HOW do you compose?

[Ballaw de Quincewold]

Do you plink away on a keyboard until you hear something you like? Do you hear everything in your head? Do you use preconstructed melodies, harmonies, voicings? How?

[Thorolf]

I use just anything. Voices in my head. Twiddleing synths, finding a nice sound, and suddenly a theme has emerged. A rhythmical idea. Some word(s). Just walking and thinking. Sitting on the piano and thinking.

But most of all: Listening to “what comes next” out of my head, when the writing has started, not disturbing that fragile silence with trying out all the time.

Thinking, thinking, thinking. And listening.

A lot of jigsawing, “what fits where?”.

Deciding how “composed” I need any idea to be. Sometimes, just the idea is enough.



Regards

[Stenir]

I just let my hand do the composing. Not because I don't want to think about what I'm putting on the page, but because if I start thinking, then I start worrying whether or not I'm doing what I want.

And, in the past, that only led me to start erasing.

So I started writing using a pen.

[ttw]

Mostly I just noodle around (on a piano or even on Finale) until I find a melody or theme or motif that seems interesting. Sometimes, I write a chord progression (like La Folia) and see if something will fit. After I get the motif, I try to expand it by the usual methods (extension, inversion, sequence, etc.)

Sometimes, I copy famous works by Bach or Beethoven or whoever to see how things are put together. I try to do something each day. (I'm in a dry spell now. Things that seem to work just don't pan out. I put them in a file for later use perhaps.)

Sometimes I look a the form or style (waltz, polka, tango, sonata, invention, etc.) and try to fit something into that.

Schoenberg's book, "Fundamentals of Musical Comosition" had some pointers. (Lots of examples, mostly based on Beethoven.)

[stevel]

Originally Posted by Ballaw de Quincewold

Do you plink away on a keyboard until you hear something you like? Do you hear everything in your head? Do you use preconstructed melodies, harmonies, voicings? How?

I offer this analogy: A novelist doesn't just sit down and start typing "it was a dark and stormy night" and write a novel. Most novelists consider all sorts of things - what it will be - murder, romance, both - how nmany characters there will be, will there be plot twists, will it be set in the past, the present, have flashbacks, and so, before they ever write a letter.

Now, that's not to say they might not come up with a phrase, or set of words that inspires them, but I'd say generally speaking, a lot of "pre-writing" takes place. Film-makers are another good analogy. They don't just turn the camera on and say "action" to make the film. They have to have a screenplay, decide on camera angles, and so on and so forth.

Sometimes, I will have an idea that can be as simple as "I bet an alto and bass flute would make a cool duet" and that gets the ball rolling. Other times, I might be playing another instrument and make a sound or pattern that interests me. That usually ends up turning into some playing around with the sounds and seeing what can be made from them, and then maybe there's enough to warrant an actual piece.

But no matter what the initial inspiration, I, for lack of a better term, usually "outline" my pieces. I decide what it's going to consist of before I start writing (though I might have a motive, or phrase, or chord progression, or set of interesting rhythms etc. that I know I'm going to incorporate).

A common story is of the student that comes in to their composition lesson and shows their work. The instructor says "these 8 or 16 bars are great, where's the rest?". And the student says "I don't know where to make it go" - or I don't know what comes next, etc. As a result, a lot of students end up patchworking together a bunch of unrelated ideas in order to make the piece "longer", but what happens is is comes out incoherent and "amateurish" sounding. The reason they can't make it go anywhere is, they haven't even thought about where it should go.

I recently wrote a work that I'll use as an example because it's pretty clear. Firstly, I had been wanting to write a small work for duet. I needed some kind of "restrictions" (one of my colleagues talks about setting up a "paradigm" for the piece - that is, the world in which the piece lives). I found a call for scores that had restrictions that helped inspire me to write something, but keep it within some guidelines. I have found (like my colleague mentions) that if you can do "anything", it's too hard to narrow down what you're going to do. So in an effort to get rid of junk you don't need, you can set up a little world in which this piece operates (just like how stories like Star Wars create their own little reality).

One of the restrictions of this event was that pieces were 100 notes or less.

