Setting all words: James Weeks interviews Michael Oesterle

Michael Oesterle’s piece all words is remarkably simple. It sets all 1,015 three-letter words in the Scrabble dictionary to music. Before performing it at BBC Radio 3 Open Ear on Saturday, EXAUDI’s director James Weeks caught up with Michael to ask an even simpler question, ‘Why on earth would a composer do this?’






James Weeks: all words sets all the 3-letter English words in the Scrabble dictionary in alphabetical order. Why did you decide to do this for your EXAUDI commission?



Michael Oesterle: I had a long look at one of the EXAUDI Vocal Ensemble publicity photos (the one where you are holding a tuning fork) and doing so I pretty much figured out the basic ground rules for writing my new piece: text? yes. language? English. how much text? loads. style? irrelevant.

Mymichael oesterle first idea was to take a random text from the internet and reorganize all the words in alphabetical order. But then I remembered the wise words of David St. Hubbins “It’s a fine line between stupid and clever.” Nevertheless, the idea to use all 3-letter words from the Scrabble dictionary came to me quickly, and it did seem like a good idea (I don’t have those very often). I could easily imagine the people in the publicity photo singing this piece.

The alphabetic words aren't the only process in the piece; there's also a harmonic process and a process of constant acceleration and deceleration. And then there are the constant changes of texture with the 8 singers coming in and out of the line.

How do all these things interact?

The 1015 words are in alphabetical order, which is a total number of 1080 syllables. The syllables I set to 1080 quarter notes (crotchets), which I divided evenly into 360 measures in 3/4 time. The 360 bars are divided evenly into 30 pages, each page has 12 bars divided evenly into two systems of 6 bars (18 crotchets). All very nicely organized -  an efficient format for the score/page setup of a piece for 8 singers.

I then wrote a melodic line in the G Lydian mode (Raga Yaman - a neutral, bright sound, useful anytime of the day), which I stretched over 540 crotchets by using a triangular number sequence (1,1-2, 1-2-3, 1-2-3-4, etc). Exactly half way through the piece I reverse the melodic sequence, but not exactly like a palindrome - I don’t actually invert the intervals and run the melodic idea backwards (too much work) - only the triangular sequence is reversed (4-3-2-1, 3-2-1, 2-1, 1). I wrote out the resulting melodic line in unison (octaves) for the 8 voices, transposing any notes that presented difficulties for  any vocal part. The result gave the piece the soundscape which I felt matched the idea for using this particular text.

I then devised a number sequence for each vocal part which determined when one would sing and when one would not sing. These number sequences became stencils which I used to give each part rests. This was tricky because the piece couldn’t have tutti rests, which would interrupt the flow of syllables/words. I tweaked this quite a bit to make it work, and it turns out that my stencils allowed for only two moments in the piece when all 8 voice sing together: on the words 'age' and 'gap'.

The overall texture became a tapestry of all possible permutations between the 8 voices. I thought this was an 'orchestration' for the piece that worked well with the given melody and text. Just before sending the score to you, I did add sprinkles of harmony - a minor/major third here and there to give the ear a bit more variety. The little changes just before sending a piece off are always my favourite parts when I hear a piece performed.

The constant speeding-up and slowing-down scheme is made with a pattern that fits the page/score setup. I felt that the music should be elastic, not simply a steady march of crotchets in one tempo, which sounded too stringent. I imagined the constant speeding-up and slowing-down like waves of words washing up on a shore. At one point, I think after I did a read-through and you sent me the recording, I did get cold feet about the elastic tempi idea, I thought it might be too difficult to perform - but you (James) seemed eager to try it, I am very grateful to you for helping me to stick with the idea.   

I think I used processes to make this piece because setting the words in a conventional manner would have been tedious with respect to this text. I imagined the steady stream of words being perceived according to a listener’s own experience with this vocabulary. For example, immediately after the premiere a composer colleague told me that he 'heard his surname' in the piece - I thought that was brilliant.

By the way, I did delete three words because of their hateful ugly meaning - so there are three profound moments of tutti quarter-note/crotchet rests in the piece. Stewart Lee would approve.


The way you write for voices is interesting – it would be easy to say it is not expressive in a conventional sense, given the text! – but my memory of performing it last time was that there was a sort of brightness and energy to the way you 'sound out' the voices that was quite joyful (as well as intrinsically funny at times).

I had a 'joyful' sound in mind for this piece, probably because I had so much fun figuring out the definition and the pronunciation of so many of the 3-letter words. A bunch of them had me thinking - come on, is that really a word? The vowel heavy words and consonant heavy words made for a nice contrasts, and I think that the 'sound' of the words is interesting throughout the entire text. Before writing any music, I read the list of words out loud many times - I enjoyed this very much actually, and the piece just emerged from there.

I thought the music for this piece should match the feeling I had that I was improving my vocabulary. But I certainly didn’t mean this to be a silly piece of sorts, I tried to give this text a setting that I felt sounded interesting and compelling. Actually, I prefer not to set poetry to music (I did a bit of that when I was younger) - a writer carefully perfecting a poetic text need not suffer the indignity of having their art ruined by my music. Had I used this text for a piece for voice and orchestra, there may have been far more giggles (and groans). Again, to me the idea for this piece seem a good match for the bright energy and fearless disposition of the Exaudi Vocal Ensemble, the people in that publicity photo.

Have you written a lot of vocal music?

I can’t say if it is 'a lot', but I have had quite a few experiences writing music for voice. Thankfully with every experience I have learned something about what composing for voice means to me, technically and aesthetically. One experience was quite difficult - some years ago I wrote the 'imposed' piece for a vocal competition. I spent a few days listening to performances by a load of fantastic singers performing the 10 minute piece I wrote for the occasion. It’s at times like these that you really come face to face with the immense genius of someone like Schubert or Gershwin, and realize that composing for voice should be something left to those who are truly gifted. Still one keeps trying.  


Do you approach voices differently from instrumental music (apart from the obvious differences of course)?

I think so. But it really depends on the idea for any given piece. I suppose at times it is about the difference between writing for instruments that require air to produce sound, and instruments that do not, for example: writing for Accordion made me think differently about composing music for voice. But one can easily focus too much on how idiomatically one should write for voice or instrument. If my music for voice generally seems more instrumental in nature, something I don’t do intentionally, it may have something to do with my preference for wider intervals, arpeggiation, and generally avoiding melismatic techniques in vocal music. Text is, for me, articulation, and I like each note articulated (he says with a Dutch accent). But this will certainly change in time, I still have plenty to learn - probably a good idea to play less Scrabble. 



You can hear Michael Oesterle's all words at BBC Radio 3 Open Ear on Saturday 14 April.
Click here to find out more and book tickets.

Find out more about EXAUDI vocal ensemble

Find out more about Michael Oesterle


Picture by Matthew Andrews