What is unfurling is a scene from Monteverdi's Orfeo. This could be from one of those trendy po-mo productions in which baroque formality is juxtaposed with the debris of contemporary urban life. In fact, it's a rehearsal for a new Kent Opera staging, and the gym is not scenery, simply a rehearsal room. When the production reaches the stage, there will be no inner-city flotsam. Still, Orpheus remains a figure for our times, whose natural habitat is opera. Over the coming weeks we have the chance to compare and contrast different versions of the tale from the first two centuries of operatic history: Kent Opera's Orfeo (premiered in Mantua, 1607); a new English National Opera production of Gluck's Orfeo ed Euridice (Vienna, 1762); and, in both a new recording, and a pair of concert performances by the London Philharmonic, Haydn's last opera, L'anima del filosofo (written for London in 1791 but unperformed in the composer's lifetime).
By definition, myth has no urtext, but, in its Greek form, the legend of Orpheus had its hero dismembered by female followers of the god Bacchus for the sin of spurning womankind. His head, the legend added, went on singing as it floated out to sea. In one version of the libretto which the poet Striggione provided for Monteverdi, the opera ends with Orpheus fleeing from the Bacchantes. It's not known if the composer ever set this episode to music; the surviving score offers a new finale, in which Apollo ushers Orpheus heavenwards, where he once more encounters Eurydice's loveliness "in the sun and in the stars". For modern audiences, this is anticlimax, as is the ending of Gluck's opera, in which Amor (aka Cupid), having restored Eurydice to life, reunites her with Orpheus to the sound of general rejoicing.
It's not something that much bothers Tim Carroll, who directs Kent Opera's new staging of the Monteverdi: "The myth is so succinct in its representation of the power of music, encapsulated in the idea of Orpheus being able to move stones and charm animals. Then we come to Monteverdi's opera, which asks `Can music defeat death?' and it answers, `In a way, yes: not physically, but in another way.' Monteverdi's ending has had a bad press, suggesting that it's a Hollywood-style happy ending tacked on because the `focus group' didn't like the original. But, in the opera, the Bacchantes would be a distraction. It's not relevant to the themes of the piece that a group of women should come in and tear Orpheus to pieces. That's a myth about denying the sexual imperative; it's not what interests me here, nor did it interest Monteverdi. His point was to emphasise what music achieves."
Meanwhile, at ENO, director Martha Clarke has been grappling with Gluck's "problem" ending. For her, "The myth is more interesting and I had a day of being enchanted with the idea of going against the music, of making the ending dark. Then I realised it was pushing it too far and getting frightfully arty." Jane Glover, conductor of the ENO performances, goes further: "I don't think Gluck liked the happy ending any more than the rest of us. You have a feeling that, after the coup de theatre where Amor returns Eurydice to life, the composer hands the reins to the choreographer and says, `OK, you finish it', and that's why the opera ends with a sequence of dances."
Surprisingly, it is Haydn's opera that comes closest to the myth: Eurydice dies; Orpheus is poisoned by the Bacchantes. As Christopher Hogwood, the conductor of the new Decca recording, remarks: "It must have been a real challenge for Haydn, writing his first opera for London, the world's musical centre, to end it with both main characters dead, a whimper, a long drum roll and lots of D minor. In fact, some people have wondered whether the opera is really finished, whether perhaps we're missing a fifth act with a happy ending."
Not that Hogwood is suggesting that their happy endings somehow compromise the operas by Gluck and Monteverdi: "Opera lives from minute to minute, and you can go the full way emotionally with either opera while knowing that, at the end, there's going to be this little rescue act by Apollo or Amor." Some recent stagings of the two works have enacted the Bacchic dismemberment in mime while the music offers the lieto fine that operatic convention demanded. Hogwood believes that Haydn's opera resists any such stylisation - one reason why it remains a rarity: "Modern producers find it hard to accept Haydn's conventions. You can't transpose a Haydn piece like this to Harlem, you can't have the heroine high on drugs, you can't have them all in boxer shorts and skateboards. So it poses problems for producers who can't take theatrical ventures on their own terms, and very few can."
Neither Kent nor ENO are offering modern-dress stagings, nor are they striving for cod baroquerie or Arcadian pastoralism. As Tim Carroll says, "Monteverdi's opera is a piece about mankind, and it requires a sense of `ancientness' only inasmuch as it provides a feeling of continuity, that this has always been the case. It's a question of trying to create a world that has the texture of being connected to the earth, of happening in a real place to real people. That's the genius of this opera: it's absolutely emblematic, and at the same time absolutely human."
It's also a work that draws its power largely from its vocal lines. Although he's an early music specialist, Kent Opera's conductor John Toll is less concerned with authenticity than with expressivity: "I wanted voices that were - can I use the word? - unspoilt. Singers get precious about the voice as an instrument, and so lose an intuitive, spontaneous response to text. The expressive force of a voice is generated by text, and that's particularly essential here, where Monteverdi is searching for a natural way of imitating speech through music. A highly stylised and oratorical speech, no doubt, but using music to boost its emotional force."
ENO's Orfeo is the counter-tenor Michael Chance, perhaps the world's leading interpreter of the role - at least in the original 1762 version of Gluck's opera, written for the castrato Guadagni (Gluck later revised the part for high tenor, while most modern stagings compromise on a mezzo in drag). Nor, says ENO conductor Jane Glover, are they "corrupting" the original text with any of the music from later versions: "If you're going to do the 1762 version, then do it. People may say, `Where's that lovely flute part? Where's the Dance of the Blessed Spirits?' But forget that those things have ever been there, and the piece unfolds in an incredibly tight way. It's so strong, we're running it without an interval."
Martha Clarke, best known as a choreographer, is using a complement of dancers, not as ornament but as what she calls "the colour-field of Orpheus' emotional journey".
"The opera is tremendously pure, so the production is simple and stark. My own direction in choreography is towards the theatrical gesture that emerges from natural movement, rather than anything based on traditional technique. What I've been trying to do is to find a common language, to merge the vocabularies of singers and dancers, so that you look at them and you can't tell who is doing what: the dancers are part of the community of Orpheus."
The community of Orpheus is a large one, embracing all opera, and all singers. Opera may have dismembered the myth for its own purposes but, as these three variants show, that dismemberment can never stop Orpheus from singing.
Monteverdi: 7.30 tonight, 3pm Sun, Theatre Royal, Margate (01843 293877), 7.45 Thu, QEH, SBC, London (0171-960 4242) and touring. Gluck: from Mon, 8pm ENO, London Coliseum (0171-632 8300). Haydn: LPO concert performances, 7pm 20, 25 March QEH (as above); Hogwood recording on Decca / L'Oiseau Lyre (452 668-2)