On Teaching Improvisation


Stephen Nachmanovitch


A talk with college and university conductors

New York, 24 Feb 2005

College Band Directors National Association


Let us begin with a show of hands. How many people here play an instrument, including voice? [All hands are raised]. Thanks, now put your hands down. Of those who just raised their hands, how many play notes on your instrument? [Many hands]. Now, please raise your hand if you raised your hand the first time but not the second time. [A few]. Why?

A: I play music.

You play music itself, not notes. We have two different things: T-O-N-E and N-O-T-E. A tone is an actual sound, a physical gesture that you make with an instrument or voice, while a note is an abstract, culturally defined notation that may be a representation of a tone and may be the stimulus for playing a tone. But it is impossible to play a note on an instrument. You cannot do it. Korzybski, a philosopher in the early part of the last century, was famous for the important epistemological statement that the map is not the territory. The menu is not the meal. If you were to go downstairs to one of the restaurants that surround Carnegie Hall, and sink your teeth into the menu, you’d be spotted as nuts. So the map is not the territory and the notes are not the music. They have great usefulness, as do all maps, but they are not music.

In the past couple of decades I have managed to be a kind of Johnny Appleseed of improvisation. A few years ago, someone asked me about my teaching methods, and I replied that I had no method – each teaching event was unique. But I’ve come to realize that was not quite true, and I hope here to spend some time thinking out loud about a few aspects the methods that have consistently worked for me. In the improvisation I teach, we go right for the sounds as directly as we can, without the intermediation of signs and symbols: no scores, no jazz charts, neither sticking to a tradition nor avoiding tradition, no agenda other than clear, direct communication with each other and with our audience. I travel around and spread seeds of this work, and then later I hear back about the things that are happening, students who are creating improvised ensembles and doing improvisational concerts, and in particular improvisational chamber music.

As improvisers, we learn to work from the territory of action and gesture, not from the map, as we create in real time. In the beginning, we naturally encounter certain fears. The maps, the scores which have produced four hundred years of great music that we want to continue to preserve and honor, provide us with a certain degree of security. People generally have an aversion to standing up on stage in front of a lot of other people and making fools of themselves. So we prefer the security of reproducing something that we know is good – though we can perform a classic score and make fools of ourselves too. In the same way, politicians often prefer to read prepared statements rather than speak directly, because in their world, mistakes can be costly. But hearing them speak directly would be such a relief! (and also lead to jail time for many if not most). My goal is to give people the experience of what it is like to get up on stage with no strings attached, with no plans or expectations, and to create quality music which is interesting, compelling, coherent, and which arises from the present mind and the present moment. I have found that you can quite reliably create the conditions for that to happen. You can’t program the music, you can’t know what’s going to happen on stage until it happens, but for a teacher it is possible to create the context and the conditions where successful improvisation can happen.

Some people think that improvisation equals jazz. Jazz is one of the manifold forms of improvised music, but what I do and what many other improvising musicians do is not jazz. When I teach I am not promoting a particular style. When I pick up a violin and perform, I have certain forms that I gravitate to and a set of styles which has become mine over the years. But as a teacher, I feel that my job is not to teach students to play the kind of music that I play, but to rather create the context so that they can create their music. The payoff is seeing the eyes of the students as they stand up on stage with nothing between them and the audience – seeing them as they experience the incredible empowerment of becoming independent creators. A fantastic alchemy happens, a thrilling sense of doing something that perhaps even a week or a day before they couldn’t have dreamt that they could possibly have done, and to discover that it is possible. The word “empowerment” has become a bit overused in our day, but I can’t think of a better word to describe what happens when you see the gleaming eyes of students who are giving their first improvisational performances.

