This Old Testament ‘annunciation’ prophesies the birth of Samson, one of the popular heroes of medieval story and art. There is a plenitude of evidence on this point, but his popularity could be deduced quite simply from the song which is the subject of this article. Samson dux fortissime is well known; it is often referred to, and it has been recorded at least twice in recent years. However, there is no adequate published edition of it, no comparative study of its sources and notations, and no analysis of the way in which the imposing rhetoric of its poetry is combined with an intricately patterned melody.1 This complex monophonic song offers a chance to examine melody and rhythm, rhetoric and rhyme, working together on a large scale in a harmonious and impressive whole.2

Among the small repertoire of early fifteenth-century ceremonial works is an anonymous lamentation motet, Romanorum rex inclite, written to commemorate the death in 1439 of Albrecht II, King of the Romans. It is thought to be by Johannes Brassart,1 the most distinguished member of Albrecht's choir, and is a fine example of the then moribund isorhythmic motet. Its music is of manifestly high quality, yet it is primarily for its text that Romanorum rex inclite is known, since this lists the names of the singers who apparently constituted the king's choir at the time of his death.2 Only two can be confidently identified: Brassart himself, who heads the list, and his near namesake Johannes de Sarto. Brassart's life is well documented; he is recognized as one of the best and most productive composers of the second

Just fifty years after Okeghem died, apparently as a very old man, Glareanus published a massive treatise entitled Dodecachordon (1547). In the last chapter, which Glareanus describes as an afterthought, he discusses the skill of certain composers, using mostly canonic works as his examples. Among them is Okeghem's three-out-of-one canon Prenez sur moy, which Glareanus describes as a ‘catholicon’. It is a song that has caught the imagination of many commentators over the centuries. Composed before c. 1470, it was presented and discussed by five music theorists of the sixteenth century and reprinted as late as 1594 – a matter that gives it a longer continuous career than any other polyphonic song of the fifteenth century. But therein lies the problem. Later writers added details that they thought would clarify the music but which only confused the issue.

The first Book of Common Prayer, published in 1549, contains no music. This is not to say that many of the liturgical texts it contains were not intended to be sung or intoned, as they had been for centuries before the Reformation. That singing as well as speaking was permissible is made clear by the rubrics,1 but notation was not supplied, an omission that can only have perplexed English priests and choirs at a time of radical and inadequately prescribed liturgical reform. As E. H. Fellowes commented in 1941, ‘very little attention has been drawn to the problems that must have confronted precentors, organists, choirmasters and composers, when the Latin liturgy was replaced by the Book of Common Prayer issued in the vernacular tongue’.2 To that list he might have added celebrants; and however great these problems may have appeared at choral foundations which maintained close contact with the principal reformers, they must have seemed almost insuperable to the priests at remote parish churches throughout the country.

When is a monody not a monody? Listen to almost any new release of medieval monophony on record and the answer will be the same: when it falls into the hands of twentieth-century performers. Few myths surrounding the interpretation of medieval song are as potent as the belief that works notated as single melodic lines are in fact latent polyphonies. Why this paradoxical state of affairs?