The process of improvisation

Even for experienced improvisers, the process of improvisation is somewhat of a mystery. Riding the edge of creation, which is what the process of improvisation feels like, is not something to be questioned too analytically. It is a bit like the centipede that was paralyzed by the question of which foot comes after which; thinking too hard can throw you off completely.

The centipede was happy, quite,
Until a toad in fun
Said, “Pray, which leg goes after which?”
This worked his mind to such a pitch,
He lay distracted in a ditch,
Considering how to run.

I usually tell people that you play one note, then another, and then another. Anything more than that often results in a mess. You make each decision based on how you “feel.” You react in accordance with how you feel the phrase should go. You can’t really think too much about the music specifically, thinking takes too much time. I have personally always felt that thinking was overrated anyway.

Nothing ever goes exactly the way you expect, and even a small change can alter the direction, scope, and effect of your original idea.

About twenty years ago or so, I ran across a scholarly article about the process of improvisation and how it works. It had diagrams and flowcharts and feedback loops with options and conditionals, etc., etc. I figure it was meant more to help someone program a computer with an improvisation program rather than to actually help someone try to improvise. It was not very enlightening, but it did get some things right. It emphasized the reactive nature of improvising as opposed to that of following a plan or focusing on an abstract or musical idea. It is a process, not a production. This is an important part of improvising that an improviser needs to fully digest. At any moment while improvising, a performer is reacting to what he or she has just heard, trying to take it where they think it should go. But before they get there, they must react to the new music. This is true, even if you are improvising by yourself. Nothing ever goes exactly the way you expect, and even a small change can alter the direction, scope, and effect of your original idea. Playing one note softer than expected changes the emotional course of a phrase, and provides (or sometimes necessitates) the opportunity to go in a new direction. Improvisers learn that no idea is sacred, and it is more practical to just avoid thinking too much. You can have the most magnificent vision and insight into the next eight bars, but by the time you’ve reached the eighth bar, you have already altered that plan at least three times. Vision wasted.

This phenomenon of continuous reaction can become a joyous experience when improvising with others. The music can be constantly enhanced by the inspiration of multiple minds, and it can also be thrilling to be part of many uniting as one in a common inspiration. Musical interaction can be thrilling when you know what everybody is doing, as is the case with Classical music. But that interaction can border on nirvana when everybody is improvising. The music is always, however, on the verge of chaos. It is only through collective effort that it stays alive. This fact enhances the thrill you receive when it works.

A dear friend of mine, who happens to also be a composer, asked me a question after I had played seven or eight chords in quick succession upon which you could have wasted fifteen pages of analysis. He asked, “Do you know what you are doing when you play those chords or are you just moving your fingers?” The simple answer was, “Yes.” I am just moving my fingers and know what I am doing. The more useful answer was that I am not doing a theoretical analysis of each chord, but I am listening hard to what I am playing and reacting accordingly.

Any improviser will say that when he or she is playing well, the music seems to be “playing itself!”

It is important, too, to not separate your brain from your fingers. Playing any instrument is a physical process and we relate the motions we make to the sounds that we hear. Any improviser will say that when he or she is playing well, the music seems to be “playing itself!” The process is flowing freely without the stifling interference of any intention. There is an unfortunate concept in classical music called “muscle memory,” which can supposedly happen when you practice a passage so much that you can play it without thinking about it. This can sometimes have unintended consequences, ostensibly because the performer is not in control of his or her fingers. This fosters the idea that one’s fingers are both untrustworthy and separate from one’s brain. The brain’s job seems to be to whip the fingers into shape and then lord over them to make sure they don’t go astray. This is nonsense. There is no way you can “think” your way through scales, arpeggios, or any other passagework without trusting your fingers to do the work. The true danger is not that your muscles will betray you, but that your mind becomes lazy. Paying attention is the foundation of any skill, and listening and guiding are attentive skills demanded of any musician. While improvising, the listening is more open and creative, and the guidance less rigorous, but it is still the fingers that do the work. When improvising, control over how you play your instrument needs only a very loose tether because there is no planning to be done. You are simply listening and reacting, over and over. I don’t think, therefore I improvise.

A Change of Heart

A couple of weeks ago, I was watching a video on Facebook about education. It was about how recent education had stopped stimulating imaginative thought through content testing, etc. The point was that imaginative thinking was more common, more interesting, and more useful than just facts. It rekindled an idea I had pondered many times, about how modern culture and technology has fostered a “cult of the literal.” How access to, even bombardment by audio, video, and constant information has suppressed abstract and imaginative thought and somehow made it seem less important.