So what I did was write 10 little mini movements of 10 notes each. I decided that I would use the duo. I decided to use 10 notes of the chormatic scale for each "movement". I decided each instrument would have a solo movement where they played all 10 notes. Then the rest I divided up. I decided NOT to do like 1:9, 2:8, 3:7, 4:6, etc. and try to have one of each possible combination. I felt that was a little too restrictive. I decided I wanted to have one piece where it was 5:5, and they played together, same rhythm. I decided I wanted to have some where one part played long notes, while the other played faster notes against it. I decided I wanted to have something like a canon where they alternate (and I'm very pleased I came up with a canon, where I "halved" the notes and gave it to the other instrument, and you have to figure out how it's a canon :-). I also decided there would be some obviosu contrasts in tempo and mood between each little snippet. I also made a decision to make this part of a larger set, and, to have a little more fun with the number 10 like make all the tempos 200, or 60, 130, etc.

All of this I decided before I wrote the piece (though I had little snippets of ideas for a couple of things). After you've done all of that "pre-composition", it's kind of like "paint by numbers" at that point. I just then thought of the kind of interplay I wanted between the two parts, and the kind of contour the lines would have, and just set down at the piano and tried to "feel" for the sounds I wanted. I'd play something, and then try to imagine things like - should the next note come fast, slow, higher, lower, close, far, consonant, dissonant, etc. Then I kind of found my notes by "trial and error".

So I typically "outline" all my pieces like this - or else I find I have a nice introduction with nowhere for it to go. Now - the outline is flexible - another piece I wrote was going to be two movements originally, but I found the ideas I had for the second movement made a nice companion to the A idea of the firs movement, so my two movements became ideas A and B in one movement only. I think if you don't allow for this kind of flexibility, you can "paint yourself into a corner". Either that or it can come out like integral serialism (and I'm not too into that).

Now, I don't compose traditional tonal music. On some levels it makes things easier - for instance, I may not have to worry about finding which harmony to accompany X note. But on other levels it makes things harder - for instance, I can't just decide on a harmony and then arpeggiate to make an accompaniment figure if I'm not using traditional harmonies. But I think the idea of "plotting out" what you want to do is a very important step that people don't learn about.

And I'm not saying this is the best method for every type of piece either. Sometimes, if you're writing in a style that has an improvisatory feel to it, it might be more effective just to fiddle around until you come across something that sounds as "spur of the moment" as you want it to. And if you're writing something highly contrapuntal and plan to use inversions, augmentations, retrogrades, etc. then you need a little more pre-planning, especially on those specific things.

Another good analogy is like a football game. They don't just go out there and play. They have "plays" they've decided on before hand. Now since there's another team involved, things change, so they have to prepare for that, but they do have a basic "game plan".

Same thing if you're organizing a wedding. You've got to do some pre-planning. You can't just show up on the day and expect the cake and flowers to be there.

Now with music, I've found that some of the decisions I have to make, or the "world" in which the piece exists might need to be specific. For instance, if I know I'm not going to be able to find a great crumhorn player, I'll probably avoid writing for that instrument. If I know I'm going to get a student to perform it, I have to write at their level. If it's a pro, I don't want to hand them something a beginning piano student could play (other than a read through). So I take a lot of things into consideration - not just the composing of the piece, but it's instrumentation, possible performance, etc.

I also wanted to address more specifically what happens to me while I'm working on a piece, because I found a lot of other people share this same experience. Once I've got a basic idea, I'll often sing it (just the basic countour and rhythm, my pitch is not perfect) while driving in the car, and then I'll come up with more general ideas - like what will come next, general direction, rhythm, speed, activity, etc. That is, I'm not super specific at this point, but it's like thinking about how the story will continue, without writing down the specific words - does that make sense?

I've found that if I just think about the music - not specific notes and things but just the general "storyline" (I'm comparing it to literature here, not saying I'm necessarily writing narrative music) and I'll come up with ideas and solutions to problems in the oddest places - in the car, at the grocery store, on the throne, in the shower, before class, etc. Rarely while I'm sitting an an instrument and playing though! Then I come back later and try to take these ideas and turn them into actual pitches.

just for example, in the car, I might find after thinking through the piece, that two sections should be bridged by upward ascending notes in chordal fashion (or generically, ascending blocks of sound). Then when I get home, I sit down, look at points A and B and decide what notes to use to make that happen. If I had a 5 note spand to cover, I might try a scale, whole tones, chromatic, thirds, etc to see what kind of distance each sound needs to be from the next (or even if that should vary as it goes - bigger jumps at first to smaller jumps as we approach, etc.). I'll try to find the sound I want - open sounds, dense sounds (this is where theory (and experience) comes in handy because you know if you want open sounds, you can try 5ths or quartal harmony, and use 2nds or clusters for denser sounds, etc.) - whether they'll move parallel, or willy-nilly, etc.

