A little while back I wrote this piece for the BBC Proms team. The original version of the post is linked at the end.

I seem to have a thing for voices. Both the pieces I have written for the BBC proms scheme have been settings of poetry (Two Boys, the competition entry, was a setting of a Whitman poem, and I’m Making a Statement, the commission, was a short suite of the words of people living with dementia). It doesn’t stop with the Proms – the majority of the pieces I have written have in some way or another have spoken or sung elements, be it by choir, soloist, chamber ensemble or even entire orchestra. Setting text and writing for voices can be a surprisingly fiddly thing to do (and it is more difficult to do it effectively), but so far (in my education and wider listening) I’ve learnt a few things from various places which I thought may be useful to put down.

1.    The ironic undercut. While some will fervently argue that the most important (nay, I tell thee, the only!) reason to set text to music is to convey (forsooth, I continue, emphasise!) its meaning, I think that is a little too simple a viewpoint. While conveying and emphasising the meaning of a text is certainly a worthy reason, it can be occasionally limiting. Sometimes, it is more fun to reinvent a text. This can be as simple as reordering it (I will get to that later) but it can be achieved without changing or missing out one word of the text. Pairing text with music which seems to jar has the brilliant effect of mutating the meaning, or undercutting it completely. For example, take The Smiths, masters of the ironic undercut. In their considerably in-poor-taste Girlfriend in a Coma, the text of what is at worst a callous boyfriend visiting their comatose girlfriend in hospital is made all the worse by the jolly music, complete with sleigh bells, in the background, turning the emotionlessness of the text into downright offensive cruelty. And it’s great.

2.    Chopping and changing. Having said that you don’t have to reorder the text, it sometimes can be great fun. Mixing texts, or cutting them up, can provide greater variety and allow for some interesting alternative readings. Nico Muhly, in his Map of the Worldfor voice and piano, combines three different texts: an excerpt of romantic poetry by Geoffrey Chaucer, John Weever’s recollection of his impression of Shakespeare’s family, and then Shakespeare’s own Sonnet 130 the well-known antithesis of traditional love poetry. Muhly’s combination of these three excerpts is quite funny – Chaucer extols his love’s virtues, Weever extols the beauty of Shakespeare’s family, and then Shakespeare slags off his own lover’s beauty and instead extols her personality. Another approach is taken by Joanna Bailie in her Artificial Environment 4, for tape and ensemble. Bailie recorded an introduction to the piece over sounds of traffic, explaining that in this ‘environment’, ‘sound seems to be operating to a different principle entirely’. This introduction (both traffic and text) is then chopped up, rearranged and so forth, with the instruments imitating the sounds playing behind them. The result is a brilliant but unnerving manipulation of the original material.

3.    It’s not just singers that can use text. There’s lots of different ways this can be applied. Joanna Bailie above proves that the composer themselves can be heard in the final composition. Steve Reich similarly uses pre-recorded text (albeit text he has recorded from other people) in many of his compositions (including his video-operas The Cave and Three Tales, the phase pieces Come Out and It’s Gonna Rain, and the string quartets WTC 9/11 and Different Trains) and often has musicians imitating the text. Peter Maxwell Davies’ Eight Songs for a Mad King wasn’t originally written for a singer, but a South African actor called Roy Hart. The instrumentalists can also deliver text themselves – a version of Stephen Sondheim’s musical Into the Woods at the Menier Chocolate Factory in London was performed entirely by actor-musicians, who both accompanied and acted. Text doesn’t even have to be sung – you may have heard of a little genre called ‘rap’. And choirs don’t have to sing either – Carl Orff and Ernst Toch (among others) wrote bookfuls of ‘Spoken Choir’ music. And it is groovy.

4.    The text doesn’t have to mean much. Sometimes, text can be used more like instrumental technique – the sound of the words can be as important (if not more so) than what they actually mean. Bon Iver’s third studio album 22, A Million, finds a beautiful mix between the two, where the lyrics seem to have meaning, but are largely carried by the sound effects used within them. Michael Nyman’s Bird List is just that – a list of bird names, devoid of any other meaning. However, it can be taken further. Claude Vivier, a composer sadly most well-known for predicting his own grisly murder, created his own language which he incorporates into many of his pieces, including Lonely Child and Journal, the ‘translation’ of which isn’t provided. Similarly, Enya and her lyrical collaborator Roma Ryan together devised a language called Loxian, which consists of nothing more than a transcription of syllables Enya sings off the cuff. Jazz scat singing is similarly off-the-cuff syllabic improvisation (Slim Gaillard’s Avocado Seed Soup Symphony is a particularly brilliant example of the humorous side of this), and in Björk’s song Venus as a Boy, she similarly breaks into a section of gibberish between two choruses. In Cathy Berberian’s Stripsody, the vocalist imitates sounds from television and radio. A lot of sounds.

5.    The voice can be as versatile an instrument as any other.Vocal extended techniques are too varied to list here. Peter Maxwell Davies’ Eight Songs for a Mad King is a pretty good place to start. If your voice can do it, chances are, someone else’s can as well.

6.    You can write your own text. Singer-songwriters working outside the ‘classical/art music’ scene know this all too well, but sometimes composers forget that there is nothing wrong with writing their own texts. Liturgical music is full of composer-writers throughout history (Thomas Campion and John Dowland are two notable English examples), but other genres have it too. Stockhausen and Wagner both wrote their own opera texts; Vivier wrote or collaborated on many of his texts as well. Even Mozart and Purcell wrote their own (admittedly incredibly rude) texts to use as the bases of rounds, grounds and canons. (Completely NSFW.)

7.    You don’t need text at all. Vocalise is a genre of music in which a singer sings a line to any sound. Rachmaninov wrote a very famous example of the genre, but many others exist; Glière wrote an entire concerto for coloratura soprano, and the soprano soloist in Glass’s 1000 Airplanes on the Roof similarly has no text to sing. And it is absolutely captivating.

In essence, the joy of working with voice is that it is one of the easiest things to experiment with. The huge variety of sounds a human voice can make is possibly unrivalled in music, and therefore allows a composer huge scope in which to work. Most importantly, as with almost any artistic endeavour, the composer should write what they want in the way they want. It may not always work, but any experiences which prove less than entirely successful will allow for future refinement. Always keep listening, and always keep writing!

The original post on the BBC Proms Team’s tumblr is HERE.