Andrew Webb-Mitchell


Poetry as music, music with poetry

The following article was first published in the poetry magazine, Mimesis, in Autumn 2008. Songs of Awe and Wonder librettist, Joanna Boulter, discusses aspects of her musical collaboration with Andrew Webb-Mitchell.

“For much of my life I have wanted to combine my two main interests and write poetry in some kind of musical form, and my earliest attempt dates back to 1959, a clumsy attempt at a chaconne written when I was seventeen. The more I considered the idea, however, the clearer it became that things weren’t as simple as I’d thought. For instance, although variation form seemed easy enough (a subject treated in varying ways, in words of different lengths, weights, sounds — in fact the basis of that early chaconne), the advanced musical forms I was being taught at school and university relied on modulation or key-change for their structure, and I could find no way of indicating this satisfactorily in words.”

“Later, developing both reading and listening on my own account, I realised that I was trying to be unnecessarily precise. A key-change equivalent could be as simple as a mood change. Many musical works make great play with repeated sections or phrases; poetry does this too, and if a repeated phrase has an impact of changed significance because of its placing or the emotional journey the piece has made, then this is surely poetry’s way of indicating a change of key, with all the formal progression that implies. I finally realised, too, that even in music form need not be as complex as sonata-form.”

[opening paras of MA long essay, Feb 2001]

These two paragraphs form the opening of the long essay I wrote for my MA in Writing Poetry (Newcastle University 2002), which describes how I set about writing my (at that stage still unfinished) first collection Twenty Four Preludes and Fugues on Dmitri Shostakovich (Arc Publications 2006).

Some twenty years before that, whilst working on a poem about a Japanese classical guitarist, I’d collected a notebook of words that could represent guitar sounds. I’d intended to build the poem from them, increasing the number of syllables with each stanza, like decorating a melody with increasingly shorter notes, but failed miserably. Clearly my original thoughts there hadn’t moved very far on from those early attempts, and this wasn’t the best way to represent music in words. In fact, whether or not it was possible to do this directly, it seemed to be beyond me. I’d always tried to make my use of words sound what is generally considered musical in poetic terms. That is, I took care to ensure that where the finished result didn’t sound beautiful, that was a deliberately intended effect, rather equivalent to orchestration. But I still kept trying to go further and failing.

I found myself working on the Shostakovich book, however, from a different angle. I’d read Elizabeth Wilson’s excellent book about the composer, and found it full of humanity and insight. The composer and his music,. Shostakovich and his music, against that difficult and dangerous Stalinist background, started to began to come alive for me and and become poems. These came first of all in pairs, third person Preludes in fairly free form indicating historical background, and Fugues in strict forms, sonnets etc, in the composer’s own voice, to represent the music. In practice of course, since music cannot speak with that degree of precision, that plan relaxed considerably as the work continued. I’d been thinking in general terms of his set of Preludes and Fugues for piano, Opus 87, but found it impossible to make a pair of poems mirror a pair of piano pieces except in the most general terms. Also, Shostakovich wrote his Opus 87 within a few months, and I needed to be able to say things about his whole life and art. So Op.87 is not a template for my poems except in a wider sense.

The book was eventually shortlisted for the 2007 Forward First Collection Prize. This generated a certain amount of publicity, and in early November 2007 I was surprised and delighted to be approached by the composer Andrew Webb-Mitchell, with a view to our working together on a major symphonic song cycle.

My immediate response was to accept, in principle, with the caveat that we should give things a trial period in case our styles didn’t suit. My feelings were that I wanted very badly to do this if at all possible. I felt I ought to be well suited to such a task; and further, that I’d been waiting a long time for such an opportunity. We were both relieved to find we should be able to work together, at least as far as style went, and we have now developed an excellent working relationship.

The work in question was to be a nine-movement piece for soprano and tenor soloists, with orchestra. Songs of Awe and Wonder takes as its theme various aspects of human inspiration – love, the countryside, the symbolism of bridges, the night sky, man’s dream of equality.

However, there was one very big difficulty. Andrew had already composed much of the music, in isolation, and was looking for a poet who’d be capable of setting the words to the music rather than the more usual way round. This seemed to me an interesting challenge although I knew it wouldn’t be easy. In practice it has turned out to be somewhat akin to writing extremely formal sonnets, something which fortunately I’m not bad at.

Another difficulty, which seemed at first worse than it turned out to be, was that I’m living in the North East of England, and he has now taken up a post as Director of Music at the British School of Beijing. This means that we work via email and MSN Messenger – no bad thing perhaps as it’s meant that correspondence and working discussions can be archived. Audio files arrive by MSN, and musical scores via email. Fortunately I could still remember how to read music, and rapidly became even better at it. However, I have had to devise a method of sending text which will indicate the bar-lines, rests, etc. There are examples below: in these the indication 3{  } shows that the enclosed words are set to a triplet figure.

