ARTISTIC OUTPUT IS ALL TOO OFTEN sacralized and mystified in order to protect the artist and their world from the reality that work is work and a lot of what one does as an artist is mundane. Not only do you have to spend many hours of toil building a thing, you also have to build a mythology around it. An image. A posture.
My posture, then, will be to de-mythologize the process, the artifact, and the artist (!) as much as I can, to draw back the curtain and show the man at the levers. By exposing its construction, my aim is to enlarge the work rather than diminish it.
I believe art should make us feel more human, not less.Tweet
Creation is a suite for orchestra and chorus. It’s the biggest, most serious piece of music I’ve ever written, and I learned a lot in the process.
Let’s dig in and see how this sausage got made!
What is it?
- 11 movements
- 1,360 measures of music
- Around 47 minutes
- Piccolo, 2 flutes (2nd doubles bass flute), 2 clarinets, 1 bass clarinet, 1 bassoon, 1 contrabassoon, 4 horns, 3 trumpets, 3 trombones (3rd bass), 1 tuba, timpani plus four percussionists, full SATB chorus, strings
- More than 48,000 notes
- 600 pages or so of printed score (170 pp) and parts
- 4 years
- 1 crazy composer
I think it must have been either hubris or bad judgement that led me to take as my topic the entire world and everything in it for my first major work for orchestra. Looking back, I think I should have written a little suite about, say, trees of the Pacific Northwest — something narrower and close to home.
A woodwind quintet maybe?
Nevertheless, the big sound of the orchestra is what I really love, and my wife Olga had suggested I try my hand at a large piece, and so I shot for the moon (and the sun, and stars, and everything beneath). I figured that around 30-40 minutes long would reasonably fulfill the definition of “big”. I also knew that wanted it to be a suite rather than a symphony, in part because I was daunted by the formal constraints of a symphony and thought a suite would be a more forgiving form (I think it was), and in part because I prefer programmatic music over absolute music, at least when I’m writing it.
I like to tell stories.
The choice of Creation in Genesis 1 as subject just came to me one day. I wasn’t going to start from absolute nothing; I wanted to incorporate some pieces I had already made. Reduce, reuse, recycle! I had one track that I thought sounded like the deep sea, and another one that sounded like flying. I thought they could be edited to be about whales and birds. I had another tracks I thought might work for man and woman — but I abandoned it along the way.
I’m a software designer by profession, so I approached it as a design problem. Go with what you know, right? The first thing I needed was a design brief, a document that describes the project in enough detail that design can get started. I made an absolutely skeletal plan for what movements I felt were necessary to get through Genesis 1, and which phrases I wanted to illustrate. I wrote about each one in prose, as if I were describing the music having listened to it after the fact.
Yes, I made a requirements list. And, like any good software designer, promptly threw it out.
From there I embarked on a process of imagining themes and identifying tropes to use to tell the story.
The difference between an artistic device, a trope and a cliché is the same as the difference between flowers and weeds: It’s how you feel about it being there.Tweet
This involved quite a lot more imagination away from the keyboard than you might think. Meditating on the problem, on the text. Imagining the subject matter as a series of scenes. How do you illustrate a mythical chaos dragon that represents the sea in musical terms? (An asymetrical 7/4 time signature, a lot of tritones and bowed gong is one answer.)
This part isn’t rocket science. Instruments and techniques naturally sound like things. You won’t be surprised that I illustrated the idea of songbirds with piccolo and flutes, or that swarming ocean creatures involve a lot of scurrying fast staccato. Abstract ideas like “light” or “resting” were a lot more challenging than concrete things like, say, animals. (I’ll talk about specific motifs more in the individual movement teardowns.)
There’s no particular reason that the line below would invoke “plants” or vegetative matter of any kind, except the context. Context is everything.
Elegance is about fit, not a measurable amount of beauty. A note, phrase, or chord may be ugly or pretty in isolation, but in context, it may fit or stand out — in the same way a color, a chair, or a swatch of wallpaper might only work in certain rooms and not others. Any element may elegantly work with all the items around it, or it may awkwardly work against them. Or it may elegantly work against them, as a contrast. The point is, it’s all a system of pieces that either hang together, or you end up with cacophony.