In teaching improvisation, I usually begin with a series of theater games, singing drones, babbling nonsense syllables, and movement exercises that create an environment for communication and listening. These activities amount to simple but real experiences of creation – and, importantly in the context of chamber ensembles – experiences of co-creation. There are many games and exercises which serve to stimulate this experience, but the underlying principle of all of them is to keep the activity simple, fun, and at a doable scale. I consistently find that there’s a profound evolutionary process that takes place when you have a room full of people, whether students or professional musicians or actors or whoever they may be, standing in a circle, tossing nonsense syllables back and forth, making gestures at each other and responding to each other’s nonsense syllables and gestures. What comes out in the first few seconds seems random. After the briefest time it evolves into patterns that seem far from random, patterns that already show some structure, interest, musicality. Randomness is an illusion. There is no such thing as randomness in this universe. Minute by minute, as the procedure of improvisation evolves, you see it connect more and more into itself so that it becomes something identifiable. At a certain point, when it seems like things are going well, I ask people to pick up their instruments and start making the same kinds of noises with their instruments at each other and to foster listening and responsiveness. By the end of the first hour, people are already playing three, four, five-minute improvised pieces of chamber music together that sound pretty good. In the second hour, and on into subsequent sessions, you already have something substantial to work with. You’ve already built up a sense of confidence that they can create on the spot, and you’re already at the point where you can stand back and critique what is happening. You can begin to record it and have people listen back and review their process of musical evolution.

One theme that comes to mind today is improvisation as a stochastic process. Gregory Bateson describes the evolution of life as the Great Stochastic Process. Stochastic refers to phenomena that have an appearance of randomness or can only be defined by probability and cannot be predicted. However, the Greek root of stochastic means “true aim” – aiming a bow and arrow and hitting your target exactly right. Innumerable little environmental and internal impacts affect a living system: whether an ecosystem, a culture, or an artwork-in-progress. The living system responds with positive and negative feedbacks, taking on form as time progresses. Soon enough, the system seems to have a shape and identity of its own, the ever-changing result of an ongoing, interdependent process of evolution resulting in profoundly complex and intricate patterns. Similarly, we can watch the evolution of an improvised piece over the span of just a few minutes. Two people look at each other and play two tones; neither of them knows what the other is about to play, but the instant those two tones have been played, they have been heard and responded to; I adjust to you and you adjust to me and we listen, respond, listen, respond. In a period of time so short that the audience can’t even detect it, we are playing coordinated, calibrated music together that is structured, organized and connected. As far as the audience is concerned, they don’t necessarily need to know whether it’s improvised or a contemporary chamber composition.


Here’s a diagram of the range of musical expression along the dimensions of structure and spontaneity.

The left end of the diamond represents the extreme of structure, playing strictly from a score. But few such performances are entirely mechanical or free of improvisation. We have all heard boring performances of great classics in which the performer expertly renders the notes, demonstrating virtuosity and nothing more. We have also heard fascinating performances of pieces we’d heard a hundred times before – fascinating because the interpreter’s personality colors the black and white notes with some of his or her own, idiosyncratic expression, some free play. They’ve transformed the known into something new and exciting.

The right end of the diagram represents the extreme of non-structure – perhaps the caricature that some people hold in their minds of improv. There is “free jazz,” which at its worst is thirteen guys standing up with wind instruments tooting as loud and as fast as they possibly can without listening to each other at all and producing a sort of random cacophony. There is aleatory music,[1] which can sometimes be wonderful, sometimes terrible. But just as no one can play a score strictly without injecting something of their individuality and the whims of the moment into it, it is also impossible to play randomly, without injecting something of their individuality and the distinct structure and shape of their personality.

The middle of the diamond represents the territory of improvisation.

The top quadrant represents improvised work that arises from traditional structures. There we can place jazz, Indian music, Iranian dastgah, the Baroque art of improvising over a figured bass, and similar traditions. This is free play within set parameters. For example, in classical Indian music, you play the same scale that your great-great-grandfather played a thousand times. There are hundreds of non-equal tempered scales to choose from, and hundreds of rhythms and poly-rhythms to choose from. The musicians decide they’re going to do Shri Rag (a raga or melodic mode) in Tintal (a tala or rhythmic mode) and they stick to those parameters. Within the confines of that rule, they express their own particular imagination and energy. Similarly, in jazz improvisation, you decide to play a twelve-bar blues cycle, or a variation on one of the standard tunes. You’re taking a limited amount of structure and weaving your way around it to discover new patterns and energies within that structure. This process can often produce extraordinary works of art.

