[The below was written for my new intro to music for Basic Books, due to come out later this year as best I can tell. I’m posting my original essays here because most of this material was cut from the book.]


When I used to teach classes in musical analysis I insisted that my students include in their essays some take on the emotional side of the piece. At first this generally got uneasy responses, because talking about emotion in the academy tends to be verboten: emotion is subjective, personal, arbitrary, therefore not scholarly. But I persisted, telling students that when a classful of musicians agree on the emotional tone of a piece, that’s interesting; when they don’t agree, it’s equally interesting. Each case tells us something important about the composer and the piece. I also think that a performer needs to have his or her own sense of what a piece is expressing; that’s what helps put a piece across. As we said over and over at the conservatory: there’s a great deal more to performing than playing the right notes and sounding pretty.

I’m a classical musician first of all because that kind of music moves me more deeply than any other kind (though I’m interested in all sorts of music). The analytical and historical aspects are absorbing, but for me the technicalities lie well behind the feelings. I’ve been known to go through a whole performance of Bach’s B Minor Mass with tears in my eyes. If I didn’t feel music that deeply I wouldn’t have gone into it in the first place, given that for most of us the field is a manifestly stupid way to try to make a living. I can’t imagine why anybody would get involved with music for any other reason than emotion and love. Well, that plus talent. As a friend said to me: “I play the cello because I’m good at it.”

I’ll say three things at the outset of what is going to be a clearly unscientific discussion of musical expression. First, musicians use the term “expression” in two ways. To play “expressively” means to play with feeling whatever the emotion at hand may be in the piece. Talking about what a pieces “expresses” concerns the particular tone. Second, like many musicians and listeners I believe that music is an emotional language beyond words, so its essence can’t be captured in words (though it can be useful to try—thus my books). Third, I like the conclusion of philosopher Suzanne Langer, who called instrumental music “an unconsummated symbol.”

The extent of what Langer means by symbol is too much to get into here, but the basic idea is that a symbol is a story, painting, image, event, etc. that we respond to in a complex emotional way rather than a directly informational way. That’s the difference between denotation and connotation. A stop sign at an intersection denotes that we should stop. At the same time, it may represent to us all the damn things in the world that tell us what to do, that get in our way, that mess with our lives. To someone else, a stop sign may elicit a comforting feeling of order, the social contract, the need to for caution. In each of these cases we’re responding to the stop sign’s connotations. In other words, we’re responding to it as a symbol.

Langer felt that our response to art and much of the rest of life is a texture of symbols, but that instrumental music, lacking words or clear imagery, is a kind of blank slate that we nonetheless respond to as if it were a tangible symbol. What the symbol is, in any given piece, is largely up to our own responses. So, “an unconsummated symbol.”

These are ideas I subscribe to. The thing is, however, that in practice music is much, much more complicated than that. In most vocal music, for example, the words tell us the subject and imply emotions, and most composers want to express the emotional and even physical sense of the words (though sometimes they might write music that inflects or even contradicts the words). In a Schubert song, when the story turns sad he usually shifts from a major to a minor key; meanwhile he jumps on every image in the text, from a spinning wheel to a tree in the wind, and pictures it viscerally.

A second complication is that most advanced musical cultures have familiar signals for emotions that listeners absorb by being part of the culture. We tend to hear music in major keys as somewhere on the happy spectrum and minor keys as more somber. But in fact, in the 18th century Bach didn’t have so strong a sense of that association; he wrote some cheery pieces in minor and some manifestly sad ones in major. Still, his time had a “doctrine of the affections” that assigned particular emotions to specific musical gestures—some having to do with pitch, some with rhythm (a joyful piece will tend to be livelier, a gloomy one slower). One famous example is the descending chromatic bass line that undergirds the crucifixion movement of Bach’s B Minor Mass. In the Baroque, drooping half steps were a standard representation of sadness unto tragedy. Beethoven was still using that convention when he wrote the Pathétique Piano Sonata, whose leading motif of a falling half step clearly represents pathos.