Our culture used to be brimming with stories, oral or written, with which you became involved. The stories taught as much as they entertained. They would exist in your mind as abstract images. They meant more as ideas than they would as literal people and places. I took my first trip to Europe this summer, and I saw places that I had heard about my entire life. But by themselves, these places were unremarkable. It’s the stories and my imaginative extrapolations that made them come alive.

I first noticed this trend several decades ago when digital watches were introduced. You would ask someone for the time and they would answer, “3:43,” where before they might have said, “It’s almost a quarter to four.” Digital watches had suddenly made precision important.

The society now has instant access to a thousand times the information that it did when I was growing up. But so much of it is a passive access. It is either video, or a search engine, or stuff roboted to your email. We don’t have to work for it or even think about it. It’s here; it’s gone. It used to take patience and skill to find things out; now if you can’t Google it, it’s not worth searching for.

The culture has changed enormously in the past couple of decades, and I suppose we really have no choice but to roll with it. But I have lived long enough to experience this immense change in my lifetime, and if I step back, I can look at how it has changed me. Because of all the tools I have at my fingertips, the role imagination plays in my work has become limited and temporary. Composers used to carry an abstract aural concept of a piece in their head for months or even years with larger works. Now our computer plays whatever we write as soon as we write it. On one hand, we don’t make big mistakes anymore; on the other hand, we don’t take any chances.

The innovation of the Suzuki method was that it taught Classical music by ear . . . Most Classical musicians at the time thought this was “cheating!”

When I was just learning music, the teaching method of Sinichi Suzuki was just becoming known in America. I can remember that my teachers were horrified. The innovation of the Suzuki method was that it taught Classical music by ear, which is the way most music is learned throughout the world. Most Classical musicians at the time thought this was “cheating!” My piano teacher advised me to not listen to any recordings of the pieces I was playing. He thought it polluted the process of creating my own concept of the piece. Today, listening to a recording is the first thing that a musician does. The Suzuki method has become standard practice. When I started my orchestral career, conductors were expected to have different interpretations of even standard repertoire. Alternate interpretations now often strike my orchestra colleagues as “wrong.”

Classical music has always lived through the imagination of its performers. Classical notation is just an approximation, and was not intended to be very precise. I didn’t think this was the case until I started using computers. Nothing alerts you to that fact more than having a computer play music from your own notation. A computer just plays what it sees – no imagination, no extrapolation. It’s dreadful. When I started recording using MIDI, I found that the computer would record the touch of my keyboard with 128 different volume gradations. It makes the seven or eight different dynamics of written music seem pretty meager. I looked at a graph of my volume values and realized I had used every one of the 128!

A screen shot of a (condensed) keyboard MIDI file showing pitches and note lengths above and volume (key pressure) in the pinkish area.  The pink areas are places where the sustain pedal is depressed.

But it is exactly that generalization of music notation that has continued to regenerate Classical music throughout the centuries. A musician can always imagine a better performance, and can always imagine improving his or her own performance. The imagined music is what makes the real music come alive; it is just like all those famous places I saw this summer.

Several weeks ago, I wrote on this blog about transcribing from my piano recordings. I had decided that the music was better represented if I transcribed it against a steady pulse, allowing the music to ebb and flow around what amounted to a stationary grid. Though I have been able to do this, and the music on the recording matches well with the music played by the computer from my notation, it does not embody the imaginative spirit that entices a pianist to sit down and play it. It creates the music with smoke and mirrors. If you play what I wrote, the music appears out of nowhere; not because you imagined it and brought it to life.

So I have gone back to notating the music rationally, by showing the rhythms and tempos in a clear fashion that gives a pianist something to re-imagine. Though I must show changes in tempo, I am not showing every rubato. I use the notation to show the structure, and let the pianist show the music. This is what musicians have always done, and what they do best.

When Improvising, Music Theory Becomes a Verb

I knew I wanted to be a composer when I was in Junior High School. My parents found me a couple of teachers my last couple of years of high school. They were both well respected composers. The first teacher showed me Hindemith’s Theory of Interval Hierarchy on the first lesson. The second teacher waited until the second or third lesson, but then brought me a short treatise on Information Theory. Though I was totally baffled, these two events were actually a pretty good introduction to how many composers think about music.

Music composers and theorists have riddled music with a litany of systematic hierarchical structures. Tonality and key, harmony, meter and rhythm, form, and many more complex all-consuming theories are all systems of organized structures meant to impart logical meaning to emotional content. It is an idea borrowed from writing, and it is a visual concept. A large written work is organized into parts, chapters, sections, paragraphs, sentences, words, and finally letters. Each letter and word has its own function and meaning, and this becomes extended to sentences, paragraphs, etc., until every part has its own place and role. Most composers think of their music in the same way, with each note, phrase, or layer having a special function and meaning, and playing a part in a larger whole.

