This recording presents the only two Masses by Josquin which are entirely based on canons. He wrote other single movements which are canons - the second Agnus Dei in his Missa L'homme armé super voces musicales is an especially complex example and is recorded on CDGIM 019 - but only here did he explore the possibilities as exhaustively as the idiom would allow. To write this kind of music may seem academic to the modern mind: who is interested in mathematical scaffolding which most people can't hear? But Josquin was interested in it - as were many later composers, from Bach to Brahms to Webern - and it is clear that, like every composer of genius, Josquin relishes the challenge inherent in being tied down to a pattern.
These two settings seem to stand at the extreme ends of Josquin's career. The Missa Ad fugam, an early work, is the easier of the two to follow; the Missa Sine nomine, which may have been Josquin's last Mass-setting before the great Missa Pange lingua, shows the fruits of his experience in mathematical writing like no other. Indeed, Ad fugam is so much more straightforward than Sine nomine that it is possible Josquin wrote the later work as a foil to the earlier, to show how much more he knew about handling this kind of composition by the end of his life. This would have been more important to him than we may recognize now: every Flemish composer up to Josquin's time had proved himself with canonic writing, Ockeghem being a leading example. If it is true that Josquin learnt from Ockeghem, it is possible that he saved up this tour de force just to show he could rival his master. In fact there does seem to be a personal note of this kind: at ‘Et incarnatus est' in the Credo Josquin quotes from his own lament on the death of Ockeghem, Nymphes des bois.
The word ‘canon' means that a single melody is stated in different voice-parts at different times so that it overlaps with itself. The mathematical element comes in making sure that one part of the chosen melody can combine with another part of it, possibly at a different pitch or in a different tempo, and still make musical sense. The easiest examples to follow are those which are at the same pitch (at the unison or octave) with the second voice following very closely on the first. One famous (and very audible) example of this is in the second Agnus Dei of Palestrina's Missa Brevis, between the two soprano parts; this work is recorded on CDGIM 008.
In Ad fugam the canon is always between the top part and the third part down, and always a fifth apart. At the very beginning of the first Kyrie, for example, the top part begins on a G with the answering tenor beginning on a C, three beats later. Since all of the five movements begin with exactly the same musical material, the canon is the same every time. Later in all the movements the distance between the two canonic voices grows wider - to three whole bars - but still the writing is so transparent that after a hearing or two the influence of the canon should be apparent. This transparency is helped by the fact that the second and fourth parts hardly join in at all.
Ad fugam is thought to be an early work partly because the canon is so rigid; partly because the common material which opens every movement lasts ten bars (substantial by Josquin's later standards); and partly because there is an original source (MS 31 in the library of Jena University) which seems to carry some second thoughts by someone - possibly Josquin - who wanted to rework the canon in the Sanctus and Agnus. Since these revisions have something of Josquin's maturer style about them, we have decided to include them on this recording as a point of comparison. It is not often in music of this period that we are given a glimpse of a revision; and certainly in this case the difference between the elongated, Ockeghem-esque lines of the original, and the sparser, tauter thinking of the later music is revealing.
Neither Ad fugam nor Sine nomine is known to be based on pre-existing material: in both cases Josquin seems to have invented the canonic melodies himself. The main difference between the two Masses is that in Sine nomine the canons are distributed all over the texture - any of the voices may take any other as its partner. And where one of the voices does not actually have the main melody, it may well join in through imitation. The opening Kyrie again is a good example. It is written around a canon between the top and second parts (a fourth and fourteen bars apart), but before the second part enters - rather late in the writing - the third part has accompanied the top part with music very obviously related to the canonic melody. Eventually the bass enters doing the same thing. The Christe and second Kyrie then go on to have their own canonic schemes. The Gloria begins with imitation between the top and second parts, which is a feint since neither of them has the canon in earnest, despite singing the opening notes of it. And so on: such subtlety of writing over the length of a whole Mass would take a short book to do it justice.
I have listened to many pieces of music from all periods with the greatest enjoyment, not knowing that they had canons buried inside them: it is not necessary to follow even the skeleton key to Josquin's mathematics I have given above to find these Masses as engrossing as those which are free. Yet in a sense any successful piece of canonic writing is bound to have an extra dimension to it. The listener will always have the apprehension, however vague, of a presence in the music which complicates and intensifies. It is quite possible that you will never get to the bottom of why a canonic Mass fascinates you, yet it is not necessary to analyse everything to enjoy it. Polyphony is always complex and canonic writing makes the most complex polyphony of all. The best polyphony does not have to be mathematically ingenious, but it should add something if it is.
© 2008 Peter Phillips