It is inevitable that any branch of historiography should in its early stages submit itself to the discipline of severe objectivity. Nineteenth-century political historians saw it as their duty to present ascertainable fact; and it is only in this century that their successors have begun to take the facts for granted, and have gone on to make inferences, to express opinion on unanswered questions, and, to some extent, to use their imagination. Musicology, being a much younger branch of the art, is still at a relatively objective stage. There are such formidable gaps in our knowledge of music before 1600 that scholars have shown a natural caution about filling those gaps with anything in the nature of speculation or guesswork. The standard work on English music before the Reformation is, of course, Frank Harrison's Music in Medieval Britain, a brilliant presentation of archival material and of liturgical and musical analysis. It is precisely that objectivity and caution that make the book so valuable. However, I am prepared to argue that, once a book of that kind has been written, it is perhaps justifiable to go beyond it and make inferences which, though they could well be proved wrong, may stimulate argument and so help to arrive at right answers. Much of this paper, therefore, is frankly discursive and conjectural, deficient in corroborative detail, and clearly open to the Scottish verdict of ‘not proven’.