Greg Sandow

I learned atonal music because I had to.

But don't worry—I'm not going to write another "composer forced to write atonal music" piece. I wanted to learn atonal music. I was hungry for it. I'd decided in 1969 to become a composer, without much formal music education. So I taught myself, and later—still largely self-taught, though I'd had some encouraging composition lessons from John Heiss—got a master's degree in composition from the Yale School of Music.

And much of what I taught myself was atonality. I knew very little about 20th century music, so I studied Stravinsky, Bartok, Schoenberg, Berg, and Webern, looking hardest at Schoenberg and Webern. I solfèged my way through parts of the Schoenberg Fourth String Quartet, practiced Schoenberg piano music, and best of all—not just because I loved the music, but also because it isn't all that hard to play—spent hours at the piano working on the Webern Piano Variations.

But though I'd listened to most of the atonal classics on recordings, I didn't hear them very often live. Which of course wasn't my fault; how many live performances of the wildly crazy Webern Op. 18 songs for soprano, E-flat clarinet, and guitar have any of you heard, or even heard about? Or of Webern's two cantatas? Or Schoenberg's Wind Quintet?

Live performance really is an issue here, because when you learn this music from scores or from recordings, you don't have to confront it. You don't have to ask what it means; you don't feel its full impact. You might not wonder (as I did, after a really good live performance at the Bard Festival a few years ago) whether the Schoenberg Serenade, delightful in small doses, isn't oppressively obsessive when you hear it as a whole. Or whether those whirling little fiendish Op. 18 Webern songs, so delicious when you hear them with the privacy of headphones, might not be too evanescent to make sense in live performance. They're over before you've had a chance to tune your ears to them, and the ratio of study and rehearsal time to the length of a performance must—for most working musicians—be just about unsustainable.

That's just one reason why I think we don't find the real meaning of these pieces—whatever that may be—until we hear them live. I say "whatever that may be" because I want to ask what meaning—impact, import, gut significance—atonal music really has. Schoenberg liked to say he wrote atonal music exactly as he'd written the tonal stuff, a statement I'd put next to his wistful confidence that someday listeners would whistle all his tunes. I've whistled some of them myself; my favorite is this one, from the Fourth Quartet (not because it has to be his best, but because it's so wonderfully whistleable):

Excerpt from Schoenberg's Fourth Quartet

But still I don't think that Schoenberg is right. I can't believe atonal music is different only in technique from tonal music, or different only in the details of its harmony. I suspect it marks a notable break with everything before it in music history (despite all that old-fashioned talk about the inevitable evolution of atonality—and then serial music—from intense chromaticism), and maybe also a notable break with everyday life. Certainly people can't be right when they say that Schoenberg can be listened to like any other music. He demands exhaustive concentration. When I try listening to him casually, in the background while I'm concentrating on something else, his music makes me nervous, even if I've sung the scores.

And even if that might not be true for everybody, there have to be huge differences between atonal 20th century music and, let's say, Guillaume de Machaut. Or Mozart or Puccini. I was amazed, last month, by a concert review in which I read that Russell Sherman played sonatas by both Schubert and Leon Kirchner with "an intensity [as the critic wrote] that rendered their differences in language beside the point." How could those differences not matter? And why on earth should we be glad to think that they didn't? Would anybody talk about painting that way? Would anybody say that—if you look at their work with enough intensity—any differences between Jackson Pollock and Delacroix are beside the point? Or that, given strong enough productions, we'd have more or less the same theatrical experience at Our Town and Ubu Roi?

I think some of us have spent too much time defending atonal music. We're defensive when we say it's just like any other music—which can't be true, since equivalent propositions would be laughable in any other form of art.

So what's the message of atonality?

To see why this question might really have an answer, consider what we say about other kinds of concert music. Music of the classical period, we say (and granted, these are stereotypes, but there's truth in them), is elegant and balanced. Romantic music is emotional. So if we move down this road into the 20th century, we might say atonal music is full of angst. Oddly, this almost sounds like heresy, as if we're insulting Schoenberg and his school. But atonal music really did emerge as a form of intense expressionist art. Schoenberg pieces like Erwartung or Pierrot are unstable, whipped toward emotional extremes.