There were more basic difficulties however, especially in the early stages whilst we were still finding a method of working together. For instance, it hadn’t occurred to Andrew, as a musician rather than a poet, that the fundamental rhythm of the English languge is iambic. Working on that first song I found myself continually pleading for up-beats. These are of course a frequent occurrence in music, but it just so happened that there were next to none in that first movement, though one or two were added in progress so as to make my job easier. There were times when he suggested making somewhat bigger changes to a musical phrase, to accommodate a verbal one that he liked but felt I hadn’t quite managed to fit into the melody, and this bothered me as I felt I’d failed to meet the given challenge. However, we have so far sorted these problems, working on the basis that there should be no question of personal pride and that our joint aim was for the overall quality of the finished work to be to the best of both our abilities.

Some of the passages so far feel triumphantly successful to us both. And there are others where I could have beaten my head against the wall as I worked, and yet in the end the setting I finally came up with was more than satisfactory. But there are one or two places where I’ve had to bite my lip and let what feels like an inadequate phrase pass, because there’s no way I can get it any better and Andrew at least is happy with it.

Many aspects of writing a libretto this way round seem to have much in common with writing very complex formal poetry, for instance sestinas and tightly controlled sonnets. As a particular example, there’s a pair of interlinked sonnets in the “24”, entitled “Double Fugue”, where I tried to use the formal musical device of the double fugue, with its twin expositions, recapitulated as an interleaved whole.

In the same way, when the music is the primary given, you lose one of the basic options of writing poetry, the possibility of moving lines about within the piece, because they almost certainly won’t fit the music in any other position. This is not unlike the lack of room for manoeuvre within a tight poetic form, and is something I choose to see as a challenge rather than a handicap.

There are problems of diction – you can’t give a singer a pinched vowel on a high A-flat – above a certain pitch it’s preferable for vowels to be mostly rounded, forward-produced sounds. That means being careful with sounds like ee.

Diphthongs, or double vowels like point, or sound, also need care – they can in practice contain 2, 3, or even 4 vowel sounds which may be separately audible when sung. And it’s astonishing what can behave like a diphthong, especially on a long high note – “now”, for instance, which can change audibly from ow to oo mid-note, and lucky if it does so without inserting a w as it changes! Things like this can pose a problem for a singer; and whilst obviously professionally trained performers will be able to handle them, I consider my job includes lessening vocal problems where possible.

As to consonants, there’s not only the question of whether they will hiss (an s at the beginning of a word following one that ends in s, for instance) or sound like gunfire (too many t sounds too near each other). I’m used as a poet to keeping an eye on that aspect; but in reading one’s own work aloud there’s time to make the little vocal adjustments and slowings-up that handle such things. In this case, writing words for already-composed music, I’ve had to ensure (by singing the line myself) that there will be time to articulate each word to its note. Fortunately I’ve done some choral singing in the past, so have something of an idea of how this can work from the inside.

These points are often not insurmountable in practice, I’m assured, but it makes sense not to add to possible performing difficulties if I can manage it. They are good examples of the differences between working this way round and the conventional, words-first way: if I’d produced the text first, it would have been up to Andrew to provide music whose pitch etc accommodated the words. Next time we collaborate, we hope to work more together from the start.

Again, there are potential problem with punctuation marks. Even light ones like commas can tend to encourage a singer to snatch a breath. So I need to be aware that they may not be seen as verbal or syntactical indicators only, and this has caused the occasional problem. It took me a while to realise that when Andrew placed a rest within a phrase of music he didn’t necessarily see it as breaking the flow of the musical line; and that therefore this didn’t necessarily mean I should or even could put a comma there. We have now realised that we need to ask each other what’s required, in passages like this. I don’t as a rule punctuate heavily in poetry, and indeed quite often use spaces and line-breaks to fulfil that function; but I feel this wouldn’t work here, at any rate until we get more used to each other’s styles and ways of working.

As I said to Andrew in an email: “I’ve been trying to approach my interest in the combination of words and music from a different angle: What was I doing in the “24”? I was trying to make the words behave as though they were music, as far I could while still sticking to the narrative, description, etc. This time I want to illustrate and illuminate particular music – yours – so that’s the angle I’m coming from.”

Email, JB to AW-M, 27 May 2008:

I’ve been finding it hard to get the verbal accents in the correct part of the bar. Sometimes the stress tries to sit down square on the 2nd beat, rather than just in front of the third, and that makes the whole thing sound really inept. I had to get it actually wrong and then analyse that, to work out what the problem was. It needs to be got right or I’m afraid the music will sound subtly skew-whiff; and not the same as syncopation either, I know about that for pete’s sake! Words, you see, carry their own subtle intrinsic stresses, which have to balance with and against poetic metre; and then we add musical stress, and counter-stress, into the equation – oh this is real fun-challenge territory!!

Here’s a simpler example, to show the sort of thing I mean:


The Artist’s Wife, bar 76, text with barlines:

Now– you say you |have me perfectly posed -.