There’s also a certain type of activity that’s labeled improv, which personally I don’t find very interesting. The audience makes requests – improvise on X, a well known piece or tune – and the pianist riffs on that theme. These performances, which are theme-and-variations rather than improvisations, are usually far less memorable than the original composition. I think the reason they don’t work is because it’s more of a parlor game or technical exercise, a way of showing off the performer’s dexterity rather coming from within the soul and making an original statement.

The bottom quadrant of the diamond represents the music I do. Let’s call it free play – improvisations that weave about a distinct structure, however the structure is not prearranged but emerges from the real-time gestures of playing. Often audience members will come up at the end of the performance and tell us that we must have had at least a conference at the beginning and talked about what keys we’d play in, or how the piece would be shaped. Or if the audience doesn’t know it was improvised, they may think it’s a fully prepared piece of contemporary chamber music. We begin with no expectations or agreements; our entire task is to listen to each other. If we adhere to that one procedure – listening – I can pretty much guarantee that we will produce works that are organized and connected to each other so that they will sound to an audience member as though they had been composed or planned. When I work with students, the recurring mantra is listen, listen, listen. Pauline Oliveros, who has done years of interesting work in this line, conducts workshops that she calls “deep listening.”


Listening is the alpha and omega of improvisational music. You need nothing else, no instructions, ideas, structures, or agreements, if you are committed to fully and deeply listening to and responding to the other musician. In fact, if you commit to totally listening to the other person and to imitating him or her, you are guaranteed an original and coherent result. If I attempt to faithfully imitate you, in fact I will end up doing something different, connected with what you do but inevitably colored by my own quirks and proclivities, by the sounds I have heard and liked, by the way my unique body and mind interact with the instrument and the environment. There is no need, ever, to attempt to be original. Attempts at originality will set you drifting off course, while attempts to listen, connect, imitate, and respond in real time will be colored by your own original nature, and pretty much guarantee that you will do interesting and worthwhile new music.

Another recurring mantra is you’re in charge. There’s a chamber music festival where I coach three daily rehearsals of two hours, followed by a performance – usually quartets with varying combinations of winds and strings. By the end of the second rehearsal, I no longer have to teach anything. The musicians are ready. They have already taken off in their own direction and developed a distinctive group imprimatur. During the last rehearsal we work on presentation. They become comfortable walking out on stage and presenting themselves. They look at the audience and connect with them, take their sweet time tuning and connecting with each other, and when they’re good and ready to play, they play. These musicians discover that they are utterly in charge of the room. They’re in charge of their own creativity. They’re in charge of their relationship to the other players. They’re in charge of their relationship with the audience, unmediated by a music stand or by any sort of plans. Their relationship with the audience is absolutely direct and connected – a palpable power of direct communication.

Another mantra is play less. When I get the students to the point where they are confident that they can play and start listening back to their work somewhat critically, it’s helpful to have recording equipment there and play short pieces, listen back, think and talk about it. Of course, there’s a whole variety of comments that you can make on why this piece could have worked better and what you could have done and even though they’ll never ever do that same piece again in the history of the universe, the next piece that they play will be to some extent based on the comments that you made on the previous piece. 90% of the comments I make are, “Play a little less than you did last time.” The less each person plays, the stronger and more distinctive the pieces become. This connects with the basic performance values of chamber music, of which the most important is, never drown out your partner. The less you play, the softer you play, the fewer notes or tones you play, the better you’ll be able to hear your partner and respond to them. Then when you come into an episode with some strength and force, the effect will be really dramatic. The corollary that arises from this is that improvisation is not a big deal. Students come into the first session feeling, “I don’t know if I can do this.” If you’re a faculty member at university and you want to start an improvisation program, the administrators are going to ask whether improvisation is “real” music. All these doubts about the reality or significance of improvisation dissolve when you realize that within a few minutes of starting a workshop, the most critical thing that you can say to your students is to “play less.” This tells us something fascinating: that the students don’t have the slightest difficulty coming up with original material. Before they walk into the room, they wonder, “Can I come up with original material?” Soon they realize that they come up with far more original material than they can possibly use in any piece – originality is simply not a problem. It’s not even an issue on the radar; instead the issue is how to focus their originality and let the ideas come out of the gate at a reasonably slow pace so that they can be absorbed and digested by the other players and the audience. The constant need to play less is clue that improvisation is perhaps far easier than you can imagine.

