Skip to content


A Private Letter
A Poet on Writing for Composers

NOT LONG AGO, I was giving a reading with another poet who has written libretti for composers. I hadn’t heard anything of his musical collaborations for a few years, and asked him if he was still working in the opera world.

“I’m doing something for television,” he said, “and I’m writing fiction, but I’ve lost interest in working with composers.”

Since I was doing just that at the time, I thought I’d better ask him why.

“It’s thankless,” he said. “No one ever describes an opera as Schikaneder’s Magic Flute, or Hofmannsthal’s Elektra. It’s all your idea; you dream up the characters, the story, the poetry of it. You inspire a composer to set what you’ve written, and then the composer gets the credit for the whole thing.”

Well, he has a point. The downplaying of the librettist’s value goes back a long way. There’s a story that Boito, passing through Verdi’s farmyard at Sant’Agata, noted with disgust that a load of manure fetched a higher price than a libretto.

But maybe this downplaying is a reflection of the way a good libretto works. T.S. Eliot famously advised Michael Tippett to write his own libretti, since “poets are always going to do with the words what your music should do.”

Perhaps the greatest of all modern poet-librettists—W.H. Auden—made a similar argument when he said that “The verses which the librettist writes are not addressed to the public, but are really a private letter to the composer. They have their moment of glory, the moment in which they suggest to him a certain melody; once that is over, they are expendable: they must efface themselves and cease to care what happens to them.”

In any song cycle, oratorio, and particularly in any opera, the music is overwhelmingly the dominant element. The words are there to inspire and generate the music, without promoting themselves. For the best libretti, and the best librettists, invisibility is a mark of quality.

For this reason, as Eliot warned Tippett, poets don’t always make the best librettists. I’ve often wondered if Auden—already a major literary figure when he began to collaborate with Britten and Stravinsky—used his partner Chester Kallman to try to achieve a measure of this necessary invisibility. Auden, who was acutely aware of the dangers of the words dominating the music, wrote all his libretti in collaboration with Kallman, and—despite the efforts of many critics to unravel how they worked, or who wrote which bit—refused to claim any line as coming from his pen alone. It’s as though—to misquote the Bible—Auden must decrease so the music could increase.

Auden and Kallman’s texts, set by Britten—such as Paul Bunyan—and Stravinsky—The Rake’s Progress—are among the finest of the twentieth century, but there are two other twentieth-century librettists who in very different, almost opposite ways have been a huge influence on me, and on many other writers who have worked with composers.

The first of these achieved such an exemplary invisibility that she’s still not featured in most dictionaries of music. Although she wrote several plays, she was not a published poet, and is little known even among admirers of Britten’s music. Myfanwy Piper was a key collaborator in the life and work of two major artists—Britten himself, and her husband, the artist John Piper. For Britten, she adapted Henry James’s novel The Turn of the Screw into one of the finest and most performed chamber operas of the last century. She made a powerful full-length opera out of Thomas Mann’s Death in Venice, again for Britten, and she worked on a number of smaller pieces for him and other composers. I admire her economy of style and simplicity of structure, and her use of different registers of speech and song, from snatches of nursery rhymes in The Turn of the Screw to a pitched battle for the soul of a child in the same opera. Perhaps above all I admire her ability to build a seamless narrative which allows the composer all the crucial elements of arias, duets, trios, ebbs, and flows. I feel it’s high time Myfanwy Piper’s talent was more widely recognized.

