Archive | Works in Progress RSS for this section

Writing for Voice, Pt. 2: Putting Together the Music

I wrote in the previous post about how to consider setting a poem to music. In this part I’m going to talk about how I actually went about putting these words to music.

So my first concern with this piece was that it should be easily learnable and singable by an amateur choir. That meant not much in the way of chromaticism, nothing particularly odd rhythmically, no large melodic leaps and nothing outside a limited range.

So I was going to have to go back to writing tunes that are singable primarily, which requires some basic human intervention – not something I usually do. I tend to stay away from intuition when it comes to melodic writing, so this on the whole a bold new territory for me. But I’m nothing if not self-challenging!

Although now I think about it, my approach to melody is different when it comes to handling music with words; since I’m of the belief that in the case of song etc., music must largely work to support the dramatic and poetic intentions of the text. Hence, through the compositional process I try to find the melody that “fits” the words  the best. It’s an approach I used in a previous song I wrote – a setting of two Philip Larkin poems – in which I first laid out the harmonic structure using a series of randomly-generated harp pedallings, and then created a melodic line using these odd, rootless modes.

In fact, I’m rather ashamed to admit it, but I wrote the main melody for the song by just writing down the first tune that came to my head while reading the poem on the tube. And the following is what came out for the first line.

Fairly simple, right? I guess it’s sort of in G-melodic. Following on from that, I had three different ideas of how to continue, so I just whacked them all together, splitting into the three voice parts.

The words don’t quite fit each line in this reduction due to the dotted crotchets happening in different places. But it’s all syllabic, and makes sense. Trust. Also, the harmonic structure is certainly odd, as a result of three different lines of thinking about melody. I could analyse it as… A7, Bb maj7, Fsus/A7, Csus2/G, Ebdim, Ebm, F#. But I’m not sure why anyone would do that.

Anyway, the point is, this slightly odd approach to harmony comes from the meandering melodic lines (pointedly free from tedious ‘rules’ of counterpoint or harmony) weaving together like STREAMS forming a RIVER. See how the text influences the process? No? Do I need to make it more tediously obvious?

Anyway, this approach continues on for a while. For the chorus, I wrote three roughly ascending melodic lines, and had them run at the same time at 3 speeds – i.e., with each syllable corresponding to a crotchet, quaver or semiquaver. The result is a simple harmonic pattern that repeats itself in often interesting ways. I guess I could have gone longer with a fourth part singing minims, but that would just get boring.

It sort of continues around like that for a while. Anyone would think I’ve gone soft in my old age with all this regular rhythm and boring harmonies. I guess they might be right. But while each line is quite easy to sing on its own, when combined they produce a nice crunchy harmony, and with the middle part lifting on the offbeat it gives the whole thing a sense of movement which I quite like.

Anyway, that’s most of how the piece works! The second verse takes the same music and the first. It’s that simple. I deliberated a long time about how to fit verse and chorus together, because they’re sort of running with different ideas (and since this is to be sung a cappella it’s not like I can just whack in a nice middle 8 or something) but I reckon they work alright together in the end.

The final part (the line “we sing / we swell”) was where I could really push the boat out on non-traditional choral techniques. I was stuck on how to end the piece in a nice way that reflected the text. I wanted to have some sort of canonin writing going on, but it’s hard to do that sort of thing without strict conducting/learning/etc. Remember, this is a choir that normally sings Take That arrangement.

So in the end, I wrote three short loops of 6, 11 and 14 beats long. They start at the same time, but obviously go out of sync fairly swiftly. The idea is then for each singer to stop singing when they lose count, get bored, or whatever, and hold the final note (“swell”) until everyone’s there, at which point the conductor can bring them off together.

Next time! I put it all together, and hopefully get a recording of it!


Writing for Voice, Pt. 1: Approaching the Text

Another project I’ve got on the backburner over the last few weeks or so has been a piece for a local community choir I was asked to write.

Now, this presents its own challenges. The choir is made up of 40 or so people, depending on who turns up to each rehearsal. As with most amateur choirs EVERYWHERE EVER, it’s largely full of female voices, and what few men turn up find it hard to sing in a classical tenor. There is an emphasis on learning music by ear rather than reading a score, so as to aid participation, and it’s not really the environment for anything particularly… well, you know. Anything a bit “new music”.

So what does this mean for me? A cerebral-type musician with a reverence for the score-as-artefact and an inability to write proper tunes? It means it’s time to go back to square one.

Of course, the first thing to decide is what the text should be. And while I’m not averse to writing words for things in certain capacities, I’ve never considered intersecting my own music and my own words. Strikes me as bit too egotistical to be honest. And always me of the following sketch.

So anyway, for this I consulted the the nearest occasional wordsmith I could find. My dad. Given the themes of the choir (and also, I think, our proximity to the river Wandle), he came up with the following:


Our roads run straight and parallel

Our homes and gardens we surround

Beneath the ground

An ancient source

Awakened by the rhythm of

The footsteps that we take along

The paths that brought us to this place.

