The Wild World of Music

What can elephants, birds, and flamenco players teach a neuroscientist-composer about music?
The illustrated head of musician and scientist David Sulzer sits at the center of a network of symbols for the brain...
After keeping his careers separate for years, David Sulzer is using science to expand our idea of what music can be.Illustration by Jérôme Berthier

Luk Kop didn’t seem to have the makings of a musical prodigy. He didn’t hum made-up tunes to himself as a youngster or shake his head when someone sang flat. He didn’t build instruments out of sticks and gourds or blow trumpet solos as a five-year-old. He had a brief moment of fame as a child actor, in the Disney film “Operation Dumbo Drop,” but grew into a sullen and ungainly teen. When the composer and instrumentalist Dave Soldier first met him, in Thailand, in 2000, Luk Kop spent most of his time eating grass and hanging around with the other elephants. He’d been deemed too truculent to mix with tourists.

Soldier was in Thailand to recruit musicians for an elephant orchestra. He had hit upon the idea with Richard Lair, a conservationist and adviser at the Thai Elephant Conservation Center, where Luk Kop lived. In the spring of 1999, when Lair was on a research trip in New York, he and Soldier stayed up late one night at Soldier’s place in Chinatown, talking about elephant art. The Russian artists Vitaly Komar and Alexander Melamid had recently taught some of the sanctuary’s animals to paint with oils by holding brushes in their trunks. The results were exhibited at the Museum of Contemporary Art Australia and auctioned off at Christie’s for more than thirty thousand dollars. One critic compared them to Abstract Expressionism. But bright colors on a canvas are easy to like; music is a harder sell. To anyone other than a parent, a grade-school orchestra sounds like a crate of instruments falling down a staircase. Why would elephants be any better?

Asian elephants have been trained by humans for more than four thousand years. They’ve learned to pull plows, carry tree trunks, clear paths, and trample armies. Some female elephants are both so intelligent and so even-tempered that villagers in Thailand have used them as babysitters. Still, the orchestra was a stretch. Studies had found that elephants could identify simple melodies and distinguish pitches as little as a half step apart. But that didn’t mean they would make good musicians—at least of the sort that play in an orchestra. When Soldier explained his plan to the elephant trainers at the sanctuary, they reacted with “slightly irritated bemusement,” he later recalled.

The first challenge was making instruments. Anything an elephant played had to be weatherproof and extremely durable. It had to be operable without hands or fingers, and it had to be very large. Working with Lair and the carpenters at the sanctuary, Soldier built an elephant-size xylophone, a drum, and a single-stringed instrument that looked like a washtub bass. The blade of a huge circular saw that was abandoned by a tree poacher in the forest was turned into a gong. Initially, Soldier had all the instruments tuned to a C-sharp pentatonic scale, so they would sound good together. Then he mixed in more instruments and tunings. A metalworker in nearby Lampang built some marimbas and thunder sheets; a Canadian artist designed a synthesizer that the elephants could play with their trunks; and Soldier brought in bells, harmonicas, and mouth organs from northeastern Thailand. Within four years, sixteen elephants were playing a full orchestra’s worth of instruments.

Luk Kop had never touched a drum in his life. And yet, when Soldier set one before him and handed him a stick, he grabbed it with his trunk and quickly learned how to wield it. Elephants tend to keep a steadier beat than humans do, a study by the neuroscientist Aniruddh Patel later found, and Luk Kop’s sense of timing was uncanny. Soldier rewarded the elephants with apples and bananas, or petted their fat pink tongues, which some elephants love. But Luk Kop didn’t need much encouragement. Soon enough, he was improvising drum solos like a giant Ginger Baker. “He was particularly talented,” Soldier told me.

After that first trip to Thailand, Soldier played some recordings of the elephants for a music critic at the Times. He didn’t mention who the musicians were. The critic listened intently for a while, then ventured that the group must be Asian. He could tell from the repertoire, he said, though he couldn’t identify the players. Soldier must have been delighted by that, but he insists that he wasn’t trying to catch the man out. He was just posing a question, in the most direct way possible, that had preoccupied him most of his life: What makes music music?

Soldier is of two minds. As a composer and violinist, he doesn’t like to define music too strictly. He prefers to mix genres, blur categories, erase the boundaries between rock and classical, melody and noise, animal and human. “Humpbacks are down at the bottom of the ocean singing every day,” he says. “Is that any different from you practicing the fiddle or guitar at home?” Soldier will be sixty-seven in November and has been a fixture of New York’s downtown music scene since the early eighties. He has composed string arrangements for David Byrne and John Cale, operas with Kurt Vonnegut, and cartoon scores for “Sesame Street.” To Soldier, it’s all of a piece. Once, in the same week, he played a gig with Pete Seeger and opened for Ornette Coleman. “It was like talking to the same person,” he told me.

“I don’t have to look at my watch to know it’s just past noon.”
Cartoon by Frank Cotham

But music is just Soldier’s night job—the thing he does when he’s not at the office. By day, he has another name and identity: David Sulzer, professor of psychiatry, neurology, and pharmacology at Columbia University. Sulzer specializes in autism and Parkinson’s disease, and has done research on Alzheimer’s with his wife, Francesca Bartolini, an associate professor of cell biology at Columbia. For years, Sulzer was careful to keep his two careers separate—his music was rarely about science, and his science touched only glancingly on music—and they came to draw on opposite sides of his personality. David Sulzer is a reductionist, trying to pinpoint the brain’s essential mechanisms. Dave Soldier is an iconoclast, trying to expand our idea of what music can be. It’s only relatively recently that the two have begun to work together—to see what science can say about music and vice versa.

