How to (Not) Teach Composition

A comment from Scott, a former Bard student, brings to mind ideas about teaching composition that I wonder if anyone else shares. I’ve never heard of anyone being taught how to teach composition, except indirectly and by example. We’re all winging it, aren’t we? I had one composition teacher who never said a word I understood, and seemed to enjoy keeping me confused; another who disliked every piece at first and thought it was lovely as soon as you copied it out neatly; another who was brilliant but considered Cage and minimalism hoaxes; another who merely tinkered with my orchestration (Feldman); one who changed my life simply by telling me about just intonation (Johnston); and another who did me a world of good simply by enthusing about everything I did. Little of that activity seemed like teaching, though by default that’s what we have to call it. I know of teachers who start out, “Here’s how you compose a piece,” and actually show you how to do it, step by step. I think that would have driven me nuts. It always seemed to me that what you mainly absorbed from an older composer was attitude, and you gravitate toward the attitudes that appeal to you.

Subsequent experience has suggested to me that there are two kinds of composition students: those who write too little music because they’re too self-critical and those who write too much because they’re not self-critical enough. (Those who write just the right amount are unheard of.) I feel much more useful with the ones who don’t write enough, because that’s the problem I worked through myself. First of all, I hardly make any suggestions on the first piece or two, because I need to learn what their reflexes and aims are before I can tell what they’re trying to do and figure out where they’re not succeeding. Then, when I see them turn down the same road a third or fourth time, I can say, “Aha! You always do that, don’t you?” Not to prevent them, but just to make them conscious. Sometimes I’ll play them, and analyze a little, pieces that do something similar. In rare cases, if they’re really stuck, I’ll assign specific compositional problems to attack (like, write a trio only using five pitches). I’ve only had one case in which that strategy really lit a fire under the student; more often, they start composing their own pieces faster to induce me to stop.

Typically I try to isolate the essential idea of a piece and write my own little variations on it, to look for ways to continue. Sometimes, like all composers, I’ll pull out a piece of my own, show the original problem I struggled with, and show how I solved it. I almost never write notes on their score, because that drove me nuts when I was a student. But I love taking their Sibelius files, “saving as” something else, and tinkering with changes. Occasionally they like the changes, but it seems to me that most of this activity simply serves to focus them and make them think harder once they leave my office. As an ongoing influence, I harangue them to turn off their superegos, to silence the constant critical carping in the back of their minds that tells them, as soon as they’ve written three notes, that those notes aren’t good enough, that compares every line they write with Bach and finds it wanting. If they can do that, they can do anything. Sometimes, if they’re obsessed with a certain composer they worship, I’ll forbid them to listen to that composer any more. I had to quit listening to Cage in college for that reason.

The students who write too much are much harder for me. They often come in with a new completed piece every week, which suggests a certain success that’s hard to argue with. It’s difficult to convince them that revision is golden, that even when it’s going well you shouldn’t always settle for the first idea that presents itself. And then, there’s little opportunity to direct the course of a piece when it comes in with the final double bar already in place. Of course, like all composition teachers, I pound away at notation, and describe at length how performers respond to, especially, rhythm and articulation markings. That’s the easy part, though they won’t always believe that you can’t start a 4/4 measure with an 8th-note followed by a half-note. My teachers’ generation often taught little besides penmanship and types of manuscript paper. Similarly, I end up teaching the finer points of Sibelius (the software).

I’ve taught mostly undergrads, and am happy with that, because they’re more pliable and need more advice. What I’m not sure is practically helpful is trying to prepare them for the way a composer’s process changes with the years. I think it seems pretty well established – though I don’t know that I’ve ever read this anywhere – that young composers tend to get inspired by a sonic image and then just start out without knowing where they’re going. At some point, in a person’s 20s or 30s, those sonic images become less spontaneous, and it seems to me that a composer has to learn to quit waiting for that inspiration. Typically, I think – and I ask this as a question – college age composers tend to have tremendous bursts of inspiration, and be almost incapable of composing when not inspired. As your psychology changes in your 20s, you start thinking less of individual moments (or melodies, or motives) and more about strategies for entire pieces (like chord progressions or rhythmic structures). Then it becomes easier to just sit down and start writing, inspired or not, and at some point inspiration creeps in and lifts the piece off the ground.

