By my own reckoning, Midnight Blue is one of the most successful pieces I have composed. I feel like I achieved everything I set out to accomplish in the piece, in terms of both form and expression. Audience response to Midnight Blue has been enthusiastic, and it is also the piece that earned me my first groupie. And isn't that why anyone goes into music in the first place?
The many splendors of Midnight Blue are the end result of an experiment which took a year and a half to carry out. It began in the spring of 2002, during my studies at Duke with Anthony Kelley. We had examined some compositions which employed additive structures -- forms in which new material is progressively added to a preexisting seed -- paying particular attention to "Call to Prayer" from In This House On This Morning by Wynton Marsalis, and to measures 7-30 of The Yellow Pages by Michael Torke. Although the additive structures in these two works are both clearly audible, they are developed in different ways. In "Call to Prayer," Marsalis starts with a very basic three-note riff, and elongates it, by adding extra notes at the end of the riff, and prepending a sequence of chords leading into the riff. In "The Yellow Pages," Torke starts out with a sparse melody in 16th-notes -- more rests than notes, initially -- and gradually fills in the gaps with more notes in an apparently haphazard fashion over successive repetitions of the two-bar cell. In both of these cases, it is easy to hear how the musical material evolves over time, and I decided I wanted to be a bit more subtle in my presentation of the structure.
The particular structure I chose to use, though clearly related to both Marsalis' and Torke's structures, was most directly inspired by a completely different sort of artist: comics artist, writer, and theorist Scout McCloud. Although he is an accomplished creator of mainstream comics, McCloud is perhaps best known for his works, both written and drawn, exploring the formal possibilities of online comics, employing what he refers to as the "infinite canvas". Though not particularly adventurous artistically, his variable-length strip Original Recipe Carl uses a clear additive structure for narrative purposes: beginning with a very basic two-panel comic, additional panels are inserted, one at a time, at various points in the middle of the strip, until we arrive at the 52-panel story of Carl's demise. While the story is hardly profound, I was intrigued at the way that the reader's perception of time can change through the many iterations of the strip. As more panels are added, impression of the passage of time, both globally -- from the first panel to the last -- and locally -- from one panel to the next -- are constantly revised, and both global and local markers have the potential to stretch or shrink -- or stay the same. In other words, the shape and proportions of the story, even the parts we are already familiar with from previous iterations, are always in flux. In contrast, the additive structures of Marsalis' and Torke's compositions both leave the shape of previous iterations intact; at any point in the evolution of either piece, you could take a highlighter to the score and reveal the seed, or any other preceding stage of the structure, in its original form.
At this point, you can probably guess the sort of musical structure I had in mind to reflect McCloud's narrative structure. I started with a simple musical idea, a 14-note melody that outlines the harmonies of the 12-bar blues form in the most skeletal way. I then inserted notes, a small handful at a time, in between existing notes, so that the skeletal line blossomed into a 35-note melody, now dancing around the blues changes. But unlike The Yellow Pages, where each iteration of the structure is limited to the same 2-bar timeframe, the boundaries of the melody grow with time. At the same time, the durations of individual notes in the melody are free to grow or shrink, to better fit the slowly swinging cadence of the blues.
It took me rather little time to work out how this melody would grow from 14 notes to 35, and even less time to work out chords supporting the melody in the piano. But successively elaborated repetitions of this melody would not be sufficient to carry the piece; I need something more. Look again to the blues tradition for inspiration, I decided that a call-and-response form would allow me to contrast different melodies, and different structures as well. I started sketching a melody for the most noble of instruments, the bass trombone. While the initial melody in the piano had a very narrow compass in both rhythm and register, the bass trombone would be largely free of such constraints. And while the piano's additive structure interpolated notes in every nook and cranny, the bass trombone would be much more straightforward: first two bars of its melody, then four, then eight, and so on. At this point, having written eight bars of melody for the bass trombone, I realized that these eight bars could easily be fit to the first eight bars of the blues form, so I added four more bars to fill out the form.
Now I had a full-blown melody for both the bass trombone and piano, and I was going to progressively build up to both melodies in alternation. But what should I do in between? The bass trombone melody was quite soulful, and could easily stand unaccompanied. But in comparison, the piano melody, even at its most complex stage, was rather weak; it couldn't stand up to the trombone on its own. So maybe I could use it as accompaniment, and write other material for the trombone on top of it. Now, just because I'm using the piano melody as accompaniment doesn't mean it's not important -- think of how a Gospel choir lays down a tune, with a soloist going to town on top. Without the tune, the soloist doesn't have anything to play off of. If I could get that kind of dynamic going in my piece, I'd really be happy.