India has one of the oldest and most sophisticated musical cultures, involving an elaborate system of emotional representation. Traditionally there are eight expressive areas, called rasas; they include sadness, love, valor, laughter, and so on. A musical work, called a raga, is largely improvised according to intricate rules. Each raga has its particular rasa or mixture of them, its rhythmic meter, its appropriate time of day, its appropriate season, and so on. Each is also based on a particular scale, with rules about how that scale must be inflected in that particular raga. Musical connoisseurs in India hear all this, I assume, in the same way as we hear a happy versus a sad piece. I am a longtime enthusiast for Indian music, but I don’t hear any of that because I didn’t grow up in that culture.

All this is to say that our culture teaches us from childhood how to listen to its music in terms of conventions of melody, harmony, rhythm, and color that are associated with emotions, tension and release, times of day (a serenade, for example, is an evening song), even landscapes (as in a pastoral work). Most composers of the past made use of those conventions in some degree or other. Beethoven was a supreme master of abstract form, but in his Missa solemnis the word ascendit is predictably set to a rising line and descendit to a descending one, and likewise for every other word in the text that can be viscerally expressed. And, remember, Beethoven wrote the Pastoral Symphony.

When I was composing my piano quintet Midsummer Variations, I wasn’t particularly thinking about its expressive direction, or rather I was feeling the expression without defining it. Only at the end did I realize that every bit of this piece written in a country hilltown in the middle of summer was about that—even more specifically, about New England midsummer, because summer here is so poignantly evanescent. All that had been unconscious, but it was still real. So if I had not attached that title to the piece, would it remind listeners of New England midsummer? I doubt it, but how listeners respond might be close to my feelings about the season. In any case, I think the tone of that piece has a strong sense of being about something.

When I’ve experimented with writing fiction I found that the events of my life external and internal, many of which I thought I’d forgotten, were flowing onto the page. That’s how fiction happens, from an amalgam of experience and invention. I believe that when we compose music it likewise comes from our lives, even if–as with my midsummer piece–we may not realize it at the time, or ever. If a work of art doesn’t come from the artist’s guts, it’s not likely to have any life of its own. Still, as with fiction, to write something happy or sad does not mean that the artist was feeling that way at the time. Famously, some of the most joyful works of Beethoven and Mozart were written when they were depressed. To write about joy or sorrow you have only to be acquainted with them. Fortunately and unfortunately, all of us are. Nor can we as listeners be expected to feel exactly what the composer was trying to convey. We have to see it through our own lens. Ultimately, the artist is supplying us with stuff to dream on.

So does all this mean that Langer is wrong about music being an unconsummated symbol? I still say that in a larger sense she’s right: beyond cultural conventions, there is still a lot that we have to fill in for ourselves. At the same time, while each culture teaches its people how to hear its music, there are potent if elusive universals. Joyous music tends to be fast, melancholy music slow; minor keys sound “darker” so are good for darker feelings. Music can mimic the sensation of movement, of breathing or sighing, of weeping, of certainty or surprise, of excitement or sexuality. The rise and fall of tension in music echoes the rise and fall of feelings. All the same, many more or less scientific attempts to define universals of emotion in music haven’t come up with much. I do think that universal responses and emotional qualities exist, but they’re hard to pin down. In any case, our response to music I believe is an intensely individual synthesis of personal and universal qualities, plus ones that have been imbued by our culture. (See “Specific Emotions in Music” below for more ideas in this direction.)

Again, the personal element is critical. A roomful of people may agree on the general tone of a piece, but when it gets down to the details, every listener will have a distinct response based on their personality, their experience, their mood, what they had for lunch, any number of factors. For me that individuality of response is one of the wonderful things about music. Sure, our responses to a poem or a painting are also individual, but with instrumental music that aspect is central if this composition in sheer tone is to “mean” anything to us. That’s what the Romantics exalted about instrumental music: it gets to our feelings directly, without having to be filtered through language or logic or story or image (though it can be so filtered, if words are involved).