I’ve interviewed dozens of composers for my articles and books, and I’ve heard one process described over and over again. A young composer will find some method to generate his music, and use it strictly. At sometime in his 30s, he or she will have internalized the method enough to quit using it consciously. My favorite example like this was Petr Kotik, who, in his early years (Many, Many Women is the famous example) based all his melodies on graphs he had found discarded in a science department that showed the rates of alcohol-level response in experimental rats. For years all his music went back to those found graphs as a kind of chance process based on “found” contours. Then, he said, at some point he learned to intuitively write melodies that had those characteristics without using the graphs anymore. I think you find the same pattern – overly strict at first, then later freer and more intuitive – in most of the good serialist composers, Boulez and Stockhausen in particular. Then, many composers also go through another tremendous shift during their 40s, in which many of the compositional ideas they studiously avoided in youth begin to appeal to them, and their music finds a whole new level as they start playing with the devices they swore they’d never use. It certainly happened to me, when around 1999 I started allowing myself to borrow stylistic paradigms from jazz and other composers. (Make sure to shy away from some great ideas when you’re young, so you’ll have great new toys to play with in old age. I have an article in Music Downtown. “Mistaken Memories,” quickly tracing this phenomenon through a wide range of composers.)

As I say, I don’t know to what extent it helps a young composer to be aware of the eventuality of these changes coming in later in life. I like to think that they’ll accept the changes more easily if they know it’s a lifelong process of evolution, that they won’t get stuck clinging to the process that always worked in the past. The most important thing of all, I think, is to get them used to the process of trying to extend their technique into some area they’ve never tried before, getting stuck, feeling helpless, and just living with their failure until some imaginative breakthrough suddenly makes them see where their imagination had been too limited. It’s a process that can cause considerable despair when young, and, as you gain experience, you learn more (or I have) to accept that “stuck” feeling as the necessary prelude to a creative breakthrough.

These, off the top of my head, are my thoughts about teaching composition – almost as much, or more, a process of emotional therapy than actual musical instruction. I am curious as to whether other teachers find this same dichotomy between young composers who write too little and those who write too much. I wonder if other composition professors muse as often as I do that training in psychotherapy might have been more helpful for this line of work than musical training has been. And I wonder if the process of “switching gears” in your 20s, relying less on sudden inspiration because it doesn’t come as often and you no longer need it so much, is as universal as I imagine. It seems to me that we composers don’t talk about these processes enough, that we all live in ignorance, blindly feeling our way, having to figure out how it works through self-examination and anecdotal evidence. Perhaps the blogosphere is the arena in which to shed some light on the categories of artistic evolution. I yield the floor.


  1. says

    Very nicely put. No one really “teaches” composition any more than I had to teach my kids how to breathe. One can teach the mechanics of it (theory, counterpoint, notation, business aspects, etc), and certainly encourage someone to trust his or her instincts. But teaching someone how to compose is about as silly a notion as teaching someone how to write poetry. We can’t teach someone how to be creative.

    I know the academics out there will weigh in that it’s important to have a supportive environment, have opportunities to get peers to perform one’s music, yadda yadda yadda. That’s all true. But it still doesn’t contradict the fundamental assertion that one simply doesn’t teach composition.

    Everything I learned about composition I taught myself, aided by self-study of an average of 10 scores per week by contemporary composers. I can’t think of a single thing I learned from my official composition teacher during high school. He was very nice and all that, and I’m sure he meant well. But after awhile, he stopped coming to teach while at the same time I stopped showing up to lessons, hanging out in the music library instead. We mutually gave up on composition “lessons,” which probably was the best thing at the time.

    Were I to give advice to a young composer, which I probably shouldn’t since I don’t feel particularly qualified to do so, I’d stress the importance of listening to as much music as possible, and diverse music at that. Western, Eastern, classical, rock, jazz, etc. Part of the problem with conservatories is that they tend to be myopic, ignoring the wealth of music around us. That’s one reason why Cage’s aesthetics were so valuable. He thought outside the box, while so many continue to learn within the box.

  2. Rodney Lister says

    I agree with just about everything you say. My experience is slightly different since mostly I teach kids (i.e. pre-college). I think what you said about young composers not having an idea of where they’re going is very true. I’ve found that I continually have conversations which start with my asking “What does this piece do?” (or what is it supposed to do, or will it do, or something…) I think it’s important that they have some kind of intention that they’re aware of–even if along the way it changes. (Apparently when Birtwistle was teaching at Harvard recently he a lot of times asked a similar question: “What is the piece?”). (At one point I did a lot of thinking about what one kind of standard one could apply to any piece–I finally decided that every piece does something.) Rate of production is tricky. Sometimes you feel that you’d like to see the student do more and do it quicker and sometimes you think it’s important to try to slow them down and think more about what they’re doing. What’s called for seems different at different times. Mainly I try to keep them working and trying to make sure that what they write down is a close as it possibly can be to what they mean.

