Chapter 5: Beyond the Performer’s Approach

The insight gained from working on the wider musical corpus represented by the motet on which Missa Inviolata is based, plus the materials from which the motet itself derives, as well as other pieces related to the same materials, allows us to outline a rich musical background for the mass:

• the last two words of a responsory (ca. 10 seconds of plainchant) are expanded with a long melisma (ca. 1 minute);
• to this melisma, a text (prosa, prosula) is supplied;
• the texted melisma becomes independent as a – still monophonic – sequence (ca. 2 minutes);
• this sequence becomes the basis for polyphonic compositions;
• among those, Josquin treats the sequence canonically as the nucleus of a motet for 5 voices in 3 sections (ca. 7 minutes);
• the motet becomes in turn the model for further polyphonic compositions;
• among those, a polyphonic mass for 6 voices in 13 sections (ca. 30 minutes).

The central realization of the present research is that all the mentioned “pieces” can actually be considered as nodes in a dense network, or points in a musical continuum, rather than individual, closed works of art. In order to further proceed with an experimental, practice-based type of research, it became clear that it was necessary to go beyond the mere idea of “piece”: the next step required to insert myself into that continuum, blurring the distinction between performer, improviser, and composer.

5.1 Extending the sequence Inviolata

A comparative analysis of the sources that transmit the plainchant sequence Inviolata shows some divergences, not only in the melody but also in the text. One of these is in its penultimate verse, acting as an invocation before the very last one: while in slightly more than half of the sources consulted, this passage reads simply “O benigna”, in some sources, this invocation is repeated up to three times: “O benigna, o benigna, o benigna”, with the same music. In yet other sources, the invocation is repeated but with additional text: the version used by Josquin for his motet has “O benigna, o regina, o Maria”. The specific source used for the transcription in Annex 3, besides its slightly different melody, expands the passage to no less than five invocations: “O Maria, mater pia, o benigna, laude digna, o regina”.

These “O benigna” variants have been taken as a chance for intervention in the sequence. Be it for devotional reasons (a litany-like expansion) or even for practical reasons (for example, when using the sequence for a procession), it is not impossible to imagine a medieval cantor expanding a sequence with a textual trope. Therefore, the first experimental intervention has consisted of writing a total of 20 extra invocations for the sequence, to be sung to the melodic fragment assigned to “O benigna”. To keep the style of medieval Latin poetry suggested by the existing variants, I have borrowed common epithets of the Virgin from various liturgical and paraliturgical sources, and organized them according to rhyme, structure, and meaning. The result is given below, both in Latin and in an English translation.

5.2 Improvised polyphony on the sequence Inviolata

The topic of polyphonic improvisation (variously identified as discantus, contrapunto alla mente, cantare super librum, etc.) has gained focus in the last couple of decades in academic discussions and in the curricula of Early Music departments, but also in informal environments such as summer courses or even online videos, to the point that it represents a sort of revolution in our modern understanding of historical musical skills⁶⁷.

Already from my time as a bachelor’s student in the Royal Conservatoire of The Hague (2007-2011), I have been involved in experiments on polyphonic improvisation, developing several strategies originally designed to train myself, but which turned quickly into workshops and projects for my fellow students⁶⁸. After that, as a teacher in the same institution, polyphonic improvisation has represented (together with other closely related practices) the backbone of the renovation of Early Music theory subjects in which I have participated⁶⁹. These developments have proved extremely fruitful, to the point that many of my students (and even former colleagues) are pursuing successful careers with the help of a rare set of historically relevant skills, gained in the context of those subjects.

As part of the practice-based aspect of the present research, a series of ex tempore polyphonic techniques have been applied to the sequence Inviolata, ranging from drones and parallel organum⁷⁰ up to fauxbourdon in three and four parts⁷¹. In order to keep with the spirit of these unwritten practices, the results have not been translated into a score, as they can be better demonstrated as part of a presentation with an ensemble or using a polyphonic instrument.

5.3 Canonic possibilities of the sequence Inviolata

As discussed in section 4.2, Josquin’s Inviolata uses a canon on the sequence Inviolata as the basis on which the motet is built. The full melody of the sequence is in the Tenor, followed at the upper fifth by the Quintus. The procedure involves working out the melody taken as cantus firmus, given that it is not originally conceived to work as a canon. There are some parameters that a composer can manipulate in order to make the specific melody work as a canon: the main two are the temporal distance of the imitation and the choice of interval. The latter is most often the octave, fifth or fourth because those intervals allow for exact imitation (i.e., maintaining the structure of tones and semitones) and are embedded in the foundations of hexachordal solfège. Additionally, it is also possible to play with note values and insert rests to solve situations where otherwise the canon would not work, making the melody fit onto itself at the chosen interval and improving the general result⁷².