In support of this, we might enlist an everyday reaction to atonal scores (at least the early ones): They make people think of horror movies. Or pain, or other kinds of instability. Some years ago I brought a date—a musical civilian—to hear the Emerson Quartet, which among other things was playing the Op. 3 Berg quartet. It sounded "drunk," my date said (which didn't mean she didn't like it).

But against this view, we can cite works like the last two Schoenberg string quartets, or his Wind Quintet. And, sure, when Schoenberg invented 12-tone writing, he became a kind of neoclassicist.

But then Theodor Adorno—surely the most powerful intellect who ever wrote about these things—thought that ceaseless dissonance embodies frozen pain. He didn't mean that Schoenberg, Berg, and Webern (who'd been his friends) were anguished people, though I can well believe that Schoenberg might have been. He meant that anguish, in the first decades of the 20th century, had become an everyday reality for just about everybody, at least beneath the froth of everyday life. (Not an unreasonable idea when the world was plunging into economic chaos, war, and massacres, though Adorno took it all a lot deeper: He thought that—quite apart from obvious catastrophes—the capitalist machine was eating away at everyone's autonomy.)

Following Adorno, we might say that early atonal expressionist works put anguish front and center. And then 12-tone music submerged it in a system, though (quite plainly in Berg and Schoenberg, more distantly in Webern) the expressionist rhetoric persisted. Postwar serialism hid the pain entirely; this, we could say, was music in extreme denial. (I more or less did say that, in an essay on Milton Babbitt I wrote in 1982.) From this point of view, it's hardly a surprise that George Rochberg—when his son died, and he wanted to express his feelings musically—abandoned 12-tone writing and went back to tonality. That doesn't mean tonality has to be any kind of paradise, or that it offers any guarantee of spiritual freedom. It has traps of its own. But at least it's easily identified with everyday emotion.

And to support my earlier contention that it marks a break with everyday life, consider Pierre Boulez, who says atonal music expresses new emotions, appropriate to the age we live in. Not, in other words, the silly old sentimental feelings older art indulged. No more raging over lost pennies! Or sighing with infinite sadness and regret when the world we grew up with disappears forever. Music—now and forever in the future—has to be above such things.

Or consider Schoenberg, in whose first atonal piece—the last movement of the second string quartet—a soprano sings a poem by Stefan George that starts, "I breathe the air of other planets." If we take those words seriously, we might decide that Schoenberg's early atonal works are more authentic than his 12-tone stuff, an opinion I wouldn't be alone in putting forth. (Boulez used to hold it, for instance, and maybe still does, though now he conducts Schoenberg works he once apparently despised.) The 12-tone pieces, after all—or at least the neoclassic ones—pretend to be plain old concert music, even using what we might call cubist versions of plain old traditional forms. But their musical language might have pointed them toward more radical paths. (That was Boulez's view.) They're in denial!

But here's another question. How do I experience atonal music?

Well, let's start with this—I've written it. Not so much in my student days; I had trouble with it, as I'll explain. But in recent years I've written 12-tone stuff, which may surprise people who've heard the first pieces of mine that have begun to circulate now that I've resumed composing. Mostly they're triadic. But I've also written a 12-tone sonatina for piano and violin, which has one movement, out of three, that I'd love to have performed. (Not coincidentally, that movement is, of all of them, the most radical.) I wrote this as a composition exercise, and in some ways that's how I think of writing 12-tone music today—it's an archaic enterprise, but useful for developing composing muscles, much as species counterpoint used to be in days gone by.

With that in mind—and also just to amuse myself—I often scrawl 12-tone musical doodles when I'm bored at conferences. I'll draw staff lines on whatever paper gets handed out, and write 12-tone melodies, canons, and chord progressions, having fun with the progressions by building them from made-up rules.

