Songwriting Strategies

by Mark Simos

  Smart songwriters learn to work skillfully with dummy lines. Dummy material is essentially a prototype of the final lyric. Dummies can also serve as rhythmic and sonic approximations of the eventual desired line. You can also work with dummy material in any facet: dummy chords, melodies, even rhythms. In other facets, little but your level of dissatisfaction distinguishes dummy from keeper material. Dummy lyrics have a distinctive quality: ideally, they sound so ridiculous we’re not likely to mistake them for keeper lines. A dummy does no harm until you put it in the driver’s seat.

  Sense vs. Sound Approximations

  Dummy lines and paraphrase lines are complementary techniques. Dummy

  lines provide templates for desired sound attributes of the line you want; paraphrase lines approximate the sense of the line you’re targeting. An alter-native to a dummy line is a simple “placeholder” marking that shows the desired stress pattern or end-rhyme sound desired for the line. You can use the strategies in tandem, iterating and/or triangulating (our strategy friends from chapter 2)

  by generating several possible dummy and/or paraphrase lines for a given line

  or bit of music to match, to get closer to both sound and sense aspects.

  Here’s an example of a practical process distinction between an inadvertent or off-the-cuff filler line, in comparison to using dummy, placeholder, and paraphrase lines techniques.

  I never thought I’d see the light. Seed lyric line. It sounds good and suggests meaning as well.

  I never thought I’d say good night. Typical “filler line” generated as an off-the-cuff couplet (an “off-the-cuff-let”?). It includes literal repetition and “chases the rhyme” as well, with a predictable rhyme and cliché. But it presents as an attempt at a “real” line, and thus might easily get lodged in the working version of the song, despite its improvised origins.

  Magenta cow that zips up tight. A “self-declaring” dummy line, catching rhythm and rhyme sounds with no attempt to match the line by sense. This will be easier to work with later.

  Da dum da dum da dum da <-ite>. A placeholder template for the line-to-be. This stress template (or “stress-o-gram”) uses the stress-marking techniques discussed earlier to lay out line length and even the specific syllabic rhythm, with a marker for the desired rhyme sound, while not committing to even a dummy line as a prototype.

  Or get a second chance to hold you tight. A paraphrase line that follows the sense of the first line. It still “chases the rhyme,” and strays from the desired rhythm to be matched, but approximates the desired meaning.

  A second chance at second sight. A potential “keeper line” that matches the rhyme sound, the rhythm, and the content of the first line in a fresh way. It’s great if you stumble onto such a line, but you’re more likely to get there by triangulating between paraphrase and dummy lines.

  Paraphrase lines, dummy lines, and placeholders allow you to work with song structure in a more flexible, nonlinear way. Instead of stopping in your tracks until you get the perfect line, you put something there and move on. (Often, there’s no way to get to that perfect line until you come back later anyway.) In the end, we want our lyrics to mean what we mean to say, and sound right and sing well, as well. But we often stumble when we try to get sense and sound aspects right in one fell swoop. We can divide and conquer rather than fight both battles simultaneously.

  The Gibberish Scale: Seven Levels of Nonsense

  Dummy lines are gibberish. But some gibberish lines are more gibber-ish than others. Paradoxically, working with lyrics as sound produces disappointing results if we allow our rational, sense-based filtering mind to intervene at the wrong point in the process, and don’t let our sound-based searcher’s ear do its work. But it’s actually quite a difficult skill to generate truly nonsensical lyric lines, especially by casual free association. Lyrical song seeds are rarely in this form; they’re typically coherent lyric fragments lifted out of context in part because of sonic aspects—very different from a true line of gibberish.