When rivers run together they make

Unfamiliar current and leave

Ripples on the surface hiding

Overpowering movement and when

Music makes us bouyant we can

Open up the floodgates float our

Hopes and our ambitions on  a

Rising tide of rhythm as we


We swell

As we

Sing we swell

SO, with that in mind. How to put it to music? I’ve been thinking a lot about Shape Notes – and specifically, the Sacred Harp tradition with which they’re most associated. Not that this is a sacred piece, of course, but the choir practice is held in a church. Anyway. My understanding of that tradition of music (with which I have little to no connection) it that it’s split into usual SATB sections, but pitch is relative and everything’s sung a capella. Listening to youtube videos of performance (which, as a side note, are generally incredible), there’s little sense of the strong topline melody against the lower voices singing harmonies as is traditionally the case in sacred (and secular) choral music. Instead, the tune seems to pass through the parts, and the blend of (largely untrained) voices singing close harmony and similar melodies creates a more unified sound. It’s nice! I like it!

But how can I use this as inspiration? And how can I address the issues I set out earlier?

The Uneven Gender Balance

This is probably the greatest obstacle to creating a cohesive choral sound. The bottom end is weak because they just don’t have the numbers, and it’s a sad truth that a lot of men are less comfortable with singing than women. Frankly, that’s why I joined the choir in the first place – to provide some male support (fnarr).

So why bother splitting up based on voice type? If I’m going for a more unified sound, it’d be best to divide on something else. How about three parts made up of an equal mix of voice types, singing at a register that is comfortable for all voices (with the allowance of singing an octave above/below)? That sounds nice. We can divide by relative skill, if need be. But I’d rather not have a hierarchical approach to parts. The theme of the text is that of unification, not division.

Coming up next! The business of settling down to actually write the music…

Thinking About Structure or: How I Learned to Stop Worrying and Love Abstraction

So, continuing on from my previous post about the as-yet-unnamed chamber piece for clarinet ensemble. This new modal way of thinking allows for crunchy discordant harmonies, but without precluding a root note; often a problem I find in writing fully chromatic/atonal music. It’s all too easy to have reams of perceptibly structureless music that go on for ever and ever, tiring an audience and shoring up the (deluded) criticism of contemporary music that it’s just a load of squeaky-bonk rubbish. I recently saw Boulez’s Dérive 2 at the Proms, sandwiched between Beethoen’s 1st and 2nd symphonies. A solid 45 minutes rushing wave of wonderfully complex music, but DAMNif it didn’t rehash a lot of the same old ideas. Seriously, there was about 10 minutes of material there that just got stretched out way too long.

Of course, there’s nothing wrong with writing reams of squeaky-bonk rubbish at all! And I’m by no means a neo-classicist/stuckist/whatevs. But this isn’t the place to have that debate. Anyway, on a pragmatic level, I’m writing with clarinet-playing teenagers in mind. And in my experience, clarinet-playing teenagers are often fiercely resistant to alternative approaches to tonality.

So what should define the root notes of each section? I started thinking about the instruments again. As a clarinettist of – oh jeez, has it been that long? – 14 years, I’d like to think I have a deep knowledge of the instrument that goes beyond my lack of technical mastery. So anyway. I picked the notes of the F major scale as played on the Bb clarinet (So Eb, F, G, Ab, Bb, C, D at concert pitch) because it’s probably the easiest scale to play. So the structure of the piece of music is going to be underpinned by conventional tonality! Hooray! So, long story short, I generated a 7-digit chain of 1364572, which trasnslates to E♭, G, C, A♭, B♭, D, F. Generating another 7-digit chain of 6174235 created an order of modes to use for each section. And so! We have created a melodic plan for the piece. It’s vague, nebulous and doesn’t really have any meaning but that’s sort of the point at this stage.

It makes sense, honest!

Sections progress left-to-right: utilised notes are shaded in grey, roots are in bold.

Can you tell I like diagrams? Anyway, the point of this is both that it proves that each section has its own, uh, ‘key’, for want of a better word. The seven notes used in each section are distinct to the other six. It also may prove useful in future in other ways too, as it reveals certain horizontal patterns, and these could easily be adapted to rhythmic figures. Notice how I still haven’t put any notes on paper! Seriously, this is how I write music. I know, right?

Now to consider the overall structure. It’s at this that I should bring in what’s really going on in my head when I write music. For each piece, I tend to have some form of image that sublimates both thematic material and compositional techniques. For a recent sextet, it was a comet flying round narrative space in a certain parabola. For a triptych of trios (the second of which is located here), it was a set of interlocking triangles rotating in three dimensions. For this piece? I see a non-euclidean seven-sided regular solid floating in space, reflecting beams of light that become the music that I’m going to write. I know, so shitting abstract it HURTS.