Elegance is about fitness to a given context, not a quantifiable amount of beauty.Tweet
I was originally going to use that motif above as a song sung by Eve, the first woman. I abandoned that idea when I couldn’t get it to work out. I didn’t throw the musical idea out, though. Instead, I recontextualized it by surounding it by all sorts of rising figures (evoking sprouts or vines reaching up) and eventually transformed it into a forest hymn. (Look for a teardown of the “Let the Earth Sprout Vegetation” movement as soon as I write it.)
It was an iterative process, and I worked on several movements at once, randomly, in pieces. I’m a doodler. When I draw, I start with an eye or a finger or a mouth and I draw outward from there. I don’t always know what I’m making until it’s made. When I write, I will often start with an opening line or a bit of dialog and then build the story around that. And when I write music, I tend to have a set of disjointed ideas and then I decide on a form somewhere around halfway through stringing them together.
So what I’m saying is: I had a plan that I ignored whenever it suited me to do so.
I have a compositional process that I’ve grown over the years that I’ll blog about some other time, but in a nutshell: I started by recording chords and themes as MIDI with a Yamaha piano keyboard and a computer mouse into FL Studio, a digital audio workstation (DAW) that has built-in MIDI sequencing. (They pretty much all do.) Then I edited the music by dragging notes around, drawing new notes around them, by copying and pasting (a lot). I listen to the music on loop, and when it matches my vision for the piece — or just as often, when the vision of the piece evolves to match what I’m hearing in the DAW — I’m done! The audio files were posted on Soundcloud and that was it as far as I was concerned.
But what if it was on paper?
I wanted to share the music with others, so I joined a few composer interest groups on Facebook and posted links. I harbored a secret daydream of having the work someday engraved and premiered, but getting plays and appreciation for the “laptop music” versions was the beginning and end of my plan.
That is, until I was approached by Davis Brown of Gusthold Music Publishers who had heard the links I posted and reached out to me about getting it published.
I should stop at this point and thank the man profusely. This work would never have been completed as a score and parts without him. I was a fool to think that I could have self-published it at some point. It would have been terrible without him, and thank the Lord, with his gentle support and guidance — not to mention hours of typographically transforming my sloppy manuscript into a readable score and parts — it’s a far better product than it otherwise would have been.
Anyhow, he naturally assumed that given the scope and ambition of the work I had (naturally) been working in notation and keeping a careful score all along, which he was interested to look at. Here’s my response:
Hi, Davis. Thanks for getting back so quickly.
[…] I have not been producing scores as I go. I did just enough scoring in college to learn that a) it’s indispensable, and b) it’s very labor-intensive. I’ve always seen two paths: one toward live musicians via paper, one toward recordings. I took what I thought of as the less impossible path.
If you are no longer interested in working with me, I wouldn’t hold it against you. You’ve already made my whole year just by considering it!
Now, I do think (perhaps foolishly) that I have been producing MIDI in a way that score production should be straightforward, that is, it’s all very tightly gridded. I have been following what I understand to be conventional instrumentation and playable articulations. I haven’t been cheating the audio by “layering” and other production tricks. I have always kept the aesthetic value of performance in mind as a guiding principle. (But see below; there are things.) I suppose I would buy Sibelius if it came to it. I’ve tested the workflow of moving from the MIDI to MuseScore and I’m pretty confident I can make that work in a pinch.
But there’s never been a pinch.
Plus, I settled on making music I could listen to and share on the internet, because for me it really is about the sheer joy of creation, that is, I really do believe that what I’m doing is ultimately for God’s glory. I believe I’ll have all eternity to sing with heavenly choirs. If the dream scenario never happens for me in this lifetime, I think I’ll be okay!