The mantra play less also applies to the teacher. I have been in situations working with a faculty group or a mixed group of students and faculty, and I will have been standing at the edge of the room, watching students create some wonderful piece; I’m saying nothing and my jaw is dropping open in wonderment of what I’m hearing, and my jaw is also dropping open in wonderment of the fact that somebody has paid me a lot of money to fly across the country to stand there and do absolutely nothing. Actually, it’s not nothing, it’s giving people permission to do what they already know how to do. Then we have a discussion and a faculty member raises his hand to say, “I’ve been trying to get improvisation going with these kids forever – I’ve tried all kinds of things and it hasn’t worked; why is it working today?” Then I’ll ask them what they’ve been trying and they’ll talk about how exercises and scales and all sorts of structures and templates. My answer is, “You’ve been trying too much.” It’s important not just for the students to play less, but for the teacher to teach less and to stand back and let the process happen. For example, two people are playing together and they’re playing out of tune with each other. If they’re playing tones that they don’t like or that they didn’t want to sound like that, they will hear it. You don’t need to be there waving your hand and saying, “Wait, let’s step back and do this again.” If you give them the space, they will hear it and they will correct themselves. The less you say, the better. The whole art of teaching improvisation is to calibrate your responses in such a way as to be supportive and nurturing – to become the room in which this takes place. The teacher becomes the room, injects as little as possible and encourages them to listen to what they’re doing. Then this great stochastic process of improvisation will take over, and you see how the organized-ness of each piece evolves over the period of the few minutes in which the piece exists, and how the organized-ness of the session or the class evolves from one piece to the next. You experience an evolutionary growth that takes place entirely through listening and responding.

Another concept that I like to play with is the two ‘bins’ of music. Every time you learn a piece, some of that learning goes into a bin of memory dedicated to that piece. But the information also goes into another bin, which is your understanding of what music itself is to you. For example, an untrained non-musician, listening to classical or popular radio, can always tell when a piece is about to end, yes? All the particular pieces that you have learned and practiced go into that second bin as well as the first, but beyond that, everything that you take in, consciously or unconsciously – all the movie scores, all the commercials, all the stuff you hear on the radio, all the things that you liked and all the things that you didn’t like – all go into that second bin of music which is your general idea and feeling for what music is – what works and what doesn’t work, what you like and what you don’t like – and that enormous fund of knowledge about sound is what you draw upon in your improvisations. So there’s a vast reservoir of information and material that students carry within them, upon which they can draw in their improvisations. That’s why it works for you as the teacher to stand back, become the room and simply give permission by setting up a context where the students can critically listen to what they’re doing by themselves without you saying very much. You can watch the work evolve and give the most minimal and simple guidance.

It becomes enjoyable to watch this great stochastic process evolve, to watch pieces evolve within a period of a couple of hours from simple utterances to interesting, organized music. I’ve talked with a few people at this conference who’ve been involved in programs that I’ve worked with and have gone off to teach elsewhere and are now programming improv pieces on their student concerts regularly, and no one in the audience knows that they’re improvised. They’re just playing quality, interesting, new music. One of the blessings of living in the twenty-first century as musicians and educators is that all of the dichotomies of old and new music, classic and modern and so forth, are obsolete. In the twenty-first century we have the entire panoply of methods and sounds available to us, across centuries and across cultures to work with, and all of these are in people’s second bin of music. They’re all present in the place from which people play. We have the extraordinary explosion of world music and cross-cultural influences which have become the dominant musical reality of our time, both in the classical and in the pop world. We’re a good century past Schoenberg so all of the notions of tonal, atonal, and all those warring categories, are long gone. Today we have a chance to play as ourselves, to be in charge of how our music is to be created, how we communicate it to other people. We can experience the liberating value of being at the origin of our creative acts, and experience the pleasure, the confidence that gives to students and performers. It’s an extraordinary experience and it’s worth trying.


