I think that if you perform with the right tools - with the correct instruments - it obviously teaches you a lot. A good instrument and a good player will tell you much about the possibilities and limitations of music-making. You can imitate the elements of Baroque interpretation on modern instruments, but it would still be an imitation. There is really no substitute for the real thing - for using actual Baroque instruments or good copies. But simply playing the correct instruments - or even playing them well - is not enough. Over the last 40 years, the way of playing these instruments has improved enormously - at least, the technique has improved; but the interest in knowing why we use them has diminished. In the 1960s and 1970s, when I studied with Leonhardt, all students were interested in reading treatises, examining the original sources; they were eager to know everything. But these days, I notice that many of the good performers are less and less interested in sources. Great musicians like Harnoncourt, Leonhardt, Brüggen, and performers like myself and others have made many important discoveries; younger players seem content with relying on those discoveries. They go off and make music, relying on what the earlier generations have taught them. I think that’s dangerous because, if we are wrong, the next generation should find out our mistakes, and correct us.

Of course, when reading old sources, you end up taking out of them what fits your personality, but you have to be careful; you cannot introduce something that is not there at all. If all the Baroque sources speak in great detail about articulation and its importance, you cannot simply dismiss this element; and the same thing goes for vibrato. You should work with what the sources tell you, and think about how to apply it.

The limitations of Baroque instruments playing in different keys - like the more covered keys or more open keys on the Baroque oboe - teach you something, and not just about playing techniques. So, although I sometimes perform Bach’s music with modern orchestras, I still think there’s nothing better than Baroque instruments for that.

How do you see your role as a conductor - are you dictating a pre-determined interpretation to your players, or do you create the interpretation in co-operation with them?

If I work with my own orchestra, it’s in co-operation with my musicians, my friends. The final decision is still mine, and everybody accepts that. But if you play with fantastic musicians, and they come up with a question or an idea, it would be stupid not to listen to them, not to take them seriously. I’m also a chamber musician and it’s great fun to work with great musicians, so obviously I listen to what they have to say.

One such case, which I remember particularly well, was in the final chorus of Cantata 63, where we had a debate about a tempo change in the middle of the movement. There’s a chromatic passage, for which there is no tempo indication. The chorus begins with an allegro, and there’s a short adagio passage; but I believe that the allegro should return after the B part, even though there’s no a tempo indication. It’s true that normally, in Baroque music, chromatic music was performed slowly. But I didn’t think this was the case here. The text of this passage was “let the devil not torment us”, but the text as a whole is celebratory, and there is clearly no sense of danger or threat. In 18th century musical thought, when the text only hints at danger, you can use a musical device which is normally associated with a very sad text - but you use it in a different way; and I think that’s what Bach did here. So I thought that we should perform this passage quickly - in the same allegro tempo as most of the movement - but in the beginning most of the players and singers didn’t agree with me, and we had a big musical and theological discussion about this. In the concerts we tried it both ways and I only became more convinced that it should be done at a quick tempo. The experiment also convinced more than half of the musicians that I might be right. In the church, we recorded it at both tempos and listened to both versions; and by then only very few musicians still thought that I made a mistake. We decided to include the faster version on the CD.

Then the CD came out. At the time, I still gave every musician a free copy. As often happened, most people came to listen to the CDs rather late. I asked the two people who insisted on the slower tempo - and one of them was a key member of the orchestra - if they had listened. “Yeah, it’s nice”. Slowly I came to that particular cantata. “And tempos?” - “Fine”. “First movement?” - “Fine”. “Last movement?” - “Yes, excellent”. They’d forgotten the argument completely. So in general I feel that the conductor should listen to the musicians, should try to incorporate their ideas - but in the end you should not be afraid to make a mistake, to do it your way.

When I work with a modern orchestra it’s just me making the decision, there’s no discussion. A baroque orchestra works in a different way. They play with you because they like you, they enjoy playing with you; you’re normally friends with each other. So that’s another way of making music, a way I prefer.When I work with a modern orchestra, there is also something of a didactic element. In the 1960s, I and other people were complaining about modern orchestras performing the St Matthew Passion with 300 people in the choir and double woodwinds; we said this is a shame, you should not do it like that. These days many orchestras reduce their size when they perform this music. Last year I did the St John Passion with the Vienna Symphony Orchestra, and I could do it with six first violins - we could really reduce the ensemble. In general, I believe there is now much more openness from both sides, and that it’s important that specialists go to modern orchestras and try to teach them how to perform this music. Of course, you are there to make music with them - not just to be a professor, but you should share your thoughts and your knowledge about this music with them. I do this quite often, and I enjoy the experience. But when you have played with people for a long time - some musicians have been playing with me for 30 years - they understand with very few words what you want them to do; you can then focus on the finer points and make the performance more beautiful, more special. These musicians have also learned to anticipate me: they know that I like trills, they know that I like hemiolas, and so they will add them before I ask them to.