Intermezzo: Impressionistic Lyric Writing
Before I proceed to a discussion of “music first” songs—not to be confused with the atavistic political slogan “America First”—I want to respond to a question from an astute reader. In response to Part I of this series, he asked what I thought of lyrics that succeed on “vibe,” as opposed to those built around character or narrative, writing:
I do think there are a lot of songs where I don't really know what's being said, but I find the "associative machinery" that the sound and timing of the lyrics evoke to be quite pleasurable.
I’m so glad he brought this up. There are, of course, many wonderful songs whose lyrics are primarily impressionistic, or in which the sounds of the words, as much as their meaning, carry emotional impact. The more we allow intuition to guide us— without intervention from our analytical mind—the more likely it is that we’ll end up with lyrics that fall into this category. This isn’t to say that you can’t analyze your way to an arresting image, but rather that the most powerful ones tend to arrive via subconsciousness.
This reader’s email also put me in mind of the broader relationship in art between catharsis and the sub/unconscious realm. Namely, that the former springs most organically from the latter. When an artist tries to consciously engineer moments of emotional catharsis, too often the result is cloying or predictable because the offending musical grammar—for example, a minor subdominant chord to signify melancholy or heartbreak—has been routinely abused. Conversely, one might argue that innovation in art can be defined as the subconscious discovery of new modes of emotional catharsis.
To be clear, the conscious mind is also an indispensable tool for art-making. The stage director John Tiffany, with whom I worked on the BAM production of The Ambassador, taught me that one can be as complex and challenging as one wants, provided that there are access points for the audience. And the construction of such access points is very often a function of the conscious mind.
Okay… picking up where we left off.
III. Music Comes First
When I write strictly instrumental music—a string quartet or clarinet concerto—I sketch with pencil from the get-go. With songwriting, it’s different. In many instances, it’s a whole lot of fumbling in the dark. I sit at the piano and improvise, often early in the morning. A chord progression materializes, then a tune. In time, dummy lyrics1 are replaced by real ones. (At this point, I might double-back to some of the strategies I described in Part I.)
Speaking of chord progressions—and a warning that this gets into the technical weeds, so feel free to skip to the next section if discussions of music theory are your go-to substitute for melatonin—a frequent question from students is this: where does your harmonic language come from?
In brief, my sense of harmony is rooted in counterpoint, which is the movement, in a piece of music, of individual voices against one another. Counterpoint unleashes invention by inviting a composer to think linearly (melodically) rather than vertically (chordally). Certainly, one can approach harmony by moving from chord to chord and arriving at something satisfying. But if you’re thinking about each shift from one chord to the next as a discrete event, your palate may be limited.
It’s much easier to be harmonically inventive, as I see it, if you follow the bliss of voices moving against each other. When chords are an outgrowth of two, three, or four voices swimming in the same stream, but not necessarily in the same direction or at the same speed, one discovers pungent, passing dissonances, as well as voicings, key relations, and harmonic motion that might not otherwise have emerged. “We Are the Saints,” which I mentioned in Part I, has a chord progression almost exclusively derived from counterpoint.
Here’s the first verse.
Now, here are the first eight bars with the arpeggios written as vertical sonorities.
You’ll notice that in the movement from one chord to the next, only one voice changes at a time, and always by a half-step or whole-step. The sequence is almost entirely determined by the horizontal movement of one voice against another. Could I have come up with this progression thinking vertically? Maybe, in those first four bars. But in the fifth bar, that C♭ on top lives in a kind of purgatory between the preceding sonority (C♭ major) and the one that follows (F⁷). It’s the gradual movement of the top line from B♭ to C, via C♭, that opens up the possibility of that fifth chord, an F⁷♯¹¹. Bottom line: if you feel that your chords have gotten stale, I highly suggest cracking open Johann Joseph Fux’s Gradus ad Parnassum, which is still the gold standard for the study of species counterpoint.
Alright, slackers in the back—wake up!
IV. Latticework
It’s rare for me to write music for a song in its entirety without words creeping in. More often, there is a kind of latticework: once I’ve got a verse and chorus’ worth of music, I’ll start chipping away at lyrics in an effort to understand what the song is “about.” I use scare quotes because songs ought not be distillable down to a “subject.” As a playwright friend of mine once said, “if I could tell you what my play was about, I wouldn’t have needed to write it.” The same goes for songs.
A while ago, I wrote about the poet Matthew Zapruder’s superb and searingly beautiful memoir, Story of A Poem, in which he asks, “was I supposed to be saying something I already knew in the most beautiful possible way, or trying to follow music, to find out what I did not know I believed?” In the case of songs where music comes first, Zapruder’s directive to “follow music” becomes quite literal. There simply is no right or reliable way to unearth lyrics to a song whose music already exists, but there are a few strategies that come in handy.