“When someone asks what my real focus is, I say it’s the basal ganglia,” Sulzer told me one afternoon. “That’s where the sensory information from touch, hearing, and vision all converge.” We were standing in his lab overlooking the Hudson, at Riverside Drive and 168th Street, staring at a plastic brain. Sulzer had taken it off the windowsill to show me, but the model fell to pieces in his hands. “If you were one of my students, I’d have you put it back together,” he said. In early photographs, Sulzer can look dauntingly cerebral, with his lean, ascetic frame, pale bald head, and heavy-lidded eyes. But his lines have softened with age; his manner has grown puckish and approachable. He was wearing a rumpled shirt and wash-worn slacks that day, and spoke in a low, deliberate voice as we walked through the lab—the steady bass beneath his sometimes bewildering talk. We passed racks of chemicals and banks of multiphoton microscopes, grad students hunched over electrophysiology data, and a break room with an atlas of the rat brain perched on top of the refrigerator. Sulzer mentioned punk rock and synaptic pruning, dopamine and country blues. “We tried to play whale songs for elephants once,” he said. “But the technology was too poor. Elephants really don’t like wearing headphones.”

When Bartolini first met Sulzer, at a club on the Lower East Side, in 2006, she found him a little arrogant. “I mean, nobody has normal conversations like that,” she told me. Sulzer was playing with his flamenco fusion band, the Spinozas, and Bartolini was in the audience. She was sure that she recognized him from somewhere, but, when she went up to him after the show, he said that he’d never seen her before. It turned out that they shared the same morning commute, on the train from lower Manhattan to Columbia. They even worked in the same building. “I guess I wasn’t very memorable,” she says, laughing. It’s hard to imagine: Bartolini, who was born and raised in Rome, was already an accomplished cell biologist, with the sharp wit and flashing eyes of a Fellini actress. But she had misjudged him as well. What she took for haughtiness was poor eyesight. What seemed like pretension was just a mind delighting in its own connections.

Those connections are often made in the basal ganglia, a snail-shaped neuronal structure perched on the brain stem like the ball end of a stick shift. You have to cut through the skull and the cerebral cortex to reach it, so it’s hard to study, but it helps coördinate some of our most complex behaviors—music-making among them. Neuroscientists know a few of the paths that a song takes through the brain: the motor cortex directs the fingers on the piano keys; the primary auditory cortex registers the sound they make; the locus coeruleus releases norepinephrine, connecting the sound to emotion. But even the plainest tune sends signals cascading through other areas, triggering memories, analysis, and all the senses. The language of neuroscience itself is rooted in music. The word “synapse” comes from the Greek synaphe: the note that connects one octave to the next, as you go up a scale—the note “that brings us back to do,” as Julie Andrews sang.

“Music is so ingrained in us it’s almost more primitive than language,” Sulzer told me. An old man with Alzheimer’s might hear a Tin Pan Alley tune and suddenly recall his daughter’s name. A young woman with Parkinson’s will stand frozen on a stair, unable to move her legs, but if she hums a rhythm to herself her foot will take a step. “I know of one man who had a stroke so severe that he could barely talk,” Sulzer said. “But he could still sing.” Music is a kind of skeleton key, opening countless doorways in the mind.

The first song that lodged in Sulzer’s mind and wouldn’t leave was from Gershwin’s “Porgy and Bess”: “Clara, Clara, Don’t You Be Downhearted.” He was seven years old, sitting in his family’s living room in Carbondale, Illinois, and couldn’t shake the sound of those lush, insistent voices—the way they lapped against one another in mournful waves. He took a few piano and viola lessons as a boy, but it wasn’t until he picked up the violin, at thirteen, that he found his instrument. Bluegrass was his first love, along with the hillbilly jazz of Vassar Clements. He learned country tunes from the bands that passed through town on the Grand Ole Opry tour, and old blues from the used 78s that he bought for a quarter—Howlin’ Wolf, Little Walter. He played in the high-school orchestra, learned to play guitar, and joined a folk-rock band. In his senior year, the band opened for Muddy Waters.

It was the beginning of his double life. Music was his obsession, but science was his birthright: his parents were both eminent psychologists. His father, Edward Sulzer, had been a child prodigy, admitted to the University of Chicago at fourteen. He dropped out two years later when his mother died unexpectedly, studied film production at City College in New York, and found a job on Sid Caesar’s “Show of Shows.” The best directors had to be good psychologists, he decided. So he enrolled in a Ph.D. program in psychology at Columbia. His wife, Beth Sulzer-Azaroff, was studying education at City College when they met. While he went to grad school, she taught elementary school in Spanish Harlem and gave birth to their three children. Then she, too, earned a doctorate in psychology. They both became professors at Southern Illinois University.

