KNOWING WITHOUT KNOWING

Pleasure in the Absence of Expertise

Two minutes and two seconds into the song “(Get Up, I Feel Like Being a) Sex Machine,” the Godfather of Soul, Mr. James Brown, famously asks his organ player and backup vocalist, Bobby Byrd, “Shall I take ’em to the bridge?”

According to my iTunes library, I’ve listened to the song over two dozen times, but until recently I had no idea what Brown meant. Music, like virtually all crafts, has its share of vocabulary and specialized knowledge, and I was able to enjoy the song even when I knew vanishingly little about the mechanics of music.

Brown’s bridge, as I now realize, was musical, not physical; he was telling his bandmates to bring on the next part of the song. A bridge is a kind of transition; it’s a new bit of music that differs from what came before but helps move the song forward; the b, for example, in a song that has an aaba structure. It relieves the listener of the agony that comes from too much repetition, yet, done skillfully, it also sets the stage for the finale.

The funny thing is, listeners don’t need to know what a bridge is in order to appreciate the music, any more than patrons of a magic show need to know the trick that allows a magician to pull a rabbit out of his hat. One can enjoy a movie or a novel without knowing anything about tension, resolution, or climax, just as one can enjoy a photograph without knowing anything about film, contrast ratios, or the rule of thirds. Most of us are just consumers, not producers, and generally that’s okay: we understand and appreciate much of what we see and hear, even if we can’t produce it ourselves.

Another great example of how we can appreciate artful composition without recognizing what is going on comes from the Beatles’ “Yesterday.” Before I began immersing myself in music theory, I had listened to the song hundreds of times without once realizing that its statements of yearning for the past (“all my troubles seemed so far away,” “I’m not half the man I used to be”) follow ascending melody lines, while statements that return to the present (“Now it looks as though they’re here to stay”) follow descending melodies. That neat bit of musical architecture undoubtedly enriches the song, yet many listeners never recognize it explicitly.

Movie viewers don’t need to know anything about acts or narrative arcs to appreciate film, and many music listeners don’t even know the names of the musical notes, let alone, say, that classical sonatas tend to repeat their second theme in the so-called tonic key, or that popular songs can be divided into components like bridges and turnarounds. How can we so thoroughly appreciate what we can barely articulate?

It’s not just audiences, though. Sometimes even performers can’t explicitly articulate the tools that they are using. Terre Roche told me about the first time she and her older sister, Maggie, went to record their own music. Terre was just eighteen, and although she was already an accomplished musician who had been touring for several years, and had even sung some harmonies for Paul Simon’s album There Goes Rhymin’ Simon, aside from that one studio session with Simon she’d almost never performed with anyone other than her sister. Suddenly Terre and Maggie found themselves in a studio in Muscle Shoals, Alabama, with the legendary Muscle Shoals Rhythm Section, a team of top-notch studio musicians who recorded hits with everyone from Percy Sledge, Rod Stewart, and the Rolling Stones to Lynyrd Skynyrd and the Oak Ridge Boys. At one point, while they were working on Maggie’s song “Burden of Proof,” the piano player Barry Beckett turned to Terre and asked her, using a common convention known as the Roman numeral or Nashville numbering system, if the song went to “the IV chord” in the second half of the first measure of the verse. Terre froze; she had no idea what a IV chord was. As she recalls:

All I knew was that I was playing an F chord. Suddenly I felt as if I was the stupidest person in the room. I was the youngest person in the room (just twenty-one) and … Maggie [and I] had never played with anyone else but each other. Listening back to those tracks, I am amazed at the music that we put together without anyone ever telling us what a IV chord was.

The point of the Nashville numbering system is that it allows you to describe a song independently of its key. But the vocabulary itself is not nearly as integral as the musicianship it seeks to describe.

Perhaps a year or two later, Terre, who had learned mainly by watching other people play, saw a girl play a song in which she suddenly jumped up to the seventh or eighth fret on the guitar in the middle of the song. Terre decided she’d write one like that, which ultimately became “Mr. Sellack,” featured on the first Roches record (the one the New York Times judged 1979 Album of the Year). As Terre recalls, “I was in the key of D, playing all of the chords I knew, then just jumped up to the higher frets and tried different shapes. Years later I learned that I had gone to a B major chord, taking me out of the key of D, but playing a secondary dominant leading to E minor. The whole thing took a little journey out of the key of D and found its way back in by the end of the bridge. At the time I didn’t even know what the term ‘bridge’ meant.”