So anyway. Now to think about how to actually write the music for each section. I initially wanted seven different miniatures that segued together with little sense of repetition (to fit in with the harmonic approach) but on further reflection, I don’t reckon that works too well: since the harmonic system is based on a recognisably “tonal” palette (i.e. there’s a root not), a free structure that floats away into space would sound meandering and crap. Instead, the structure as it stands in my head is based on a partial movement of the aforementioned regular heptoid: swinging in a vaguely parabolic manner through are four distinct styles of music.

Sections 1 to 7, described with relative length, with thematically linked section in similar shading.

  1. Slow dream-like chords with a free central melody on the lead instrument
  2. Increased polyphony; flurries of ataxic pitch clusters
  3. Fugal development between instruments
  4. High energy, violent.

Speaking of a ‘lead insrument’ (i.e. the part that’s going to get the prominent melodic line within each section) that’s another thing that can be randomly decided. A quick roll of the dice leads to the 7-digit chain 5473261, which means the seven section look thus:

  1. Alto Clarinet, mode 6, root E♭
  2. 3rd B ♭ Clarinet, mode 1, root G
  3. Contrabass Clarinet, mode 7, root C
  4. 2nd B♭ Clarinet, mode 4, root A♭
  5. 1st B ♭Clarinet, mode 2, root B♭
  6. Bass Clarinet, mode 3, root D
  7. E♭ Clarinet, mode 5, root E

This already creates images of the internal structure of each section, and the manner in which each lead instrument’s going to feature. How to distinguish between B♭ Clarinets? How will the mirrored pair section (1&7, 2&6, 3&5) counteract each other?

With this in mind, it’s almost time to write some actual notes. That’s for next time, though!

How to Begin Writing Music. Or Not.

So I was recently asked to write a piece for a clarinet ensemble. The brief was wide open: something for teenage players of a fairly high standard, playing everything from the Eb to the Contrabass.

Of course, this gave me absolutely no point to start writing from. Of the many neuroses I’ve been developing over the past 3 years, my doubt in the ability to put one note in front of another has been the most prominent. What authority do I have, just an ordinary human, to ordain that an arrangement of sounds should be just so? I mean, I can’t hold Wagnerian levels of control over any music I create because a) I live in a 21st-Century interconnected world full of a billion musics and b) I just don’t have kind of ego. So I tend to use chance methods  to create musical rules that write pieces themselves, as well as compositional techniques that allow my subconscious to come to the fore.

So I began with the ensemble itself. Seven distinct parts (Eb, 3 Bbs, Alto, Bass, and Contra). Immediately I was drawn to the number – days of the week, deadly sins, hills of rome, dwarves etc. You name it, the number seven has been culturally attached to it. But most prominently for my purposes was perhaps the most obvious: the amount of notes in a diatonic scale. Of course, this is no way to begin writing music. The seven notes in the major scale have been done to death over the last 350 years. And don’t get me started on minors.


How then to use a seven-note harmonic system in a new and exciting manner? I began to think of modes. Oh, poignant Dorian. Roguish Lydian. Really-hard-to-use-well Locrian. They’re all well and good, but being Caesar-cypher rotations of the original Major tonality I couldn’t really sustain much interest to use them for my own purposes.

It occurs to me at this point that I should break off and briefly clarify both for the sake of anyone at the back of the class and to lay the groundwork for the way of thinking about tonality that I will soon be considering. The major scale is an arrangement of 7 notes picked from the 12 defined within even tempered Western Music, and can be defined as a series of intervals either 1 or 2 semitones in size.

The usual modes, then, can simply be thought of as rotations of this pattern when starting on the same note:

  • I – Ionian – TTSTTTS
  • II – Dorian – TSTTTST
  • III – Phrygian – STTTSTT
  • IV – Lydian – TTTSTTS
  • V – Mixolydian – TTSTTST
  • VI – Aeolian – TSTTSTT
  • VII – Locrian – STTSTTT

A simple analysis of these when starting from the same note shows that the note a perfect 5th above the start is the most common across the board (except in the Locrian) – evidence of the dominance of V – I in Western music? Who knows.

So I began to think about other arrangement of 7-note scales. Using only tones and semitones, there are only 2 other sets of modes that repeat at the octave: rotations of TSTTTTS (the ascending melodic scale) and TTTTTSS (a whole-tone scale with a bit on the end). Discounting the latter because its rotations immediately sound boo-o-oooring, I was struck by the distinctly alien set of modes creating by starting on the melodic ascension:

  • I  – TSTTTTS
  • V –  TTSTSTT

(note: the prevalence of the 5th is not so strong in this set, lacking as it is in modes III, VI and VII)

With these seven modal areas defined, I decided the piece of music should have seven distinct sections, each one based on each modal area arranged in a random fashion (decided by rolling a dice). But where we go from there is a story for another post…