And indeed Stephen Oliver (1950-92), the most prolific British opera composer of the last century, first came to notice with his version of The Duchess of Malfi at the age of 21....

Gosh. I have a suspicion this is correct in a generalized way, but only because great poetry (and surely this is what WS's plays are, in addition to their dramaturgical qualities) does not usually make for grateful sung text. Why does anyone, composers especially, imagine that a coherent bit of drama or poetry will make for a coherent libretto? It's nonsense. Britten, Mozart - and da Ponte and Boito - realised this.

Oops, I missed that. Er... Turfan Fragment stole my glasses!

I don't think "mediocre" is really the right word though. I think of it more like this: a composer has to feel that there's something unique he or she can say about the text, look at it in a subtly or radically new way, deepen or widen the viewer's understanding of it, rather than "add to" or "complete" it. This is very difficult to do with Shakespeare because it's all been interpreted and produced to heck, but presumably Humphry Searle thought he could do it with Hamlet and it seems he couldn't, though I don't know that work myself. Nevertheless I don't think it's impossible. I can't think of any examples of composers actually succeeding in doing it, but Edward Bond's Lear is an example of the kind of reinterpretation, but on a musical level, that fulfils the conditions I've outlined. It does tell us something about Shakespeare's play, specifically about its characters, and much more besides.

Yes, r, and plays like the Marowitz Hamlet, where the play is de- and re-constructed. Sadly the Searle, despite the modernity of his score, was a very matter-of-fact paring-down of the text, like a shortened school version, and the production was very traditional, resembling theatre productions of the piece twenty years previously. I saw the premiere production of Bond's Lear, and know exactly what you mean. I've heard the Reimann, but never seen it.

To return to Stephen Oliver for a moment: one of his forty or so operas was a Timon of Athens. 

It's had four productions though (Munich, San Francisco, Paris and London) which isn't bad for a composer whose work is hardly ever performed outside Germany and not really very often there either. The London production, which I saw twice, was quite powerful.

So many fascinating comments here. The problem (for me) is the sheer volume of Shakespearean operas and how to say something semi-coherent in the 10,000 words allowed by the commission - ohime! Anyway, in terms of plot, like many operas, plot is in many ways the least interesing thing about Shakespeare's plays, many of which are lifted wholesale from other sources. Rather, what makes the plays (and by analogy operas) is the music of the language (or, conversely, the language of the music). I'm also interested that no one yet has said anything about the copious amount of music in Shakespeare's plays: songs, obviously, but also instrumental music. Obviously music was an important feature of early modern drama and Shakespeare is far from being the only writer who used it. But he has some particularly interesting things to say about music and I suspect that he wouldn't have found opera to be a completely alien medium...

I think that Shakespeare's A Midsummer Night's Dream is a thing of wonder, but I am also a big fan of Britten's treatment of it. It was my way into Britten's operatic output. I think I'm right in stating that Britten only added one line of his own to Shakespeare's text? Prudent to plunge straight into the magic of the wood, with that strange disorientating opening. I wonder why he chose to have Puck as a spoken role rather than sung? To do with the text, or for dramatic reasons?