Some years later, the tables were turned. Terre eventually became deeply interested— and deeply fluent— in music theory; she began teaching music theory and started taking courses in complex jazz harmonies. At around the same time, she found herself working with a good friend who was a talented yet totally unschooled musician, capable of reproducing entire Beach Boys and Beatles records in his home studio entirely by ear. One day he and Terre were working on one of his songs, and he wanted Terre to sing a harmony in a particular way, but Terre wasn’t sure that her friend had really meant for her to do what he was asking. “Yesterday you told me to sing [my note at a] third [above yours],” she said, attempting to clarify, “and now you want me on the ninth,” which in C major would be nine keys apart rather than three. Her friend’s response? “Why don’t you just shove all that book learning you’ve been doing and just sing the friggin’ song!”

As the composer and mathematician Dmitri Tymoczko has put it:

When composing, I make various choices about chords, scales, rhythm, and instrumentation to create feelings of tension, relaxation, terror, and ecstasy, to recall earlier moments in the piece or anticipate later events. But I do not in general expect listeners to be consciously tracking these choices. Listeners who do (“ooh, a dissonant #9 chord in the trombones, in polyrhythm against the flutes and inverting the opening notes of the piece!”) are like professional magicians watching each other’s routines— at best, engaged in a different sort of appreciation, and, at worst too intellectually engaged to enjoy the music as deeply as they might.

In short, the formal elements of music theory have their uses, but one can certainly appreciate— and even play— music without them.

In the early 1990s, a French cognitive psychologist named Emmanuel Bigand set out to find the differences between experts and novices in how they listened to music. There was at that point already a sizable literature on the basic sensitivity of novices and experts to pitch and rhythm. Simply put, experts’ ears are more sensitive to fine-grained details; they are more prone to noticing when something is slightly out of key and (as mentioned earlier) better able to recognize precisely how far apart two notes are. Experts are also quicker to detect deviations in a rhythm, and they can often identify the individual notes in a chord, something that novices can rarely, if ever, do.

But a lot of what Bigand found was surprising. He went in expecting to find huge differences between musicians and nonmusicians but instead in many cases discovered that nonmusicians knew far more than anyone could ever have anticipated. The usual terminology in this literature distinguishes between experts and novices, but Bigand eventually started to realize that all of his adult subjects, trained or otherwise, were experts of a sort: they were all expert listeners.

Even without musical training, Bigand began to realize, the average listener comes to know implicitly many aspects of the formal structure of music. Untrained listeners won’t necessarily be able to articulate that knowledge; professional musicians can often describe what’s going on in a song with far more precision than mere listeners can. And professional musicians are obviously (by definition) much more capable of producing music than nonmusicians are. To paraphrase the great baseball legend Yogi Berra, “You can hear an awful lot, just by listening.”

For instance, one of Bigand’s first studies focused on a theory that had been introduced by Ray Jackendoff and his collaborator Fred Lerdahl, aimed at understanding how sophisticated listeners interpreted complex pieces. The idea was that in any given piece, some notes figure more centrally than others; if you were a jazz saxophonist improvising changes to the melody in “Over the Rainbow,” some notes might be safely changed, whereas modifications to others might leave the listeners scratching their heads. Bigand set up an experimental test of Ler dahl and Jackendoff’s theory about which notes were critical and found that expert musicians more or less behaved as Lerdahl and Jackendoff had predicted. But so did the non-experts; ordinary listeners seemed to have much the same intuitions, if in less precise form, as the experts. After that, Bigand conducted dozens of other studies— looking at aspects of music such as resolution (whether a composition returns to the root of a key)—and found much the same pattern: experts were invariably better than non-experts, but non-experts always did a lot better than mere chance. A novice might not be able to say which particular key a musical piece was in, but experienced listeners who lacked formal training could still tell when a musical composition broke the rules. Other studies showed that nonmusicians were often nearly as good as trained musicians in their capacity to recognize the emotion of a piece from a brief excerpt, and that untrained listeners could often do a fairly credible job of learning to recognize new pieces.

Looking at the same question from a neural perspective, Bigand found still further evidence that untrained listeners were able to extract a great deal of the structure that underlies music. As he put it, although there are surely differences between musically trained listeners and untrained listeners, “the human brain is already intensively trained to music through everyday life experience: adding supplementary training in music schools makes it possible to acquire specific skills [like reading sheet music or naming musical intervals] indispensable to be professional musicians, but [formal training] is not what determines the musical ability of human beings.”

Listening is what first gets us into the game.

In the late 1970s, two cognitive psychologists, Richard Nisbett and Timothy Wilson, wrote a groundbreaking paper— still worth reading thirty years later— called “Telling More Than We Can Know,” which reviewed dozens of studies showing that human beings often have no idea why they do the things they do. It is possible, occasionally, for a human being to articulate exactly what he or she is doing and exactly why, but it’s also eminently possible for average people to act without any particularly clear awareness of why they do what they do, justifying their actions after the fact, rather than thinking them through beforehand. In one famous study, hundreds of shoppers were asked to look at an array of nylon stockings and decide which they liked the best. Subjects tended to browse the array from left to right and, far more often than chance would allow, chose the pair of stockings farthest to the right. None of the subjects mentioned position as a reason for making their choice, and when position was offered as an explanation, everyone denied it played any role; in reality, people often tend to take the last choice simply because whatever comes last is easiest to remember.