  3. says

    This post and the previous one are not only great, but timely. I’ve been teaching now for a couple of years and this fall I’ll begin the real test – being completely in charge of the composition department at SUNY-Fredonia. Obviously these questions are always there, but they’re currently very much on the forefront of my mind and it’s comforting to hear similar sentiments from you.

    As far as composition pedagogy, what you’re saying is pretty much where I come from as well; it’s not as much a matter of actually teaching how to compose, but rather to allow the student the opportunity to compose a lot of music in the right environment with enough models to choose from (both in process and in attitude). This can be assisted with the existence of many performance opportunities as well as strong collaborative attitudes from the the other music faculty and, through them, their students.

    Working with new composers can be a challenge, but I can still put myself in their shoes. I was writing for many years before I had a single composition “lesson” and by that time I had picked up a lot of different tools from jazz and film scoring that seemed to help when applied to writing concert music. Writing the notes wasn’t the problem – getting over the huge conceptual hump of believing I was a composer with a upper-case “C” was much more difficult (due in no small measure to my background as a jazz saxophonist/arranger). Once I got over the initial self-doubt and the more opportunities I had to write at a high level (with the accompanying high expectations from instructors/performers/conductors/etc.), only then did my own chops blossom and I started having some serious fun.

    Your techniques of “save as…” in Sibelius (I do the same in Finale) and giving the students some different options in real time is spot on with what I’ve done in the past. With students who write quickly and a lot – if they bring in a piece already completed, sometimes I’ll ask them to think about revision, but not before they’ve already heard it once in performance. I consider all of my premiere performances rough drafts before the final edit, since it’s only when I hear it performed live with an audience that I truly experience the work as it honestly is, scars and all.

    If there’s one concept I think I can get across to the students, it’s the idea that how you say something musically isn’t nearly as important as having something personal to say in the first place. As they age and mature, they’ll transcend the “how” and focus on the “what”…if I’ve done my job. It’s not deep, but it’s a start.

  4. Paul Beaudoin says

    At the large university here in the Northeast where I teach composition, students are given pvt. lessons in composition for 2 years. If they “stick it out” the following year is spent in a composition class. Most of my students have very little experience in actually writing music. They have worked with software that allows them combine loops and sequence events but there often isn’t much thought about harmony, form, or counterpoint – though it can often have some very exciting textural qualities. But I often wonder what motivated them to become a “composer”? It has also motivated me to think about what I mean by the word “composer.”
    I think many decide to become composers after significant musical experiences – of any kind. I remember wanting to become a composer after having heard Stravinsky’s Sacre for the second time. This first time I heard it I didn’t “get it” and gave it a second hearing. That experience was far more powerful than the angst of Black Sabbath. Here was a music that moved me at a very deep level and I wanted to learn how it was done and learn how I could do it for myself. I spent many Saturdays at the public library listening to all kinds of music and reading as many scores as I possibly could. I wouldn’t have that kind of experience again until I heard Mahler’s Third Symphony in Boston in the 1990s.
    More to a point is that I think young composers write music similar to that which they hear. My first advice to my students is always to listen voraciously to all kinds of music and as often as possible with a score. This allows them to see how to make “things” happen. Model their work on the kinds of music that gets them excited. I often have the student analyze the modeled work in tandem with their own composition. But unravelling the threads of the model we can learn how it was assembled. Hopefully we can reassemble the threads in our own effective way.
    On the other hand I often feel that today’s young composer needs to be as versatile as possible. I think one of the best things I can do is to encourage them to experiment in various styles so that their compositional toolbox is well supplied with the kinds of tools the composer can use with some sufficiency. Write the “rock tune” but then arrange it for a jazz band. Write the new age piano piece (whatever that might mean) and then score it for a chamber group or make it an underscoring for a film clip. Do something according to a chance operation and then subject yourself to the rigor of serial music. This is one of the (many) things I admire in the music of Bjork. Her ability to recontextualize her tunes in various panoramas is, for me, one of the splendors of her sonic world.
    I agree with the sentiment heard here that these encounteres eventually strengthen ones “original” voice and along the way introduces you to methods of musical expression that you can use when they are the best choice for that expression.
    We don’t (usually)use a saw to change a tire.
    In the end, as a composition teacher, I can only respond to the music brought to me by my students and reflect back to them how I experience their work. This always leads to questions that I hope focus the task at hand and support its continuation.

