Alice Parker: Finding the Music in the Text

Which comes first—the melody or the lyrics? Composer Alice Parker explains her creative process.

I am constantly searching for texts to set. I love lyric poetry and the kind of prose that approaches poetry (the King James Bible, for instance). I can tell immediately as I read whether the melody hidden in those words will reveal itself to me. I feel the sounds in my throat as I read, and they either "sing" or do not.

The lyric brevity of Emily Dickinson is music to my ears, but the expansive parabolas of sound in Walt Whitman (which I also love) are not. There's no accounting for it. We are each different and respond to different cues. (Thank goodness!)

When I start to set a poem, I begin by copying it out in longhand, noting all the punctuation, line indentations, verse separations, and spellings. Then I memorize it, repeating aloud until I know it by heart (what a wonderful phrase).

Often I will then test myself by writing it out again from memory, checking that each jot and tittle is correct. I want to put myself in the mind and voice of the poet, feeling as though I myself had created these lines.

As I speak, the phrase curves grow more and more familiar, and the separating pauses take on their own pattern. I'm digging down through the surface of the poem to its inmost structure—rhyme, assonance, accent and cross-accent, pause and rush forward. The text is the rhythm; the spoken sound becomes, in our inadequate notation, the sung rhythm.

Here is a poem by Emily Dickinson. It is full of her own punctuations—dashes. Do they mean expressive pauses? Can you glimpse her vision? Can you read the poem aloud in a way that transmits it to a listener?

Exultation is the going
Of an inland soul to sea,
Past the houses—past the headlands—
Into deep Eternity—

Bred as we, among the mountains,
Can the sailor understand
The divine intoxication
Of the first league out from land?

—Emily Dickinson (#76)


When I was working on my opera The Ponder Heart, setting a delightfully comic story by Eudora Welty, I had to find a musical "voice" for the narrator. Edna Earle, of course, has wonderful words supplied by Ms. Welty—whole paragraphs of them as she tells the tale. Could I set them in a way that kept all their southern speech flavor without distortion?

I found I had to copy them onto manuscript paper under a blank staff and then literally notate the speech rhythms as I read aloud. Not surprisingly, the basic meter was 7/8 (we speak in little groups of twos and threes, not a constant 3/4 or 4/4) with fairly frequent variants. Even the rests between phrases, sentences, and paragraphs were measured before I ever added any pitches. And even these came from the text! The easy rise and fall of speech became the widened scale of song with extreme highs and lows reserved for very dramatic moments.

Take that simple phrase, "Hello, how are you?" Imagine it set by Mozart (yes, in English). By Britten. By Stravinsky. It's always different, isn't it? No matter who is speaking or singing, in whatever dialect, the phrase changes. If we had the needles and gauges to record all the variables, could we ever say it exactly the same way twice?

Accentuation is enormously important. Listen to people around you speaking—on a bus or at the other end of the room. The rise-and-fall of the pitch may be limited in scope, but the pressure of the accent is amazingly variable. In trying to present a visual image, I think of someone bouncing on a trampoline with gauge lines painted on a wall behind.

It takes great control to get a steady rhythm going, and that's not the aim in speech. Here, the basic rule is: No two adjacent sounds are ever the same. The jumper on a trampoline depresses the surface a different amount with each pass, taking longer with the deeper levels to go down and come up, moving more quickly with the shallower falls.

That amazing variability of speech has to be built into whatever language is sung. In our page-driven age, we tend to shove the word into the note value, as if the quarter note were an ice cube with the syllable frozen inside. I'd much rather think of the notation, both pitch and rhythm, as being at the service of the word or phrase as naturally spoken, with all the subtlety preserved. My inner vision is of a word, in its lump variability, being bathed in pitch and duration as though these were precious oils being poured over it.

If we really believe that speech underlies the song, we teachers, singers, and conductors would approach songs completely differently. We would begin with the text, lovingly read aloud and learned well in the mood and rhythms of the song. We would find the music in the text first, as though we ourselves were going to set it, then notice the specifics of what the composer had done.

We would observe and act on any helpful hints in tempo indication, expression markings, fermatas—all indications for subtlety, for deviation from the prison of the page. We would know that only when the solo singer or chorus can read the text beautifully in the rhythms of the song is it time to add the pitches. Last, not first.

The pitches come last because, willy-nilly, pitches have duration. If we learn them before we have the rhythms of the song clearly in mind, then we learn them incorrectly and have to change. This goes a long way toward explaining much of the dull singing around us. I'm totally opposed to learning the notes and rhythms before adding the words—they can never, then, regain their primacy. Words first!

This article is adapted from The Voice, Winter 2007-08.
Excerpted from The Anatomy of Melody (G-6765) by Alice Parker. (c) 2006 by Gia Publications, Inc. Used by permission. To order:, or call GIA at 1-800-442-1358.