The Sulzers were revolutionaries in establishment dress. Disciples of the psychologist B. F. Skinner, they believed that almost any behavior could be learned or unlearned through stepwise training. Sulzer’s father went even further—he was a “radical egalitarian,” his son says, convinced that conditions like schizophrenia were largely social constructs. As the psychiatrist Thomas Szasz put it, in “The Myth of Mental Illness”: “If you talk to God, you are praying. If God talks to you, you have schizophrenia.” Sulzer’s father knew Timothy Leary and was an early user of LSD. He did much of his research in penitentiaries, learning how to rehabilitate people in prison by offering them rewards for small changes in behavior. Sulzer’s mother helped pioneer the use of behaviorist techniques to teach severely autistic children. The medical establishment considered her patients incapable of the simplest tasks—even dressing themselves or brushing their teeth. “But she got them there, step by step,” Sulzer says.

Sulzer’s double identity seems modelled on his parents—one part establishment figure, one part revolutionary—but it’s more compartmentalized. His scientific career followed a fairly straight path at first. After high school, he majored in horticulture at Michigan State University and earned a master’s in plant biology at the University of Florida. He gathered wild blueberries in the Everglades and crossed them with domesticated plants to breed varieties that could be farmed in Florida. He told himself that he would be the first person to use recombinant DNA in plants. Then, one summer, he went to hear a lecture by William S. Burroughs, the writer and former heroin junkie. Burroughs foresaw a time when synthetic opioids would be so powerful that they would be addictive after just one or two uses. Sulzer couldn’t get the idea out of his head. Like the issues that preoccupied his parents, addiction was a behavioral problem rooted in the mind’s inner workings. It connected science to society, and society, through some of the musicians that Sulzer had known, to art. When he began his Ph.D. program at Columbia, in 1982, he had a fellowship in biology. But his focus quickly shifted from plants to the brain.

His musical career was even more unpredictable. As a college student, he took composition lessons with Roscoe Mitchell, of the Art Ensemble of Chicago, and played in honky-tonk and blues bands. In Florida, he played rhythm guitar with Bo Diddley and joined a bluegrass group that opened for auctioneers. When he first moved to New York, in 1981, he had yet to be accepted at Columbia. So he found a room for a hundred dollars a month in Red Hook, Brooklyn, and joined any band that would have him. In the first year and a half alone, he performed with roughly a hundred groups. He wore cowboy boots and leather vests to the country shows, black jeans and T-shirts to the avant-garde performances, a tuxedo to the lounge acts and Mafia parties. “It was a point of pride that you never turned down a gig,” he told me.

Sulzer sometimes wrote out parts and simple scores when he performed with jazz and classical groups, and he went on to compose pieces of his own. In 1984, he founded the Soldier String Quartet to play them. To shore up his technique, he took night classes at Juilliard with the composer Jeff Langley. It was a humbling experience. “Someone in the next room would be playing a Tchaikovsky concerto better than I could if I’d practiced for twenty years,” he told me. “And I’d open the door and the kid inside would be nine years old.”

Sulzer’s strengths lay elsewhere. His quartet had the usual violins, viola, and cello, but they could be joined by bass, drums, and singers, depending on the piece. He wanted them to be able to play anything from Brahms to Earth, Wind & Fire. “Like the more famous Kronos Quartet, the Soldier navigates waters outside the chamber music mainstream,” the Times critic Allan Kozinn wrote in 1989. “But the Kronos’s unpolished performances leave one suspecting that it adopted its repertory to avoid comparison with better quartets. The Soldier seems to be the real thing—a virtuosic band given to iconoclastic experimentation.”

The records Sulzer made never sold many copies. Yet they represent a kind of shadow history of New York’s underground rock and classical scenes. He seems to crop up in every era in the company of the city’s most daring musicians: Lou Reed, Steve Reich, Richard Hell, La Monte Young, Henry Threadgill. Still, he had little interest in being a full-time musician. “I just looked at all the guys between forty and sixty, and I didn’t know a single one who had a stable home life,” he told me. “Not even one. They were on tour all the time. Every marriage was broken up. Everyone had kids they didn’t know. And touring can just get really boring. Sitting around the concert hall for five hours after the sound check. Playing the same hits every night. Spending all your time with the guys you just had breakfast with. Even if you like them, you end up hating them.”

On weekday mornings, sour-mouthed and stale with smoke from another late-night gig, he would throw on his grad-school grunge and head north to Columbia to do lab work. He knew better than to mix his two careers: neither his uptown nor his downtown peers had any patience for dilettantes, much less crossover artists. “You could either do minimalism or serial academic stuff,” he says of the classical-music community in those days. “I did neither one, so I got harassed a lot. I was in a no man’s land. Now that no man’s land is called ‘new music.’ ” The scientific community was even more single-minded. When Sulzer was working on his doctorate, his adviser forbade him to play gigs. That’s when Dave Soldier was born. “He wasn’t fooled,” Sulzer told me. “We were in the office one time when the phone rang, and it was Laurie Anderson’s office asking for me. He was, like, ‘Dave, you fucking asshole, you’re still making music.’ ”

“Painting was her love, but framing was her passion.”
Cartoon by John Klossner

Early one evening last year, in a building on West 125th Street, a man sat in a chair with electrodes bound to his forehead. The electrodes were wired to a laptop, on which Sulzer and Brad Garton, the former director of Columbia’s Computer Music Center, were monitoring the man’s brain waves. His name was Pedro Cortes. Heavyset and fierce-looking, with a jet-black mane and deeply etched features, Cortes is a virtuoso guitarist and godfather of the flamenco community in New York. As the computer registered the voltage changes in his brain, he chopped at his guitar in staccato bursts, like the hammer strokes his grandfather once made as a blacksmith in Cádiz. Beside him, his friend Juan Pedro Relenque-Jiménez launched into a keening lament, but Cortes abruptly stopped playing.

