A chamber opera in one act with libretto by Marilyn Bowering based on the last night of Marilyn Monroe's life.

For soprano (Marilyn), baritone (the Men) and two smaller choral parts "The Tritones" (tenor and bass). The score is for a small ensemble of eight players: 2 violas, cello, double bass, clarinet/bass clarinet, horn, basson, percussion, with the addition of an on-stage jazz trio: tenor saxophone, piano, double bass

Note : #1

Marilyn Forever

The opera was developed initially in 2010 through a workshop in Banff, Alberta and was eventually performed in Victoria BC on September 13 and 14 2013 with the Faroese singer Eivør Palsdottir in the title role, and with Danish singer Thomas Sandberg as "The Men. It was commisisoned by Aventa, whose director Bill Linwood has conducted all the performances to date.

It has been subsequently performed three times in Adelaide in February/March 2015, with a different cast; and shortly afterwards by Long Beach Opera in San Pedro, California in a different production, which used two singers for the title role.


The opera examines Marilyn Monroe's relationship to love, death and ambition as these are revealed during her passage from life to death, August 5th, 1962. As the work progresses, the performance interweaves what is taking place on stage with the trajectory of Monroe's life through relationships, fame and myth. Ultimately, the characters of the performers and men in her life fuse with the forces that have led to her death.


It is the night of August 5th, 1962. Marilyn's bedroom is entered by men who view her body and search through her belongings. Once they leave, Marilyn arises and quickly runs out. When she next appears she is on a rehearsal sound stage where waiting for her is the Rehearsal Director and the backing group, the Tritones, who have been reviewing her notorious history. The rehearsal begins. Marilyn reveals her feelings about love and beauty in a lyrical announcement to the performers of who she 'really' is. These are feelings she clings to despite the actualities of her life, and they set up inevitable contradictions as she obliterates her original identity as Norma Jean and creates the figure of Marilyn Monroe. Success comes with a price and the price is a slippage of identity and an engagement with the darker side of the movie business.

Marilyn's search for love persists alongside an ambition born from the memory of her mother's madness and abandonment of her as a child. The Rehearsal Director reflects her failed relationships and marriages, and as the contradictions she is living increase, Marilyn has a breakdown.  Her subsequent relationship with Arthur Miller gives her a chance to reform her life. He nurtures her intellect, and the countryside in which they live reawakens her love of nature. Most importantly, she believes that having children will allow her to be 'normal' and find fulfilling love. When she suffers a series of miscarriages and Arthur also returns to his work, the relationship is doomed.

To counter her sense of not belonging, Marilyn devotes herself to creating a home. Although she is trapped by and subject to the desires and dismissals of the powerful she retains a hold on her touchstones, her ideals of love and artistry. But she is also a Pandora's Box: the secrets of influential men to which she is party must not be released. The forces that have long threatened Marilyn overwhelm and destroy her.

Only after death are Marilyn's mythic and touching childhood selves free to find each other. It is only in death that she is seen with clarity and receives the pity and love for which she has longed. Her death seals her gift to others; she becomes an icon-a screen on which unlimited love and desire and beauty are projected: but of course it is a triumph in which Marilyn cannot participate.

1. Victoria performance


2. Long Beach performance


February 23, 2002 (premiere)

The other 2002 dates were:

February 27
March 8, 15, 25, 29
April 2, 24, 25
May 7, 12


Staatstheater Mainz
Gutenbergplatz 7
55116 Mainz

Note : Gavin Bryars on writing the opera

Gavin Bryars on writing the opera


G is an opera based on the life of Johannes Gutenberg. It was commissioned by the opera house in Mainz for the year of the 600th anniversary of Gutenberg's birth, 2000, although due to delays in the rebuilding of the opera house it has been postponed to 2002.

The libretto is by Blake Morrison, with whom I wrote Doctor Ox's Experiment and who wrote the text for Doctor Ox's Experiment (Epilogue) the concert piece, written in 1988, which was a draft for the opera proper. We have also worked together and continue to do so, most recently on my First Book of Madrigals for the Hilliard Ensemble. In the course of conducting research for the libretto Blake wrote an entire novel on the subject, The Justification of Gutenberg, which was published in August 2000.

There is little reliable information about Gutenberg and no likeness from the period. The only images of him date from more than 100 years after his death, and show him with the well-known forked beard. In reality he was almost certainly clean-shaven! During the greater part of his life he was called Johannes Gensfleisch and only used the name Gutenberg in his last years. For that reason, or at least partly for that reason, we have called the opera ≥G≤, but have given it a lengthy subtitle, in the manner of an 18th century English novel:


Being the Confession and Last Testament of Johannes Gensfleisch, also known as Gutenberg, Master Printer, formerly of Strasbourg and Mainz

This absence of hard information enabled Blake and myself to develop many facets of Gutenberg's life and character in ways which are not precisiely given by historical records. We do use the facts that are to hand - his various court cases and documentary evidence where it exists, but we have felt able to be more inventive than might otherwsaie have been the case.

For example I decided to approach the characterisation of Gutenberg in a, perhaps, unusual way. In developing his character I thought that Gutenberg could be a very different person in Act 1 (in Strasbourg) than in Act 2 (back in Mainz). In Strasbourg I portray him as an opportunist, entrepreneurial, even quite shady, businessman, whereas he becomes more high-minded, serious and idealistic in the second act. A consequence of this is the idea to have a different singer for Gutenberg in Act 2. The two are of a similar vocal type, are be dressed and made up to look the same, but there would be a physical difference. The opera need not always be done this way and it is written in such a way that a single singer could sing the whole part, but it is my ideal and this is the way it will be done in Mainz.

One source for this idea came about when I was in discussion with Hans-Jurgen Syberberg about a possible collaboration in the mid-1980's. In his film of Parsifal (which is mimed to playback of course) Syberberg uses two different people to play Parsifal - the first is a teenage boy who could not possibly have Parsifal's tenor voice, the second is a girl. Both are present in the Epilogue, and I have them sing together, but not always in unison. We also see the moment when the second takes over from the first - in Act 2 scene 2 while the Good Angel sings.

At the end of the opera G - the more mature one - sings alone on stage (with the other G in the shadows) - acco,mpanied only by a group of baroque strings, on stage with him, rather as if he were singing a Bach cantata. This gives a progressive reduction in the forces, reinforced by occasional cadences played by the Wagner tubas, who remain in the pit.

In Mainz the opera will be staged by Georges Delnon ( intendant of the Mainz Opera House) and designed by Atelier Rosalie. Rosalie is one of the most interesting designers working in German opera and theatre, as well as exhibiting paintings. The conductor will be Gernot Sahler.