But then I also wrote—quite seriously, taking lots of time—three 12-tone sections in a piece I'm just now finishing, a long set of variations for string quartet on the theme of the last movement of Mahler's Third Symphony. Most of the variations are tributes to things I like, musical and otherwise; the 12-tone ones are tributes, no surprise, to Schoenberg, Berg, and Webern. I wrote them more or less in the styles of those composers, amusing myself, in the Schoenberg variation, by reconstructing parts of the Fourth Quartet from memory. (I knew, of course, that I hadn't memorized enough of it to make my reconstruction anything but a new piece of my own. But see how all that solfège came in handy?)

I guess the Webern variation is my favorite of the three; it's a double canon in inversion (elegant and airy, if it comes off the way I'd like it to), with each voice making one statement of the 12-tone row.

But is this music full of angst? Beats me. I didn't feel any when I wrote it, or when I hear it now. The Schoenberg variation sounds a bit ungainly to me, but then so does Schoenberg. And in any case I wrote these variations at a distance, so to speak, looking through Schoenberg's, Berg's, and Webern's glasses. Maybe that helped keep the angst at bay. Though it does strike me that all three 12-tone variations are linked to fragmentation in the larger progress of the piece, and the Schoenberg variation—coming after a theme and eight previous variations in D major and D minor—sounds like a breach.

But what's more important, I think, is that all three 12-tone variations sound a bit abstract to me, even the Berg variation, which is meant to be a lively little dance. This brings us back to my comments many months ago about how atonal music is in some ways a musical equivalent of abstract art. Whatever arguments I made in favor of that at the time, I see now that one reason I believe it is, very simply, that I think atonal music sounds abstract.

What do I mean by that? I remember the first time I heard the Schoenberg Fourth Quartet live (which, maybe not so incredibly, was only a few years ago, again at the Bard Festival). It didn't strike me like a structured chamber piece. It reminded me, instead, of a mobile—something static that revolves in space.

If you draw lines around the aggregates—statements of the full 12 notes—in a Schoenberg 12-tone piece, you see them show up in a reasonably regular array, at the approximate rate of one aggregate every one or two bars. In aural terms, that means that clumps of all 12 notes are always present, or, more specifically, that in the mid-ground motion of the piece, all 12 notes are always sounding simultaneously. And that means the harmony can't go anywhere; it's always saturated with all 12 notes. Certain pitches might get momentary emphasis, but the longer any 12-tone piece goes on, the more those emphases cancel themselves out.

Which makes impossible, in Schoenberg's 12-tone works, something he loves about tonal music. In one of his letters, if I remember this correctly, he complains about bad performances of tonal works, performances that don't react to modulations. The harmony changes color, he says, but the performance doesn't change with it. That makes him feel lost; he knows the music has gone somewhere new, but he can't tell how it got there.

But how can he move to any new harmonic area in a 12-tone piece, when all 12 notes are always in effect? How can we even define harmonic areas, when all 12 pitch-classes are all in play at once? How can we change harmonic color? I'm not saying 12-tone music doesn't have harmony; of course it does. Notes have to be combined together, and how you do that matters; I spent a long time, in my Berg and Schoenberg Mahler variations, picking the exact notes I wanted in every chord. And this harmony will have some kind of texture. The notes in the chords can, for instance (to give just the most obvious examples), be widely spaced, or crunched together.

But this harmony will very likely have no color, no depth, no real location, even if analysis reveals fancy pitch configurations, and despite any contextual devices (rhythm, most obviously) that might define a destination for each phrase. In the 12-tone repertoire, one exception to this might be parts of Lulu, where I sometimes do hear real harmonic movement, for example in Lulu's scary rising cries for help at the end of the first scene of the second act. But then in other places Berg seems to cancel the harmonic motion, for instance in the "Lied der Lulu," earlier in the same scene, which from the first day I heard it, many years ago, has always seemed abstract to me.

