Movement 1: Kyrie
Movement 2: Gloria
Movement 3: Credo
Movement 4: Sanctus and Benedictus
Movement 5: Agnus Dei
Although this technique is not quite unique, the Missa Alma redemptoris mater is by far the most extensive example of its use in Renaissance music. (Other pieces of this kind include a canonic motet also by Moulu and discussed—alongside another such piece, by Costanzo Porta—in an early seventeenth-century treatise on music of Lodovico Zacconi; a fantasia con e senza pause by the English-based Fleming Philip van Wilder; and a setting of the Marian antiphon Regina caeli by Pierre de Manchicourt (c1510–1564) in which the second-highest voice is derived from the highest by removing the rests. Manchicourt’s motet has been recorded by The Brabant Ensemble on.) Moulu’s Mass thus fits into a long-standing tradition of generating musical compositions that obey mathematical constraints, frequently of considerable complexity. The existence of such pieces underlines the close association of mathematics and music in late-medieval conception: the singers who would have performed such complex works were highly specialized professionals, often in direct contact with the composer, and well able to deal with the challenges of cryptic notation. However, by circa 1520 when Moulu was active, the new technology of music printing was bringing compositions to a much broader public than heretofore, and consequently not all potential recipients of his music could be expected to comprehend the canon or resolve it correctly.
The problems caused by Moulu’s canon are evident in the Mass’s sixteenth-century sources (which number an impressive fifteen when one includes those such as theoretical treatises or collections of two-voice pieces which transmit only sections of the work). Although in a modern score it is necessary to notate the two versions entirely separately, the sixteenth-century practice of notating voices singly—whether in separate partbooks or large choirbooks with each voice in a block in one corner of the double page—allows both versions to be notated at the same time. The singers then simply ignore the longer rests when singing the short version. Because the number of such rests is not the same in each voice, however, some voices (in one or other version) will need to finish before reaching the last note, so as to cadence together with their colleagues. The places where this is to happen are indicated by the use of a signum congruentiae or sign of congruence, similar in appearance to the ‘dal segno’ symbols found in later music. This system evidently caused considerable confusion: partly this is due to Moulu’s ingenuity in varying his practice—sometimes it is the short version that must omit certain parts of the music, sometimes the long version, and no source ever indicates which. Further problems were caused by the editor of the earliest printed source (Rome, 1522) failing to reproduce correctly this and several other aspects of the Mass, necessitating remedial work by those who relied on this edition. The consequence is that Moulu’s Mass took on a number of different forms as the century progressed. For some scribes, for instance, the dual nature of the piece was beyond what they felt their performing forces could manage, and the sources that they copied avoid the issue entirely by presenting only the ‘short’ version of the piece, without the rests. One of these can be shown to have been copied from a complete version, which together with the omission of the third Agnus Dei—where the texture is expanded from four to five voices—gives a clear indication of where the scribe saw his choir’s abilities to end.
from notes by Stephen Rice © 2010