The whole matter of expression is still more complicated, but I can only touch on further issues. For one, some composers are more emotionally direct than others. Beethoven set out to express human passions and expected his listeners to be sensitive to what he was saying. When I write about Beethoven’s music it tends to give me good adjectives, even when the tone is ambiguous; I’ve written about Beethoven’s touching major-key pieces as “minorish majors” and their opposite, “majorish minors.” He is, in other words, intentionally transparent in his expression and in his implied narratives (though less so in the late music). Brahms sometimes and Mozart much of the time I find more elusive, often poised exquisitely between qualities. With those two composers I often have to resort to adverbs: tenderly, excitedly, and the like.

At the same time, every piece is constantly being renewed in performances, so its expressive quality evolves. That is one of the great strengths of classical music: the silent notes on the page are brought to life in constantly new versions. I remember coming out of a performance of the Schoenberg Violin Concerto by Christian Tetzlaff, James Levine, and the Boston Symphony saying, “I’ve known that piece for thirty years and never realized how weird it is! There ought to be new performance indications for it: insectoid, fidgetissimo.” Then I heard the celebrated Hilary Hahn recording and it sounded like a Romantic concerto. In an interview I asked Hilary if she deliberately set out to play it that way. Her response was basically, “Well, duh, it is a Romantic concerto.” It is, if you make it work. Both those versions of the piece worked for me. Recordings, of course, negate that renewal, which is why I prefer live performances. Meanwhile, as we change and evolve, so too does our sense of a piece when we come back to it. The greatest, most lasting works are ones that maintain their fascination through this process, that we feel we never get to the end of.

I think music is good at conveying basic emotions: happy, sad, angry, and so on. It’s not so good at, say, “wistfully regretful.” The more subtle emotional territories are where our individual responses set in. Meanwhile much of the effect of music may not be a matter of happy and sad and so forth, but of visceral appeal. Plenty of pieces I love I can’t put an emotional label on at all; they grab me because they grab me. I sniffle through Bach’s B Minor Mass not because it’s all tragic—some of it overflows with joy–but because I am moved by the beauty, the brilliance, the craft, the depth of expression, the power and glory of the human imagination at its most profound.

A couple more points. The very idea of emotional qualities in music was out of fashion in 20th century academe. The idea of “abstract” or “pure” music ruled the scene. I was once slapped down by a teacher for suggesting that the piano accompaniment of Schubert’s Erlkönig evokes the galloping of a horse. Of course it does. Program pieces illustrating a stated story were tolerated in the classroom, but with a pitying smirk. The celebrated music writer Donald Francis Tovey said that Beethoven’s Pastoral Symphony had essentially nothing to do with nature at all; it was an abstraction like every symphony. Sorry, Donald, that’s rubbish. Beethoven geared the melody, harmony, form, orchestral colors of the Pastoral to express its program of a day in the country, with its brook, its birdcalls, its peasant dance, its thunderstorm, its warm sunset. (See my Beethoven bio.)

So of course emotion in music is present and vital. I’ll go beyond that and point out something I don’t know has been said before: every one of the central revolutionary pieces of the last two centuries was based on a story, and found its innovations in the composers’ determination to paint and express that story: Beethoven’s Eroica Symphony was originally titled Bonaparte and is about Napoleon from beginning to end; Debussy’s Afternoon of a Faun evokes the eponymous time of day, that amorous creature, and Mallarmé’s poem about them; Stravinsky’s Le Sacre du printemps got its singular energy and style from the ballet’s story about a primeval human sacrifice; Schoenberg’s Pierrot Lunaire gained its stunning originality of voice from capturing the fin de siècle weirdness of its lyrics.

I’ll let Schoenberg have the last word regarding the necessity and centrality of expression. In the preface to Pierrot he warned musicians not to add expressive notions of their own but rather to stick to the notes as written. Famously at one moment in the piece where the text says “Pierrot scratches on his viola,” Schoenberg pointedly avoids using a viola even though he has one in the ensemble. In a later article called “This Is My Fault,” Schoenberg points out that younger composers took his preface as a manifesto saying that expressing a text was out, and in ignoring the meaning of the words they wanted to “surpass me radically.” But if music and text are independent, Schoenberg asked, then why not have Wotan, King of the Gods in Wagner’s Ring, stalk across the stage to a boogie-woogie? Of the anti-expression movement he concludes, “What nonsense!”