That's usually how it goes.
So I hope that helps. I know methodology can be very personal, but I also know this is a subject that's not explained well a lot of places (becuase it's so hard to explain a lot of times - especially if some people are simply "inspired"). This is my experience with it. But if you (or any people who've bothered to read this far) find yourself getting "stuck" after 8 or 16 bars all the time, I'd seriously recommend making an outline, or storyline, or plot, or whatever you want to call it, and think through what you want the piece to do, and what kind of paradigm it's going to exist in. I think you'll be rewarded with compositions that virtually write themselves (not really, but it's better than continually throwing away unfinished ideas that don't want to budge - and if you do they sound forced).

Best,
Steve

[stevel]

Originally Posted by Thorolf

But most of all: Listening to “what comes next” out of my head, when the writing has started, not disturbing that fragile silence with trying out all the time.

Regards

This is a very good example of what I was mentioning. I find the "what comes next" is easier to find if I have my little storyline in place (and again, not inflexible). ANd each time you need to move on, or bridge two sections, the "not disturbing the fagile silence" (well put Thorolf) by trying to "force" the next notes is the best way to fill things in. Too much of that trial and error can destroy the "flow" if you've got the flow going. And the reverse is also possible - if the flow is not there, sometimes a little plunking can get it going. But, one has to be careful of just beating an idea to death - then it often seems like there's no solution - sometimes you need that "fragile silence" and use it to come back with a fresh approach.

I wanted to add, I'm also very specific about what I compose. I don't like to put in anything just for the sake of putting it in. I have a reason for everything and I make sure every single note is what I want. I may want if for some stupid reason that only I understand, but it's my artistic vision. I don't want to feel like I "have" to do X just to do it (like a lot of people feel like they "have" to write a Symphony, or in multi-movements, or using serial techniques, just because of something they heard somewhere and haven't fully thought through).

Once I got myself into that mindset, I found that I was much happier with the music I was producing.

Best,
Steve

[Keith]

i write Program music, so its a matter of reading the original poem or story, then i can use the literary devices of the author to create the rythmn. then play a siplified version of the most basic idea and improvise constantly for days until something sticks. improv alone works wonders for absolute music. if a section is good, you'll end up playing it every time. treat every theme like its a real person and let it decide how it wants to be used.

[sawscape]

I generally go with the noodle-then-expand approach, as mentioned above. Once I have something in Finale, I will listen to it over and over until I have a good idea of what I need to fix or add--this includes structural decisions. The listen-fix-listen-add-listen cycle will go on for a long time until I think that the piece is done, although almost everything is open to further modifications.

[reith]

I've recently had to ask myself this question. In trying to fight a "symbolist" tendency that relies a lot on motif and fairly short phrases (indeed, pieces), I started a symphonic-styled work a couple of years ago. I'm now tackling the 3rd movement - it seems necessary to hear quite long stretches in my head before I can write then check it at a keyboard. Sometimes I get something wrong - the harmony doesn't work, but if the contrapuntal elements and flow are ok that's usually easy to adjust - at least, I note something and can revise it later...as long as I can move on.
If I've notated the current part of the theme badly, I have to guard against overworking it as that leads to frustration and I end up putting it down for a while. I don't know if others find that overworking something has this effect.
In that event, I work on something else (as long as it doesn't look back to the more impressionist stuff though I won't deny giving-in sometimes).

Reith.

[stevel]

Originally Posted by reith

I don't know if others find that overworking something has this effect.
Reith.

Yes, overworking is the Devil incarnate. I've learned (I hope) to catch myself early and walk away (or work on another spot). But I remember this one piece where I knew I wanted this contrapuntal bridge between two sections - basically the piece was finished except for that bridge. I worked and worked the counterpoint, tried various kinds of entries, tried more consonant/less consonant, tried imitation, tried free cpt -didn't matter - no matter what I came up with felt forced and it became this struggle to put something in.

Then one day I'm going over it in my head, and looking at some other composer's music, and it hit me - THREE part counterpoint instead of 4 part! I got home and it virtually wrote itself using virtually the same material. I was trying to over-compose it, and in doing so, it was causing me to over-think everything and not see the forest for the trees.

Steve