You mention the length and complexity of the work itself. Yes! It’s quite involved. And maybe familiarity breeds contempt, but I think it’s still very much in need of an edit that I can’t do on my own because the defects I hear at this point are probably not what should be fixed. It’s all about blind spots. I mean, I could call it finished today and just never write the final movement (God rested on that day, why shouldn’t I?) but — do you really think it’s good enough to be published? Some days I feel like that whole second movement is unlistenable garbage. Some days I wonder if the animals movement fits with anything else. Some days I think there’s a shocking lack of development and all the structure is ad hoc. Too much timpani, not enough trombone. And there are passages that are so self-indulgent, like the one (one!) bass flute line in the sun, moon, and stars movement. Those things would have to be edited away, I imagine.
So anyway, have I thoroughly discouraged you yet? I’m definitely “willing for you to consider publishing the suite.” You seem great, and I really like that you started a company instead of just going away bitter. (My day job is in electronic publishing; I know how rough it can get for creatives.) I support small business. I’d be willing to work hard, just because I’m stubborn, but it would be a lie to say I have anything like a plan for self-promotion.
I could attempt scoring some of it to see if it’s a workable product on your end, and to establish a timetable for score production and editing to proceed in parallel. That’s a heck of a lot of work to put in on your end.
So as I say, you’d be taking a chance on me, not the other way around.
Soli deo gloria,
How’s that for a pitch letter? Shockingly, he didn’t run away. I asked for a year to do the edit. He suggested I buy Finale not Sibelius. It made no difference to me, so I did. (These days I’m coveting Dorico, though.)
The next year was a deliberate march through the movements one by one, paying penance for the many, many sins I had thus far committed.
Editing, or “The long dark night of the soul”
I decided to do all my editing and rearranging in the digital audio workstation (DAW), just because I knew the tool better and it was faster that way. I use FL Studio, mostly as an accident of history — it was my younger brother’s preference, and I stuck with it. It has tools to quantize the performances as needed, to stretch, fold, spindle, and mutilate them — as sound events on a timeline, not graphical symbols on a page.
I use EastWest Hollywood and Quantum Leap orchestra libraries (via a subscription to their Composer Cloud) as my voice banks, which I both love and hate. As a tool for making orchestral mockups, it’s pretty good.
First I had to develop a new DAW template that divided tracks in such a way that one track in the DAW was one staff in Finale, and I moved over all the notes from the old project files to the new template, track by track. The template was important for two reasons: One, it created a “box” to put all the music into that facilitated the MIDI export, and two, forced me to think through the exact shape and size of the ensemble that this work entailed.
The music projects that I started with weren’t terrible, but it wasn’t very close to usable, either: Yes, I had gridded all the notes with quantization, to make sure the MIDI exported as sensible music, and not the kind of hot garbage MIDI conversion where everything is 1/32 note off the beat.
Quantization: Sort like taking people walking randomly down the street and turning them into a Japanese precision walking team.
All the while I had been thinking about a natural-sounding ensemble. It was real music, for the most part, and not … trickery. But the ensemble had drifted in my mind as the draft unfolded over the three or so years from when I started. Was there a harp or not? How many of each kind of woodwind? Sometimes the chords implied two horns, sometimes six.
There were times I was lazy and put a low note into a horn track instead of switching over to a trombone or something more appropriate to the range, or I made those decisions based on the timbre of my sound libraries and not based on the proper physical characteristics or idiom of the actual instrument.
I might have been writing with a live orchestra in mind, but the one I was actually working with came out of a tin can.Tweet
The chords were structured without much regard to voicing. When you are only making sounds for your own pleasure, you can afford a lot of divisi in whatever string or horn parts you want. I’m embarrassed to say that my normal method of writing chords is to just play them on a MIDI keyboard and not worry too much about how many lines there are. I have ten fingers, don’t I? Many chords were just blocks and needed to be broken up.
I had always decided to fix it later, if I ever decided to publish. (I was always thinking I might self-publish.) Now it was later, and I found myself picking through the weeds and doing things like breaking up “Strings” tracks into first and second violins, violas, celli, and basses; revoicing woodwinds to accomodate two flutes and piccolo, not three (or in one place, four!) flutes; moving technically-in-range-but-please-don’t notes from trumpet to other players, or switching octaves, or rearranging chorale portions so the melodic lines would sit more comfortably within the ranges of the instruments.
(If you’re doing this, I recommend you buy an instrumentation book. Almost doesn’t matter which one. Here’s a good list to get you started. I used Adler in college.)