Gutenberg I (in Act 1, Act 2 scene 1 and part of scene 2 and Epilogue): bass baritone
Gutenberg II (in Prologue, Act 2 scene 2 and Epilogue): bass

Fust: character tenor
Schoeffer: counter tenor
Nicholas of Cusa: bass
Evil Angel: tenor (sung by Fust)
Good Angel: bass (sung by Nicholas of Cusa)
Ennelina: soprano - preferably with quite a pure, early music voice
Beilbeck: low mezzo soprano/ contralto
Christina, Schoeffer's fiancée: soprano (similar to Enneline)
Elliwibel (Ennelina's mother): coloratura mezzo soprano
Jorg Dritzehn: baritone
Andreas Dritzehn: baritone
Claus Dritzehn: bass baritone
Matthias Heilmann: baritone
Judge: bass baritone

(These singers all come from the Mainz company)



2 flutes (both doubling piccolo)
3 cor anglais (1 doubling oboe, 1 doubling oboe d'amore)
1 clarinet (doubling E flat)
bass clarinet
2 bassoons

5 horns (four doubling Wagner tubas)
2 trombones (both doubling alto)
bass trombone

2 percussion

celeste/ harpsichord (one player)

strings (the Mainz string strength)

The composition of the orchestra is intended to be relatively conventional, although the fact that 3 of the orchestra's 4 oboes play the cor anglais gave me the opportunity to reduce the input of an instrument which I dislike and to vary the palette of double reeds (I have available 1 oboes, 1 oboe d'amore, 3 cor anglais, 2 bassoons and one contrabassoon in this department). In addition, of the orchestra's 8 horn players 4 also play the Wagner tuba. As the first horn does not, I decided to write for 5 horns, four of whom double Wagner tubas, with the first horn being a soloist. In addition as both tenor trombonists also play alto trombone I include this slightly old-fashioned instrument too. These factors of instrumentation also give the piece a somewhat historical character.

The opera was completed the piece in June 2001 ready for the February 2002 premiere.

Gavin Bryars.

Review by Della Couling


Mainz is the logical place for an opera on its most famous son, Johannes Gutenberg, inventor of the printing process in the late fifteenth century. The English composer Gavin Bryars and English poet Blake Morrison as librettist are a less logical choice to be commissioned for an opera to celebrate the reopening of the renovated Staatstheater - particularly so, as the libretto was translated into German anyway. (An English spinoff has been Morrison's novel, The Justification of Johann Gutenberg.)

This is the second opera collaboration between Bryars and Morrison, the first being Dr Ox's Experiment. In spite of its hybrid origins, 'G' is largely successful, thanks too to director Georges Delnon and set and costume designer rosalie (sic), who created a late medieval/modern style and some clever visual inventions: the clerks whose livelihood is threatened by Gutenberg's invention protest carrying huge pens like spears.

There are a lot of long monologues, beginning with Gutenberg (Elmar Andree) suspended high up, at first alone. Little is known of Gutenberg's life, so Morrison in novel and libretto used his imagination to flesh out the characters. In the opera there are also a good and a bad angel, who speak to Gutenberg, and heighten the drama, as they underline Gutenberg's personal dilemma of choosing between God and Mammon, and cleverly indicate the inconceivable and uncontrollable revolution unleashed by his invention. It also raises the opera above the level of just another dramadoc with music, a present fashion largely to be deplored.

Given that this is a remote and unfamiliar world, the characters are dressed as bizarrely as creatures from another planet: bald, with white masks (with the exception of the two singers - Elmar Andree and Hans-Otto Weiss - who play Gutenberg). The music is in those long, languid lines those familiar with Bryars' work will recognize. There are allusions to the musical styles of Gutenberg's time, at times discretely underpinned by minimalism. There is one perfect moment, musically and dramatically, when Ennelina (Kerrie Sheppard), Gutenberg's jilted fiancée, now a nun, returns. She is slowly drawn across the stage on a 'paper' boat. Her vocal line dominates; the orchestra supports it, so that the voice is always clearly audible. Those composers who spent their years in the galleys did this automatically. Today, the poor voice, which should be king in the opera house, is left mostly to fend for itself, in spite of what the orchestra is doing. You've proved you can do it once, Mr Bryars; in your next opera, please do it all the time. But all in all, this is a work that, like Gutenberg's, deserves to travel well beyond Mainz.


Opera; libretto by Blake Morrison (after the novella by Jules Verne)

Duration: 2 hours

12 soloists (2 sopranos, 2 mezzos, 2 counter tenors, 2 tenors, 2 baritones, 2 bass baritones); Chorus (SATB); Orchestra: 2 (2), 2 (oboe d'amore, cor anglais), 1 + bass-cl, 1 + contra; 4. flugelhorn.2 + bass.0; harp; electric keyboard, Percussion (3 players) (see Percussion Notes below for details); Strings: minimum (1 amplified)

First performance: English National Opera, London, June 15th, 1998.

Note : Programme notes by Gavin Bryars

Programme notes by Gavin Bryars

Doctor Ox's Experiment

Opera, libretto by Blake Morrison (after the novella by Jules Verne)

Duration: 2 hours

12 soloists (2 sopranos, 2 mezzos, 2 counter tenors, 2 tenors, 2 baritones, 2 bass baritones)

Chorus (SATB)

Orchestra: 2 (2), 2 (oboe d'amore, cor anglais),1 + bass-cl, 1 + contra; 4. flugelhorn.2 + bass.0; harp; electric keyboard, Percussion (3 players) (see Percussion Notes below for details); Strings: minimum (1 amplified)

First performance: English National Opera, London, June 15th, 1998.

The Genesis of the opera

When Robert Wilson asked me to work with him on an operatic version of Euripides' Medea in 1981, I had written nothing for orchestra, nothing for the human voice and nothing for the stage. Although I knew some operas in an academic sense, it was a genre which did not impinge on composers from what was called "the experimental tradition". Indeed the only opera that I had seen live by that time was Gunther Schuller's The Visitation which I saw when I was working with John Cage in Illinois in 1968. I did, additionally, see a production of La Bohème at the Metropolitan Opera during a workshop period about halfway through the 10 months that it took me to write my first opera. Admittedly I looked to certain models during the course of the composition both for orchestra with voice and for dramatic music but essentially I learned about opera by writing one.

Although the planned first performance of Medea in 1982 at La Fenice was cancelled the period working on the opera in Lyon and Paris in 1984 when it was eventually produced was immensely stimulating and satisfying. As I was also one of the five composers for Bob Wilson's abortive CIVIL WarS operatic project, during this period I found myself working with some very fine singers - in Medea Yvonne Kenny sang the title role and a youthful François Le Roux was the Messenger; and in the sketch rehearsals for CIVIL WarS I was involved with people like Donald McIntyre, Hildegard Behrens and, briefly, Jessye Norman. It was a long way from the composer/performer ensembles of English experimental music.

Towards the end of the performances of Medea in France, which were very successful, I felt that I wanted to write more operas and immediately came up with three subjects. These were Flaubert's Bouvard et Pécuchet, Thomas de Quincey's The Last Days of Immanuel Kant, and Jules Verne's Doctor Ox's Experiment. Although I drafted an outline of the dramatic treatment for the Flaubert piece, the first music that I wrote in relation to these possible operas was a piece for solo jazz bass and chamber orchestra called By the Vaar, which was commissioned by the Camden Jazz Festival for the great American bassist Charlie Haden. The "Vaar" is the name of the river in the Verne story and, apart from the piece's autonomous existence, I saw this as a sketch - a kind of backdrop to future music - for the scene by the river where the two lovers Frantz and Suzel pass the day in an apparently innocuous fashion with the sort of non-activity which characterises the town's ethos. This scene comes from the chapter in the original story which had struck me first as crying out for some kind of dramatic treatment employing the sort of expanded time that I had encountered in the work of Bob Wilson. When the opportunity arose for an Arts Council tour with my ensemble in 1988 I planned to write the last scene of the Kant opera but the tenor - either Heinz Zednik or Kenneth Riegel - who I wanted to sing the piece, both of whom wanted to do it, were not available at that time. So instead I decided to move in a different direction and write a work for soprano and ensemble which would involve an overview of the Ox story but viewed from a point when the story had ended - effectively an epilogue to the whole narrative. I worked with Blake Morrison for the first time on this project and the resulting work, Doctor Ox's Experiment (Epilogue), was given 6 performances in this country and 2 in France by Sarah Leonard and my ensemble. Some elements from this epilogue and from By the Vaar appear in the opera itself.