So music is expressive of emotion, sometimes in more concrete ways and sometimes in less concrete. Some of that response is cultural, some of it innate. After all, even one-celled animals respond to sound. I suspect our response to music starts at the cellular level and travels up to the higher brain functions. And the most important part of our emotional response is particular to ourselves. We can sometimes agree on what a piece expresses, but we’ll each fill in the details differently, and we’ll never fully understand how music moves us. What we feel from music is like what we feel from a sunset. The sunset contains no emotion; it’s a physical phenomenon that has nothing to do with us. Maybe the dinosaurs enjoyed them. In any case the feelings are ours, some of them universal to humans, some individual. Ultimately the source of such responses is a matter of magic and mystery, and so music echoes the magic and mystery of the universe.



This is a companion piece to the essay above. There I was making larger points about musical expression, mainly that our response to expressive qualities in music is partly personal, partly universal, partly according to the ways our culture has taught us to respond. Here I want to examine some pieces that seem to me to be going after particular emotions. As always when talking about such things, a certain amount of this will be personal. But I maintain that it’s not entirely personal, that definable qualities can be portrayed in music (most manifest when joined with words). As someone who’s highly emotional about music and a composer who is determined to write expressive music, I’m on familiar ground with these matters.

As I also said in the above essay, I think music is good at expressing broad emotional categories, not so good at more subtle ones, which we have to supply for ourselves. To get into specifics I’m going to start here with vocal music, because in that case we can know with relative confidence what the composer intended to express.

I’ll start with a remarkable juxtaposition of qualities in Bach’s B Minor Mass. The central part of any mass setting is a juxtaposition central to the faith: Christ is crucified for our sins and resurrected from the dead. These mass movements are the Crucifixus, always tragic, and the Et resurrexit, always joyful. Here Bach paints the crucifixion with a variety of colors, some traditional and some invented.

His Crucifixus is set over a chromatically descending bass line, which is a Baroque convention for representing sorrow. Meanwhile the bass repeats over and over, which suggests something relentless and inescapable; the steadily repeating rhythmic figures work to the same end. Over the bass, the lines of the chorus are drooping, mostly set low in the voices, and the instrumental parts also lie fairly low, so the texture and color of the music are shadowed. The harmonies over the chromatic bass are dissonant and unsettled. In the voices the opening phrases are scattered, broken. So every aspect of the music—melody, harmony, rhythm, color—is geared to expressing tragedy.

The Crucifixus ends with a falling and fading away, unmistakably representing a descent into the darkness of the grave. From that moment of despair the Et resurrexit erupts with a torrent of joy. Everything about this movement is the converse of the Crucifixus: the rhythm whirling unto delirious, the melodies soaring and dancing and laughing, the colors bright with high strings and winds and ecstatic trumpets. That transformation is one of the most stunning I know in music, and Bach maintains the joyous celebration from beginning to end of the movement.

There’s joy and there’s comedy. For the latter I’ll present three examples. Beethoven is comic more often than he gets credit for. One of his more overt bits of humor is the last movement of the Second Symphony, which begins with a gigantic hiccup—or maybe the cry of a jackass—that turns out actually to be a primary motif of a high-spirited and waggish romp. Stravinsky’s Renard the Fox, from 1916, based on a Russian folktale about a clever fox who gets his comeuppance, begins with an outlandish march; the entry of the voices, imitating the animals in the story, is brilliant and uproarious. From that point the tone oscillates between comedy, parody, and irony.

Brahms is not generally noted for humor in his art, but that’s a bit of a bum rap. One of his comic outings is the finale of the String Quintet in F Major, which is not thigh-slapping stuff but still suffused with wit, especially in a refrain that comes back over and over, a goofy tune scored for high violin doubled by cello octaves below. In his person Brahms was a subtle ironist and also a practical joker, and now and then his music reflects that.

Charles Ives believed there was not enough laughter in the concert hall, and he intended to do something about that. Part of his humor comes from realistic portraits of amateur music-making; as he noted to a copyist, “Bandstuff! They didn’t always play right & together and it was as good either way.” Putnam’s Camp, the middle movement of his Three Places in New England, is an affectionate satire of town bands falling off the beat, playing in the wrong key, and generally losing it on the march.