  5. Bob Jordahl says

    How exhilarating all of this is! Why not expand it and open the discussion to all composers?
    KG replies: Well, I yielded the floor. Expand it to whom? Composers who don’t read my blog? How do we get them here? Leave a trail of bread crumbs from The Rest Is Noise?

  6. Roger Ruggeri says

    This has been an excellent and insightful exchange about concepts that, at least for me, have too often remained internal.As a longtime orchestral bassist, program annotator and occasional composer, I’ve rarely come upon such honest candor about the compositional process. Bravi!

  7. says

    I get the sense that your comment, “What I’m not sure is practically helpful is trying to prepare them for the way a composer’s process changes with the years” is in some way a reaction to my statement about how I like to present many different ways of composing to students so they can learn not to rely on inspiration alone. The thing is, in my experience, the problem of not knowing what to do next is most certainly not limited to older or more experienced composers. I taught a classroom full of high school students a month or so ago, who were working on writing some simple little songs. Almost all of them didn’t like what they had written very much, except for the beginning. So I asked them how many of them sat down to write, started at the beginning, and then got stuck soon thereafter, and everyone but one student raised his/her hand. Then when I asked them how they eventually finished their songs… well, some of them never did and gave me pages of unfinished sketch. Some of them said they just scribbled down anything to finish the project. And some of them said, “well, I just kind of repeated the part I had written until the song was over.” Only about 2 kids out of 25 had written complete songs that they were satisfied by.

    So I think there’s a great use to teaching various techniques for approaching the blank page. Of course not every student will be helped by this, but not every composition student will be helped by anything, so the best we can do, I think, is to give them a huge array of knowledge and let the 50% (or whatever) that works for them stick, and don’t worry about the rest. Again, I believe quite strongly that more knowledge can never hurt. Best-case scenario, they’ll find their own way to understand and incorporate it into their thought processes; worst-case scenario, they’ll ignore it and resent me a little for making them learn it. And even then, chances are that something about it will sink in a few years down the road.

    The issue of whether or not it’s possible to “teach composition” seems to be a contentious one. For what it’s worth, I feel that yes, you certainly can. Of course, there’s a component of being a composer that can’t be taught; the ever-elusive Inspiration and Creativity. But that’s the same in any field. To be a great chemist, you need the creative inspiration to come up with a brilliant experiment that’s timely and effective. Without that inspiration, you can’t be more than a mediocre scientist. But that doesn’t stop people from teaching chemistry, because all that knowledge enables one’s creativity to come to the fore.

    I like to think of composition teaching like acting teaching. Sure, there’s an element of acting that just involves having the right spark. And the public tends to assume that acting is all about talent and nothing else. But there are techniques to acting, and I’ve never met a very successful actor who didn’t use them. You need to know how to analyze a script. You need to learn to follow every impulse without blocking. You need to understand that everything you say must be motivated by a specific goal, and that goal must be active rather than passive. You need to learn to never move a body part extraneously, but dedicate your entire body to the task at hand, which gives you stage presence. You need to understand the specific impluse behind each word you say. And so on…

    I think this is directly analogous to teaching composing (and painting, and poetry writing, et cetera). There are many sets of skills, from the very practical (orchestration, voice leading) to the more abstract (following impulses, not censoring yourself, learning to take risks). These are skills that make one a better composer, which is teaching composition. Sure, we can’t teach an uncreative person to be creative. But I don’t think there are entirely uncreative people. Some people just need the creativity teased out of them, need help not being embarrassed by it, need help trusting it. And in the end, some students who study composition won’t be able to access their creativity, and won’t become great composers, just as some students who study chemistry will still turn out to be bad chemists.

    In the end, I just think it’s a big fallacy to mistake being inspired for being an effective composer. Many of the “you can’t teach composition” arguments seem to be implicitly saying “because composition requires inspiration, you can’t teach it.” But again, that’s like saying that because being a succesful chemist requires inspiration, we can’t teach it. But no one ever questions chemistry teaching… because it’s understood that while we can’t teach inspiration, we can teach someone to know how to handle that inspiration if/when it strikes.