The other librettist I want to mention could hardly be more different. Quite how different was brought home to me during a trip to the Isle of Skye a couple of years ago. I had been commissioned by the Welsh National Opera to write a libretto for a James MacMillan opera. As a preparation for this, James and I spent a week in a house on Skye, to lay the groundwork of plot, subject, and character. We would work most of the day, then in the late afternoon watch DVDs of operas we admired, or thought might be relevant. On one particular afternoon, through accident rather than design, we watched The Turn of the Screw back-to-back with Strauss’s Elektra. Myfanwy Piper’s subtle, understated text was in astonishing contrast to Hofmannsthal’s superheated expressionist outpouring. Myfanwy Piper opens The Turn of the Screw with the enticing lines: “It is a curious story. I have it here, written in faded ink—a woman’s hand, governess to two children, long ago.” Hofmannsthal’s text begins with all guns blazing: Wo bleibt Elektra? Ist doch ihre Stunde, die Stunde, wo sie um den Vater heult, dass alle Wande schallen, which roughly translates as, “Where’s Elektra? It’s the season, the season when she howls for her father, and all the walls echo again.” If Strauss wanted to hit the emotional heights and depths with his music, Hofmannsthal was going to get there first with his words. At the beginning of their long collaboration, Strauss told Hofmannsthal not to consider the music when writing his libretti. “I’ll see to that,” said Strauss. But he needn’t have bothered making the point. Hofmannsthal replied that he had no intention of considering the music. “Rest assured, my dear Strauss,” he said, “that over the whole text I shall rely upon myself alone, and not at all on the music; this is indeed the only way in which we can and must collaborate.”

Perhaps not surprisingly, given the force of his personality and the power of his work, Hofmannsthal is one of the best known of all librettists. I admire him because he seems to be a genuinely equal creative partner with Strauss, not just a supplier of words. Rather than seeking invisibility, wearing plain clothes so the music can dress in bright colors, he holds nothing back. His approach seems to call on the composer to match the intensity of the words. He seems to be saying “I’m holding nothing back, so if you want to set my text, you’d better raise your game to meet it.” They may have had markedly different styles, but Piper and Hofmannsthal each inspired some of the finest work from their respective composers.

In my collaborative relationship with James MacMillan, he has never asked me to hold anything back. I’ve heard lots of horror stories from other poets about working with composers. The most common complaint is that the music rides roughshod over the inherent musicality of the poetry. “He might as well have set a page from the telephone directory,” said one disgruntled poet of a composer collaborator.

I think I was fortunate, in that my collaborative relationship with James developed in the right way. It began when we met at a radio recording about fifteen years ago and liked each other’s work, but without any thought of working together. Then we became friends, and our families became friends, and then we started to work together. By the time the work began, we already knew each other well and trusted each other’s judgment. So far, we’ve produced a song-cycle, two choral works, an oratorio, a chamber opera, and a symphony with a choral coda, and we recently completed an opera for the WNO—The Sacrifice—which was premiered in Autumn of 2008. Through all these pieces, I’ve never yet felt any pressure from the composer’s side to write in a different style from my poetry, or to hold anything back. However, over the years I have begun to realize that poetry and libretti have different voices, different forms, and though in previous years I have included libretti in my books of poems, I now regard them as a separate part of my work.

My first collaboration with James was a piece called Raising Sparks. This was the only moment in our work together where the text existed before any discussion of the music, and was made without any such consideration. I wrote a sequence of six short poems, inspired by a Hasidic Jewish story of creation and redemption. This story had come to my attention in an eighteenth-century work called The Light of the Eyes, by Rabbi Menahem Nahum of Chernobyl.

The story depicts God holding back his power and light to make space to create something other than himself—an act of self-limiting or withdrawal called Zimzum. Then into this space God shines his light of creation, but that light is so intense that it smashes the clay vessels intended to capture it—a cataclysm known as Shevira.

This cataclysm results in fragments of divine light—sparks—being scattered across the world, landing in accessible and inaccessible places. Sometimes these sparks are concealed by shards of the clay vessels.

The purpose of life then becomes a redemptive one, to find and raise the sparks, and make the divine light complete again. According to Hasidic teaching, these sparks may lie in trivial encounters or in major challenges. They are as likely to be found in the eyes of strangers as in those of your own children.