Lyrics should “taste” good in your mouth. They should be a pleasure to sing. It also helps if the sound of the words, and particularly the vowels, carry emotional weight in tandem with their meaning. (This hearkens back to the reader’s question about impressionistic lyric writing.)
As lyrics fall into place, that latticework may shift back toward the music. If I have a “tic” as a songwriter, it’s re-harmonizing chords underneath a melody as a song progresses. See, for example, “Veda,” “Chemex,” “To Be American,” “November,” or “Empire Liquor Mart.”2 Other times, a lyric will suggest that the form—the number of bars in a phrase, the number of beats in a measure—should evolve. (See, for example, the discussion of “We Are the Saints” in Part I.)
Needless to say, this kind of latticework is not an exact science. It requires a great deal of trial and error, and a willingness to surrender to a near-constant state of not-knowing, which we’ll return to at the end of this essay. But before we get there, I want to shift our attention to one final mode of songwriting, in which the song’s raison d’être is to help us master a particular musical challenge.
V. Song As Étude
An étude, broadly defined, is a piece of music that confronts a technical challenge not through rote, mechanical repetition, but with beauty, expressivity, and a sense of musical architecture. (For the uninitiated, the Chopin Études for piano are a great place to start.) While we often think of études as compositions for a solo instrument, there’s no reason that they can’t be conceived for a singer-songwriter. And that’s precisely what I’ve done, when, faced with a problem of coordination or facility, I’ve employed song form as an exercise toward mastery.
In “Bradbury (304 Broadway),” for example, I set myself the goal of playing a 7 against 4 polyrhythm on the piano (seven beats in the right hand against four beats in the left hand) while singing a languid melody. First, I wrote what was more or less a complete piano piece, and then went back and added a melody and lyrics (drawn from images I’d gathered from the film Blade Runner, much of which was shot in the Bradbury Building in downtown Los Angeles, hence the title.)
At other times, the étude addresses a technical challenge related not to performance, but to the writing process itself. With “The Basement Engineer,” I gave myself the prompt of writing a song whose bass-line was derived from a twelve-tone row. In other words, each of the twelve tones of an octave would appear in a strict sequence before any would be repeated.
Here, I used a process called “cataloging,” which I adopted from my teacher Ken Frazelle. The basic idea is to divide a piece of manuscript paper into four or more sections, and then to fill in each of them with a different “solution” for whatever musical problem is at hand. The advantage of working this way is that it offers relief from the pressure of writing from beginning to end, freeing us, if only temporarily, from the anxiety of having to “get it right.”3
Think of it like a Pinterest mood board for songwriting: after generating a surfeit of material through this process, we are now free to pick, choose, and arrange what we like best from all that we’ve created.
If rampant self-criticism is the biggest obstacle to being creative, then cataloging musical ideas is a way of temporarily silencing the inner critic. To suspend the practice of writing in sequence is to give oneself permission to be generative without knowing when or how you’ll use a particular bit of music or turn of phrase. Put another way, it is an invitation to get lost in the process. And there is pleasure, I swear, in surrendering to the not-knowing.
Coda
Access the not-knowing in other ways, too. Give yourself assignments that take you outside your creative comfort zone: write a song about a favorite tree, or begin with a line from a poem you've pulled at random from the shelf, or depict a snow-capped monastery in Vermont, or write a sonnet to a beautiful stranger on the subway and then find the tune. Create musical puzzles. Write for new instruments. And when the critic visits, remember, as the poet Robert Hass said, that “you can’t revise nothing.”
What I’ve written here is in no way an exhaustive list of my working methods, nor does it begin to account for the ways in which you might dream a song into existence. These practices are, and must remain, malleable. The most important directive for all of us is to remain open and alive to what the song wants to be.
P.S. If you are a longtime reader of this newsletter, and have not already done so, I’d ask that you consider upgrading to a paid subscription. It’s my intention to keep this newsletter free, but your voluntary support allows me to justify spending more time crafting these essays. And thank you, as always, for reading.
Dummy lyrics are just that: something to stand in for the real thing. Many lyricists talk about trusting the emotional payload of particular vowel sounds when they work to replace dummy lyrics. If the dummy line is “oh, it’s a yellow balloon,” a songwriter might end up with, say, “oh, I fell off the moon” in an effort to preserve the long “ooh” sound.
In the case of “Basement Engineer,” this meant writing out the tone row in its initial form, in inversion, retrograde, and retrograde inversion, and then inventing various harmonies and contrapuntal gestures, while also varying the speed of the bass-line.