“It’s kind of out there,” Cortes said, glancing at the lines zigzagging across the screen. “But it’s kind of awesome.” Sulzer grinned up at him from the laptop. “The skull is like an electrical resistor wrapped around the brain,” he said. Cortes was the guest speaker that night for Sulzer’s class on the physics and neuroscience of music. The students met every week here in Columbia’s Prentis Hall, a former milk-bottling plant that was later home to some of the earliest experiments in computer sound. (One of the world’s first synthesizers sat in a room down the hall, a sombre hulk of switches and V.U. meters, silent but still operational.) Sulzer’s class was based on his book, “Music, Math, and Mind,” published in 2021. He wrote most of it on the subway, on his morning and evening commute, and filled it with everything from the physics of police sirens to the waggle dances of bees. It was both a straightforward textbook and a catalogue of musical wonders—Sulzer’s first attempt to commit his strange career to paper.

Cortes was here as a musician and a study subject. He had told the class about the origins of flamenco in fifteenth-century Spain. He had demonstrated the music’s complex rhythms and modal harmonies. Now we were hearing how playing it affected his brain. The Brainwave Music Project, as Sulzer and Garton called this experiment, was an attempt to have it both ways—to join music to analysis in a single, seamless loop. First, the electrodes recorded the activity in Cortes’s brain as he played. Then a program on the laptop converted the brain waves back into music—turning each element of the signal into a different rhythm or sound. Then Cortes accompanied the laptop on his instrument, like a jazz guitarist trading fours with a saxophone player. He was improvising with his own brain waves.

The music coming from the laptop was nothing like his guitar work. It was thin and herky-jerky and oddly ersatz, like something you’d hear at a dive bar in a “Star Wars” movie. But playing along with it was a good deal more pleasant, for Cortes, than earlier experiments of this kind. Human brain waves were first recorded by the German psychiatrist Hans Berger in 1924. Berger sometimes used his children as research subjects. He knew that the brain generates bioelectricity, so he placed electrodes on their scalps and amplified the signal enough for a machine to draw a line across a piece of paper. When he had wired up his daughter Ilse, he asked her to multiply 5⅕ by 3⅓ in her head. Stroke by stroke, a jagged pattern appeared on the page: beta waves, we now call them. When Berger’s subjects were sleeping—still with electrodes on their scalps—their brains often generated longer, more slowly oscillating signals: delta waves.

Brain waves tend to reflect your state of mind. The higher their frequency—from drowsy deltas to jittery gammas that oscillate up to a hundred times as fast—the more alert and focussed your thoughts usually are. But brain waves measure only the electrical fields on the brain’s surface. They say nothing about the myriad signals coursing underneath. (The auditory nerve alone has thirty thousand axons on each of its two branches, Sulzer points out, all of which carry their own electrical charge.) In the nineteen-thirties, a neurosurgeon in Montreal named Wilder Penfield began to probe those subcurrents. He suspected that epileptic seizures were caused by rogue electrical surges in the brain, so he used electrodes to locate their origin on a patient’s head. Then he carved out a piece of the skull in that spot—the patient was wide awake during the procedure—and stimulated the exposed brain. When he’d zeroed in on the source of the seizures, he could remove the malfunctioning tissue and prevent the problem from recurring. That method is still used.

Penfield and others went on to map the whole surface of the brain’s motor cortex. They found that, depending on which spot they stimulated, a patient’s upper lip might contract, the left eyelid would blink, the right index finger would curl, and so on. The same was true of the auditory cortex, situated on the temporal lobes above each ear. By stimulating an area called the lateral sulcus, Penfield could make patients think that they’d heard a sound—a knock or a buzz or a clear tone. Nima Mesgarani, a neuroengineer at Columbia, and others have since shown that certain neurons in the cortex respond to specific consonants and syllables in our speech. By seeing which neurons are activated, you can reconstruct the sentence that a subject just heard. You can even predict which note someone is about to hear in a song: the brain can tell where the melody is going, so it seems to activate neurons in anticipation.

Yet music’s path through the brain is never straightforward. It’s less like sound travelling over a speaker wire than like data flowing across the Internet—every phrase, every rhythm and pitch, is subdivided, distributed, and reassembled over an infinitely complex network. It’s hard to even isolate the signal. When the synthesizer down the hall was first invented, Sulzer told his class, the sounds it produced were too mathematically perfect to be musical. “A pure sine wave is just so damn boring,” he said. “They had to build circuits to dirty it up.” A modicum of noise is essential to any instrument’s sound, it turns out. Reeds rasp, bows grind, voices growl, and strings shimmer with overtones. In West Africa, musicians attach gourds to their xylophones and harps to rattle along as they play. Music, like most beautiful things, is most seductive when impure.

The line between signal and noise has only become blurrier over the years. “Music is undergoing the same kind of growth as neurology,” Sulzer told the Times in 1999, two years before the first iPod was released. “We listen to so many kinds of music now, from medieval music to music from Asia, Africa, South America, and all over the world. We can use any sound, any rhythm, any kind of polyphony or phrasing.” Since then, streaming services and home studios have sent music sprawling so far outside its old categories that it has spun back around to fundamental questions: What is a song? What distinguishes it from other kinds of sounds?