The late-20th-century master Gyorgy Ligeti has the reputation of a fearsome modernist, but he was in fact a passionate humanist and also one of the funniest composers—though his comedy tended to have an edge. His Adventures is a chamber opera with no words, only inarticulate sounds from shouting to laughing to purring. I had to see it live and acted out to realize how hilarious it is. More to the point is his series of Nonsense Madrigals on poems by Lewis Carroll. Those bits of surreal whimsy have never been better realized. Try “A Long, Sad Tale” for a sample: a collection of bizarre vocal manifestations adding up to a little scene both comical and shivery. His opera Le Grand Macabre is a hyper-pastiche of apocalyptic import. I suggest watching a staged version on YouTube. Also try the hyperbolic Simon Rattle/ Barbara Hannigan version of the excerpts Mysteries of the Macabre, figuratively and literally a scream. There’s a lot of laughter in Ligeti, even if much of it is uneasy.

When it comes to tragic pieces there are inevitably lots of examples in religious music. One of the most powerful I know comes from the relatively obscure Italian Giacomo Carissimi, the final chorus of his oratorio Jephte, from around 1650. The story is of the Biblical king who promises God that if he wins a battle he will sacrifice the first person he sees afterward. These promises usually don’t go well; it turns out to be his beloved daughter Jephte, and this time, unlike with Isaac and Abraham, God doesn’t come to the rescue. The piercingly tragic and beautiful final chorus, Plorate, filii Israel, is a lamentation by her friends. The music is sustained wail of grief, building to a climax on a chain of heartrending harmonies. Chorus singers report having trouble getting through the piece without choking up.

A madrigal is a piece for usually four or five singers, designed mainly to be done by enthusiasts at home. The English madrigalists of the Elizabethan era were expert at minutely portraying every image and shade of emotion in a poem. The Silver Swan by Orlando Gibbons presents a literal swan song, its mournful course ending, “O Death, come close mine eyes! / More Geese than Swans now live, more Fools than Wise.” Like many mournful pieces, it gets much of its effect from falling vocal lines and poignant dissonances. Also from that era, John Dowland wrote many songs for voice and lute, most of them on his own verses. His signature lute song, one of the great hits of its time, is the exquisite Flowe my teares.” Its lament builds steadily to its devastated last verse: “Hark! you shadows / that in darkness dwell, / learn to contemn light. / Happy, happy they that in hell / feel not the world’s despite.”

Beethoven said that even thinking about the Cavatina in his late String Quartet in Bb Major, Op. 130, brought him to tears. Toward the end, in a section marked “anguished,” the music seems to break into sobs. As in so much late Beethoven, the suffering here is not simple but complex, blending moments of hope, regret, universal human grief.

Among the 20th-century Viennese sometimes called chilly intellectuals, it was Anton Webern who after the death of his mother wrote the devastated Six Pieces for Orchestra. I know of no more naked depiction of anguish than its fourth movement, a funeral cortege that resolves into a mounting roar of percussion punctuated by shattering cries in the brass, ending on a sudden and devastating silence. This is not sorrow; this is agony.

Let’s touch on a couple more qualities. First, heroic. The obvious choice here is the first movement of Beethoven’s Symphony No. 3, which he titled Eroica (originally called Bonaparte, so it is about Napoleon). The first movement is a kind of abstract portrait of a battle or a campaign, searching and unstable, building to grand perorations. The second movement portrays the aftermath of battle–a funeral march. I’d add the finale of Beethoven’s Fifth Symphony, which breaks out in a brassy song of triumph and sustains a festive and heroic tone throughout. (Beethoven appears in this essay more than anybody else partly because I think his work is some of the most emotionally intense and also expressively transparent in music.)

There’s love music everywhere, it’s much of what we sing about, but how about pieces that shade into the erotic? One example is the exquisite slow movement of Mozart’s Piano Concerto No. 21 In C Major, which for me has always evoked a languid evening of lovemaking. (In the 1960s it became a hit accompanying the lovers in the movie Elvira Madigan.) The most graphic example of sex in music, no doubt, is Wagner’s Prelude and Love Death from Tristan und Isolde. The opera is about a fatal amour fou. The Prelude starts with a series of sighs representing the rising of desire; it builds to an earth-moving climax. (In a letter Wagner compared the opening gestures to the Hindu breath of Brahma that creates the universe—a lovely image of creation as the desire of Unbeing for Being.) In the Love Death, Isolde expires in ecstasy on the body of her lover Tristan, and the music does justice to the image.