I also spent a little bit of time thinking about key signatures while I was still in the DAW. I know you can transpose in Finale, but I do a lot of my work by ear, and as I say, I’m comfortable with what I’m comfortable with. In the initial creation phase of the project, I sat down to a keyboard and made the sounds come out the way I wanted them to. This resulted in spending quite a bit of time in the lesser-used keys. This is sort of unavoidable if you decide to structure a piece around movement by thirds, for example C (no flats or sharps) – Ab (four flats) – E (four sharps), not so bad, but D (two sharps) – F# (six sharps!) – A (three sharps).
This poor F horn player:
Thankfully, the F# major passage (which is C# for the horns!) doesn’t last for very long.
I remember being a performer in college and dealing with four or five sharps on a page. I’d look at the name at the top of the page and think, “Who is this jerk, and why do they want me to suffer?” Now I’m on the other side of that coin, and … I’m sorry? I tried my best to balance the needs of the performance with the needs of the music and in the end, ¯\_(ツ)_/¯ ? Trust me, the alternatives were worse. I opted for short bursts of intense pain over longer stretches of less pain.
I don’t know if art is suffering, but artists are definitely insufferable. Myself included.Tweet
Anyway, the idealized process for each movement was:
- Export MIDI from the DAW.
- But wait! My DAW doesn’t export time signatures into MIDI very well, and doesn’t export multiple time signature changes at all, so I had to use a little freeware MIDI hacking tool called Aria Maestosa to hack the time signatures in.
- Import the MIDI into Finale. For each movement it seemed a different set of import settings worked best to produce readable music, so this always took a few times.
- Make the “clean” movement from the publisher’s template also in Finale. I found it best to add empty measures and put in time signature and key signature changes in the empty file first.
- Then copy and paste the staves from the imported-from-MIDI Finale file into the template. (This was faster and much more accurate than trying to conform the imported MIDI to the template.)
- What that gave me was measures with notes in them on the proper staves. So then began the painstaking process of adding all the performance amenities, or about 40% of the music that MIDI doesn’t capture very well: tempo markings, rehearsal marks, articulations and slurs, dynamics and hairpins, technique instructions, and anything else that goes into real-deal honest-to-goodness performable music.
That was the ideal process. In reality, I did a lot of massaging at every step, normalizing, cleaning, pruning, editing. It was basically a big database migration — software people know what I mean.
In the end, this process took around 2,000 hours, off and on over four years to finish. Seems longer than that, to be honest.
In my defense, I didn’t sit down to a blank page (or in my case, screen) at the beginning of this work. I started with three short pieces I had alredy completed as the basis for three movements, and slotted them into the narrative of Genesis 1 as best I could.
I took as my model orchestral suites like Holst’s The Planets and Mussorgsky’s Pictures at an Exhibition. Not that I’m on their level! I just mean that I meant the movements to both 1) cohere as an entire unit and to 2) stand alone if necessary. The movements are between two and seven minutes long, give or take. In some ways, you could imagine it almost like an album in contemporary terms: eleven songs, around a radio hit’s length each. In fact, that’s originally what I was aiming to do.
The movements generally follow the path laid out by the text of Genesis 1. Each one is inspired by a word or phrase in that narrative. And so it starts with “Let There Be Light” and ends with “The Seventh Day.”
As they progress, the movements get longer and more involved, less minimal, more polyphonic and more adventurous in terms of harmony — in the same way that creation itself gets more complex and involved as more categories of thing are added to it.
(These counts will vary slightly from the published work. I wrote a computer program to parse Music-XML out of Finale to count notes, which worked off of the manuscripts I sent to the publisher, not what eventually got on paper. But it’s close enough for blogging. And if you count ten bars instead of eleven, it’s because the tenth movement segues directly into the final movement, so it’s one file.)
Now, one possible structure for the work as a whole would be to organize the thing into seven movements, one for each day in the narrative. There is a very clear “and there was evening, and there was morning, the Xth day” formula that divides the action of the narrative into seven sections. So the logical choice would be seven movements.
So how did I end up with eleven?