I owe a great deal to Dennis Marks who expressed interest in a film version of the work when he was at the BBC and who, when he moved to English National Opera, had the courage to commission the work. I cannot forget too that it was he who suggested that I might consider contacting Atom Egoyan about directing it. I had seen Atom's film 'Exotica' and I had been struck by many aspects of the film which made believe that he would be ideal as director of the opera, aspects which were reinforced by his most recent film 'The Sweet Hereafter'. I enjoyed his very subtle awareness of the relationship between music and sound design and of the use of music itself. I was struck by the very clever way in which the narrative was slowly revealed and how the viewer's early assumptions were undermined by new possibilities in the story. I was aware that he had successfully directed his first opera in Canada but it struck me too that his way of working in film - with his own ensemble in the manner of, say, the early Orson Welles - is my own preferred way of working in music and is also similar to the way that an opera company like ENO operates.

That the opera now exists is thanks to Dennis's encouragement and vision, to the tenacity of my manager Jane Quinn and to the careful supervision of my editor Sandy Brown.

Jules Verne

For many years I have been interested in the work of Jules Verne.

I have used texts from Verne for vocal works and taken inspiration from his imagery for other pieces over the last few years. For example I have used three different texts from 20,000 Leagues Under the Sea for quite different vocal works: for a section of Effarene (for mezzo-soprano, 2 pianos and 6 percussion) in 1984, for The Black River (for soprano and organ) in 1990 and for The White Lodge (for mezzo-soprano and electronics) in 1991. Similarly his story The Green Ray, especially certain key visual elements, was the inspiration for my saxophone concerto written in 1991, and entitled The Green Ray.

It is easy to be complacent about Verne and about his position in literature. He is often, for example, thought of as merely the father of science fiction, just as Poe was the 'father' of the detective story. But his work is elusive and can be viewed from many perspectives. Certainly French writers, such as Roussel and Queneau, revere his work for the resonance of its word play, but also for its ambiguity and for the existence of many levels of significance. In the recent past a number of English authors, such as Andrew Martin and William Butcher, have written about Verne in a similar spirit. His work has always been important too for those associated with the Collège de 'Pataphysique and there are several remarkable discussions of Verne's Voyages Imaginaires from a highly original perspective in the publications of the Collège. There is one stunning analysis of his Voyages Extraordinaires which demonstrates that each one deals with an obsession for a single direction: travelling north, travelling up a meridian, travelling from the earth's surface downwards, travelling upwards into space and so on.

In addition to the obvious "scientific" aspect of Verne there are many other aspects which are fascinating. For example, as a person Verne was a solidly bourgeois, conservative middle class town counsellor yet in his works we find him supporting other, quite extreme, political positions - anti-colonialism, a sympathy for anarchism. Similarly there is his love of wordplay which appears not only as decoration or wit but rather as a key element in the narrative, something which Roussel was to take to extraordinary limits in his writing where wordplay becomes the generative force for the whole work. At times, in fact, Verne's science could be incredibly factual simply moving one or two steps forward in time, or sideways in geographical location by drawing out the implications of what was already present, or at least latent.

Doctor Ox's Experiment (musical treatment)

I had come across the novella Dr Ox's Experiment in the library of Lord Berners in the late 1970's when I was working on his biography, a project which I had to give up once work started on Medea and which has recently been completed by Mark Amery. Berners' copy was an English first edition that he had been given while at prep school towards the end of the last century. Doctor Ox's Experiment is an apparently straightforward narrative which could be seen to have the concept of 'tempo', relative pace and the play between musical time and chronological time as a structuring device. Its chief protagonist, Ox, is in the anarchist tradition of Verne's heroes like Captain Nemo or Robur the Conqueror, but at the same time displays the single-mindedness of, say, Captain Hatteras (the explorer trying to reach the North Pole who, even in his terminal madness, sought only to walk to the North). In the book Ox himself is a rather shadowy figure and we have chosen to develop his character more substantially with these considerations in mind.

In transforming the story into material for an opera there were many things which needed developing, especially in terms of the characters in the story. Verne is chiefly interested in the town and the ensemble of people within it, who are strikingly uniform. He tells us very little about Ox and his assistant Ygène and both Blake and I felt that they needed to become real characters - Ox is, after all, the title role. There were others who are background figures in the story, Aunt Hermance for example, who we also wanted to give more substance and even to give her some key material.

I also made decisions about orchestration, vocal types, choral disposition in order to clarify what could become otherwise a confusing ensemble piece. I decided that I would associate vocal types with types of characters so that the adult townsfolk, most of whom are male, generally have low voices to reflect their inherent gravity and slowness of demeanour. The lovers, on the other hand are high voices (2 counter-tenors and two high sopranos - we decided to add mirror images of Verne's solitary pair of lovers) reflecting both their youth and innocence. As Ox is clearly from another place and even dimension I wanted his voice to be unlike any of those in the town and at one stage, before I started writing the music, I discussed with Tom Waits the idea of his taking this part. In the event I made Ox a high, agile lyric tenor and Ygène the baritone equivalent and neither of these would be found, in normal circumstances, among the inhabitants of Quiquendone.

Two examples may serve to show how this may be used to clarify the dramatic situation. In the second scene of the opera, after Ox's arrival in the town, to show 3 family groups in their everyday life I let the chorus dissolve into an ensemble of 13 solo voices showing the 3 different levels of tessitura as the voices move from fathers to mothers to youthful offspring. For the chorus in scene 6 where we have workers singing quickly and the townsfolk moving at their usual pace I give the slow music to the female voices and the fast music to the men, the inversion of what would have been the case earlier but showing that the transformation is underway.

I also made decisions about orchestration which would clarify further these distinctions. The orchestra as a whole is not a large one and, apart from the electric keyboard and extensive percussion, is no bigger than an orchestra from the early classical period. There are, however, some instruments which are chosen for very specific reasons. In the case of the oboe, an instrument which I normally dislike, both players double on other instruments: one on the cor anglais the other on the oboe d'amore. As I wanted the scenes with the lovers to have something of the purity of early music they tend to be accompanied with relatively light orchestral textures and so an åold' instrument, the oboe d'amore, can be used for obligato material. The amplified jazz bass is an obligato instrument too, for the love scene, and perhaps an unlikely one but, given the nature of the town, quite apt. This part involves a certain amount of improvisation during the lovers' scene, but when it returns in the epilogue although still amplified it no longer improvises. In addition I use a flugelhorn rather than the trumpet in the brass section for its mellower, less assertive character. There is also a major part for the solo bass clarinet.

This comes about for a number of reasons. In the first place I have a very fine bass clarinettist in my ensemble and I had included this as a solo instrument in the concert Epilogue and to double the solo bass in By the Vaar, adding resonance and sustaining qualities to the legato singing pizzicato of Charlie Haden's bass (and mine in the Epilogue). However, there are both musical and Vernian reasons for the instrument's prominence in the opera. The major incident which demonstrates the extent of the disruption to the town's life, at the end of the first act, comes at a performance of Meyerbeer's Les Huguenots and the section which Verne refers to in his narrative is in Act 4 of that opera. At the beginning of Meyerbeer's next act he writes an extended and quite famous solo for the bass clarinet, which then accompanies the ensuing aria. Meyerbeer was only able to compose this solo because Adolphe Sax had developed the instrument's mechanism beyond its earlier rudimentary state. Sax, of course, came from Belgium, the country where Verne had located with extraordinary geographical precision, although now appearing on no map, the town of Quiquendone.