I noticed this atonal abstraction when I was getting ready to write this column, and listened to chunks of atonal music. The Schoenberg Third and Fourth Quartets sound limpid too me, though they bang around sometimes, and certainly are witty (or maybe zany is a better word; the zany high point, in these two works, comes in the second movement of the Fourth Quartet, where the sounds get really weird). But the music—for all its references to standard forms—doesn't walk down any streets I know. Movements end sometimes with a thump, hiding their lack of any audible harmonic destination with a violent rhythmic blow. Or else they just vanish into air, like a solid substance dissipating into gas; their harmony dissolves, and they end with phrases that are shaped like conclusions, but which finish on pitches that don't sound tied to those that went before.

But doesn't the row give these final notes a context? Sure. They're not unexpected; we get used to the progression of the intervals that make up the row. But on the other hand, these final notes don't sound especially final; they're just the notes that come next. Coming last confers no special privilege; it's just another, none too special, way of being next.

So what these conclusions offer, in my view, is a kind of discipline. "Accept the notes I give you," the row says, in effect. "You'll be a better person—more rigorous, less self-indulgent." I feel that I can hear Webern doing this at the end of his Op. 15 sacred songs. That's not a 12-tone piece, but it's rigorous in another way. It's a canon, and as the last remaining voice takes its measured steps to finish, the notes don't sound to me like pitches Webern ever would have written if the canon hadn't offered them.

He lets those pitches shape his music. And I hear much the same thing in the chorale (or sort-of chorale) that ends his quietly radiant and modest second cantata. I'm sure the chorale has some fascinating 12-tone structure, but the notes don't come to anything I'd call an audible end; they simply stop. Again—such discipline! The final movement of the Piano Variations seems like it's going to ground itself, on its final page, on repeated low E flats (which peal out softly like a deep bell ringing), but in fact the movement ends in a spray of dust, a scattering of notes that go a different way, because the row requires them.

Which leads me to another thought. If 12-tone structure defines a context for each note, that context is different in every piece. Whereas in tonal music, the harmonic context of each note is something shared, common ground for all of us. Maybe that's why atonal music sounds to me as if it lives in some place not defined by ordinary life—the quantum realm, perhaps. (I might add that the boundary between tonal and atonal music isn't fixed, and that the harmony in free atonal pieces doesn't work the way it does in 12-tone works. In free atonal music, pitches can be reserved for special purposes. A musical phrase might use 10 of the 12 pitch-classes, introducing the last two only as it ends. This can create harmonic motion, though mostly of a fairly subtle kind. Free atonal harmony, if I follow this line of thought, might really be a heavily chromatic kind of tonal writing, leaving 12-tone harmony as something very different—though its historical descent suggests, once more, that the angst associated with free atonal music isn't all that far away.

But—I have to say this—I like ordinary life. I don't think that music should abandon it. I listen to Elliott Carter's Fifth Quartet, and really love the conversation of the instruments. But what are they saying to each other? Better not to ask. The answer might turn out to be what the characters in Carter's opera What Next? talk about, which is nothing.

To say this is unfair to Carter, in one way, since the libretto of the opera is embarrassing. Paul Griffiths, who wrote it, can be an acute critic, but he's not remotely a literary writer; anyone who writes "her hair, the clear red of sunset" (I'm quoting from memory, and might not have that absolutely right) hasn't looked very hard at either hair or sunsets; he hasn't properly distinguished between earth colors (which hair has), and the colors of light and clouds.

But then Carter accepted this libretto. (I'll note in passing that when rarified atonal music does encounter normal life, the results, when Berg isn't the composer, can be embarrassing. To me, the dance of the golden calf in Moses und Aron and the chant in A Survivor from Warsaw sound a little crude, more so certainly than the rest of the works they're part of.) And thus it's Carter's fault that the entire premise of the opera is embarrassing: People in the aftermath of something undefined—they keep calling it an "accident"—wonder who they are and what they're doing. Apparently they're trapped together? Have we seen that anywhere before? How about Endgame? No Exit? Other plays and novels from, God help us, the 1950s?