Neuroscience hasn’t been much help. For all the multiphoton microscopes in his lab, Sulzer can still seem like Galileo, trying to infer the positions of planets from pinpricks of light in ground glass. “We know a lot about audition, and the pathway from the ear to the midbrain to the thalamus and cortex,” he told me. “But what gives meaning to sound? That has been pretty impenetrable.” The brain processes sound in the auditory striatum, where signals from the auditory cortex and auditory thalamus converge with dopamine. But only recently have neuroscientists learned how to identify the exact neurons involved. Adrien Stanley, a neuroscientist in Sulzer’s lab, is using a technique called fibre photometry to trace the process in mice. His animals are bred to have a special protein in their auditory striatum that fluoresces when certain neurons are activated. Stanley trains the mice to associate particular sounds with safety or danger. (A safe sound means nothing will happen; a dangerous sound means the mouse will get a mild electric shock.) Then he sees which neurons fluoresce as the mouse reacts. How does behavior follow from sound, and where is that connection processed? “That’s auditory learning,” Sulzer says. “And auditory learning is music to me.”

The ability to process sound sometimes breaks down in people with Parkinson’s and Alzheimer’s, Sulzer says. That may be why music has such a dramatic effect on them: only a very strong signal—a beloved tune or a rhythm that they hum to themselves—can bridge the gaps in their neural circuitry. But processing sound is just a start. To make meaning out of music, the brain has to make connections that Sulzer still can’t trace in his lab at Columbia. He has to look to other scientific fields. When he flew to Thailand to start the elephant orchestra, in 2000, he went as Dave Soldier, musician. Since then, his work with animals has been done mostly as David Sulzer, collaborating with experts on birds and apes.

Animals inhabit a sonic world separate from ours. Their voices and hearing are tuned to different sounds and frequencies. (The human voice ranges from a rumbling low of about eighty hertz to an ear-splitting high of three thousand; an elephant can go four octaves lower, a bat more than five octaves higher.) Yet animals and humans tend to process sound in much the same way. “The circuitry is similar in anything with a cortex,” Sulzer says. When Nima Mesgarani recorded the brain activity of ferrets, he found that certain sounds trigger their neurons just as they do ours. “Ferrets can hear human speech and pull it apart into phonemes,” Sulzer says. “Which is just nuts.” Some species are extraordinary mimics. An Asian elephant named Koshik, in a zoo in South Korea, could utter five words of Korean by sticking his trunk in his mouth. A beluga whale named Noc, captured by Inuit hunters and cared for by the U.S. Navy, learned to mimic the voices that he heard around his tank. One day, he chortled “Out!” so convincingly, by forcing his voice through his nasal tract, that a diver left the water. He thought he’d heard his supervisor.

When Sulzer first began working with elephants, he noticed that their trainers used the same techniques that his mother used with children. Like them, the elephants quickly outstripped their instruction. “We are proud when our dogs can understand five or six commands,” Sulzer told an interviewer a few years ago. But, to the villagers in Thailand, elephants seemed nearly as responsive as four-year-old children. They couldn’t understand as many words, but could carry out equally complicated verbal instructions—“Take all these logs and arrange them in a pyramid-shaped pile,” for instance. More than that, Sulzer found, the elephants were instinctive musicians, with a sense of timing and tone so deep and clear that it seemed intrinsic to their biology.

Every elephant in the orchestra had its own peculiar talents and interests. Mei Kot couldn’t stop playing the gong. Phong preferred the ranat—a kind of giant marimba. (When Sulzer was recording the orchestra’s second album, Phong walked up to the ranat with his mallet, improvised a long, intricate solo, then dropped the stick and walked away.) Prathida had excellent timing—some thought it was even better than Luk Kop’s—and a gift for finding an instrument’s sweet spot, where it resonated best. One day, as an experiment, Sulzer replaced one of the bars in Prathida’s ranat so that the note it played was badly out of key. “She hit it once and then avoided it,” he told me. “But then, after five minutes, she started playing it over and over.” Like a punk rocker or a modern composer, he later wrote, Prathida had discovered the joys of dissonance.

Other species seem to be equally musical. A dolphin can learn to play an underwater keyboard with its beak and imitate the sounds it hears, the psychologist Diana Reiss found. Then it can use those sounds to communicate with its trainers. A bonobo named Kanzi, studied by the primatologist Sue Savage-Rumbaugh, learned to improvise on a piano well enough to jam with Peter Gabriel. (When Sulzer tried a similar experiment with the bonobos at the San Diego Zoo, they preferred tossing the instruments to playing them.) Three years ago, the philosopher and jazz clarinettist David Rothenberg released a double album of music he had made with nightingales in Berlin. His recording method was simple: he waited for the birds to congregate in trees, set up his trio under the branches, and spent the evening trading licks with them. Still, when animals make music with humans, the result can be hard to judge. Is it art, mimicry, or irritated compliance? Are the nightingales really singing with the band, or straining to hear their own song above the noise?

