The Masses of Frei Manuel Cardoso (1566-1650) and his Portuguese contemporaries follow an interesting pattern in their Kyries, providing two distinct polyphonic settings of the Christe. This Kyrie–Christe–Christe–Kyrie sequence is only exceptionally found in the Spanish and Roman etc. schools, but was standard on the Iberian Peninsula in the 16th and early 17th centuries. The plainchant printed in the editions of Cardoso’s two Requiem Masses shows that the expected method of performance was 9-fold, i.e. 3x Kyrie, 3x Christe, 3x Kyrie, with the odd-numbered invocations sung to plainchant. This, in turn, required two polyphonic settings of the Christe.

Only Cardoso’s Requiem Masses print the plainchant alongside the polyphony; for the others we have no information about what chant was sung, nor instructions to help choose suitable chant. For the modern editor and performer this raises the question of what plainchant to use when none is supplied and the Mass does not use a plainchant theme.

More and more plainchant sources are being restored, digitised and made available online (see the Portuguese Early Music Database). These are of great interest since use of chant throughout the Church was far from standardised in the 16th century and the uniformity desired by the Council of Trent (1545-63) by no means immediately implemented. Chants developed fluidly as the aural tradition gradually came to be written down. This, combined with the multiplicity of copyists copying multiple sources, frequently resulted in [usually small] regional variations; a Kyrie in Lisbon would not necessarily be the same in Braga, for instance. To this confusion we add the tragic loss of so many treasures in the earthquake and subsequent fire in Lisbon in 1755 which destroyed the famous music library of John IV and reduced Cardoso’s monastery to the ruins that now remain.

The use of plainchant diminished and all but vanished in the 18th century. Publications in the second half of the 19th century (see are clearly the first steps in recovering a lost heritage. Where sources in the 16th century are broadly in agreement, here there are profound differences between publications produced only years apart. It is not until the 1905 Solesmes Kyriale is published that the current forms are solidified.

Today there is no way to arrive at a solution which is certainly authentic; choosing plainchant is best viewed in practical terms, for example, identifying one that is modally/tonally in agreement with the Mass.

The Kyrie of this week’s Mass setting, Cardoso’s Missa de Beata Virgine Maria from his third book of Masses, follows the expected 9-fold pattern; however, although the polyphony is derived from the Marian Mass IX Cum iubilo making the choice of plainchant apparently easy, both Kyries quote the ‘wrong’ invocation to fit the alternating plainchant/polyphony model. I have produced a PDF to illustrate this and provide an overview of the Kyrie as given in some Portuguese sources (click to open):

 Illustration of the changing forms of Mass IX Kyrie and the fragments used by Cardoso in his Missa de Beata Virgine Maria

Illustration of the changing forms of Mass IX Kyrie and the fragments used by Cardoso in his Missa de Beata Virgine Maria

Here, although the plainchant is known, Cardoso plays his usual trick of blurring the lines between modality and tonality; the 2nd note of the second Kyrie is raised a semitone to function as a leading note which makes for a decidedly awkward oscillation between chant and polyphony.

Perhaps this was a feature of Cardoso’s now lost source? Tonal adjustments are found in some early modern editions, for example, in the 1857 Pustet Kyriale version of the current Mass I Lux et origo:

 Agnus Dei from Missa Tempore Paschali, Pustet Kyriale, 1857

Agnus Dei from Missa Tempore Paschali, Pustet Kyriale, 1857

The strict Tridentine requirement for a 9-fold Kyrie is not retained in the Ordinary Form, (though is not forbidden). In both forms, some leeway is anticipated when singing polyphony since the words are recited more than once in each Kyrie or Christe section, i.e. the repetition is inherent. For the majority of Masses sung as part of Cardoso450 we are singing just the polyphony since arbitrarily choosing plainchant to interpolate adds little or nothing to the music and, within the liturgical schedule, a lengthened Penitential Rite causes the music to be overly dominant.

Researching early plainchant sources gives a good overview of the changing and flexible nature of something that is, nowadays, well-established and fixed. Perhaps being similarly flexible in performance and resisting combining plainchant and polyphony which do not happily sit together is the best way of letting the genius of Cardoso’s music speak clearly.