This creation and redemption story has a dark twin in the chilling parallels—broken vessels, shattered light falling—between Chernobyl’s Hasidic tradition and its twentieth-century infamy as the scene of a nuclear disaster.

At the time I wrote that poem, around 1993, James and I were in regular contact, and discussing our work in progress. I sent him the new sequence to see what he thought of it, and he replied that he wanted to set it. He’d been commissioned by the Nash Ensemble to write a piece for them, and he wanted to combine that with a commission for the mezzo-soprano Jean Rigby. He had studied the poems in detail, and already had a strong sense of how he wanted to set them. The next stage was for us to meet and spend a day talking about the ideas and images I’d used. Apart from this general discussion, he asked me to read the poem aloud to him, as many as six or eight times during the day, so he could pick up the nuances of my music. I didn’t realize it at the time, but I believe that was quite a rare experience for poet-librettists. Many composers aren’t remotely interested in the music of the poetry. Some seem to measure poetry by the yard, rather than listen for the measure of its metrics. The Russian composer Aleksandr Mossolov took this approach to its logical conclusion by setting randomly chosen small ads from the newspapers. But James wanted to understand the rises and falls of the poem’s rhythm.

After that day, we spent several more days, or parts of days, where he played and sung draft sections of the piece on the piano to me. Because the poems were written without singing in mind, and because I knew something of the complexities of breathing, pronunciation, and enunciation for singers, I expected a lot of problems and alterations when it came to the detailed word-setting. In fact, only three things changed: Firstly, he suggested that I might write a refrain, which he felt would help him to knit the different sections of the poem together musically. I wrote the couplet—“Zimzum, Shevira, shattered light falls as dry rain / static in my hair.” I thought it worked well on the page too, so I kept it in all versions of the poem. Secondly, the phrase “by a fraction” was not set, in a line which read “when hands dropped, sky had lightened by a fraction.” Thirdly, the phrase in the last poem—“the silent line that queues up by the railway track to welcome its Lord”—was set as “rail track,” rather than “railway track.” That third change was made to suit the rhythm of the music at the end of the piece. The second change, I have to confess, was made accidentally because James was working from a slightly out-of-date draft when he set that particular section. Although there were only these minor changes, I decided to allow two fractionally different versions of the text to exist—the poem as published in my collection called Raising Sparks, and the libretto as published in CD and program notes.

I don’t read scores very well, so I had no real idea of the fruits of this collaboration until the piece was rehearsed at the Royal Festival Hall on the day before the premiere. I was delighted at how faithfully and diligently James had dug into the poems and brought up every nuance, every image.

I was struck by the use the music made of those wonderful, mystical words from the Hasidic tradition, Zimzum and Shevira. I was also struck, and positively, by the way he would repeat phrases and lines from the poems, sometimes in a different order, making a different shape out of them. But perhaps the greatest surprise for me on that rehearsal day was the way James had woven a kind of Russian folk melody into his setting of the last poem. It’s so appropriate to my vision of that poem, and I find it so beautiful, that I rarely include this poem in readings any more. I can’t read this last section without thinking it ought to be sung rather than spoken.

About a year after Raising Sparks, I worked with James MacMillan again. This was a very different kind of project, in that it was collaborative from the outset. Before a word or a note was written, we hatched a plot together. He approached me to say he had a commission from the BBC Proms (the leading British classical music festival) which allowed him to assemble huge musical forces—the BBC Symphony Orchestra, the BBC Symphony Chorus, the Hilliard Ensemble, and the Westminster Cathedral Boys’ Choir. With forces like that, it seemed obvious that we should try to create a big choral piece, an oratorio, so we needed to decide what our subject would be.