The climactic disorder towards the end of the first act which occurs with the accelerated and dislocated performance of Les Huguenots, where the townsfolk are drawn into the opera itself, leads to the fastest tempo in the whole work. This is mirrored in the second act where an explosion resulting from Ox's inattention and his struggle with Ygène leads to the loudest dynamics. As part of the explosion, as the sound decays the strings emerge as 24 solo instruments, like debris fluttering to the ground in the aftermath. In each act the climax is followed by an aria for Aunt Hermance, a shadowy figure who, as chaperone for the lovers is the personification of the town's character, but who can also be transformed, like litmus paper in a chemical experiment, to signal the move to a new state. At the end, however, her previous charge Suzel at least has achieved something close to self-knowledge and even enlightenment. In her epilogue she is supported by both the jazz bass playing in the low and middle register and by the bass clarinet playing altissimo.

Doctor Ox's Experiment is dedicated to my mother, who was also present at the last performance of Medea in Paris.

Gavin Bryars

Programme notes by Blake Morrison

Jules Verne, Gavin Bryars and me.

An English translation of Jules Verne's novella Dr Ox's Experiment was first published in London in 1888. In France, the book had appeared over a decade earlier, in 1874, immediately after the most famous of all his novels, Around the World in Eighty Days. Verne's work-rate being prodigious, and the text of Dr Ox running to only a little over 100 pages, it's likely he bashed it out
very quickly. Neither critics nor book-buyers seem to have noticed it much: there are so many other Verne novels to choose from, after all - Journey to the Centre of the Earth, Twenty Thousand Leagues under the Sea, The Green Ray and Five Weeks in a Balloon among others, several of them made famous by film versions. Dr Ox's Experiment remains one of his least-known works.

Certainly I'd never heard of the book when, in May 1988, exactly a hundred years after its publication here, Gavin Bryars wrote to me with the idea of turning it into an opera. At that point, not being well up in contemporary classical music, I hadn't heard of Gavin Bryars either, but I liked the sound of him - and when I heard his music I liked the sound of that, too. We met in a restaurant near Chelsea Bridge, close to the new Observer building where I worked as a literary editor. I'd recently published a book of poems, The Ballad of the Yorkshire Ripper, and it was that which had prompted Gavin to propose a collaboration. Over lunch we discovered we had things in common - Yorkshire childhoods, bibliomania, an interest in sport - and could get along. By the time coffee was served, we had shaken hands on the project.

I had never written a libretto, but I could see at once why Gavin believed Dr Ox could work as an opera. The plot is devastatingly simple: the mysterious Ox, along with his assistant Ygene (Ox + Ygene), offers to instal gas lighting in the sleepy Flanders town of Quiquendone, while secretly planning to pump the gas about the town and to observe its effect on the inert inhabitants; chaos ensues.

The neatness of the storyline; the potential for fast and slow tempo; the presence of an opera within an opera (Oxís great breakthrough with his experiment comes during a performance of Meyerbeerís Les Huguenots at the town theatre); the play of wit and whimsy on the surface, while deeper themes (morality and scientific progress) lurk below; and, most important, the sense of fantasy and magic - all this naturally seemed to lend itself to opera.

(Indeed, we later discovered that Dr Oxís Experiment
had previously been adapted by Offenbach, in 1877, though Verne disliked his version - not lively enough - and audiences evidently felt the same: despite Offenbach's popularity, it closed after a short run.)

Gavin's agent had warned that it was "unlikely that the opera would be produced before 1990", something of an under-estimate as it turned out. And the summer of 1988 was a busy one for me, since as well as a day job at the Observer I had a two-month old baby in the house and 120 novels to read for the Booker Prize. I was eager to make a start, nevertheless, and by the autumn we had produced a concert-piece, in effect an "epilogue" to the novella, for one of the principal females, Suzel.

In November 1988, at the Queen Elizabeth Hall, the soprano Sarah Leonard sang the part, and Gavin played the piece with his Ensemble. The reviews were good, and one critic even suggested, as weíd hoped, that someone should have the courage to commission a full-scale opera.

A long silence followed. Gavin was busy with other works - including a new version (with Tom Waits) of his piece Jesus Blood Never Failed Me Yet, a besteller in 1993. I too had other things to work on, including an abortive musical version, with the composer Howard Goodall, of Wuthering Heights (there was some interest, at first, but once Cliff Richard announced his intention to play Heathcliff, in a rival version, the phone stopped ringing).

Years passed. The Eighties became the Nineties. But Dr Ox didn't fizzle out. Dennis Marks had shown some interest in the project while working at the BBC. And when he arrived at the ENO, he duly commissioned us to produce the opera for the Coliseum.

If there's a simple rule for how librettists and composers should collaborate, we didn't discover it. Gavin's home is in the country near Leicester, mine is in London, and though we met many times during the ten years of Dr Ox's evolution, most of the communing was done by fax and phone.

We advanced by trial and error, with Gavin sometimes chucking out words of mine he knew wouldn't fit and sometimes asking me to write new ones for scenes that proved dramatically or musically richer than we'd anticipated.

David Pountney gave us much useful advice, especially with the dramatic structure of the piece, and Dennis Marks lent steadfast support. But it was never easy to predict what would or wouldn't work musically. For Gavin, the lyrical high point of one section came with the phrase "with the gasworks".

It was chastening for me to discover that banal words can sometimes inspire beautiful music. Not that I set out to create banality, but some of Verne's scenes - a long, nitpicking council meeting, for example - are unavoidably prosaic, and that's part of their comedy.

Other scenes, notably that in which the young lovers, Frantz and Suzel, pursue their slow courtship by the River Vaar, required a more elevated diction, with an erotic subtext.

I used a good deal of rhyme, both with the gentler, more romantic scenes and for the farcical ones. I also tried to contrast two kinds of belief system, tradition on the one hand, science on the other: each has a sweet logic which, in extremis, turns into nonsense. Just as there are two tempos in Doctor Ox, so there are several linguistic registers.

Poets are used to working alone, their only struggle with themselves. A librettist has responsibilities to others - composer, director, choreographer, orchestra, cast - and this means being willing to make cuts and changes to the text, for the greater good of the whole. Far from feeling compromised by this process, I loved the companionship (a welcome break from the isolation of my basement) and the discipline which collaboration imposed.

I'd thought of the libretto as a dramatic poem, but it isn't: it may exist on the page, but it only comes into being when sung. However clear the singing, some words will usually be lost, a reality the librettist doesn't relish but has to accept. Even when every word is heard, the attention of the audience may be elswhere, with the gestures of the performers, or the scenery, or the orchestra. Itís in the nature of opera for words to count for less than they do in poetry. But that doesn't mean working any less hard at them.

In the case of Dr Ox's Experiment, dialogue had to be written that doesn't exist in Verne's text. The descriptive passages of Quiquendone had to be replaced by a chorus of townsfolk. Instead of being psychologically analysed by their novelist-creator, the characters had to be seen (and heard) to interact. Debates became duets. Introspection became song. Above all, the personalities of Ox and Ygene - and the dynamics of a master-slave relationship - were fleshed out.

Gavin and I didn't agree in every detail about the tone and shape of the opera, but there were no serious rifts, let alone slanging matches. One principle uniting us was a wish to honour the spirit of Verne's novella and the questions it raises about scientific and political advance.

At the heart of the story is a battle between the impetuous Ox, who hates the traditionalism and stupor of Quiquendone, and the sceptical Ygene, who comes to believe the townsfolk are better off as they are, without the benefits of his masterís invention.