So, I want to come up with something for the bass trombone to play on top of the piano, but at the same time I want it to be a part of the additive structures I've been building up. I notice that, as the bass trombone's unaccompanied melody goes on, it introduces new pitches not previously heard at key points. While this is not particularly striking on its own -- if you take a close look at this now long-winded description I've been writing, you'll find that each paragraph includes new words I haven't used previously -- it still seemed like something I could use. The bass trombone, when it plays on top of the piano, is at first limited to the very first note of its melody -- D -- and with each additional segment of the unaccompanied melody that it plays, the new pitches that are introduced can be used as well. Thus, the available pitches for the bass trombone to play on top of the piano keep growing: D, then A and C after two bars, B and F after four bars, G, G#, and F# after eight bars, and E and C# after the whole 12-bar melody has been stated in its entirety. And this also suggests an ending for the piece: a cadenza in which Bb and Eb, the two missing notes, finally become available to the trombone. I go nuts writing this cadenza, adding numerous flourishes to the basic 12-bar melody I had already sketched out. Then a cascade of all 12 chromatic pitches, tumbling across the full register of the trombone, and then borrowing the chords from the piano's initial statement-- wait. Chords? I can play some two-note chords on my horn, but the chords I want are fuller and more agile than I can manage with my multiphonic technique. Maybe this isn't a bass trombone piece after all. Maybe it's...*sigh*...a cello piece. Guess I'll have to write something else if I want to be the star.
But, yes. Cello. And piano. I've got the piano part all worked out. I've got the concept of the cello part all down, but the details need to be worked out. I'll start with the full twelve bars of the cello melody, unaccompanied, just to let everyone know where this is headed. This melody ends on D, the same note that it starts on, so I'll treat that as the first bar of the melody for structural purposes, and when the piano comes in with the skeletal form of its melody, the cello is limited to that D. I've allowed myself to use any octave, so it isn't too monotonous, but there's only so much I can do in this context with just Ds. The first stage of the structure is complete. Then the piano lays out again and I write the first 2 bars of the cello melody, with a little elaboration, but there's only so much elaboration I can put into two bars, especially as I want to stick to the pitches in the first two bars of the original melody, of which there are 3. Then the piano comes in, and with 3 notes to pick from, the cello can have a little more fun on top. That's stage two. Then four bars of cello solo, and we're up to five pitches. Piano comes in, cello plays with those five pitches, getting even more intricate.
At this point, I run into a wall. After three iterations of the various processes underlying the piece, and with only five available pitches in the cello part, I'm running up against the limits of what I can hear in my head and write down. Well, the cadenza I have planned for the end is even wilder than what I just wrote, but that's unaccompanied. Whatever I do with the cello part up to that part has to work with the piano part, which in turn is getting more active and more syncopated. It's taken me two or three weeks to get to this point, and it would take more than a year before I could go further. By now I've graduated from Duke, and am beginning my studies at Tufts with John McDonald. My unfinished sketch for this piece -- still unnamed -- was one of the things that got me into the M.A. program, so one of my first projects is to finish the darn thing. Somehow, I manage it. Maybe it's the impetus of having it performed that semester, maybe my mind's ear is sharper now, but some way or another, I pick up where I left off and see it through. I flesh out two more iterations of the process, each one wilder than before. As the pitch inventory of the cello grows, so do my options -- I can arpeggiate chords both "inside" and "outside" the piano harmonies, I can emphasize the blue notes, I can throw in trills and mordants and other classic figurations. And I go absolutely wild with my rhythms -- syncopations within syncopations, or hemiolas against a beat which is only implied in the piano. I want this music to swing and sing, and I go through a lot of effort to meticulously notate the desired rhythmic nuances, so the cello can swing even if the cellist can't. But I was lucky to have a great cellist, Emmanuel Feldman, premiering the piece, and he did a great job with it.
I wound up calling the piece Midnight Blue, for a few reasons. First, the title alludes to the blues form that I love so much, and which forms the backbone of the piece. It also describes the instrumental color of the piece; the combination of cello and the lower register of the piano (the piano never goes higher than the E above middle C, and spends most of its time and octave or two below that) is very dark and rich. And Midnight Blue, especially with the melodic outpourings of the cello, has an emotional urgency which, for some reason, I associate with the middle of the night. Like there's something on your mind keeping you awake at night, and it gets so heavy that you just have to get up, find the highest spot you can, whether it's a mountaintop or the roof of your apartment building, and just shout out loud. That's what Midnight Blue is to me.
Midnight Blue is not a perfect piece. It's got things I'm not entirely happy with. A few days before the premiere, I realized that I broke one of the rules of my additive structure -- that the cello would always be limited to the pitches from the last segment of its melody -- in two or three places, using notes that hadn't been properly introduced yet. If I had caught those "mistakes" while writing the piece, I would have written something else, but it was too late to change anything for the premiere, and, frankly, I couldn't come up with anything that sounded as good using the "right" notes. Even more troubling is the way I go to town with the cello part -- for all that I am suspicious of showy displays of virtuosity, Midnight Blue is the showiest piece I have ever written. Maybe that's part of why it has gone over so well with audiences. But even if it doesn't fully align with my own musical standards, it's still a fine piece, and I'm proud to have my name on it.