In my own life I have found that one possible way to gain maturity is to look increasingly inwardly, to increasingly put things in a broader and ‘milder’ context. Take for example the St. Matthew Passion, which is increasingly performed in concert halls and throughout the year, not just at Easter. I regret that, but at the same time accept that you probably can’t avoid it anymore. Perhaps with times being as they are we should be happy that people play the St. Matthew Passion throughout the year. But the only way of respecting the essence of the work is to go to that essence, not giving in to fashionable ways of doing it, or simply try to please the audience. One thing that is clear is that you’re not going to learn it from books about interpretation and so on. You have to draw on your own inner experiences to express yourself. For instance, we walked the pilgrimage route to Santiago. That kind of experience helps me to be drawn into what I feel is the centre of Bach’s music.

In your notes for a disc of three of the cantatas, you make clear that you have increasingly become an adherent - although by no means a dogmatic one - of Joshua Rifkin’s arguments in favour of a Bach “choir” consisting of single voices to a part. Can you explain something of your personal reasons for adopting such an approach?

I have to confess that when I first heard about Rifkin’s research some twenty years ago, I thought “Oh, my God, here’s some pushy American who has come up with something new”. But I have to admit that I was wrong, because I judged without reading it. That’s never good.

I don’t think you were alone in that. Rifkin came in for some ferocious attacks, some by people one suspects had not read his work.

That was to be expected, of course, because he was attacking a sacred subject. He was very daring to publish his findings, but had I been him I would have done the same. It was necessary. But I did not originally follow the whole story; I’m not a subscriber to Early Music and such scholarly journals. Then Graham Nicholson, the trumpet player and builder, drew my attention to the controversy going on between Rifkin, Andrew Parrott and Ton Koopman and said it was so interesting that he would send me copies. And when I read what both sides had to say, I thought the arguments of Rifkin and Parrott were so superior to those of Koopman. So I started to become interested, but I have to admit that I have not conducted my own comprehensive research on the topic. Why should I? You could say that’s weak, because I’ve just accepted something someone else has said. But I’m not a musicologist and I’m not going to become a musicologist.

But you must have your own practical reasons for accepting Rifkin’s theory? My practical reason is that I had already lost faith in the fact that, for example, when you have a choral fugue you’d have four or five singers in unison on the first entry. I was already doubtful about such things. And then when you start to think about the fact there was virtually never any distinction made in Bach’s scores between solo and tutti, such things fell naturally into place. So Rifkin’s findings were in some sense confirmation of concerns I already had. It has proved that my feelings about fugal expositions were right; they were never meant to be performed by a group of people. Even in rare examples where Bach uses two singers to a part, such as the Leipzig version of Cantata 21, the fugue entrances are still marked as solo and the ripieno quartet only enters, doubled in this instance by trombones.