2
Somewhat pervertly, in all extant sources of the piece, even the frame interval common to both
species, that is, the fifth, is further distorted by the accidental of b molle instead of of b durum at
the highest notes of the cantus firmus, always underpinned with a B flat or G in the bassus.5 This
was probably just to avoid an infelicitous melodic progression. While the chant, in its original mode,
sounds quite naturally, in its transposed version it progresses in the melodic tritone, also known as
diabolus in musica, that would result from the melodic movement b natural–g–f.6
Why did Obrecht insist of transposing such a melody patently unfit to the Phrygian mode?7 Let us
have a look at another composition that must have originated around a decade later, Josquin des
Prez’ Nymphes des bois/Requiem aeternam. Josquin’s piece is also a lament, this time on the death of
the arguably greatest composer of the preceding generation, Johannes Ockeghem (in 1497). Like
Obrecht, Josquin cites the Introit to the Mass of the Dead by transposing it in the Phrygian mode.
Though Josquin rhythmicizes the chant in a different way than Obrecht, he also uses a b flat in the
fifth degree, at least according to the more authoritative source, Tielman Susato’s Le septiesme livre
of 1545.8
Finally, when Josquin himself died in 1521, one of the most important composers of the next
generation, Nicolas Gombert, wrote a lament Musae Iovis/Requiem that uses the chant melody in
exactly the same rhythm (but without b flat),9 thus referring to Josquin who was referring to
Ockeghem (and probably to Obrecht), effectively building up a kind of genealogical line.
5 Of the three sources of the motet, Segovia has only one b flat at the first melodic peak, Florence C at the second
peak, and Petrucci, Motetti C, which writes out (somewhat superfluous) the resolutiones of the two different readings,
has both in the first reading and only at the first peak in the second. Taking account of the fact that compositional
accidentals often got lost in transmission, this is quite convincing evidence that the piece was originally intended with
a bflat in the tenor voice, though only locally notated. Otherwise, the Phrygian mode is very much emphasized and all
the other voices use the b natural throughout.
6 However that may be, the result is what theorists called a commixtio tonorum, or mixture of tones, at four points in the
motet, a local flavour of a sound more related to the original Lydian mode, or even a Dorian mode transposed on G,
a strange cross-relation that should inspire composers of the next generation. This kind of modal oscillation should
become a trademark of compositions in the Phrygian mode, and especially Pierre de la Rue was very fond of it. See
Meconi, Art. La Rue NGr2, quoting the ending of the Sanctus of his Missa Nunca fué pena mayor. See also his Regina
celi, and the anonymous setting of Dulces exuviae in the „big“ chanson album of Marguerite d’Autriche (B-Br 228).
Scholars have tended to ascribe this composition (also in London Royal 8.G.vii) to La Rue and I concur. These two
motets are, as far as I know, the only compositions around 1500 in E phrygian than incorporate one part–and only
one–with a B flat signature.
7 Transpositions were far from uncommon in both monophonic and polyphonic repertoires. To quote a secular
example, the famous L’homme armé tune was used both in Lydian and in Dorian, sometimes even in Mixolydian versions
in the many masses composed on this cantus firmus. But there are few mass settings with the tune in Phrygian mode,
and one of these is by Jacob Obrecht. Depending on datings, Obrecht’s mass may be the first to use the tune on E.
Alternatively, the credit for this goes to Loyset Compère. Two later settings in the Phrygian mode are by Robert Carver
and Cristobàl Morales. I am indebted to Kirsten Pönnighaus (Weimar) for discussing the modal issues in L’homme-
armé masses with me.
8 Le septiesme livre contenant vingt & quatre chansons a cincq et a six parties, Antwerpen: Tielman Susato, 1545.
9 At the relevant points, Gombert always has an E in the bassus, thereby making a b flat impossible or at least
improbable.
