Well, I think you have to give the librettist some of the credit for that. It’s too easy just to say that Handel was a great dramatist. Of course that’s true, but as we know from the correspondence between Charles Jennens (the librettist) and the composer there was a kind of complex alchemy going on. Saul’s a particularly interesting role, because we regard it as particularly psychologically apposite, but it went through several layers of change, especially in the instance of a huge aria in Act 2, which was excised, greatly changing the character of the role. Sometimes these masterpieces have come to their final form as much by accident as design.

You can play Bach on the piano, a symphony orchestra or a quartet of saxophones, but let’s stop this silly, childish business of knit your own musicology, where you just take one line out a document to seek to attempt some sort of spurious historical justification to keep employing your choir.

During the summer I heard you give a wonderful performance of Acis and Galatea at the Beaune Festival. What do you think it is about this work that has enabled it to remain so fresh?

One of the things that always amazes me about Handel is that he goes to Italy and composes Aci, Galatea e Polifemo, which is a fantastic piece and as Italianate as it comes, and yet ten years later writes a piece in England on the same subject that is more English than Dryden or Constable. It’s drippingly pastoral in a glorious way. And I don’t know whether it’s me being a sloppy romantic, but you can just smell the English countryside in that piece, which maybe why it’s so appealing for us. There’s this rather naïve use of almost folk-like material. It’s an incredibly strong piece, yet it’s so slight. Acis is like a butterfly, a bit like Dido and Aeneas in that respect—if you overcast it, it can’t take the weight.

Have we any idea how much Purcell was known to Handel? There does seem to be a very strong influence in Acis.

It’s a very interesting question. Purcell’s songs, church music and some of the theatre pieces were almost certainly still being performed in the early 18th century, but quite how widely is difficult to know. Certainly I think that Acis is the work that Purcell might have written had he lived to become an older man.

It’s interesting to consider the history of music and how it is disseminated. For example, we’ve recorded Josquin’s Praeter rerum serium in a reconstruction of a seventeenth century Roman service, simply because we know that Josquin was still one of the most frequently performed composers in 17th century Rome. The interesting thing is that when you perform Josquin in a 17th century context you have to have a completely different approach to things like musica ficta. We’re all interested in performing music in the way that it might have been done the day after the ink dried, but what about repeat performances even in the early years of a piece?

Well, it was common for works to constant undergo change on repetition. One need go no further than Handel’s operas, where changes were constantly made to accommodate new singers. Such works we’re living organisms, constantly undergoing alteration.

Yes, they were re-worked. But there’s another interesting paradox there, and I think it’s very much at the root of the relationship I have between musicology and music making. I get very irritated, for example, about the way Minkowski mixes different versions of Rameau operas on recordings. It’s not a personal thing, because I like Marc. To me it’s one thing to say that changes, which were sometimes improvements and sometimes necessity, were made, but quite another to say it’s possible to do a little bit of this version and little bit of that version. There has to be some sort of integrity within the dramatic framework—you can’t just pick a particular high note from one version and a particular low note from another. You have to be scholarly; the fascinating thing is that we’re applying archaeological principles, but you have also to apply them to the way scores are used in performance.