Ox sees himself as a benefactor. The gift he brings isn't just gas but light, speed, music, democracy, modernity,
self-fulfilment. To Ygene, the gift is a poisoned chalice. It may make the people less repressed, but it also makes them unhappy - and distorts the ancient rhythms of the town.

"Are virtue, courage, talent, wit, imagination - are all these qualities or faculties only a question of oxygen?" Ox believes they are; Ygene disagrees. Theirs is a debate which goes on to this day: are we masters of our own personalities and destinies, or are they determined by factors (genes, chemicals) beyond our control?

Operas are more famous for exploring passions than ideas, but much of the passion of Dr Ox comes from its ideas and we wanted the piece to reflect this.

Though the story of Ox's gas experiment is a fantasy, at least one American expert, Hubertus Strughold of the USAF Aerospace Medical Center, has assessed its scientific plausibility. He suggests that Verne may well have been familar with the pioneer studies of the French physiologist Paul Bert, who observed the behaviour of animals when exposed to pure oxygen under a barometric pressure of several atmospheres (a state of extreme excitation was followed by convulsions, then death). He also makes comparisons with high-altitude sickness, and notes that ozone, rather than oxygen, does induce some of the effects Verne attributes to his oxygen-like gas. That a twentieth-century scientist should give serious attention to Verne's science fiction is a mark of how far ahead he was of his time.

Politically, too, the story of Dr Ox seems pertinent, and even prophetic: when the Quiquendonians, "liberated" by Ox, become warlike and nationalistic it's hard not to be reminded what has happened in Eastern Europe since 1989.

Verne was interested in music as well as science and politics, and it's no coincidence that Ox's gas experiment takes place at the town opera-house.

(No doubt Verne would have known that the first gaslights on stage in Britain were introduced at the English Opera House, at the Lyceum on the Strand, in 1817.) Having written songs and several libretti early in his career, he was regular opera-goer both when visiting Paris and at home in Amiens, and he thought of music as a wonderful stimulant (just like oxygen). Among his favourite composers were Wagner, Mozart, Gounod, Berlioz, Rossini, Verdi, Beethoven and Haydn. Several of his fictions have plots involving music, including the novel Propellor Island and the extraordinary short story "Mr Ray Sharp and Miss Me Flat". All in all, turning one of his books into an opera seems a natural step.

In Quiquendone, where young couples are engaged at least 10 years before marriage, nothing happens quickly. So perhaps it is appropriate that it has taken a decade for the opera of Dr Ox's Experiment to reach the stage. Verne would have seen the justice of that. And being a man who liked to pun, he would have loved the idea of Ox and Ygene being directed by someone called Atom.

Blake Morrison

June 1998.

Doctor Ox's Experiment reviewed in the Sunday Times by Paul Driver

The two-act opera that Gavin Bryars and poet Blake Morrison have adapted from Jules Verne's novella Une fantaisie du Docteur Ox is the first full-scale new work to be premiered at either of the London houses in several years. Its gestation has been long and its eventual production by English National Opera at the Coliseum was deferred from last season to last Monday, but the upshot is something of a triumph. Doctor Ox's Experiment is a real, singable, musically self-justifying opera, with many of the medium's traditional strengths, if always at a deceptive angle to tradition. It is a work of pungent originality that is never "experimentalist" in the way of earlier Bryars works, but makes a surprisingly close approach to mainstream repertoire.

Texts and imagery from Verne have informed various pieces (including a saxophone concerto) by Bryars, whose sources are as diverse as they are arcane. The story of Dr Ox is one of Verne's least known but exists in a beautifully witty (unattributed) Victorian translation. Set in the little Flemish town of Quiquendone, whose trade is "the manufacture of whipped creams and barley sugar on a large scale", a place where nothing has ever happened, everything runs smooth and slow, municipal decisions are infinitely deferred, and courtships last a decade if a day, Verne's satire imagines a situation in which such civic tranquillity is insidiously undermined.

The anarchic outsider Dr Ox, who uses people in his experiments as other scientists use rats, offers to modernise the town, free of charge, with gas-lighting. But, helped by Ygène (Riccardo Simonetti), he instead floods it with oxygen (Ox-Ygène), looking gleefully on as the Quiquendonians become uncontrollably excited, disrupt a performance of Meyerbeer's Les Huguenots, declare war on a neighbouring town over a 700-year old dispute about a cow, and seem to prove his point that human qualities are only a matter of oxygen levels, "mere atoms, like everything else".

The tale's operatic potential, of which the Meyerbeer is an emblem, has to do with its compactness as a fable, its graphic contrasts, its chorus-like involvement of townsfolk, but its appeal to Bryars may also lie in the way that hyper-ventilated Quiquendone sinks into catastrophe as imperceptibly as the great liner in his tape and ensemble piece The Sinking of the Titanic (1969-97) is engulfed by the ocean (to a gurgling of hymn-tunes). And it is difficult not to see an invitation to a composer noted for lugubrious tempi in Verne's evocation of Quiquendone's opera theatre, where no more than two acts of an opera were ever got through in an evening, and "composers would never have recognised their own works, so entirely changed were the 'movements' of the music".

Such sea-change is at the heart of Bryars' idiom. Though his music is all too plainly built from tonal chords, they always sound subtly unfamiliar as though reverberating under water, and the rhythmic repetitions of his admittedly gentle minimalist manner rarely break free from a sort of sea-swell engulfing the texture. The duos, trios and other standard forms abundant in this score - there is a fully-fledged if dreamily lethargic love-duet, with catchy jazz obbligato for double bass (Bryars' own instrument); a sonorous "council chamber" trio for low male voices; a lovers' high-voice ensemble parodying early music to an accompaniment of glittering harmonics - all come touched with the dead-pan irony that is Bryars' peculiar note and apt enough here; while the choruses have a suspended, other-worldly quality right for the phlegmatic Flemings.

In a sense, Bryars has his operatic cake and regurgitates it. His gambits work on the stage even as their element of glacial stylisation threatens to sink the whole dramatic enterprise. His vocal lines are shapely and strong, his orchestral writing often lyrical too, but despite some parallels, this is not the world of Verdi, Janacek, Britten; and though the first act rises to a traditional operatic climax, it is achieved by throwing "opera" - in the form of the riotous audience behaviour at Les Huguenots, complete with Meyerbeer quotation - back in the audience's face. In this production by Canadian film-director Atom Egoyan the spotlight that I took to signify the oxygen flow is trained on us at the start of the interval.

We return, suitably stimulated, to find the Flemings inflamed by erotic passion and their world turned upside down. Whereas the earlier scenes were done with softly swaying choreography and characters who, having emerged from the opening wedding-cake tableau, moved in great white crinolines (designed by Sandy Powell) like chess pieces across the stage, now the people are half-stripped and frantically coupling, enmeshed in the wires of a newly automated whipped cream industry. The second act, at 40 minutes, is half the length of the first for obvious reasons, and the macro-contrasts are well expressed in a production, attractively designed by Michael Levine, somewhat darkly lit by Rick Fisher, that makes the most of such simple means as rope, ladders and falling glitter. Lighting gear nicely serves as the mad scientist's apparatus. There are no mandatory fascists in leather coats, no Mengeles or Husseins to underscore any allegory.