  You can practice and develop lyric-by-sound skills, just like any other aspect of songwriting or musicianship. To do this, we need to make a more intentional technique out of what great songwriters are doing intuitively as they roll their eyes to the ceiling and spill out seemingly random word-sounds. As a starting point, I’ve distinguished a seven-level “scale” of gibberish, that moves progressively from lines mixing sense and sound aspects towards purer lyric sound-music. This list is by no means exhaustive or all that systematic. I’ve tried to list the different styles of nonsense in order of increasing “nonsensicality,” since I believe it takes increasing skill to be bolder with these sound aspects of your lyrics. There are various routes into these kinds of lyric experiments. I’ll illustrate with a rhythmic source, generating successive nonsensical matches to the syllabic rhythm “/ / ‿ / ‿ / ‿‿ / / ‿ /.”

  Nonsense. Real words put together in surprising, nonsensical ways. “You’ve got a harem litany skewed by a string.”

  Wonder-puddleful. Words put together in ways that jumble grammatical categories. (Apologies to the poet E.E. Cummings, who was famous for lines like these, as in his poem that begins “in Just- spring …”) How about: “Green as the always turning to angrying night.”

  The neologist twist. Made-up words that sound like words that mean something. These are more convincing if they appear in grammatically familiar contexts. “I’m talking candalicious if you’ll swoodle too.”

  Faux-foreign. Singing in a faux impression of a foreign language you don’t speak. (Go watch old Danny Kaye movies to see a master of “faux-foreign” speaking and lyric-spouting at work.) “A lora della piccola bon bella bon.”

  Word buds. Think of this as speaking in something that sounds like a language, but you don’t even know what language it is! The syllabic sounds may not even sound like plausible words, but are made up of consonants and vowels with the right textural feel for the line. Here’s some faux-Klingon: “Jak tultu cha-deh kitana forl paka tran.”

  Syllable buds. Word buds stretched until you’re not even sure where word boundaries would fall. Syllable buds can include vowel contours, consonant contours, and combinations of these. “Fa ne do ka ne dinala yo va ne sa.”

  The howl, the mumble, songwriter satori. Where you open your mouth and let sounds spill out in a free improvisational stream. (You try it!)

  As you move through the Seven Levels of Nonsense, results may at first sound very funny. We associate nonsense lyrics with humor, and that’s not a bad instinct. Humor is not an emotion in and of itself, but actually a way to stay loose, to avoid preemptive commitment to one specific emotion. The French philosopher Henrí Bergson, in his book Laughter, points out that for us to find something funny, we need at least a brief moment of emotional detachment. If we feel sorry for the clown who does the pratfall, we won’t allow ourselves to laugh at him. Allowing yourself to generate humorous lyrics is a temporary distancing—a way to keep the emotional content of your final lyric open and malleable. Remember, nonsense lyrics are dummies, so you need not wind up with humorous lyrics at the end.

  Make no mistake, though: this is potentially powerful work. Take your nonsense seriously. You’re accessing the unconscious mind’s greater wisdom, to get to lyrics you’d be far less likely to find with a more “rational” process.

  Exercise 4.3. Lyric-by-Sound Strategies

  Here are ways of working with Songwriter Gibberish and the Seven Levels of Nonsense.

  Practice moving seamlessly through these levels of nonsense. The examples above and on page 82 are rhythmically but not sonically related. However, the technique is probably most useful when each level serves as a sonic approximation for the next, honing in on vowel and consonant sounds as well as rhythm. You might start at the gibberish end and work your way toward the sense lines. Alternatively, you could start from a meaningful dummy line that’s still not the right line, deconstruct it into its sonic nonsense version, then pivot back to a new sens
e line.

  Generate lyric sounds in response to material in another facet: rhythmic, melodic, and/or chordal material. In working this way, it’s often most effective to begin at the “formless” end of the nonsense spectrum. For example, in setting from a rhythmic pattern to a lyric, begin with a vowel or consonant contour, move to syllable buds and then word buds, and finally to real words and dummy lines.

  Use an existing, sense-based lyric line as a template for associative generation of new lyric lines. Here again, practice extracting vowel or consonant contours, or a combination of the two. You can try generating response lines linked thematically (sound + sense), or connected only sonically.

  Create lyric syllable-bud patterns using structural templates. This is more a skill-building technique than a specific writing strategy, unless you plan to incorporate nonsense lyrics directly in the song. You can apply one pattern for leading consonant sounds, another for vowels.