Our lack of awareness of our brain’s internal operations similarly extends to many athletic endeavors. The act of catching a fly ball in baseball, for example, can be calculated in terms of a complex stew of calculus and tangent lines, but there are thousands of Little Leaguers who have yet to take a calculus class and still manage to catch fly balls just fine. Just because the actions of a musician can be described in the abstract mathematical terms of music theory doesn’t mean the musician needs to be explicitly aware of those abstractions.

Philosophers sometimes distinguish between “knowing how” and “knowing that.” There’s a difference between knowing how to do some thing, sometimes referred to as muscle memory (really a kind of memory in the brain), and being able to explain what you are doing.

The brain uses distinct systems for encoding facts and for encoding actions. Facts, stored in declarative memory, have been associated with a section of the brain known as the medial temporal lobe (centered on the seahorse-shaped hippocampus) and (after they have firmed up) with the temporal and frontal cortices. Action programs (the not always consciously accessible product of what I referred to earlier as “automatization”) are largely associated with procedural memory, encoded in a loop that centers on structures such as the cerebellum and the basal ganglia. One reason for the decoupling of formal musical knowledge and practical musical skill is that the two are treated by the brain in fundamentally different ways. Declarative knowledge has been optimized for conscious knowledge; procedural knowledge has been optimized for rapid reflexes that aren’t necessarily accessible to consciousness.

If we sometimes find it hard to consciously articulate what our muscles are doing, it’s because the memories that feed our conscious thoughts are largely separate from those that control our muscles. It is only with considerable conscious effort and reflection that one can really learn to traffic back and forth fluidly between the two.

Bill Evans, one of the most influential pianists of the twentieth century, most famous perhaps for his piano solo on Miles Davis’s “So What,” first made his mark largely by changing the composition of the chords he played, altering what are known as the voicings of a chord. Drawing partly on impressionist composers like Debussy and Ravel, Evans began to experiment with looser interpretations of chords, omitting notes, extending them over time, interspersing others while still retaining the overall flavor. At the time, nobody else could figure out what Evans was doing, and Evans himself might have been hard-pressed to fully articulate his techniques. Even Evans’s own brother Harry, who was a well-known pianist, was initially puzzled.

Although Bill must have had some idea of what he was doing, the bulk of what he was doing was procedural; his muscles and basal ganglia knew how to create these novel chord voicings, but in the early stages of his experimentation that knowledge had yet to be translated into declarative facts. Eventually, though, someone did figure out what Evans was doing, and it wasn’t so complicated that nobody else could understand it; instead, once the translation was made, “Bill Evans’s Chord Voicings” became a standard part of the jazz curriculum and a tool in every jazz pianist’s kit. What happened in that particular case is not uncommon; improvisers work out innovations by feel alone, and only later do those innovations get codified into theory that can be explicitly encoded into declarative knowledge. The real virtue of music theory in a case like that is not to forge new ground but to provide a language for sharing discoveries with other musicians.

A lot of jazz, in fact, is about blending the declarative with the procedural. The jazz pianist Marilyn Crispell, often compared to the great Cecil Taylor, and longtime member of Anthony Braxton’s group, explained to me that in her improvisations she often blends conscious, explicit knowledge of the overall frame or direction of a piece with a kind of unconscious playing in the immediate moment. I’ve overheard Pat Metheny express much the same. To some degree, classical compositions can be played fairly well by sheer rote (drawing entirely on procedural knowledge), and a lot of rock can, too, but satisfying jazz almost always requires a kind of mixture. On the one hand, the jazz player must depend heavily on highly practiced habits that allow the musician to rapidly ascend scales or otherwise play lightning-fast riffs. On the other hand, to be truly compelling, jazz musicians must also integrate declarative knowledge in the form of plans, ideas, and concepts—schemas that help shape a piece and yield music that, as Crispell put it, feels “composed, and not merely improvised.”

Classical used to be this way, too. Bach, for example, was a legendary improviser, capable of creating four-part fugues on the fly. And until the nineteenth century, when the spirit of improvisation began to fade in favor of orthodox interpretations of canonized great composers, performers in the classical tradition improvised frequently, especially in the sections known as cadenzas— forerunners to modern guitar solos, filled with fast, virtuosic passages. The idea of an entire piece being written out note for note, fully annotated with markings, is a relatively recent tradition. In global and historical perspective, the improvisational impulse behind jazz is the norm rather than the exception. And the best players have probably always been those who could mix the declarative and conscious with the automatic and unconscious.

In the real world of guitar, sometimes musicians don’t even know the names of the chords they are playing.