Because Raising Sparks told a story, albeit a strange and cosmic one, I was keen that this piece should go in a different, less narrative direction. At the time we were both fairly new fathers, so we started talking about that. I’ve always been fascinated by the metaphysical poets’ use of the word “quickening,” and I discovered it again in pregnancy handbooks as the moment when a mother first feels the movement of the child inside her. Rather than a linear narrative, we decided we would try to create a series of impressions of the moment when life begins. From the outset, I had in mind a quotation from John Donne which later became the epigraph to the poems. Donne wrote that “A fly is a more noble creature than the sunne, because a fly hath life, and the sunne hath not.”

This time, the process was much more genuinely collaborative than on Raising Sparks, when the poems were complete before the music began. With Quickening, there was a steady stream of phone calls, letters, and meetings in which we would each hand over ideas or fragments of work to allow the other to move to the next stage.

We decided that to make proper use of three very different choirs, we should give them different voices, so broadly we defined the BBC Symphony Chorus as the central narrative or descriptive voice, and then we created a dialogue for the other two voices. The remarkably pure sound of the four singers in the Hilliard Ensemble became the voice of a new or expectant parent, an introspective mother’s voice addressed to her unborn baby. The idea for this voice arose partly from our own families and partly from the composer Olivier Messiaen’s wonderful story that his mother, the poet Cécile Sauvage, commissioned him to be a composer by reciting specially written poems to him in the womb.

The boy choir—Westminster Cathedral Boys—would be the voice of the child itself. I put this last voice in italics, to set it apart from the others. We decided we wanted the three choirs to interact, to interrupt each other sometimes, to pick up from each other and move the piece in different directions.

At various points, you hear a kind of babble or background texture of voices. We wanted something that sounded pre-verbal, as a counterpoint to the poetry. For this we used a rough phonetic version of the Lord’s Prayer, spoken in Aramaic, its original language.

In the last section in particular, “Living Water,” I was aware of writing very much with the possibilities of the music in mind. I wanted to write about water coming to life, water after drought, and then to connect that with an image of John the Baptist. This was an example of the libretto directly affecting the music.

The idea of a water charmer, who could conjure music and bring water to life by rubbing the edge of a great bronze bowl, appealed so much to James that he asked the percussionist to find one, and to learn to play it.

Quickening was the last libretto I included in a book of poems. Even then, I felt it sat slightly uneasily with the other poems. It’s not because I had to hold anything back in the making of the libretto. There was no conscious flattening of the language to allow the music to flourish. James responded positively to the richest, most densely packed language I could come up with. I think the reason I feel they don’t belong in the same book is more to do with it being a collaborative venture from the outset. When a piece begins—as Quickening did—with a conscious sense of musical possibilities, a sense of what the composer might be able to do with this, somehow it changes. Perhaps Quickening is a better libretto but a lesser poem than Raising Sparks, because Quickening was—in Auden’s words—“a private letter to the composer.”

Because a true libretto has no existence without its music, there’s a profound sense in which it doesn’t matter. That Russian composer setting the newspaper small ads was merely making explicit what is implicit in the work of most composers. Stravinsky, in preparing to write Oedipus Rex, didn’t set Sophocles. He took a French translation (by Jean Cocteau) of the Greek drama, then got a young theology student to translate the French translation into Latin. So, in the end he set a text from one ancient language and took it to another ancient language via a contemporary one. He explained that this cumbersome process would endow the music with “a certain monumental character.” For “monumental,” I’m tempted to read dead. He needed to kill Cocteau’s libretto to let his music flourish. Whatever his motive, Stravinsky’s Oedipus Rex illustrates a timeless truth about libretti. They can be in any language, and about anything. Their only purpose is to generate great music.

My only work (to date) that arose directly from academic theology is another James MacMillan collaboration, a piece called Parthenogenesis. Its starting point was a rather strange commission. A Cambridge University venture called Theology through the Arts decided to commission a range of artists with theological interests to produce new pieces in collaboration with academic theologians and philosophers. James and I were paired with one of the most brilliant theologians of the last fifty years, a poet himself, who was, I believe, the youngest person ever to hold an Oxford University chair. Rowan Williams has since become Archbishop of Canterbury, but we worked with him a number of years earlier, when he had just left the university to become a bishop in Wales. The deal with Cambridge was that the three of us must commit to spend fifty hours talking, spread across six months or so. Out of that conversation, we should produce some kind of work, or several different works, with their origin in the conversation itself.