I started with an idea that I’d divide the work into days, but as I got into it, the different material for each day didn’t really cohere. And if you look at the narrative, the length of the days is quite asymmetrical.
Generally speaking, there is not a lot of what you’d call motivic or thematic development like you’d find in Beethoven, say. Instead the development of each movement and of the piece as a whole is driven by the story, and it doesn’t dwell too long in any one place. This was mostly intentional. Biblical narrative is famously dense, almost to the point of being cryptic. There are no long meditations or explications about the meaning or consequence of God creating light in Genesis 1:3. He says, “Let there be light,” and ipso facto there’s some light, which causes a separation between light and darkness — and that’s that, on to the next thing.
The music itself proceeds in this way, as well. In lieu of motivic development, there is more like a sequence of moods. I wanted it to seem like the music was telling a story, and for that story to be tightly woven and not very repetitive.
That’s not to say there are no recurring motifs. The text of Genesis 1 breaks into two sections, roughly speaking, the first where various separations of raw elemental materials — light, water, earth, wind — create spaces to be filled, and the second where those spaces are filled with living things. To create linkages between the movements, there are a few motifs that are first introduced in the “container” movement.
For example, “I. Let There Be Light” introduces a “Light” motif (leitmotif, get it?) which is restated in the corresponding “filling up the container” movement, in this example, “VI. Sun, Moon and Stars“. Both movements open in very similar ways, with a pedal tone in the strings and a perfect fifth horn call. This call is then transformed “XI. The Seventh Day” into an explosive call for Creation to praise its Creator. It starts and ends with light.
Motifs are created by an organic process of imagination above. Sometimes, though I resort to other means. That particular motif for “light” was derived using a musical cryptogram from the Hebrew text of Genesis 1:3, illustrated below.
Okay, this is embarrassing, but I said I would go for full disclosure, right?
The first statement of this theme is wrong. It puts a C#, and not an E as the last note. The reason? When I set up my grid for translating the Hebrew alphabet, I transposed the letters tsadi, qof, resh with the letters samekh, ayin, pe. That meant that resh, that last letter, fell on the C# instead of on the E. Worse? I built the entire first movement around a D major 7th chord, spelled D-F#-A-C#. So it appears as A-D-C# throughout.
So, any future grad students out there writing papers about my orchestral works: It was just a mistake. You’re welcome.
As Bob Ross used to say: We don’t make mistakes, we have happy little accidents.
I don’t do cryptograms very often, because the results are not usually very fertile ground for musical development. This one happened to work out, and ends up being a thread that runs all the way through the work, which I think is particularly fitting.
I’ll discuss more motifs in the individual movement teardowns, linked below.
I wanted the work to be interesting enough to not just be a pastische, but also remain broadly accessible. One of the complaints I have about a lot of 21st century concert music is that it’s too challenging. On the other hand, I know that just writing pure imitation of earlier works, while throughly enjoyable, has to compete with 500 years of of Western Music’s Greatest Hits. I mean, if you have a choice between programming Beethoven and a Beethoven wannabe, which do you pick?
So I wanted to find a sweet spot with enough familiar/conventional/common practice elements to be on sure footing as far as audience appreciation goes, but with enough new/surprising/challenging elements to be interesting.
This worked itself out in a few different ways.
Repetitive Figurations. I’m fascinated by the minimalists, Glass, Adams, Reich, even if I don’t actually want to be a minimalist composer. But I have to admit, the minimalists, especially Philip Glass, found a way to make stuff that is accessible but doesn’t sound like it came from the 18th century. I’m more into the so-called “mystic minimalists” like Pärt, Rautavaara, and fellow Pacific Northwesterner, Alan Hovhaness.
But as I said, I’m not too interested in actually doing minimalism, so much as appropriating it in small pieces. And so you’ll find some of what I call “little machines” — figurations of repetitive interlocking rhythms that aren’t quite arpeggios, which create a sort of embroidery effect. I don’t know if this is technically a minimalist technique, but I was certainly inspired to do it by listening to a lot of Glass. There is a lot of hemiola, or 2-versus-3 in these textures, which I love because it creates a shimmering, unsettled effect. Here’s an example in the woodwinds from “Let There Be Light.”