Perhaps the easiest song ever to play on guitar is America’s “A Horse with No Name.” There are only two chords in the song, and both are easy to play, requiring just two fingers each on the fretting hand.

One of the two chords is the common E minor chord, in which the middle and index fingers hold down the second frets of the fourth and fifth strings. The other is obscure but equally easy to play; instead of holding down the second frets of the fourth and fifth strings, the guitarist switches to holding down the second frets of the third and sixth string—still two fingers, and hardly any motion involved (and as any guitarist realizes, the less motion required, the better).

But what is the second chord called? You won’t find it in any list of the top forty most common chords. As the guitarist and bass player David Hodge points out, the other chord could go by about five different names, not one of which would be familiar to the average guitarist. You could call the second chord in “A Horse with No Name” a Dadd6add9, an F#dim7 (add 6), or an F#m7 (no 5)(add 4)(add 6); you could also think of it as E11 or Bm11. But the guy who wrote the song— Dewey Bunnell— quite likely didn’t think of the chord in any of those terms; more likely, he was just fooling around on the fretboard, and once he stumbled onto the Dadd6add9 (or whatever you prefer to call it), he had the good sense to trust his ears, even if he couldn’t explain in exact technical terms what he had created.

In the mid-1990s, the journalist Daniel Goleman wrote an influential book called Emotional Intelligence, which argued that there was more to intelligence than the sort of puzzle solving you’d see in an IQ test; Goleman argued that people’s social and emotional savvy was as important to professional success as their raw analytical intelligence. There can be little doubt that social and emotional insight matters for music, too.

Goleman illustrates emotional intelligence with a story he borrowed from a friend who happened to be one of the first Americans to journey to Japan to study martial arts. On one particular afternoon Goleman’s friend found himself on a subway in Tokyo when a drunk and hostile passenger got on. The drunk started screaming— and eventually taking swings— at the other passengers. Finally, Goleman’s friend had had enough; he stood up and steeled himself for battle. But before he could enter the fray, an old man dressed in a kimono cheerily shouted “Hey!” as if he had spotted an old friend. By startling the drunk, the old man postponed the imminent violence.

“Why the hell should I talk to you?” the drunk asked.

The old man replied not with an answer but with a question, “What’cha been drinking?” When the drunk said “Sake, and it’s none of your business,” the old man began to turn things around. “Oh, that’s wonderful, absolutely wonderful,” he said. “You see, I love sake, too. Every night, me and my wife (she’s seventy-six, you know), we warm up a little bottle of sake and take it out into the garden, and we sit on an old wooden bench.” Painting a picture of the persimmon tree in his yard and pondering the nature of sake, the old man added, “Yes, and I’m sure you have a wonderful wife, too”—at which point all the drunk’s remaining aggression melted away.

“No,” wailed the drunk, “my wife died. . . .” Soon the drunk was sobbing, explaining how he had lost his wife, his job, and his home; by the end, the drunk had cradled his head in the old man’s lap.

In defusing the situation, the old man displayed a spectacular command of emotional intelligence— or what cognitive scientists often call “intuitive psychology” or “theory of mind,” the art of reading (and anticipating) the beliefs and desires and needs of others. And it is that intuitive psychology, more than the formal apparatus of music theory, that really matters most: the ability of a performer (or songwriter or composer or artist) to anticipate the pull that a work will have on its audience, and to figure out how to get that audience to share in some mood or feeling. Perhaps the Boss, Bruce Springsteen, said it best:

I always write with an audience in mind. If I feel that [connection] coming back at me then I feel like I’m doing my job. That’s why people come to my music— for some emotional experience or a perspective, either on their own lives, or on the world that they’re living in.

Some of the best, most in-depth discussions on the art of songwriting I have ever encountered have been interviews conducted by the author and songwriter Paul Zollo. In one, Zollo interviews James Taylor and after some easy questions starts delving deeply into advanced topics in music theory, asking Taylor questions about technicalities such as diminished chords and twelve-tone scales, elements that might lie outside the average pop star’s range.

Taylor, however, is unfazed and has no trouble keeping up. He explains, for example, that he learned from Paul Simon how to use diminished chords as a way of “escaping from a melody, or from [one] harmonic sort of context and jumping into another one.”

From there Taylor moves on to his former mentor Paul McCartney, and a discussion of an unusual chord McCartney played in “Michelle”:

Paul said that was the only jazz chord [he and John] knew. They used to go down to a record store in Liverpool and there was somebody there who played guitar and he showed Paul and John this [unusual] chord. So the second chord in “Michelle,” under “ma belle,” that second chord is a very unlikely chord. … And you wouldn’t expect to see it.

… But, boy, the way he [Paul] bounces one onto another. It’s really very much like cubism … it represents so much [emotion] in just a simple line.

And that’s what music is all about. What really matters is not how many chords you know, but how you use the ones you do know.