Are there any questions?


Q: What are some of the resources for learning theater games and other exercises for getting the students started?

There are many resources. You might look at the reading list on my website, www.freeplay.com. Some of the sources on theater improvisation include Viola Spolin, Improvisation for the Theater, Keith Johnstone’s work, and the work of Augusto Boal from Brazil. Gurus of improvisational comedy like Del Close also have much to teach us. There’s an enormous number of theater games that you can try or make up on your own.

Del Close made one comment on comedy improvisation which speaks strongly to the spirit of chamber music: “Your job as an improviser is not to come up with a brilliant line. Your job is to make your partner’s shitty line sound good.”

I have a particular love of gibberish, the universal language. Let’s say you have a circle of 20 people. You have them pass syllables and gestures around the circle from one person to another, asking them to pass the syllables seamlessly so there are no gaps caused by excess thinking or worrying about doing something good. Then, you ask them to pass a syllable and a gesture to anyone in the circle, and that person passes it along to someone else. The person who does the passing has to conduct their sound and the gesture so that it is completely clear who is to receive it, and the potential receivers have to be completely alert so they can transmit the material without any hesitation.

Since the students are talking nonsense, they cannot possibly be wrong. Music students have been, for generations, terrorized about playing the wrong note, making mistakes or making fools of themselves. But if you begin with nonsense, you’re already in the realm of the foolish, so there is no problem making a fool of yourself.

Then you graduate from passing syllables around to continuing your syllable while the other person is making their syllable. The gibberish syllables begin to harmonize. You don’t say “harmonize,” you just say “keep sounding your syllable while the other person is sounding theirs.” You put a quartet of people in the room and you have them sing to each other’s gibberish. The sounds spontaneously relate to each other as each participant adjusts to make it more interesting. The sounds may harmonize, they may be atonal, they may deliberately clash, they may be sweet or gritty, but they will never be random. There is an automatic entrainment that occurs when people connect in this way, and as the session goes on, the sounds evolve into more organized and meaningful forms.

In the gibberish work, we establish an increasingly interesting facility with sound, and with reactions to sound. We gradually make the transition to real pieces. We begin by improvising 60-second pieces with a beginning, middle and end. One minute is a non-intimidating amount of time. Building up this material in non-intimidating units guarantees a success. What is a beginning, a middle and end? Everybody knows what it is, though each individual interpretation is different. Then do it again and make the connection between the beginning and the end even stronger. It’s useful to play with percentages or incremental amounts, because if you say, “Do something original,” naturally they’ll be stymied. They are not stymied if you say, “Next piece, play 10% less,” or “Make it 10% more musical, however you choose to understand that word.” Take some quantity that is already there, and dial that quantity up or down to adjust the flow of sound to become more appealing or intense. Quickly we begin to generate vocal pieces that are quite interesting, even before the students pick up their instruments. When they then pick up their instruments to play together, they have already built up a framework for listening and responding with immediacy and confidence. Having done 60-second pieces, we can then do pieces of several minutes that retain a sense of wholeness and completion, and evolve from there to long-form improvisations.

When possible, it’s nice to do this work in a big space with no furniture, like an empty performance stage or a dance studio, where people can move around, because then you can incorporate movement and space into the process.


Q: Do you go ahead and start with everyone doing it in a large group and then have them break into smaller groups?

No. I like to keep everyone focused in the same space at the same time. I find it is not helpful to break them into groups because their concentration on each other is broken. It’s much more helpful to work with a small number of people in the center and all the people who are not playing are focused on them and supporting them.

