano gives us the despair and terror in their pure form outside of a concrete story.

In fact, many of Schubert's piano parts are onomatopoeic, in turn representing:

• The relentless galloping horse carrying father and dying child in “Erlkönig” (with popular virtuoso transcriptions for solo piano and violin).
• The rippling stream with a darting trout in “Die Forelle” (with popular variations for piano solo, piano quintet, and Samsung washing machine).
• The steady drone of the hand-cranked Hurdy-Gurdy in “Der Leiermann” (The Hurdy-gurdy Man)
• The swaying waves of “Auf dem Wasser zu singen” (To be sung on the water)

In every case, the voice provides the melody, and the words provide the story. But the piano gives it a character that goes beyond mere accompaniment. The two interact as equals. ^📖 (This is one reason behind the recent push to re-brand Accompanists as Collaborative Pianists, a term which commands a bit more respect.)

Songs Without Words

It didn't take long for composers to look at Schubert's model and think, “I can basically do what a song does even without lyrics.” And so siblings Felix Mendelssohn and Fanny Hensel (née Mendelssohn) developed a new genre of piano pieces in this style—which Felix just called Songs Without Words and Fanny called Songs for Piano. (This probably didn't help teach any piano students that you can only call something a song if it is actually sung.)

They mostly sound like someone wrote a song, and then took out all the words and made a piano transcription. But once you start looking through this lens of music-as-song, you also see this song-like sensibility emerge in music not called “songs” at all.

Johannes Brahms even converted some of his songs into a sonata for violin and piano, which, obviously, can't talk.

We could easily just call this as a faster transcription of Brahms' Rain songs, the piano playing the pitter-patter of the rain.

We've lost the imagery of the nostalgic language -

How blissful then it was to stand With naked feet in the flow!

but I think the more interesting question is what have we gained. The melody gives us a motif through which the entire sonata can travel. It can transform from character to character, drip like tears or cascade in a downpour, and we can fill in our own nostalgia.

There is a trick to finding that storytelling that music teachers love but students hate: adding lyrics back to music without them. It's supposed to teach students that even in absolute music you can have all the trappings of a song—characters, desires, etc. As a student I would always come up with something banal like “I love you” or “I hate you.”

But sometimes it feels like the music just doesn't want to cooperate. How would one even approach the Badinerie of Bach, which whirls and dances. It is played here by Hannah Schlubeck on pan flute:

Incredibly catchy melody, and clearly full of character, but is it really possible to go from that to a specific scene? Yes, in fact, Alfredo Chacón and his father use the notes to argue with each other. Or in a completely different interpretation Achinoam Nini finds an “I love you” with quite a bit more pizzazz:

There's a great meta lyric in there that captures this approach to music and life—“Life's a song that we are playing… you can change it but it's never really wrong.”

With abstract music like this Bach, we can fill in any interpretation we like—it's never really wrong. In the next chapter we will go a bit more abstract and leave lyrics behind, gaining another way to interpret music, and life: through dance.