Only the concluding scene, which diverges from the novella and possibly the libretto (it depends on the meaning of a final "Ah!"), moves in the direction of generic modern opera production. One of the love-duettists, Frantz (well taken by countertenor David James) appears to have been killed in the gas explosion (rather muffled in this staging) that puts paid to Ox's scheme; and the other, Suzel (the splendid Valdine Anderson), sings what is in effect a Liebestod over him. Though inertia returns to Quiquendone, in the operatic version life is never going to be the same: we are presented with a landscape of bleak recrimination tinged by a hope of renewal. In the story no one dies, everything is as before, and though the staged denouement has a poetic power, part of me wished that the opera, like the satire, might simply end with a bang and a laugh. The cast, led by clarion tenor Bonaventura Bottone, diabolical as Ox, is altogether fine, though verbal clarity comes and goes in curious phases. James Holmes conducts impressively.

© Paul Driver 1998

Instruments needed in percussion section

Doctor Ox's Experiment (1997)

marimba, vibraphone, glockenspiel, crotales, tubular bells, cencerros, bass drum, tam-tam, sizzle cymbal, suspended cymbal, mark-tree, Chinese bell-tree, wind-machine, timpani (4 or 5 drums)

(3 players)

. . . The fact that the bulk of Bryars' first opera is set in Greek automatically lends the work an air of detachment - in much the same way that Stravinsky's use of Latin gives an objective quality to his sacred music for the opera Oedipus Rex.

In Medea, that intrinsic quality is emphasised by Bryars' depiction of the wronged and betrayed eponymous 'heroine', not as a vengeful hysteric, but as supercool, almost dislocated from all around her. . . Crashing headlong into this detached objectivity was the music itself, hour after hour of sumptuous, tonally gorgeous music.

This music, despite the pulsing minimalistic nature of much of it, is saturated with a sensuous romanticism that derives directly from the most voluptuous music of Richard Strauss. The local lines, endlessly soaring, were of an unremitting lyrical and emotional intensity.

The collision between the two poles - absolute detachment and the most intimate subjectivity - produced a tension that was so intoxicating it was almost tangible. . .

— from Michael Tumelty's review after the concert performance of Medea in Glasgow in 1995

Original version ( 1982, revised 1984).

Duration: c. 3 hours 45'
Opera ( libretto after Euripides.) Direction and design: Robert Wilson. Dedicated to Richard Bernas.

7 soloists (soprano, contralto, tenor, 3 baritones, bass).
Chorus (SATB). Orchestra: 3 (piccolo, alto). 0. 3 (E flat, 2 bass clarinet). 2 (contrabassoon). ; 4.0.1 (bass).1.; 2 saxophones (alto/soprano, alto/tenor); 2 harps, piano; timpani + 5 percussion (see Percussion notes below for details); strings (no violins; 10 violas, 8 or 10 cello, 4 or 6 basses

First performance: Opéra de Lyon, France, 23 October 1984. Subsequent performances at the Théâtre des Champs-Elysées Paris (co-production: Opéra de Lyon, Opéra de Paris, Festival d'Automne).

Revised version (1995).

Duration: c. 2 hours 45'

First (concert) performance of the final version: Tramway, Glasgow, 3 November 1995.

Note : Medea by Gavin Bryars

Medea by Gavin Bryars


Medea is my first opera and was first done in collaboration with the American director/ designer/ writer Robert Wilson. Medea was the first work in which he used something other than his own writing for the text. He told me that when he had once been asked what works other than his own he would be interested in directing he had replied Medea, Parsifal and King Lear. In due course he was to do all three, and had already started work on a production of Parsifal which did not come to fruition, but Medea was the first. The project was a leap in the dark for me given that I had written virtually nothing for the human voice, nothing for orchestra, nothing for the stage, and my sole experience of opera had been to attend a performance of Gunther Schullerπs The Visitation in Illinois in 1968!

Gavin Bryars and Robert Wilson (photo by G. Ansellem)

He had got in touch with me in 1979 when I performed in the Festival dπAutomne in Paris. In the event he was not able to come to the performance but I learned subsequently that he had asked Benedicte Pesle to let him know her view on the possibility of our working together. He eventually called me and we finally met in April 1981 when I spent a three days talking with him about his work and looking at video recordings. It was at that time that he asked me if I would be interested in writing the music for his production of Euripidesπ play, which he had already adapted and for which another composer, the late Arthur Russell, had written music for a workshop in Washington. He was not happy with this music and wanted to see what I might do. He said that, for the play, he would like to have the possibility of certain passages being sung rather than spoken - some of Medeaπs own speeches, perhaps the chorus and so on. I recorded some draft ideas, which he liked, and then he asked me to look through the whole play and see what lines could be sung and what had to be spoken. By August of 1981 what had started out as a play with incidental music and some singing, had become an opera. However, the original planned date for the play production, for La Fenice in Venice, September 1982, still stood and so the music had to be written by June 1982....

Even before a note was written it was necessary, for reasons of budgets and planning, to decide questions of orchestration and the role of the choir. From some rapid research into what we know (or at least, knew then) about ancient Greek music I made a number of choices. In the first place I used few brass instruments (no trumpets or tenor trombones) and the references to rudimentary xylophones in the literature encouraged me to use a large body of tuned percussion (in the event 5 players, plus timpanist). Then I made the decision to have no violins and have strings only from violas downwards, a decision which has had consequences for subsequent works and for the ultimate formation of my ensemble. In addition I replaced the oboes, my least favourite instrument, with saxophones. Berlioz talks of ways of using the viola - the top string, for example, being associated with åreligiousπ or the åantiqueπ. In Act 3 scene A (in the version rewritten for Lyon, see below) the long melody on the violas is entirely on the G and D strings, following Berliozπ recommendation! Saxophones were chosen because of their division into vocal families corresponding to the ranges of the human voice and used in the first instance to support the chorus. My original idea for the chorus was to use only altos and tenors, in the registers where the voices overlap, but this proved impractical (from an administrative point of view...). The very low first aria of the Nurse for example, in the original version, draws on the hypothesis that, in the past, voices were accompanied from above rather than below.

The nature of the language used in the play affected the music of course. The greater part of the opera is in the original Greek of Euripides. For the Nurseπs first aria the rhythm of the music was derived almost entirely from that of the spoken Greek. Later (for example in Act 2 scene A) the Greek language affects rhythm but chiefly within recitative. But there are other places where a metaphorical approach to language is introduced - in Act 3 scene B for example. In Euripidesπ original play at this point there is a hymn to Athens, which is in a poetic style quite different to the rest of the play. So I wrote this chorus in a quite different musical language from the rest of the opera, a language which is much more like the conventional operatic choral set-piece than the way the chorus is used elsewhere in Medea - to commentate, to interpret, to interject.


Acts 1 and 2, plus Act 4 scene C (written in New York) and Act 5 scene C (the ending) were performed, with two piano accompaniment at the end of February 1982 following rehearsals at City College, New York with a mixture of students, semi-professional singers and singers, notably Wilhelmina Fernandez, who was originally cast as Medea. This gave us the opportunity to see how the piece worked, and to give guidelines for the operaπs completion. Although I left New York prior to the public performance, but at the end of the rehearsal process, I had learned a great deal about the practicalities of writing for the human voice and about how to approach opera. I also managed to see only my second live opera - a production of La Bohème at the New York Met. I had the chance to work in detail with individual singers, like Wilhelmina and others - the virtuoso scene for the Messenger (Act 4 scene C) which uses a very wide range (two and a half octaves, including substantial falsetto) was made possible by being able to work directly with the baritone singer in New York. The other substantial bonus was the opportunity to spend a good deal of time with the conductor Richard Bernas, who was my choice for the opera, who eventually was to conduct all the performances in Paris and Lyon, and who is the workπs dedicatee. His knowledge about opera is encyclopaedic and ultimately I owed learned more about opera from him than from any other single person.