What Next? comes off like a watery take on Samuel Beckett, and it's almost weird that Carter's music sounds at times so powerful, so truly operatic. The music seems to live a life of its own, heedlessly adopting operatic traits (and even creating stereotyped operatic characters, like the busybody mezzo). But the drama is definitely abstract. The people aren't anybody in particular. They're abstractions, who seem to live (in spite of references to what might have been their daily lives) on one of Schoenberg's other planets.

(Parenthetically, I'll add that this piece has—or ought to have —about the same profile in our current scene that Strauss's Four Last Songs had in Samuel Beckett's time. It's a survival from the past, an example of a style that's no longer current, not even remotely. So why don't we see it that way? Why we do we treat Carter as somebody whose work we have to seriously confront, instead of as a beloved older figure, as much out of touch as Strauss seemed in 1948?)

Compare what happens when I listen to Michael Torke (his six CDs of previously recorded work, for instance, which he's reissued on his own label, and which make me smile). Or Steve Reich, or Phil Kline's exhilarating new CD Zippo Songs, or David Garland's new CD On the Other Side of the Window (I'd like to give a shout to Garland, who has quietly been writing and performing songs for 20 years or more, and does it with a homemade kind of quiet force). Or lots of other current music... I hear it, and I'm on familiar streets. I might be surprised; the music might do nothing I expect. But I know where it's doing all those things. I know where it lives. I can draw the line from it to me—which, with atonal music, I can't so easily do.

Or at least I can't, and still stay inside normal life. Though I did make one connection that surprised me. I've been listening to Strauss's opera Intermezzo, which strikes me as a neglected gem, and also offers some of Strauss's most complicated music, approaching Schoenberg in complexity—not vertical complexity (it doesn't have Schoenberg's many-legged counterpoint), but complexity that's horizontal, or sequential. So many ideas follow each other in such rapid succession that I get dizzy listening. It's no surprise that this is one later Strauss work that Schoenberg liked.

But periodically throughout the opera, and above all at the end, Strauss pours out his familiar huge, enveloping orchestral Technicolor. This can sound sentimental. The opera shows us a husband and wife (based on Strauss himself, and his wife Pauline), having troubles, and then coming back together. To accept the music at the end, you have to allow enormous emotion, even transfiguring emotion, at the heart of marriage, and also that this emotion, while in some ways cosmic, is also deeply ordinary. I can do that; someone else might find it too extreme.

But Schoenberg, too, wrote an opera about modern marriage, Von heute auf Morgen, a piece that's hardly ever staged. I have a CD, but when I started to think about all this, I'd only listened to the first few minutes. After hearing Intermezzo, I wondered what Schoenberg's ending might be like. The situation is in many ways the same—there's been a quarrel, and now there's reconciliation. But in Schoenberg, the reconciliation sounds ironic. So give his 12-tone score some props; it makes at least this one sophisticated point. I can see Schoenberg's characters, at the end of this opera, certainly better than I can see Elliott Carter's. They're like figures in a modern painting, sitting in a room full of modern furniture, the whole thing slanted at an angle that's odd, and not entirely comfortable.

That got me interested, so I went back to the start of the opera, and quickly got bogged down. There are wonderful details in the music (an amused, exasperated musical shrug from the wife, for instance). But then of course there are wonderful details; this is Schoenberg. My problem was that the concentration I had to muster to follow what was going on—in music that, in the end, tells a pretty simple story—just exhausted me. There's a big disproportion here between means and ends. We have to work too hard for a fairly light reward; Schoenberg asks us to exhaust ourselves over and over again, beating us limp over a longer span. It's no surprise that the opera is almost never staged—or that Michael Gielen, who conducts the CD I have, writes in liner notes that the piece has elements of horror. Elements of horror, I might add, that Schoenberg never intended. Here we run again into Adorno's argument—Schoenberg's dissonance is frozen angst—perching, like a great black bird, on a roost where it seems right at home.