Cartoon by Ali Solomon

“Anthropomorphism is always on my mind,” Sulzer told me. It’s easy to mistake ordinary animal behavior for something more expressive. The elephants in his orchestra had a highly developed sense of rhythm. But if they preferred percussion to wind instruments, the trainers later told him, it was because they were worried that a snake might be hiding in a mouthpiece. When the elephants played together for longer stretches, they seemed to find a groove, flapping their ears and twitching their tails to the beat. Sulzer assumed, at first, that they were moving to the music, but they were just getting overheated. “Elephants only have sweat glands in their toes,” he says. “So they have to flap their ears to cool off. And swinging their tails, frankly, means a little bit of boredom.”

One morning in May, Sulzer and I took a field trip north of the city, to Rockefeller University’s Center for Field Research in Ethology and Ecology, near Millbrook, New York. I asked Sulzer to do the driving, so that I could take notes as we talked. He seemed a little rattled at the wheel, slowing down for green lights and crawling up the Taconic Parkway at forty miles an hour. Like a true New Yorker, Sulzer hardly ever drives. It took a catastrophe to get him to the research center the first time. On September 12, 2001, the day after the Twin Towers fell, Sulzer fled to Millbrook to escape the dust and despair that engulfed lower Manhattan.

“I was just ten blocks away—I watched the second plane hit the tower,” he told me. “I thought the neighborhood was going to blow up, honestly.” The fires at Ground Zero burned so hot that Sulzer feared the gas lines might ignite beneath them. “So I kissed my cat goodbye and came up here the next day,” he said. His girlfriend at the time was studying with the neuroscientist Fernando Nottebohm, then the director of the research center, so Sulzer drove up to join her. He ended up staying a week—though he went home every day to feed his cat.

The research center lies in a secluded glen of hardwoods and sunlit meadows. Its quaint, half-timbered buildings look like a Disney set for “Beauty and the Beast”—they were once the gatehouse and stables for an estate owned by an heiress to the Standard Oil fortune. When we arrived, we were met by Ofer Tchernichovski, an animal behaviorist and Sulzer’s sometime collaborator. Born in a village near Tel Aviv in 1963, Tchernichovski has a stocky build and a brusque manner, almost childlike in its directness. He has a moon-shaped face, a mop of white hair, and eyes that squeeze into cheerful slits as he talks. He led us across the grounds in long, eager strides, talking as he went, then crouched down suddenly to look at something in the grass: a small frog. “I just love it here,” he said, watching it hop away. He’d recently seen a snapping turtle lay a clutch of eggs at a pond nearby, he added, then pointed to a deer emerging from the edge of the woods. “See how it turns to us? They’ll lift their tail and run, but first they always turn around and stare. They’re saying, ‘I see you. But all you’re going to see is my ass.’ ”

Tchernichovski says that he has never had a good hypothesis in his life. But he is a tireless observer. Behavioral research is all about “letting the animals tell you how they understand the world,” he told me. When he was working on his doctorate at the University of Tel Aviv, he built a giant rat compound on the roof of the zoology building, then spent months watching the inhabitants colonize it. The rats, he found, didn’t establish a single home base, as people assumed. They built a network of small shelters, like safe houses, and shuttled between them like covert operatives. “When you look at a species, you always ask yourself if you can generalize from their behavior,” he said. Some behaviors hold true across species—most animals prefer to defecate in private, for instance—but others do not. Only turkey vultures and a few other birds like to shit on their own feet.

Of all the world’s humming, squawking, buzzing, and growling creatures, birds may be the most single-mindedly musical. They sing when the sun rises and again when it sets. They sing to find mates and to claim territory. They sing to soothe their chicks and to sound alarms. At one point, Tchernichovski told us, he compared the songs of forty-five thrush nightingales with human songs from six cultures. On average, the birds could keep a beat and steady tempo as well as people could, but they processed rhythmic changes much faster and more accurately. “Birds really are the world champions,” he said. Still, it’s not clear how much we can generalize from that. Do birds take pleasure in singing, or is it just utilitarian? Do they share our sense of beauty in music?

When Sulzer first came to the research center, in 2001, he was in the middle of a bird study inspired by Tchernichovski’s work. Tchernichovski had outfitted a zebra finch’s cage with a lever that a bird could push with its beak to trigger a recording. If the lever was programmed to play the song of a male finch—female finches make simple, expressive calls, but they don’t sing—a baby finch would press it over and over. Sulzer wondered if other songs might have a similar effect. For his study, he built rows of levers, like miniature bird pianos, each of which triggered a different recording. “I thought, Birds like to sing, but do they also like to play?”

He began by programming the levers to play birdsongs of various species. At first, the finches wanted to hear only their own songs, but they slowly began to branch out and play others. Then Sulzer replaced the birdsongs with human music. He never trained the finches or offered them any rewards for pecking the levers. And yet, little by little, they began to gravitate toward certain recordings. Trumpets and flutes were predictably popular, but so were the gongs and xylophones of an Indonesian gamelan orchestra. The birds didn’t take to the gamelan right away, Sulzer told me. “They hit that lever very rarely at first. Then, a couple of days later, they hit it hundreds of times.”

There were only two levers that the finches avoided. One played a recording of a canary—a large, threatening species. The other played a song by the Oblivians, a noisy garage-rock band from Memphis. The first time they heard the band, the finches shrieked and jumped away from the lever. They never listened to the Oblivians again.