    Once again, long, scattered comment. But I’m just fascinated by this topic, and am enjoying all the discussion. Thanks for opening the floor!
    KG replies: I wasn’t actually reacting to your statement at all, just talking off the top o’ my head.

  8. says

    To be a great chemist, you need the creative inspiration to come up with a brilliant experiment that’s timely and effective. Without that inspiration, you can’t be more than a mediocre scientist. But that doesn’t stop people from teaching chemistry, because all that knowledge enables one’s creativity to come to the fore.
    Not to split hairs, but there’s a huge difference between the sciences and an artistic pursuit like music composition. Yes, one needs some inspiration when designing experiments, and sometimes even some creativity. But the vast majority of chemistry, physics, mathematics, etc. remains grounded in objective theory and evidence supporting the theory. Richard Feynman was both a great physicist and amateur artist, but he kept the two very separate and while he could appreciate the beauty within nature (indeed, he was very eloquent about it), clearly maintained that physics and mathematics were more rational and fact-based.

    Nothing about composition is fact-based. I was taught how to do differential equations and how to devise and perform experiments in molecular biology. I was never really taught how to compose, and I have no idea what that even means. The term “teaching composition” has zero meaning to me, whereas it is very clear how people can teach someone how to do scientific experiments. The insight for good science is not necessarily taught, since there has to be an innate skill and vision in that regard. But most science really is not so much creating something new as it is making an observation and trying to explain it. Composition is, by definition, creating something new.

  9. says

    Not to split hairs, but there’s a huge difference between the sciences and an artistic pursuit like music composition … most science really is not so much creating something new as it is making an observation and trying to explain it. Composition is, by definition, creating something new.

    Thanks for the very thoughtful response, David. Personally, I just don’t think there’s as huge a difference as people tend to think between the arts and the sciences. Obviously, composition is more creatively-focused than chemistry, but they also have a great deal in common. I’m just not sure I see the huge difference between studying differential equations to become a lab chemist, and studying (for instance) practical music theory to become a composer. Neither of those skills is directly applicable in the actual doing of the task at hand, but both of them provide a strong mental framework that allows the experiment/composition to go ahead more smoothly. One could have a great lab experiment without knowing differential equations (like Alexander Fleming discovering penicillin on accident) and a great piece of music without knowing music theory and other musical strategies. But it sure does help to have a base of know-how.

    Also, while it’s not the whole name of the game, there’s a strong element of the scientific method (observation, hypothesis, experiment, results) in composing. I observe that “Der Leiermann” from the end of Schubert’s Winterreise moves me. I hypothesize that it’s the incredible stillness and repetition after so much tumult that creates a beautiful, somewhat creepy mood. In a future composition, I design a piece that ends using a similar tactic. And then I observe, in my mind, and perhaps in talking to an audience, whether that compositional strategy is effective in the way I want it to be.

    I’m making it sound a bit cold, perhaps, but I think this is a process we constantly use as composers. We hear things we like, we guess as to what’s so cool about them, and then we try to extrapolate ideas, lessons, and theories from them, which we later absorb into our own compositional technique. And in order for this to be more and more effective, we have to learn things. We learn music theory to better understand the music we’re hearing, and to help construct our musical experiments. We learn strategies of approaching the blank page so our experiments don’t stall. We learn to feel comfortable in lots of musical environments, so we can have a full range of “lab equipment” when we embark on a musical experiment. And so on…

    Of course, there are successful composers who don’t have all this knowledge at hand. Heck, I don’t think anyone really has all of it. And there are chemists who aren’t that good at figuring out reaction mechanisms and equations, but have great success in the lab. But all the knowledge you can get helps. A lot. And teachers help us acquire that knowledge. Whether we call it “teaching composition” or “teaching skills that help composers,” it’s the same thing, and a great thing, in my opinion.

  10. Dan Schmidt says

    The one technique thing I actually learned from a composition teacher was MAKE USE OF REGISTER! All of my pieces had tended to sit within a radius of an octave from middle C.

  11. says

    Scott, you made my point for me exactly. One can teach theory and other mechanistic, common core subjects necessary but not sufficient to be a composer. But you can’t teach someone how to actually compose. Let me put it like this—there’s a difference between learning how to write notes (theory, counterpoint, etc) and how to compose. At least in my uninformed opinion. Thanks for a very detailed response, however.