A subset of the matter of expression is what are called musical topics. These are pieces evoking a fairly specific image by use of musical gestures the culture has developed for those purposes. The most familiar topic is a march. Marches can be created in all kinds of styles, but certain things will be consistent: it will be in two-beat or four-beat meter (because we have two feet), and it will be in some kind of marchable tempo; the melodies and gestures will be vigorous, the overall style something associated with its culture. (This includes the “Turkish” style used by Beethoven and Mozart, heavy on bass drum and cymbals, which came from Turkish military music. See Mozart’s famous Rondo alla Turca, for piano; the finale of Beethoven’s Ninth also has a Turkish march.) Marches, after all, are used to move troops around and sometimes even accompany them in battle. They will be geared to give people courage and strength, and even when translated into concert music they will still have some of those qualities. In the Classical style of Mozart and Beethoven, most concerto first movements are in a more or less military mode.

All dances amount to a topic of some kind, each with its characteristic rhythms, gestures, forms, and tempos: minuet, englische, waltz, and so on, their rhythm, style, and form geared to the steps of the particular dance. In turn, every dance has its social overtones, from the courtly tradition of the minuet, to the more democratic associations of the 19th-century englische (a contradance in which everybody danced with everybody), to the sexiness and delirium of the waltz (it was the first dance in which couples embraced), to the vigorous two-beat polka, originally a Polish folk dance. Brahms described his Fourth Symphony thus: “Oh, once again I’ve just thrown together a bunch of waltzes and polkas.” He was partly, but not entirely, kidding; the third movement is sort of a polka, and all the movements have dance overtones. Much of Chopin is founded on Polish dances such as the mazurka and polonaise.

Another familiar Western musical topic is the pastoral. This music evokes fields and woods, perhaps inhabited by amorous nymphs and shepherds and such. From Handel’s Pastoral Symphony in Messiah to Beethoven’s Pastoral Symphony, this topic is going to feature a gentle, warm atmosphere, folkish tunes, bagpipey drones, simple harmony and rhythms from gentle to vigorous folk-dances. A sort of subset of the pastoral is “hunt” music, leaping and rhythmical, often with hunting calls in the horns; the finale of Beethoven’s Violin Concerto is a hunt piece. Often in the pastoral mode there will be a suggestion of the ancient, imagery of Greece and Rome. Mozart’s Concerto No. 21 above is in a serenade topic, meaning night-music.

Other topics include the passionate and exotic “Gypsy,” aka “Hungarian” style, used by composers from Haydn to Brahms, who wrote a collection of popularistic Hungarian Dances, but also used the style for all his concerto finales. Another late-Classical topic is Sturm und Drang, meaning a highly expressive, intense, bold style in some cases, inward and subjective in other cases. Not all emotional expression is covered in the list of topics, but a lot are. Mozart, to name one, was usually wielding some topic or other in his works. We may no longer recognize them, but many in his time did. A whole piece does not have to stick to a topic—a symphony movement, say, may have marchlike or Sturm und Drang moments en route.

And so it goes. Music expresses particular emotions sometimes with the help of words, sometimes with the help of cultural traditions, sometimes in more mysterious ways having to do with our innate responses to rhythm, tempo, harmony, the colors of sound. Some music is transparently expressive, some not. Ultimately each of our responses is particular to ourselves, and that individuality of response I call one of the most splendid things about music. What I’m encouraging here is not to take my responses as any sort of final word on what any piece “means,” but rather to cultivate your own responses in kind. The first time you hear a work, maybe it’s best to take it in without thought. But as you return to pieces, begin to find your own resonances, your own paths into the work, and watch them evolve as you evolve. As we change and grow, the things we love change and grow with us.


Leave a Reply

Your email address will not be published. Required fields are marked *