For the purposes of the practice-based aspect of this research, a similar procedure has been undertaken as an experiment, with the result offered in Annex 5. The chosen imitation interval is the octave (as opposed to Josquin’s fifth), and therefore a series of different decisions were taken to make the canon work, specifically in terms of note values and inserted rests. Other dissimilarities with Josquin’s canon are related to the specific chant source used, which in this case is the one offered in modern transcription in Annex 3. The two-part structure resulting out of this experiment could serve as the basis for a fully-fledged composition akin to Josquin’s motet.

From the perspective of the present research, this experiment has served to understand the challenges in using a cantus firmus in canon, and to better appreciate Josquin’s solutions to those challenges.

5.4 Completing Missa Inviolata

My background as a composer is originally rooted in the compositional training I received as part of my studies in Music Theory (degree awarded in 2002). This training was complemented with private studies in composition but it has been radically transformed by my ulterior work on ex tempore techniques since my transition into Early Music and the development of practical skills such as basso continuo, partimento, and most importantly for the current topic, contrapunto alla mente.

The fact that Missa Inviolata does not set the complete text of the Ordinary of the Mass (as discussed in section 1.3) opens another avenuet for further experimentation: composing music for the “missing” sections. As part of the practice-based aspect of this research, two sections have been composed: a duo for the Pleni sunt and a six-part Hosanna, thus partially completing the Sanctus. Both sections adhere to the procedure of parody, thus using materials from Josquin’s motet.

The first of these is a duo for Tenor and Bassus, setting the text “Pleni sunt cæli et terra gloria tua”, which follows immediately after the Sanctus. Given that the author of Missa Inviolata uses material from the first part of Josquin’s motet for the Sanctus, the material chosen for the Pleni sunt has been extracted from the second part, which precisely starts as a duo. This material has been expanded with several interpolations and developments into a complete section. The resulting composition is in Annex 6.

The second compositional experiment is a six-part section setting the text “Hosanna in excelsis”, which follows the Pleni sunt. Continuing with the plan established, the material for this section has been extracted from the third part of Josquin’s motet, which has been reworked into a fuller six-part texture in ternary meter, following the standard conventions of 16^th-century sacred polyphony regarding the typical meters used for the Hosanna sections. The resulting composition is in Annex 7.

These compositional experiments have allowed me to gain a hands-on awareness on the parody technique, its possibilities and the many decisions involved in working with (or re-working) pre-existing musical material, even if at a preliminary level. Beyond the artistic merits of the results, and on a practical note, the additional sections composed are particularly useful when using Missa Inviolata for liturgical occasions, and, by extension, in concert situations where a full liturgy is reconstructed.

5.5 Further possibilities

The experiments described in the previous sections represent only a few among many possible ones. It would have been possible, of course, to focus on other aspects, or to pursue entirely different directions while working on the same aspects. One experiment that attracted my attention during the process of research but had to be omitted from the project due to falling outside its scope, is monophonic improvisation – i.e. developing the skill of constructing a plainchant piece to set the text of a responsory such as Gaude Maria Virgo.

Although this piece has been considered for the present research as the furthermost item in the chain of derived repertoire, it is itself derived from techniques involving the deployment and recombination of a corpus of formulas and gestures related to a specific modal space. Such experiments would have placed us in the role of the early, pre-Gregorian medieval cantor – that is, before the fixation of Gregorian chant as a repertoire of finished “pieces”⁷³.

Notes for chapter 5

67. A comprehensive bibliography on improvised polyphony would be quite extensive, out of the scope of this text. The monograph Canguilhem 2015 is a fairly recent, in-depth discussion of the topic, with a focus on Renaissance music, and includes such a bibliography.

68. The first public presentation of those experiments took place in the context of the project “Music around the Immaculate Conception” (December 2010), led by Peter Van Heyghen. I was in charge of preparing a part of the concert program, which included a Magnificat featuring Gregorian chant and improvised polyphony (parallel organum and fauxbourdon in three and four parts).

69. These are Musica Practica and Historical Keyboard Skills. For more information, see the Curriculum Handbook for the Bachelor of Music, Early Music: <https://www.koncon.nl/storage/documents/Curriculum-Handbook-Bmus-Early-Music-22-23.pdf>.

70. Parallel organum is first described in the Enchiriadis treatises: see Erickson 1995.

71. Fauxbourdon and related techniques are described in the treatise De preceptis artis musice et pratice compendiosus libellus, written by a Guilielmus Monachus (ca. 1470). For an edition, translation, and commentary, see Park 1993. For a practical introduction into these techniques, see Janin 2014.

72. See Rice 2007 for more details on this procedure.

73. More on this perspective on the formation of chant, comparing the technique of the early cantor and that of the poet of orally transmitted epics, can be found in articles by Leo Treitler from the 1970s, 1980s and 1990s (most of the relevant ones are collected in Treitler 2003.