Following this I returned to England and completed the writing of the opera by the end of June 1982. At the beginning of July there was a very long meeting in Venice at which we analysed precisely the state of the opera and its readiness for production. It was concluded, after a very difficult and occasionally stormy meeting that the production was not ready (we did not get round to talking about the music) and so the performances were cancelled.......


Jean Pierre Brossman at the Opéra de Lyon committed himself to producing the opera, in partnership with the Opéra de Paris and the Festival dπAutomne. In the course of rehearsing the opera, however, a number of things were changed. In the first place Act 3 scene A, where Medea has a friendly exchange with Aegeus - the only scene in the whole opera where she is relaxed, had originally been done in a semi-jazz idiom (Aegeus was to have been played by a black American baritone, to partner Wilhelmina Fernandez). In Lyon, however, Medea was played by Yvonne Kenny (white and Australian) and so the jazz idiom was slightly ridiculous. I therefore re-wrote this entire scene during the rehearsal period. Further, those parts of the opera which still used the spoken word were in English and the view was expressed that, with an opera directed by an American, composed by an Englishman, conducted by an American, with performances only in France, surely this should be put in French. I agreed, but this also meant that those lines which were sung in English had to be sung in French, and the vocal line re-written...

Shortly before the dress rehearsal it was realised that, scenes having been rehearsed individually rather than in sequence, there was both a scene change and costume change before the final scene, and so extra music had to be written an hour or so before the dress rehearsal, copied and put in the players parts before the second half started. Given the extreme length of the opera I held a meeting with the orchestra, in the first instance to apologise but also to discover who would not object to playing extra music. All four horns, the saxophones, the tuba, the bass trombone and percussionists put up their hands. This 90 second interlude remains in the final version and is a testament to their fortitude!! I spoke with the principal horn, Pascal Pongy, after writing the music at speed (all of it involving transposing instruments!) and said that I hoped the music ≥was playable≤. His words, which I treasure, were ≥Gavin, le moment où elle est écrite, la musique est toujours jouable≤.

Two scenes were cut from the opera before the first performance 3 days later, one for orchestra alone, the other for chorus, due to problems with staging. The lengthy prologue, which was essentially a series of tableaux for with Bob had said he needed little music, proved to need more material and so he collaged texts by the East German writer Heiner Müller on top of the music. This entire prologue was cut when I revised the opera in 1995 and the cut scenes reinstated.

The opera was in 5 Acts, each one with four scenes (except Act 5 which had 3) preceded by a four scene prologue. In the event Act 4 lost one scene (with chorus) and Act 5 lost one (with orchestra).

There were 11 performances - 6 in Lyon and 5 in Paris, all of which played to full houses. Working with the orchestra in Lyon, which was very highly motivated and comprised mostly young players, was an immense pleasure. By contrast the orchestra in Paris was, with a few notable exceptions, aggressively negative and it was chiefly the fact that they were baffled by the extreme (New York) sarcasm of Richard Bernas that the battles were won........


Nurse: Marie Marketou

Tutor: Frangiskos Voutsinos

Medea: Yvonne Kenny

Creon: Steven Cole

Jason: Louis Otey

Aegeus: Pierre-Yves le Maigat

Messenger: François Le Roux

Off-stage soprano: Liliane Mazeron

Conductor: Richard Bernas

Structure of Original version


Scene A (full cast) tableau of daily life in Colchis -mid morning, olive grove

Scene B (full cast, off stage chorus) tableau of travel through the Anatolian mountains - midnight, rocks

Scene C (full cast) tableau of death, Medea stabbing her brother - midday, dais

Scene D (Medea, Nurse, children, men) tableau of departure, Medea throws overboard pieces of the body to delay pursuit - - dawn, ship

Act 1

Scene A (Nurse) aria in which she tells of the deplorable things that have taken place, and expresses her anxiety about what will happen now that Medea has been deceived and abandoned by Jason.

Scene B (Nurse, Tutor, children) The Tutor joins in and tells further the news that Creon has banished Medea.

Scene C (Nurse, Tutor, children, Medea off stage) Medea is heard

Scene D (Nurse, Medea off stage, Off-stage soprano) Medea sings of her suffering and of being torn between her love for her children and her anger against Jason

Act 2

Scene A (Chorus, Nurse, Medea off stage) The chorus offer their support to the Nurse and to Medea

Scene B (Medea, chorus) Medea describes to the chorus the condition of women in marriage and demands that they be silent.

Scene C (Creon, Medea, Chorus) Creon orders Medea to leave Corinth. Medea persuades him to allow her a delay before leaving. To the chorus she speaks of her initial plans for vengeance.

Scene D (Medea, Jason, Chorus) Medea reminds Jason of what she has done for him, he defends his actions, offers material support, justifies his new liaison with Creonπs daughter. The chorus judge him to be wrong.

- interval -

Act 3

Scene A (Aegeus, Medea, Chorus) Aegeus is unable to produce children and has consulted the oracle. Medea tells him of Jasonπs treachery, her banishment and promises to cure him of his sterility if he gives her a safe haven in Athens. He agrees.

Scene B (Medea, chorus) Now assured of a safe refuge Medea sets out her plans for vengeance, not on Jason but on his new wife and her children. The chorus sing a hymn to Athens, implying their disapproval.

Scene C (Medea, Jason) Medea feigns submission and asks that her children might remain in Corinth rather than share her banishment. She gives a present for his bride.

Scene D (Jason, children - instrumental only) The children bear the gifts - the chorus know that they will lead to death.

Act 4

Scene A (Chorus) The chorus sing of the impending disaster

Scene B (Tutor, Medea) The Tutor says that the gift has been accepted. Medea receives the news sadly. Knowing of the imminent death of her children she is torn between love and pride.

Scene C (Messenger, Medea, Chorus) The Messenger is shocked by Medeaπs joy when he tells of the agony of the princessπs death, the veil which sticks to her skin and burns her, of Creonπs death when he tries to help his daughter.

Scene D (full cast) Scene of the childrenπs murder. Break in the action, seated at a long table the cast speak directly to the audience.

Act 5

Scene A (Orchestra, Off-stage soprano) The death of Creonπs daughter, depicted by a puppet

Scene By (Jason, Medea, Chorus) Jason, seeking his children, learns of their death. Medea refuses even to let him bury them. She escapes in a chariot bearing them to the sanctuary of Hera.

Scene C (Jason, Chorus) Jason appeals to the gods. Corinth is in flames.


Act 4 Scene A and Act 5 Scene A were not in the 1984 performance, though included in the final dress rehearsal.

The Prologue and Act 4 Scene D were omitted from the revised version in 1995, Act 4 Scene A and Act 5 Scene A were reinstated, and Act 1 Scene A was re-written.

Letter to Jane Quinn about a DAT recording of the dress rehearsal in Lyon.

Billesdon March 18th 1994

Dear Jane,

Thanks for your call asking me for a tape of Medea. I enclose 2 DATs (I assume that DAT is suitable) being the two halves of Medea. This was recorded at the final dress rehearsal before the first performance at the Opéra de Lyon -this explains why there are many camera clicks at the moment where scene changes happen. I have put a track ID at the beginning of each act rather than for each scene.