But it is impossible to retain this kind of detailed attention to multiple levels of musical structure while improvising. You can keep track of what is going on for a while but at a certain point you reach the Too Much Information point and things fall apart. With no pre-planning, the amount of information, levels, and functions add up until you are overwhelmed. At that point, you just give up and think about something else, because, after all, you are creating everything on the spot, not just the structure.

So when improvising, the amount of actual musical information you can pay attention to is limited. This is why many improvisers prefer to improvise on given themes or motives, because when doing that, the emphasis shifts to creative manipulation for which improvisation is excellent. But all of traditional musical theory treats music as information, so when freely improvising, you have three choices: 1) limit the amount of information you keep track of, 2) not care, or 3) keep track of your music in a different way. Choice No. 1 very quickly becomes a precondition. Limiting means making choices, usually in advance. This is how most of the world improvises, with limits of key or mode, tempo, meter, harmonic choices, etc. It narrows the playing field so that the improviser can concentrate on expression and creativity. Choice No. 2 pollutes the imagination and eventually becomes a choice. I believe that a lot of Free Jazz or Improvisation has become like this, where making an organizational choice of any sort becomes taboo.

It is not the music itself but its relationship to the surrounding music that gives it meaning.

I found that, as a composer, the only choice open to me was Choice No. 3 because I didn’t want to make pre-conditions and I did care. So instead of treating bits of music as information, I started treating them as relative values on the sliding scales of several simultaneous musical “fields” or parameters. Every bit of music I improvise is louder or softer, higher or lower, faster or slower, darker or lighter, more or less resonant, more or less dissonant, denser or more open, flatter or sharper, etc. than the music which is sounding simultaneously or adjacent to it. It is not the music itself but its relationship to the surrounding music that gives it meaning. Because emotion is perceived as a specific oscillation of intensity and release, and oscillation in many forms is a fundamental component of music, these relationships make excellent expressive tools while improvising. Most importantly, because their focus is very specific, they become tools that are actually possible to use while improvising in an intelligent way.

Because music has multiple dimensions, and these dimensions can be creatively manipulated simultaneously, the creative possibilities are nearly endless. An improviser can also create his or her own expressive parameters just by juxtaposition (i.e. more temple block and less triangle). Thinking about oscillating parameters encourages an improviser to concentrate on what he or she is doing at any given moment rather than creating the mental distance it takes to keep track of specific information over a longer period of time. This is exactly the mental attitude it takes to improvise well!

Of course, identifiable ideas do emerge. But instead of treating them as informational Legos to build things with, I treat those ideas as relational focal points that are altered and changed at every appearance. The information relates to itself. Large-scale structures do emerge too, but I don’t often think about them consciously. Sometimes I will deliberately bring back an idea from earlier within an improvisation if it is particularly prominent (and I remember it). Sometimes I will bring up an idea from a different improvisation or even another piece. But I seldom find that when I do it has any real or special impact. I find I can use and trust my musical instincts to control large-scale shape, contrast, and function. In other words, I play it by ear. It usually works. If it doesn’t, oh well, I’ll try again next time.

Turns Out My Improvisation is Composition After All

For nearly two decades now I have been campaigning for the virtues of improvisation. Actually, I have been doing it most of my career, but since 1998 I have been putting my music where my mouth is and turning out albums of solo piano improvisation.

It is important to me that my music is created spontaneously, but for many others, it is of no consequence. Many musicians misunderstand what improvisation is, especially at the compositional level. Even such a creative icon as Miles Davis was quoted as saying he had “no idea” what Keith Jarrett was doing when he performed his solo improvisations. I have heard people say improvisation is “real-time composition” or, one of my favorites, “composition in motion,” but this is not really the case. Many people have said to me, “Well, at some point, all composition is improvisation.” Unfortunately, this couldn’t be further from the truth.

Composition is about ideas. It is about the methodical construction of music directly from those ideas, whether the ideas are musical or non-musical. Sometimes the ideas are subtle and seemingly unimportant, sometimes they are the whole point of the music. The ideas can be motivic or harmonic, or they can be philosophical. They can be a self-driven process, or they can follow a script, film, play, or dance. They can be about social comment or be completely introverted. Or they can be all of the above. And all of these ideas guide the choice of musical material, how it is developed, and what happens to it. Emotion and expression come into play, of course, but they are nearly always part of the overall plan. Composers develop musical plans, structural plans, and emotional plans. It is the same with writing a book or creating a movie, it is about “constructing” a work of art. Naturally, there are many moments of inspiration, some of them you never hear, but mostly the process falls into the category of Edison’s “one percent inspiration, ninety-nine percent perspiration.”

