Arts: Libretti, equality, fraternity: Composers and librettists seldom get on. Michael White reports from Kent on a bid to change that / Opera

SURRENDER is the sting in the tail of creativity. You have a child, it grows up, hates you, and leaves home. You paint a picture, sell it, and it's never seen again. Worse still, you write a text and someone turns it into music - a creative trauma that doesn't necessarily wipe your work off the face the earth although you might well wish it could if you can't live with the result.

A few years ago I wrote an opera libretto for a reputable composer, and I remember the cold, quasi-paralysis that crept down me like hemlock as I read, for the first time, the finished score: not by any means a bad score but a brutal massacre of my defenceless words. Or so I thought; and there ensued a Scene, the like of which can only happen between composers and librettists with 400 years of operatic precedent to draw on.

Collaborations in opera have always tended to be volatile affairs, distinguished by a handful of fortuitous marriages - Mozart and Da Ponte, Strauss and Hofmannsthal, Verdi and Boito, Stravinsky and Auden - but more failures, where one side (usually the librettist) lets the other down.

There has always been a disparity of status between composers and librettists. Until the mid- 18th century the libretto generally took precedence, with the same revered texts set repeatedly by composers whose work was considered comparatively ephemeral. Then the composer emerged as tyrant and what passed thereafter for 'collaboration' was rarely more than an arm's- length service where the writer delivered his text and the composer took it away to adapt as he pleased. Even in modern times, the instances of composer and librettist working intimately together have been rare. Britten and Ronald Duncan shared a desk when they wrote The Rape of Lucretia, but Britten was very much the boss - as he was with Montagu Slater, who was so unhappy with the changes made to Peter Grimes that he had his original, unadulterated libretto published independently of the score.

You might conclude that some guidance in the processes of collaboration would be useful; and, 400 years late, guidance of a sort has arrived in the form of Opera Lab, a project that straddles the objectives of a creative marriage-broker, country- house party and salon. For 10 days last month a group of composers and writers gathered in a farmhouse in Kent. They were guests of Opera Lab, a private trust, invited on the recommendation of theatres, television companies and teaching institutions; and they were there - well, they were there to find each other. To make relationships.

Some had come with specific plans for an opera that only needed the input of another mind - which had words but no music, or music but no words. Some had come to experiment and assemble ideas on the spot. Others were there simply in search of the necessary confidence. They ranged from the playwright David Benedictus flourishing ink-wet sketches for a piece about the man who invented Monopoly (opera? musical? he hadn't decided) to Cindy Oswin, actress/author of one- woman shows, who had been mulling for seven years over the legend of the Sumerian sky goddess Inanna and hadn't produced a word ('I know it needs music, somewhere'); from the composer Julian Johnson with ideas on Christina Rossetti, to composer Rhian Samuel working on comic scenes with Robert Young, an engagingly oddball protege of the Royal Court Theatre.

The support came from a group of singers (old hands at new music such as Jane Manning and Fiona Kimm), pianists, designers and a director who were available like house-doctors, more or less around the clock, to try out whatever got written. There was also a small tutorial team under the composer Robert Saxton to offer technical advice. But otherwise there was no pressure to produce, beyond an unspoken element of competition that generated lunchtime queues at the photocopier (the regime was writing in the morning, showing in the afternoon). The unanswered question hanging over the whole enterprise was: what happens next? Could you really expect these fragments, nursed into existence in the unreality of a creative greenhouse, to live on outside it?

The very idealism of Opera Lab makes it hard to believe in. As a marriage market, it doesn't offer much choice - maybe half a dozen candidates for the creative partner of a lifetime - and as an introduction to the world in which opera gets made, it's pretty soft. In the real world you don't have singers, directors and designers on call to help cultivate the great idea you had at breakfast.

But maybe you should. Susan Benn and Nicky Singer, Opera Lab's founders, admit to being 'unashamedly ambitious'. 'I hope,' says Singer, 'that we're creating a precedent here for how things could be. We run a lab like this for screenwriters, and the support we've had from the film industry proves that this approach is viable - not least because it saves time and money by ironing out problems early on, before you're on location and find you have to start rewriting. I want the powers-that-are in opera to see what we're doing and realise it could work for them, too.'

In fact, they are already seeing. Last month's Opera Lab pulled in Jeremy Isaacs, Elijah Moshinsky, Sir John Tooley, Nicholas Payne (the new opera director at Covent Garden) and Ian Ritchie (the new general director of Opera North), which would be an impressive guest list for any Kent farmhouse. It clearly has credibility. And it can claim results. Last month's was the second Opera Lab. The first took place in 1992, and fertilised a seedbed of ideas and relationships that fed through to this year's Royal Opera House Garden Venture and ENO's Contemporary Opera Studio projects.

Talking to the Opera Lab participants, the immediate gains are much as you'd expect: acquiring - at an accelerated pace - techniques of writing for the lyric stage. Composers learn that singers need to breathe occasionally; writers learn to cut their imagery to the core and discover, the hard way, why Andrea Maffei said that libretti must be miniature dramas that the lens of music magnifies. More broadly, they learn about what it takes to break out of their customary closed existences. 'Which is why,' says Nicky Singer, 'it's absurd that they're expected to just get together and create by divine order. Here are two people who are normally in control of their medium; they have to learn how to deal with someone else's input.

On the back of that, we're aiming to subvert the rule that says the composer is God, and give the writer the right to collaborate as an equal. I think that's important.'

W H Auden thought otherwise. His essay on the subject in The Dyer's Hand insists that libretti are private letters to the composer, their sole function to suggest a melody. That done, they are 'expendable, as infantry to a Chinese general: they must efface themselves and cease to care what happens to them'. Opera Lab may be about to prove him wrong.

Details of future Opera Labs: 071-839 5677.