August 15th, 2015

eyes black and white

Why I didn't create that song I wrote

I spent the other day yet again fixing bugs in the accompaniment for the song I authored on a famous poem by William Ernest Henley, Invictus. Clearly, no one else wrote that exact song before me; and so it's mine. Or is it? Frankly, I just wrote down a song already implicit in the poem I was putting to music. And I'll prove this claim by analyzing in detail the start of the song.

Now let's consider the rhythm of the first verse. The words "Out of the night that covers me" already have the rhythm and stress pattern: the syllables "out", "night", "co" and "me" are stressed; the other syllables are unstressed. Generally, the stressed monosyllables are longer than the unstressed syllables, but in the two-syllable word "cover", the stressed "co" is short and the unstressed "ver" is long, whereas the unstressed "that" should remain longer than the "of the" and "co". Let's use the simplest grid for rhythm, a 4/4 bar, and fit the stressed notes to the stressed beats, first and third. We can do it simply by using a crotchet for the longer syllables and a quaver for the shorter ones, except that we'd have to stretch the "ver" of cover for "me" to fall on the beat; actually, it sounds rude, whiny and pretentious to say "me" on the beat, so let's not stretch that "ver", and introduce a syncopation, which also gives more life to the poem. There, the rhythm was totally determined by factors outside my control.

Did I at least choose the melody and/or harmony? Not at all. The topic is gloomy, so it obviously calls for a minor key; let's arbitrarily pick G minor so I can sing it — it also has only few alterations (two flats) on the key signature, and happens to be quite playable on guitar (I didn't play any guitar at the time I wrote the song, but common influences from musical history had already shaped both my mind and the guitar); you can of course transpose to your key of choice, and it will remain the same song, so only really matters the fact that it's minor, and that's imposed by the mood of the poem. The first verse is an introductory verse, and it is building up a tension, so it obviously has to be in ascending tones. Since it doesn't end the sentence, the verse shall end up in suspension; the simplest way to achieve that is to end with a dominant chord. We also want to state the tonality, with a tonic chord somewhere, and since we end with the dominant, it's simplest to state the tonic as the first chord. So our stressed syllables will see an upward progression from tonic to dominant. There are four stressed notes; first is a tonic (G), last is a dominant (D); in an ascending progression, it is natural to fit a subdominant (C) between the two, for it has the next simplest chord. The simplest progression would be tonic, mediant, subdominant, dominant, where underlying the mediant would be a repeat the tonic chord; but since we're building tension, we have to do something more than this simplest progression; so we do the next simplest progression: tonic, subdominant, leading tone to the dominant, dominant; not only is it the next simplest, but the leading tone in itself contains tension that we seek. Since we're building up tension, let's also change the mood on the last chord, and make it major while the tonic and subdominant chords were minor; considering the meaning of the poem, it also makes sense to give a brighter color to the word "me" than to the other words and use a major chord. (Actually, in the poem we're adapting itself, the word "me" is deliberately ending the first verse, precisely for this effect, that we're translating to music.)

Now for the short notes: they could be anything that isn't dissonant, but better remain low to provide greater contrast to the upward progression of the stressed notes; and the simplest solution is to just repeat the tonic, G. However, since we have two short notes right after the initial tonic, we can make the first of them the mediant while still on the tonic chord, and it fits perfectly both with the upward progression of sounds and with the fact that the stressed notes set the chord. As for which chord underlies the leading-note-to-the-dominant, we find that a E flat minor seventh, which is just one note half a tone away from the previous chord of C minor seventh, is the simplest solution. Tada! The constraints natural in the text and in western music completely wrote the first verse for us, including every single note, its pitch and duration, and for each note the underlying chord. Of course, I didn't have to explicitly think about these constraints when I wrote the song — the melody and its implicit harmony just "fell into place", because I've internalized these constraints, and this is indeed the simplest solution, by far.

I could go on with the rest of this song: the next three verses have their rhythm decided likewise. A repetition of the melodic pattern of verses one-two in verses three-four is the natural expectation, with one-two ending on a dominant and three-four on a tonic. Verse two (and four) need their melodic line to be descending to use the tension previously built. Their first note needs to be above the last note of the first verse, to use its trampoline; it also needs to start with a tonic chord. However, for variety as well as to allow for the three note variant in verse two, it's better to start on a mediant from which you can go to the tonic (via the supertonic as a transient note), rather than starting from a tonic and have nowhere down to go within the chord. Filling the melody with stressed notes that are different from the (notional canonical chord) bass line also makes for a nice response to the previous verse where the stressed notes in melody were mostly the same as the bass. The words "Pit" also suggests going lower than what is simplest (F♯), and even lower than what is natural (B♭ instead of B), in another interesting interaction between word meaning and pitch (I don't even know how to name the resulting chord, because the natural B is also present; or maybe it's actually a C♯ in an G chord with major seventh and augmented eleventh? does it matter?). Similarly, "pole to pole" suggests going up and down around the "equator" of the tonic, while following the general downward pitch progression of the sentence. I could go on and on, but I'll leave it as an exercise to the reader to figure out all the reasons why this song is indeed the simplest solution to the problem of putting this poem to music. Of course, the idea of music scores as puzzles isn't new: Johann Sebastian Bach and his contemporaries famously wrote the scores of some canons and fugues as puzzles to be completed from scant indications: a theme is stated at the beginning, with indications on what symmetries and transformations to use to fill the blanks with modified copies of the theme, and seasoned baroque musicians would be able to reconstitute the entire piece.

Now, consider the set of rules and constraints, some hard, some soft, some formal, some informal, some purely musical, some related to the syntax of English, some related to the meaning of the words, according to which these solution is simplest; together they constitute a sense of aesthetics, that I am indeed bringing on the table. Of course, I probably invented not one single of these rules: I inferred them from a large body of extant western music that I've been exposed to, and the small amount of codified theory that I actually learned. In this sense, the fact that I developed my sense of aesthetics from cultural precedent only illustrates how all creative work is derivative. Still, I've been exposed to a different set of music pieces and music theories than any other person, this set of rules is less unambiguous than those followed by Bach, and wherever there are soft or ambiguous rules or constraints, I probably give them slightly different relative weights and different interpretations than other musicians would; and there are probably many rules I am altogether unaware whereas each other musician might also have additional or missing rules compared to mine. Thus, this sense of aesthetics still constitute a personal, though marginal, input of mine, to the "creation" of the song — as does the will to work at composing a song, and to work on this poem rather than another one: this song would have remained a mere potential if I hadn't cared to make it a reality. My input, small as it might be, was thus nevertheless existential for this song.

My ultimate point is that, inasmuch as I "invented" the song, this "invention" does not at all fit the modern mythology of artistic creation ex nihilo of something that didn't exist, and does completely fit the etymological meaning of "invention", to find: finding something at your feet, coming by something that already exists and that you notice, being first to discover an eternal truth, etc. There is of course merit in going where no one has dared or cared to venture before, or in noticing what no one noticed or cared to notice before though it was in front of their lying eyes; but it is not at all the merit of creating what didn't exist before, it is the merit (and demerit) of being me and having traveled, inside and out, all the way to where I've been — what I've created is myself. And the implications in terms of law, politics, economics and ethics are very different. But that will be the topic for a different essay.