But it also turns out that improvisation is difficult to define and almost impossible to discern. If a musician is not playing from music, he could be playing from memory, or he could be playing by ear. Even if he is improvising, how much of what he is doing is spontaneous and how much is planned or familiar. It is impossible to tell. I’ve heard that Louis Armstrong worked out and practiced his solos in advance. I have played with several jazz soloists who played the same or nearly the same solo every night. I’ve talked with other musicians who have said that even during free-improvisation sets, the group will, over time, revert to those things that have worked before. This was true for me as well when I tried to do free improvisation gigs early in my career. So the only real way to tell whether a performer is improvising and truly creating new music on the spot is to ask him!

Most people consider improvisation to be a technique of performance. Often the standard by which to judge the quality of an improvisation is to decide to what extent the music does not sound improvised. But this means that the improvisation must sound “familiar,” which has a tendency to be rather inhibiting. Many suggest that improvisation is merely “stream of consciousness,” and some I have heard certainly is. But good improvisation is no more stream-of-consciousness than meditation is sleeping. Improvisation takes intense concentration and focus.

Improvisation differs from composition in that it is spontaneous. It is not “about spontaneity” (an idea); it is truly spontaneous. So what difference does that make? Primarily it means that you are listening to an “experience” instead of a presentation. It means that the focus and depth of the music is happening right now, and has not been reflected upon, perfected, and polished. This much is clear enough, but how does that make the music itself different? I had only a partial idea until I started transcribing my improvisations many years ago. Improvisation, indeed, handles the music differently. Instead of the music being “deduced” from another musical idea, it evolves within itself. It uses material that is actually played rather than referring back to material that was chosen beforehand. The focus of the music changes as the music evolves. The improviser “discovers” his or her material, and memory is not always perfect. Ideas, musical or otherwise, are induced and synthesized from the actual music itself. This suggests a different philosophical concept of time, cause and effect, specifics, and abstracts, and it also suggests the idea that change, relation, and juxtaposition is more fundamental than any abstract idea. As a performer, improvisation appeals to me primarily as a vehicle for expression. Music that is conceived in real time is as honest as it gets.

In the sample below, (“Place of the Butterflies”, from my album Night Drift) listen to how each musical phrase draws upon the previous phrase and feeds the one that follows. In Japan, there is a form of oral poetry called “linked verse” in which new stanzas of poetry (such as haiku) are linked to the last stanza in some way.

Though how my music is created is very important to me, it does not mean that I am haphazard or casual about how I treat it. Though every single note is spontaneously conceived, that does not stop me from editing the MIDI files or adapting them to different piano samples. I don’t use the same sounds while recording that I do when I am mastering so I must adapt my MIDI files to the samples and to the response of my keyboard. But I not only edit for my equipment and software, I edit to make sure the music is exactly what I want. I am a composer, and this is my only shot at the material. This has involved me making two (or more) shorter pieces out of one longer one, starting at a more interesting spot than I did originally, or even making cuts within a take (cutting 10 seconds can make a world of difference). Though these techniques are all common in both classical and jazz recordings, I have drawn heat from many improvisation purists for using them. To me, it is not about the performance, it is about the music. But I have finally decided that instead of trying to change the world, maybe I should just try to get people to listen to the music for what it is, and not for how it was conceived. So I no longer am going to call my music improvisation.

The final tipping point in my decision came not from the improvisation or jazz world, but from Classical composition. New Classical Music now readily accepts music that only exists as a recording. Many composers put out recordings with electronics, samples, field recordings or samples from other composers, real world sounds, etc. Some composers write site-specific works, even site-specific operas, and the imagination for what is included in music these days is vast. Improvisation in New Classical Music, with certain limitations, has become rather commonplace. My concern about the acceptability of my piano improvisation within this genre has become almost silly.

I mentioned before that I have transcribed my improvisations and performed them live. I have also transcribed and adapted them for other instruments, including orchestra. At that point, these works can no longer be considered improvisations by any stretch of the imagination. I have also come across other works that have been conceived as strictly for recording but have also since been adapted for live performance. One of my favorites is Steve Reich’s Violin Phase (1967), which was originally done with two tape recorders playing the same violin melody on two slightly different length tape loops, but the piece has been adapted and is now often done live. Here are two versions, one done solo with a computer Steve Reich, Violin Phase (solo violin with computer), and a second done with four solo violins Steve Reich, Violin Phase (four solo violins).