  Gah Tay Low Goo

  Sah Nay Koh Noo

  Tie Lie Nee Kee

  Tah Lay No Koo

  Patterns may be reminiscent of musical or other forms including: scat

  singing, Gaelic mouth-music, auctioneer lingo, nursery rhymes, nonsense refrains of folk songs, even speaking in tongues or glossolalia, the cooing of babies and those smitten with them, and last but not least, the raving of lunatics (or the Nashville version, “former hit songwriters who lost their publishing deal”).

  Sonic Contours

  Just as we used a lyric line as a rhythmic model or template and pivoted back to new lines (see exercise 4.2), we can also derive sonic contours from existing lines, matching one or more aspects of these contours in new lines. End rhyme and internal rhyme are special cases of this kind of sonic contour matching. Note, though, that exact rhyme relies for its effect, not only on matching syllabic sounds—vowels plus consonant sound(s) at the end of the rhymed syllables—but also on contrast or nonmatching of the respective initial consonant(s) of rhymed syllables. (Without this latter condition, we get identity or false rhyme: e.g., “lease” and “police.”)

  The sonic contour of the line as a whole can be worked with as a template to be imitated, strictly or loosely, in subsequent lines, using sonic matching, free association, and thematic focus. In theory, you could even work from an abstract vowel or consonant contour, but these patterns are so sonically complex that we generally need to model from contours of a real line. Some hip-hop lyricists develop considerable skills in creating new lines that match the rhythmic and especially sonic contours of lines—even improvising such matching lines in real-time free-styling:

  Talk about guessin’ there’s a real mess ahead of us

  We keep messin’—the wheel toss will skid the bus

  Oedipus, Sisyphus, octopus omnibus…

  (This is why aging former classic civilization majors like me shouldn’t try to rap….) Note there are four internal rhymes in the first two lines (guessin’/messin’, real/wheel, mess/toss, head/skid). This sonic matching is impressive and can also produce surprising and fresh results. But it’s best treated as a special effect technique for the purposes of general songwriting skills.

  In old-school songwriting, with lyrics set throughout to vocalized melody, we often build up the song by setting from starting lyrics to new lyric material. In this scenario, the close matching of the sonic contour for the whole line as described on page 84 is actually too close for our purposes. (I’ve seen accomplished hip-hop lyricists struggle with just this problem; their ears are trained to build too many sonic connections to adjacent lines.) Instead, in conventional song form, when we generate a second verse from a first verse as “template,” we want to match lyric rhythm (more or less closely, depending on genre) and the rhyme scheme—while not repeating the specific lyric sounds used in that first verse. Tighter sonic matching would tie our hands too much, both sonically and in terms of theme and subject matter. Still, songwriters in all styles can benefit from the lyrical skills to isolate rhythmic and other sonic aspects of an existing lyric line, and to pivot from these as templates back to new lyric lines.

  Exercise 4.4. Match a Lyric Line’s Sonic Contour

  Starting with a model line, identify key vowel and consonant sounds of the line. Use these contours to generate new lyric lines, sonically matching the original in terms of the selected contour. As with the earlier lyric generation in exercise 4.2, resulting words need not relate to the original line. You can try this with vowel contours or consonant contours derived from the model line. There is, however, a curious asymmetry in how intuitively we can hear and imitate vowel vs. consonant contours. Vowel contours seem to be easier to hear; consonant contours require more intense mental effort, and more quickly depart rhythmically from the model line.

  You can use this exercise starting from any “model” line of a song that delights you. Using a familiar or even famous line (as in the example below) isn’t a drawback. You’ll quickly leave behind the familiar associations. For a bit of extra fun, try this with a partner or in a group setting. Pick a lyric line secretly and see if your partner/the group can guess it from the sonic contour! For example, consider the following vowel contour:

  AH – EE – I – IH – AH – AY – UH – OO(Recognize it?)

  (Luckily, as of this writing, songwriters cannot actually copyright sequences of vowel sounds, so Dolly won’t be comin’ for ya any time soon!)

