‘What we are doing with early music is genuinely authentic to such a small degree that the word loses most of its intended meaning’

One of the many promising developments in attitudes towards the performance of early music is that an increasing number of scholar-performers are prepared to admit that most of what they do in constructing a performance style is-of necessity-of their own invention.1 Historical research may provide us with instruments, and sometimes even quite detailed information on how to use them; but the gap between such evidence and a sounding performance is still so great that it can be bridged only by a large amount of musicianship and invention. Exactly how much is required can easily be forgotten, precisely because the exercise of musical invention is so automatic to the