You |need to un-derstand- |

that there’s so- much more in |me than meets—the |eye–.| //|

How– to show a |wo-man just as she is-?

I’m |more than cheek- and hand,|

more than curve- of hip and |breast alone- could |be–.| //|


‘So’, for example, isn’t on a strong beat, but it can take the sense-emphasis because it’s a longer note, coming directly after a rising interval. However, ‘perfectly’ feels a bit uncomfortable to me, in a way that ‘just as she is’ in the corresponding passage doesn’t. This seems to be a question of the comparative weights of the syllables, and also the comparative impact of the consonants. It’s that c and t coming together in the middle of the word I think – they make the syllable altogether too long and the word too clumsy for the rhythmic figure.

[However, perfectly did pass Andrew’s critcal ear, so was permitted to stay.]

This first movement was psychologically complex, with its emotional exploration of the artist’s wife herself, Edith, wife of Egon Schiele. It proved deeply satisfying in the end, but very difficult to work on initially, especially as it was my first effort.

The second movement, this time a tenor aria in contrast to Edith’s soprano, posed much less of a problem. For one thing, its form is more straighforward, indeed almost architectural. For another, I’d begun to get used to working this way. The main theme has a strongly rising element, and gives an open-air mpression. Discussing it, we thought of Elgar and his famous cycle-rides around his home in Malvern. Here, the rising motif in the melody Andrew had written reminded him that Elgar used to fly kites on the hills there, and me of his comment that the Malvern countryside was full of music, for him to take as he needed. Those things combined made me think of hamonics on stringed instruments (Elgar was an excellent violinist), and also of sympathetic vibration, as found in certain older double-strung instruments such as the viola d’amore.


Kites over Malvern, bar 110 – 119, text with barlines:

The |string is taut,- harmonics |haunt the air–

vi- |bra-ting- in the |at-mos-phere–

and |now- they re-so-nate in |me,—

and I re-|peat their song,–

be-come a |soun-ding board- for what I|hear.—


This pattern of balancing lines, with or without some degree of rhyme or assonance, has also been turning up in the second movement, Kites over Malvern. It’s a shape I often use myself when writing loosely structured verse, so it’s very interesting to find Andrew’s phrase-patterns using it too. That is, a longer line in a 4-line verse, or a short line interposed to make a 5-line verse. It’s a structure-device which appealed to my ear aged around 15 when I met it in Wordsworth’s Immortality Ode. But thinking about it now, I suspect I found it also in the symphonies and quartets of Brahms.

This is part of a conversation we had via MSN, some time in early August 2008:


Kites over Malvern, bars 33 – 37, text with barlines:

and now – it’s |3{riding the} ris-ing- |3{- air,

the} 3{sway of the} |3{breeze and the} 3{swell of its}|breath. – -|


A: That line seems masterful to me – sway breeze swell breath. So clever.And riding rising – superb – in my ignorant opinion!

J: I was pretty pleased with it, yes!

A: But I don’t know what is GOOD good – I just know if I like it. In this case I’m delighted by it. I don’t know what snobby poet critics will say!

J: That’s the point, that we should like it! And this is not a situation where I’m having to make poetic concessions to the fact that this is a libretto. That is EXACTLY the sort of thing I would have writtten if I’d been on my own with it. In fact, I am using the same poetic skills.

J: By the way, I was thinking last night about what Elgar said – that if you cut that music it would bleed.

A: Yes.

J: I suspect he meant not only that he loved it, as you said, but that he himself had already trimmed out all the extraneous stuff, till nothing was left but what really mattered. Now this is something that’s difficult for me – if you’ve written music there, I have to give it words, find SOMETHING for it to say! Working the other, more usual way round, I would give you the essential poem, and you would have the option of expanding the piece with interleaved music – which is not an option open to me here. Never mind – a challenge!!

A: Sometimes I feel it would be handy to have you about as I craft the vocal line

J: I love that term – “craft the vocal line”.But I fear we might come to blows! don’t you think?

A: I think you like to be given the finished article!

J: I do know what I’m doing, then. And if we were working the other way around, would YOU prefer the finished article?

A: The finished article would be a problem. I’d probably prefer a sketch from you, because I’d struggle to tie myself down to the lines of the poem. The music would want to wander off somewhere

J: It’d be interesting to do it that way for once though.But that’s my main problem, you see – the balance between not putting in too many words just to fill the musical phrase, and making sure I’ve said enough for the meaning, the progression, to be clear and to the point. And making sure it’s what Michael Symmons Roberts in a similar context has called, rather than outright poetry, “poetic language”

A: Yes. Luckily, my structures are pretty tight.

J: Let’s carry on as we are, at least for now.

We now (September 2008) have seven movements left to complete of a nine-movement work. Most of what remains is of course libretto, since I got started well behind the composer; but there are two movements still to be composed and in several the orchestration is not yet finalised.


Unless we do come to blows, it seems unlikely that this will be our last collaboration.

en_GBEnglish (UK)
zh_CN简体中文 en_GBEnglish (UK)