  As you work back from the selected vowel contour to new lines, try different qualities of consonants: more plosive and harsh, to more sibilant and liquid. Dive in boldly, or sound your way along the Seven Levels of Nonsense scale toward sense lines:

  FIG. 4.7. Nonsense Types

  As you try iterative attempts, you can gradually loosen your commitment to the exact original contour sounds, especially as you get a promising idea for a real line. But be careful for spots where the thought takes over and pulls you too much from the flow of the vowel or consonant contour you are writing to.

  Sonic contours are also a way to set from a lyric line to a melody and/or rhythm, or from other facets to lyric. Vowel contours tend to suggest melody, consonant contours rhythm (though again, these correspondences are still somewhat speculative).

  Setting from Rhythm to Lyric

  We first made use of sonic constituents of lyrics—syllabic rhythm, vowel and consonant sounds, and contours—in setting from a lyric to rhythm. Here, our main challenge was cultivating rhythmic flexibility in hearing a given lyric line in different rhythmic ways. We’ve also looked at techniques for moving between lyrics by sense and by sound, and for matching lyric lines. We will rely on the same intuitive connections going the other way: setting from a rhythmic phrase to a lyric.

  We deal more rarely with pure rhythmic song seeds than lyric ideas. We also follow this path—from rhythm to lyric—when we align new lyric material with material already written, and want that alignment rhythmically tight. This could be, for example, a chorus of rhythmically matched lines, or—if writing in a strict phrasing and sectional matching style— a later verse matching the lyric rhythm of a “template” first verse.

  In these cases, you generate and match new lyrical material to a rhythm that was itself derived from a lyrical line. The new lyric must match rhythm but not specific lyric sounds such as rhyme. The match is more satisfying to the ear, and creates better sectional differentiation, with contrasting lyrical sounds. By abstracting rhythms from the source lyric to generate the new lyric, we can be less influenced by either lyric sounds or meaning aspects of the source lines.

  Rhythm to Lyric by Sound

  In principle, we can set from a rhythmic pattern to a lyric via alternate paths: by sense or by sound. Generating a lyric from a rhythmic phrase strictly by sense would involve, in effect, first framing the rhythmic phrase. While lyric rhythms alone can carry associative meanings, in practice, we rarely frame from rhythmic material only. The pull toward the “by-sound” route is much stronger. Rhythm provides so much structure directl
y suggestive of lyric sounds that these connections tend to dominate. The more strongly patterned the rhythmic aspect, the more it invites exploration by sound.

  Setting from rhythm to lyric by sound is also challenging. You’ll be tempted to fall back on familiar words and phrases, cliché lines, or free association and memory for scraps of coherent lyric, whether these fit the rhythm or not. These words will have their own innate rhythm, more often than not mismatching to the source rhythm in slight or extreme ways. Ironically, this process produces both many poorly set lyrics, and many thematically unfocused lyrics.

  To avoid these pitfalls in setting from rhythm to lyric by sound, give yourself creative permission to generate lyric-like sounds that are not yet actual lyrics or even words. However, the critical skill is to hold yourself to rigorously honoring the exact source rhythm. Even if you catch yourself in moments of fear-based searching or clinging for referent meaning, keep returning to that source rhythm and its sound associations.

  The following example shows some—admittedly speculative—stages in exploring this “proto-lyric” space. Our goal is to pull a lyric “out of the air” suggested by or at least matching well to a rhythmic phrase. Here’s where you’ll need those Songwriter Gibberish skills! Ready to sound temporarily ridiculous?

  We’ll start with a simple rhythmic phrase for a single lyric line, beginning on an upbeat syllable, shown in figure 4.8 as a “dry rhythm” (as described in the “Rhythm” chapter). Note that in working from a rhythmic phrase rather than a groove or accompaniment rhythm, we’re already working from a rhythmic pattern captured as seed material, or composed, to potentially “take a lyric.” (The example shows the sustained rhythmic durations of the “bagpipe” vs. “woodblock” style of rhythmic notation.)