“It’s not just that some birds have inborn tastes in music,” Sulzer said. “They develop taste.” But how does it shape their singing? Do their songs change over time? “Dave really inspired a lot of these ideas with his work with the elephants,” Tchernichovski said, when we reached the cottage that housed his lab. “Science is all about crazy stuff. Scientists who are very level-minded—those aren’t real scientists.”

The front room of the lab was equipped with a large computer server, a row of analog-to-digital converters, and a pair of screens that were monitoring live recordings. The recordings were coming from a small room in the back, filled with insulated boxes with wires sticking out of them. They were Tchernichovski’s version of recording booths. He made them out of ice chests—“You can buy them premade for two thousand dollars, but mine are two hundred”—and fitted each one with an air-circulation system and lights that would rise and fall like sunlight over the course of a day. He added a mirror, so the birds could see themselves and not get lonely, and a lever. “It’s a world in a box,” he said.

To see how zebra finches learn their songs, Tchernichovski took male chicks that had been raised entirely by females and isolated them in the boxes for two months. Perched in the corner of each box was a fake bird taken from a Christmas-tree ornament. If the chick pecked the lever, a hidden speaker in the fake bird would sing a male finch’s song. Tchernichovski ran the experiment with three hundred chicks. He recorded them continuously and analyzed more than a million sounds per bird. The results, plotted on the monitors next door, looked a bit like old-fashioned sing-alongs: follow the bouncing ball. The chicks’ songs began as single tones, like syllables, represented as colored dots. They slowly established a rhythm, gathered into clusters, like syntax—long, high-pitched sounds; short, low-pitched sounds—and finally developed repeating motifs.

“They build a word, a sentence, a story,” Tchernichovski said. “It’s like embryo development—like a body, head, and limbs.” The first time a chick hears a male finch’s song, it doesn’t make a sound. It just falls asleep immediately, as if knocked cold by the revelation. When it wakes up a few minutes later, it plays the song again and again. By morning, the bird can sing it by heart. “We knew from behavioral studies that there were huge changes overnight—that in the morning something crazy happens,” Tchernichovski told me. At the University of Chicago, the neuroethologist Daniel Margoliash and the neuroscientist Amish S. Dave recorded the brain activity of zebra finches as they slept and dreamed. It had the same pattern as when they were singing. “The birds were doing playbacks of their songs in their brains,” Tchernichovski said.

One of the recordings that he had analyzed was from a box that held both a male and a female finch. Onscreen, the female’s calls were represented by red dots, the male’s calls by blue. At first, they clustered in separate patterns, like children in day care playing side by side on the carpet. Then, day by day, the two sets of dots began to mirror each other, to repeat the same patterns. By day four, the two calls were fully in synch. “You can really see where they fall in love,” Tchernichovski said.

“I just really thought this was going to be the lip gloss that changed everything.”
Cartoon by Liz Montague

In the wild, zebra finches typically live in colonies of between four and twenty birds. The Rockefeller research center has more than six hundred. When Tchernichovski opened the door to the room where they were kept, a wall of sound tumbled over us, like a rock-concert crowd on helium. The birds flitted from corner to corner in their cages, quick little sprites with black-and-white breasts and flame-orange beaks. They called back and forth to one another in fluctuating patterns: Taka tow tow, taka tow tow, babadoo babadoo babadoo. Tchernichovski grinned, basking in their voices. “I love zebra finches,” he said. “There is so much drama. I can tell if they’re excited or looking for something or interested in sex. It’s like their state of mind is pouring out of them.”

The birds knew him by his white hair, he said. They liked to fly over and pull at it when they were out of their cages, thinking it might make good nesting material. “There is a story about a zookeeper who takes care of his birds every day, and they know him well. Then one day, after twenty years, he goes in to feed them and they panic. It takes him a while to figure out that he’s wearing a new hat.” The colony was in constant communication with itself, he said. A single tissue of thought. “Silence is the real signal. The moment someone stops calling, they know something is wrong.”

He walked to the middle of the room, flanked by tall racks of cages, and smacked his hands together. A hush fell over the room. “Now watch this,” he said. He whistled a high, clear note—two kilohertz, he later told me, the finches’ preferred frequency—and waited. For just a beat, the air in the room seemed to tense up around us, as six hundred birds held their breath. Then they exploded into a raucous cheer. “Listen to them!” Tchernichovski said. “They’re so excited. You can feel the soul of the animal.”

We think of birds as creatures of habit, singing the same songs day after day. A few species, such as phoebe flycatchers, do seem to repeat the same innate calls all their lives. But others are as delighted by novelty as we are—the odd note, the new rhythm, the impromptu cadenza. Finches are thought to be especially set in their ways: their songs hardly seem to change after the first ninety days. Tchernichovski thinks we’re not listening closely enough. When he records three-year-old birds that he also recorded as chicks, their songs seem to have been subtly revised, remixed, layered with new rhythms and motifs. “An older bird might add Tadadam tadam bababam, tadadam tadam bababam,” he said. “There is more complexity, a higher level of organization.” He has no idea at what point in life that complexity is acquired. But it’s there.

A song is never as simple as it seems, Sulzer and Tchernichovski say. It’s both signal and noise, message and meaningless pattern. It can seduce and repel, say “I’m with them” and “I’m not like them” with equal conviction. It’s how we define ourselves against ourselves. From the moment we’re born, we’re taught to sound like our parents. But who wants to sound like their parents? So we make our songs our own.