There’s a reason for the success of the string quartet or the jazz trio or quartet or a rock’n’roll band – all of them chamber-size groups, three, four or five people. Bigger groups can be a prescription for chaos at the beginning. until the people get more comfortable with each other. After some experience bigger groups are possible. I have found that the sweet spot is a quartet, which is nice because you can create many kinds of bilaterally symmetrical games. The rest of the class forms a circle of support around them. They surround the performers with their attention and support in a nurturing way. I often work with 15-25 people in a workshop, with four at a time playing in the center of the circle. As this work has gotten more popular, I’ve had groups of 150 or 200. Then it becomes a master class. I might have four playing, 20-30 around them in a circle, and the overflow audience around the inner circle. The inner and outer circles can switch places after a while so as many people as possible have a chance to play. This metaphor of the circle of support is effective. They don’t have a teacher with the baton to whack them if they play the wrong note. They have is all their peers around them listening to and supporting what they’re doing. The quality of the work seems to evolve and the mistakes seem to disappear of their own accord.

I’ve made up some rough rules of thumb: in an ensemble of five musicians, no one play more than 50% of the time. With four people playing, no one play more than 60% of the time. With three people playing, no one play more than 70% of the time. With two people playing, no one play more than 80% of the time. In a solo, play no more than 90% of the time, because the rests are important, too. Playing a solo improv is an exhilarating experience, but you have to remember the power of silence – let the phrases sink in and shape themselves if they are to be meaningful.


Q: How do you want the improvisers to see the audience, to see the observers in this process? Because certainly it must change the sense of what’s going on.

As the students become more comfortable playing within the circle of support, they can come to see an audience in a concert hall as a circle of support too. The great thing about playing without a music stand (or a mental music stand if you play a memorized score) is that you can relate directly to the audience. You can look at them and establish a real connection.

Training in improvisation also affects students when they later sit in an audience themselves. I’ve had students perform improvisations in a festival, followed by other chamber ensembles playing Beethoven or Shostakovich. My students came up to me after the concert and said they now heard the Beethoven and Shostakovich differently. Before, thinking as instrumentalists, they would listen to other people’s performances with an ear for technique, dynamics, phrasing, and all the other qualities of interpretation. After improvising, they now hear other music from the point of view of the composer. How were the pieces structured, how well timed was the introduction of new ideas and the repetition of old ones, how did the composer’s mind work? It’s empowering – I have to use the word again –– when they learn to think as composers. In classical music there’s an intimidating barrier between the audience below and the skilled musician on stage, and an even more intimidating barrier between the musician and the (usually dead) composer sitting up there in the ethers. Through improvisation, we dissolve some of these barriers and open people to the workings of the creative process. Creativity becomes something that we all own, and of which we all partake.

A final note on the subject of teaching: Teachers and conductors are accustomed, when directing an ensemble or giving lessons, to stopping the play and pointing out mistakes. Yes, there are certainly mistakes in improv – we make sounds that don’t work in the context, that meander without focus, that drown out our partners. I have found it useful to let a piece come to a conclusion and then encourage the students discuss it. Interrupting their improv often does more harm than good when the biggest task is building self-confidence and a sense of flow and wholeness. Students seem to consistently hear their own mistakes and engage in enough self-criticism without some extra punishment coming from me. I’ve found it far more helpful to offer feedback about what worked, to encourage and boost the confidence of the students. The mistakes begin to dwindle away of their own accord. My friend Al Wunder, a wonderful guru of improvisational movement and theater in Australia, wrote a fascinating paper called “Positive Feedback.” He points out that when a one-year-old takes his or her first steps, the parents and friends applaud with delight and encouragement. This in spite of the fact that the kid is constantly teetering and falling down. We’re happy and communicate this joy to the kid. We don’t say, “Now Johnny, if you could only hold your back up straighter and put your feet down with a steadier rhythm, you’d walk even better.” We trust the process.



© 2005 by Stephen Nachmanovitch, all rights reserved, www.freeplay.com










[1] John Cage used random number generators to create some of his pieces. One day I visited John and asked him about this. Cage was, in addition to being a composer, an avid mushroom collector and expert. Do you know how he got to be a mushroom collector? When he was a student at Pomona College, his teacher said, “John, you’re too single-mindedly focused on music. You need to broaden your perspectives.” In his typical fashion, John looked in the dictionary and the word before “music” was “mushroom.” I asked him, “John, when you go out and collect mushrooms and you decide which ones to eat and which ones not to eat, do you toss coins or throw the I Ching, or do you use your knowledge of and feeling for?” He gave me a big sweet smile and said nothing.