At our first meeting, we talked about the weather, and the news, and our journeys to Monmouth for the meeting, and we all agreed that we didn’t want to have formal discussions about themes or issues. What we needed was a story, or a character, to kick-start our conversations. In my former life as a documentary filmmaker, I had made a film about genetics, and I’d come across a remarkable story in the process of researching the film. I’d wanted to write about it for some time, but hadn’t yet found a way. I told them the story, and that gave us our subject. Here it is in a nutshell:

It’s 1944 in Hanover, Germany. A young German woman is caught on the streets in an Allied bombing raid. She later recovers from the minor injuries inflicted by the blast, but nine months later gives birth to a daughter with identical fingerprints, blood type, and other indicators to the mother. Medical tests support the woman’s claim that she had not had sex. Doctors hypothesize that the shock of the bomb may have jarred a dormant body cell within her womb, triggering parthenogenesis, or nonsexual reproduction. Effectively, the daughter was a natural clone. This was not a made up story, but an unverified one. I checked with geneticists. It could have happened. No one can be certain whether it did or not.

Well, that was our fifty hours’ conversation mapped out. We talked about genetics, cloning, about the incarnation and the annunciation, about what it would mean to create a work that was a negative print of a key Christian doctrine—a virgin birth with human evil (a bomb) as the unseen father, rather than God. At the end of the process, Rowan produced an essay about the need to create “parody versions” of the central stories of the Christian tradition, to allow us to renew our sense of the originals. James and I produced a music theater piece with the catchy title Parthenogenesis. It requires a small orchestra, two singers—soprano and baritone—and an actress. It is a kind of parody annunciation based on the Hanover story, with the virgin mother (soprano) being wooed by the angel (baritone), constantly interrupted and challenged by the spoken voice of the actress, playing the cloned daughter, bitterly seeking an identity of her own.

Of all the pieces we have done together, this strange, angular, difficult piece has been the most performed. It’s been translated into Spanish, Italian, and most recently into German, in a fully staged performance in which the mother and daughter characters wore identical clothes and identical towering wigs, and the daughter was heavily pregnant—despite the fact that geneticists believe a human clone would be sterile. I love the fact that we weren’t consulted about that or any of the other performances. I think I’d be much more anxious about my poetry being staged or changed or reinterpreted, but because these music collaborations are collaborative from the start, I’m very happy for that collaboration to extend to others, to exclude us, so the pieces have a life of their own. Parthenogenesis has recently been revived again, in a production at London’s Royal Opera House this past spring.

My work with James MacMillan continues, with another short opera in the pipeline. And that Welsh National Opera commission planned on the Isle of Skye came to fruition last year under the title The Sacrifice. Most recently, I was commissioned to produce a new set of singable English versions of the poems from Schubert’s Winterreise for a new production by the tenor Mark Padmore. These new versions, in an innovative staging by director Katie Mitchell, interweave the Winterreise songs with sections of poetry and drama by Samuel Beckett. The resulting production—having excited and divided critics in the U.K.—will arrive at the Lincoln Center in New York this December.

The backbone of my life and work is still poetry. Like a number of poets, I regard my other life as a librettist as a kind of accidental career. I don’t play an instrument with any accomplishment; I’m a poor reader of an orchestral score. I didn’t set out to forge a career in music. But as long as this work with composers remains exciting and difficult, and especially as long as new ideas keep arising from my ongoing friendship with James MacMillan, then I’m delighted for this accidental career to continue.

The Image archive is supported in part by an award from the National Endowment for the Arts.

If you like Image, you’ll love ImageUpdate.

Subscribe to our free newsletter here:

Pin It on Pinterest