Which brings me to the exhaustive—and exhausting—complexity of Schoenberg-influenced atonal music. Though "complexity" might be the wrong word (Mahler is complex and Beethoven is complex, but not like this); maybe we should talk of Schoenberg's density. There's a lot going on in his music, both sequentially and vertically. You have to concentrate—force yourself to concentrate quite furiously—to grasp even the surface of the music. I'm sure that's why it doesn't work for casual listening; if I listen without full attention, the notes keep jabbing at me, and I can't tell why.

And it's hard to know what all that density really gives me. What emotions do I feel? Webern touches me (he sketched his later pieces, when he first planned them, as successions of feelings, not as 12-tone structures). Berg shakes as well as touches me. But in Schoenberg's neoclassic 12-tone pieces I hear just two emotions, strain and wistfulness, which of course might be my own limitation. But I wonder. In Carter's Fifth Quartet, I feel abstract emotion. When all the instruments come together in a hush, I'm moved, but I'm not sure by what. That's not a complaint; I don't have to know, any more than—in Antonioni's film L'avventura—I have to know what happened to Anna, the character who disappears early on. Critics with conventional minds couldn't believe that Antonioni didn't solve this mystery. Where's Anna? But it was at least subliminally clear to me (even when I was in high school, and first saw the film) that we didn't need to know, and in fact shouldn't know. Anna's leaving opened up a space that revealed the wasteland in the lives of all the people who remained.

There's something similar in Carter (or I suspect there is), a cluster of emotions that are more evocative if they aren't named. Thus we avoid sentimentality—but also leapfrog over everything that actually happens in our lives, with all the feelings inevitably attached. Antonioni doesn't leapfrog like that, despite his formidable modernist credentials. There's a huge gap, I think, between musical modernism, and modernism in the other arts; the other arts never lost touch with ordinary life, no matter how stringently they examined it.

I'd love to ask Boulez—since he thinks atonal music depicts new emotions—what he thinks happened to the old ones, the love, let's say, of a husband for his wife, or a mother for her child. Are these unworthy now? Are they unfit to be addressed by our wonderful atonal music? And, Mr. Boulez, is your extreme reticence about your personal life related to your feeling that music shouldn't deal with everyday emotions?

If he's right, I'm sorry—I opt out, as I guess I did in graduate school, when I couldn't bring myself to write very much atonal music, as all my teachers expected me to.

But I'll end with this. In spite of everything I've said, I love the density—the genuine, if sometimes almost screeching complexity—of atonal music, and as a composer, as well as very simply as a musical person, I've been inspired by Schoenberg and the rest, no matter how hard I've campaigned in the past against the influence of all their heirs. I doubt that there's a bar in my music that's not influenced in some way by them.

I just listened, while I read over what I've written here, to the first act of my opera Frankenstein, which I had fun writing in the style of traditional Italian opera. But the orchestration changes constantly. Where did I learn those color changes? From Bellini? Not a chance. And not even from Strauss or Wagner. They're Webern; that's where I learned them. We can't pretend atonal music never happened, or that it's not important, or that because it happened in the past, we can all ignore it now. It's a crucible that helped to form us, even if any of us don't remember that it did. We can't escape it, and we shouldn't try.

So I let it go on inspiring me. I bought a digital music player, an iPod equivalent, with a 20 gig hard drive, and I've been filling it with music. Much of the music that I'm putting on it is atonal. Schoenberg, Carter, Webern, Berg—that's part of my answer to the pointed question a gadget like this poses: "What music do I want to have, ready to play all the time?" The Beethoven symphonies? No way, much as I've learned from them. I do have some of them, made taut and alive under David Zinman, on my player, but I'll delete them when I'm through with them. Likewise I'll delete that reissued Pagliacci, despite the jawdropping cast of Jussi Björling, Victoria de los Angeles, Leonard Warren, and Robert Merrill.

But the Schoenberg quartets, the Webern cantatas, and much else that's 12-tone and atonal—these works are staying. I can learn a lot from them, and use it in everything I write.