There are a few small differences between this and what was actually performed 3 days later. The choral scene which begins Act 4 was cut because the choreography of the dance which accompanied the scene was so banal (like a vernacular taverna dance), and the orchestral music for the opening of Act 5 was also cut. This was because the scene used a puppet falling from a cliff in flames and it simply looked ridiculous. I think the orchestral music is actually not bad (Richard Bernas, the conductor, thinks it is one of the best things in the opera!). In addition, I have taken out the whole Prologue from this tape. If ever Medea is done again I would re-write the Prologue. Bob Wilson had said that the Prologue was to be a series of visual pictures with only occasional musical fragments - and that was what I did. In the event, the pictures went on for so long (about 25 minutes) that the music assumed an importance disproportionate to its acoustic merits. I have left in, however, the lengthy spoken scene at the end of Act 4, where there is a sort of academic discussion about the piece by the cast (this was a particular theatrical device which Bob included, but which was under-rehearsed and, here, does not work very well at all). In the actual performance it was much tighter but I decided to leave it on the tape. Here, in fact, it was the first time that the spoken text and the music had been put together and Richard Bernas decided to play all of the music, even though the text was much shorter than the music. If you want to pass over it, fast-forward to the ID for Act 5.... The gaps between sections, where they happen, are due to temporary problems with scene changing and did not happen in performance. It was curious to listen to it again after all this time and I enjoyed quite a lot of it. The language moves between Greek and French - originally between Greek and English but I had to re-write several sections in French shortly before the first performance, and the first scene in Act 3 was completely new for that occasion. As you know, it was originally to be done for La Fenice in Venice in September 1982, but was cancelled at a late stage. For that production Medea was to have been sung by Wilhelminia Fernandez and the opening of Act 3 was to have been a relaxed pseudo jazz style piece with two black singers. When it was recast for Lyon and Yvonne Kenny (a white Australian) sang the part this scene did not work at all and so I wrote a new one, directly in French. Two of the cast are, in fact, Greek (the Nurse and the Tutor). Yvonne Kenny and the French singer (François Le Roux) who sang the part of the Messenger have now become quite famous. The full cast was: Yvonne Kenny (Medea), Louis Otey (Jason), Stephen Cole (Creon), Pierre Yves le Maigat (Aegeus), François Le Roux (Messenger), Maria Marketou (Nurse), Frangiscos Voliotis (Tutor), with the chorus and orchestra of the Opéra de Lyon conducted by Richard Bernas.

Gavin Bryars.

Note about the revisions made for the 1995 concert performance.


For the concert performance of Medea by the BBC Scottish Symphony Orchestra the following modifications were be made. With these changes I feel that the final state of the work has been reached. It was altered several times up to the first performance in Lyon in October 1984, some of these were for the good others not so but following the Lyon and Paris performances I have reflected at length and the Scottish performance gave the opportunity to allow the piece to find its best shape.

1. The entire Prologue is cut.

The opera starts, therefore, with Act 1 scene A. The raison d'être of the original prologue was to give a sequence of visual tableaux telling the story of Medea up to the beginning of Euripides' play. However, as Bob Wilson realised afterwards (sic) the opening aria of the Nurse in 1A tells exactly the same story - admittedly using words! The way Bob had spoken with me about this prologue originally was to say that the visual pictures were everything and therefore no music was necessary, or at best only the most unassertive background music should be used. In the event more was needed, the music (composed to be incidental) was thrust into a foreground space it was ill-equipped to occupy. Bobπs solution was to superimpose texts by Heiner Müller, all of which has now gone leaving only the original Euripides material as far as libretto is concerned. Now the opera begins with the spoken Greek monologue of the Nurse with a few bars of music preceding her newly written aria, a female chorus is superimposed on her singing and the orchestration is modified.

2. Act 4 Scene D is cut except for the opening chorus (vocal score P104-5)

This is the scene where Medea murders her children. In Lyon and Paris it was done in such a way that the murders happen off stage and the scene comprises a spoken "play" with all the singers in the opera appearing in everyday dress seated at a long table reading fragments from news reports about things like child abuse and so on. This scene was created very hurriedly in Lyon and, for the music, again I was to provide a quiet background (which ironically the conductor Richard Bernas liked very much!). Its pace is that of Ives The Unanswered Question, using very slow quotations from many different operatic versions of Medea.

In the revised version the music now goes from the end of the chorus at the beginning of 4D directly to the "kneeplay" before 5A leading into 5A itself, an orchestral interlude, which would serve exactly the same function as the cut 4D i.e. an intimation that events were happening off-stage.

3. Act 5 Scene A is reinstated.

This orchestral interlude, although played in the final dress rehearsal (October 20th 1984) and therefore on the tape recording of the rehearsal, was cut. This scene had not been rehearsed until then and a puppet, used to show a figure - Creonπs daughter - falling in flames from a cliff, did not work.

4. Act 4 Scene A is reinstated.

This scene was cut from the production because the dance which accompanied the choral singing was not interesting, though again not looked at fully until the final dress rehearsal. It is a scene for divided male chorus and was given extensive music rehearsals. This, again, is present on the 1984 dress rehearsal tape.

5. At the beginning of Act 3 there was a spoken text preceding the music which begins the act. This has been cut. There was also spoken text during an orchestral fermata in Act 3 Scene A interrupting the duet between Medea and Aegeus. This text (which is not in the score) is cut.

This produces the following structure:

Part One

Act 1

Scene A Nurse, Chorus

Scene B Nurse, Tutor

Scene C Nurse, Tutor (Medea off stage)

Scene D Nurse, Tutor (Medea off stage)

Act 2

Scene A Chorus, Nurse (Medea offstage)

Scene B Medea, Chorus

Scene C Medea, Creon, Chorus

Scene D Jason, Medea, Chorus


Part Two

Act 3

Scene A Aegeus, Medea, Chorus

Scene B Medea, Chorus

Scene C Jason, Medea, Chorus

Scene D

Act 4

Scene A Chorus

Scene B Tutor, Medea

Scene C Messenger, Medea, Chorus

Scene D Chorus

Act 5

Scene A Orchestra

Scene B Jason, Medea, Chorus

Scene C Jason, Chorus

This alters the running time as follows

Part One 1 hour 10 minutes

Part Two 1 hour 25 minutes.

(total about 2 hours 35 minutes)

It is possible to break up Part Two into Act 3 (about 35 minutes) and Act 4/5 (about 50 minutes) making a three part opera - particularly if two intervals were needed.

Gavin Bryars.


Programme note by Sarah Walker.


First performance of the final version: Tramway, Glasgow, November 1995.

Nurse: Patricia Bardon contralto
Tutor: Gidon Saks bass
Medea: Majella Cullagh soprano
Off-Stage soprano: Eileen Hulse soprano
Creon: Iain Paton tenor
Jason: Richard Halton bariton
Aegeus: Nicholas Folwell bariton
Messenger: David Barrell baritone

Conductor: Martyn Brabbins

The enormous success of Gavin Bryars' five-act opera Medea, produced in 1984 by Robert Wilson at the Opera de Lyon, was the cause of consternation to certain factions. . .

Instruments needed in percussion section

Medea (1982, rev. 1984)

Xylophone, marimba, vibraphone, glockenspiel, tubular bells, bass marimba, 4 tuned gongs (B a 9th below middle C, C sharp below middle C, D below middle C, E above middle C), crotales, 7-12 roto-toms, tam-tam, orchestral bass drum, tambourine, 5 temple blocks, 4 woodblocks (Japanese preferred), 4 suspended cymbals ( ride, 14" Turkish, Chinese, sizzle with good sustain on rivets), cabaça, chocolo, 2 triangles, 2 maracas, 2 flexatones.

(5 players, in addition to timpanist)