Sulzer has yet to settle on his own sound. He has spent the past forty years at the restless edge of the avant-garde, never committing to a style long enough to claim it. At one point, as if to show how arbitrary our tastes can be, Sulzer and the Russian artists Komar and Melamid recorded an album called “The People’s Choice: Music.” It had only two tracks. Both were based on a survey that asked five hundred people which musical instruments and themes they found most appealing and which ones they found most unappealing. “The Most Wanted Song” was a love ballad scored for guitar, saxophone, bass, drums, and piano. “The Most Unwanted Song” was a cowboy tune for bagpipe, accordion, tuba, and children’s voices. The latter, ironically, proved far more popular. “It has lots of fans,” Sulzer says. “Over a million plays on YouTube.”

Still, he began to feel as if he had painted himself into a corner. The destruction of the two towers had left him reeling, in need of something more from his music—some sense of how it could help heal the world and not just comment on it. He worked with groups of children in East Harlem and Guatemala, improvising hip-hop tunes and Mayan mountain music. He immersed himself in flamenco, inspired by the music’s extravagant passions, its roots in a rare confluence of Romani, Moorish, and Jewish exiles. He wrote gospel songs on the theme of St. Francis, lover of animals, with words in the saint’s ancient North Umbrian dialect. “I wanted more emotion,” he told me. “I thought, How can I work with professional musicians where I get the same deep feeling that I get from the children and elephants?”

We were sitting in the living room of his apartment in Chinatown, late at night, after one of his classes at Columbia. All around us, the desks and bookcases were covered with the tools and detritus of a working musician: keyboards and monitors, piles of sheet music and empty instrument cases. A lyre from Nairobi lay on a table in the vestibule, next to some panpipes from Vietnam, a hand-carved kettle drum, and a banjo made from old 45 records. The neuroscientist in Sulzer seemed nowhere in sight. Then he stepped over to one of the keyboards and showed me his most recent score.

Of all his compositions, this one probably came closest to joining his two halves. It was a four-part motet based on Johannes Kepler’s “Harmonice Mundi”—“Harmony of the Worlds.” First published in 1619, Kepler’s treatise was both an abstruse work of mathematics and a vision of the universe as a kind of celestial music box. Kepler worked out the planets’ elliptical paths around the sun with remarkable accuracy, then compared their motions to notes in a chord, ringing in perfect harmony. In the final book of the treatise, Kepler urged the composers of his era to set his equations to music. “To him who more properly expresses the celestial music described in this work,” he wrote, “Clio will give a garland, and Urania will betroth Venus his bride.”

A number of composers had taken up the challenge over the centuries, Sulzer said, but they’d all fudged the mathematics. He was determined to play by the rules. Was it hard to do? I asked him. “Fuck yeah,” he said. “But it was also kind of fun.” In his piece, as per Kepler’s instructions, the parts of Saturn and Jupiter were sung by basses, Mars by a tenor, Earth and Venus by altos, and Mercury by a soprano. Their notes cleaved closely to Kepler’s calculations: Saturn’s part ranged from G to B and Jupiter’s from B to just above D, for instance, but Venus, with her more circular orbit, could only oscillate between E and E-flat. Kepler wanted listeners to feel as if they were standing on the surface of the sun, hearing the harmony of the spheres as the planets circled around them. The closer each planet came to the sun, the higher its notes ascended.

Sulzer opened a MIDI file on his computer and played me a passage. Its synthesized voices were a poor substitute for celestial singing, its harmonies as eccentric and stubbornly mathematical as Kepler’s theology. But later, when I heard a vocal group called Ekmeles perform the piece in a studio, I found it strangely moving. The music wasn’t luminous and ethereal, as I had expected. It was earthy and heavy-footed, full of steady, stomping forward motion. It was like an angry crowd that slowly, grudgingly joins in a folk dance. When the ethereal harmonies did come, they flashed through the music and quickly faded, like the sun’s rays at the edge of an eclipse. “That’s what Kepler was looking for—a moment of consonance in the universe,” Sulzer said. “Usually it’s not there. But, when it is, it’s evidence that God did something right.”

The world is full of music we can’t hear, Sulzer says, hidden in messages and melodies, patterns and harmonies that move through and around us all the time, beyond the range of our perception. It’s in the high harmonics of the swirling atmosphere and the subterranean chords of shifting plates. In the voices of creatures that communicate at frequencies far above and below our speech. Mice that squeak to one another ultrasonically as they move through our walls on padded feet. Birds that flicker by so fast we barely hear their songs—it’s only when we slow down their melodies that they sound like ours. Whales that sing song lines so leisurely they last for hours and transmit halfway across the ocean before they’re done.

When Sulzer was working with the elephant orchestra, he knew that the music they played wasn’t really their own. It was just an approximation, as foreign to them as fiddling on a cricket’s wings would be to us. The orchestra went on to record three CDs, including Sulzer’s arrangement of Beethoven’s “Pastoral” Symphony for elephants and marching band. They played for the Queen of Thailand and the BBC World Service, and appeared in a Moment of Zen on “The Daily Show with Jon Stewart.” But no one could hear what the elephants were humming to themselves, in the deep subsonic of their own frequency, as the drums clattered and gongs crashed. “We are just at the beginning,” Sulzer told me. “There is a whole auditory world around us that we’ve ignored.” Not quite the harmony of the spheres, but music enough for this one.