Neotonality? That is, mostly diatonic, with some spice. I’d say it’s more like the soft neotonality of contemporary cinematic music, and not the hard neotonality of the 20th century art composers. Throughout Creation, most of the chords are triads or triad-adjacent, with a few jazz chords thrown in, a few borrowed chords, a modal scale here and there, some whole tone scales, and tritones — lots of tritones. As the piece progresses, it moves from one or two keys per movement, to three or four keys, to several keyless movements, and back again to end on a strongly diatonic dominant-tonic ending. (For the applause!) I’ll dig into the harmonic language of each movement in the individual movement teardowns, later on.
Programmatic Storytelling. One of the things I told myself while working on this piece was that I didn’t have to advance the state of music, I just had to tell a good story. That’s a Tweetable, hang on:
Hey composers: You don’t have to advance the state of music every time. Sometimes it’s okay to just tell a good story.Tweet
In fact, the story had already been told, and it was a best-seller. All I had to do was illustrate that story. What I ended up with wasn’t quite as involved as impressionism but not fully abstract, either. I’m hoping that you find the movement about birds or whales or plants to be evocative of the idea of those things, rather than necessarily mimicing the sounds of those things. (Although there is some of that, too — but only as necessary to spark the idea.)
Occasional Chorus … in Latin? At one point, I had an idea to follow the text beat for beat — for example, to have a particular texture of brass and timpani every time the text had God speaking. I soon abandoned that approach, because it became too mechanical.
Transparency to the source material is one thing; slavishness another.Tweet
I also thought about setting the entire text with the chorus, but instead decided to only grab phrases here and there, and to grab them from various parts of the Hebrew scriptures (and one part from the Greek text of the old testament). I tried setting the texts in English, in Greek, and Hebrew, and finally settled on Latin. This may seem a curious choice, given that I wanted the piece to be accessible. But then again, I don’t want it to just be a Bible story. I want the juxtaposition of music and text and the frission it creates to be primarily musical, not textual. And so I chose Latin in part because it has a simple syllable structure which is relatively easy to set as music, and in part because most people don’t speak it and won’t understand the singing portions as primarily textual, but as primarily musical.
Text and Subtext. Generally speaking, I wanted to hew close to the text, but not too close. I mean, if you just want to read Genesis 1, it’s freely available all over the place. And music isn’t a straightforward way to encode a story anyway — cryptograms aside. And so I hope you find a lot of interplay between the text of Genesis 1 and the subtext created by the music. There are some very conscious choices I made to create harmony with the traditional understanding of certain passages, and to create dissonance with others. (Wait until you hear why the third movement is called “Leviathan and Behmoth” instead of say, “Sea and Dry Land”.) Hopefully the give and take between the language of the stories and the language of the music will spur you to think about the story in new and exciting ways.
A little word on theology. I don’t personally subscribe to any particular interpretative theory about the Creation story. Literal 24-hour days? Days are metaphors for long ages of time? It’s a literary structuring thing? History or mythology? Scientific or not? Directed evolution? I am not willing to fight about it. I expect to arrive in heaven and to find that in the FAQ section of the new arrivals packet!
I’m treating the account as primarily a literary artifact, because that’s where its power and beauty lies for me. I care very much about the exact historicity of some facts of the life of Jesus — death, burial, resurrection — because my Christian faith is founded on them (1 Corinthians 15). But I’m willing to be, well … jolly about how you read Genesis. So long as you do! My intention with this work is not to try and indoctrinate anyone into a particular viewpoint on the creation event, but rather, to celebrate it in all its multifaceted glory.
Why this? Why anything at all? I don’t know, but I’m sure glad to be alive, and I hope that you are, too.
Art is supposed to make us feel more human, not less.Tweet
This has been a real labor of love, and I hope you enjoy experiencing it as much as I enjoyed making it!
This post is part of a series. Read them all: Creation • I Light • II Separation • III Sea & Land • IV Plants • V Sun, Moon & Stars • VI Swarms • VII Birds • VIII Whales • IX Animals • X People • XI Sabbath