Conference Agenda

Overview of the Session

Citations: The Renaissance Imitation Mass (CRIM; crimproject.org) focuses on an important but neglected tradition, in which short sacred or secular pieces were transformed into five-movement cyclic settings of the Mass Ordinary. If counterpoint is a craft of combinations, then the Imitation Mass involves the art of recombination on a massive scale.

How can digital tools help us understand this remarkable repertory? In part, through controlled vocabularies and structured data. In the CRIM Project platform, analysts can annotate scores in precise ways; to date an international team has assembled over 2500 ‘relationships’ that detail connections across a corpus of some 50 Masses and models. But we have also developed a robust set of Python tools (CRIM Intervals, itself based on music21) that help us analyze these contrapuntal patterns on a scale that is both vast and detailed.

Training this mechanical ear has required that we discipline ourselves: asking questions about our assumptions, and generally formalizing expertise that often remains undocumented. In this session, participants in the CRIM project will share results of their efforts, and point the way forward for further learning and research with digital tools. Attendees will learn how to participate in the larger project and how they can adapt CRIM tools (which are open-source and freely available) for their own work.

In “Hidden Forms in Palestrina’s First Book of Four-Voice Motets” (2007), Peter Schubert established three “presentation types” used in in Renaissance imitative polyphony that use repeated contrapuntal combinations (“modules”). My recent work on Verdelot’s madrigal, Ultimi miei sospiri (c. 1528) and Padovano’s Missa Ultimi miei sospiri (1573), has led me to believe that Schubert’s presentation types can have formal functions in sixteenth-century Renaissance polyphony. Imitative Duos (IDs) are used at the beginnings of works, movements, or major sections, Non-Imitative Modules (NIms) are used at the end, and Periodic Entries (PEns) and Fuga tend to appear in the middle. Schubert’s analyses of the Palestrina motets confirm this hypothesis: eighteen of the thirty-five motets begin with an ID and fourteen end with a NIm. One reason NIms work at the end of a piece is because they often have a sustained or repeated pitch in one voice, as in Burmeister’s supplementum.

I will test this hypothesis with reference to the CRIM corpus of fifty Masses (250 movements) and forty-two polyphonic models (chansons, madrigals, and motets), using both personal score analysis and the tools developed in CRIM intervals to identify the presentation types at the beginnings and ends of each of the pieces or movements. I will also look at additional imitative motets and madrigals c. 1530-1570. For motets I will focus on the eighteen motets with thirty or more sources listed in Jennifer Thomas’s “Core Motet Repertory” (2001). For madrigals I will focus on five- and six-voice madrigals (which are more likely to be imitative) by Verdelot, Willaert, Rore, and Lassus.

An overview of the types of imitation used to begin and end Renaissance imitative compositions will result in a new understanding of Renaissance compositional strategies.

Because of our cultural biases, we’re always in hot water when we analyze music of the Renaissance (or, really, any other historical era). We can’t unlearn common-practice harmony and become “authentic” experts in psalmody and counterpoint. To address this issue, we decided to start as much as possible from the “neutral” level hypothesized by J.-J. Nattiez, in which data is gathered without prejudice. We proceed with one of us as the “blind” analyst, who uses CRIM Interval’s toolbox to find melodic elements that we define minimally (e.g., the most often-occurring 5-note melodic string). The other is the “sighted” analyst who tries to interpret this raw data in terms that another musicologist can understand.

We have applied this method to a madrigal by Rore (“Quando lieta sperai”) and to the Kyrie of the Baccusi Mass based on it. The blind analyst asked the computer to find the melodic strings five notes long that were the most often repeated in the madrigal and in the Mass, either exactly or diatonically transposed, disregarding rests, rhythm, repeated notes, and text. (Two prejudices that underlie these decisions are that melodies are important, and that repetition is important.) Then the sighted analyst looked at where they occurred in each piece.

Both strings consisted only of “generic” stepwise motion, with no notable ”characteristic” skips, suggesting that it was mere “filler” material. However, the string that occurred most often in the madrigal presented itself in clusters, showing that it had a structural function. The one that occurred most in the Mass (41 times) occurred overwhelmingly with the same solmization syllables, showing modal consistency. This string, however, only occurred seven times in the madrigal, revealing how Baccusi extended the original material. These two melodic strings, that we might expect to see everywhere, were adjacent only once in a single voice, dovetailed by Baccusi at the climax of the Kyrie. Most striking was the fact that both Rore and Baccusi often hid these strings inside longer melodies—to be motivic, then, a string need not appear at the beginning of a phrase or be associated with any text.

Studies of imitation masses have traditionally focused on the rearrangement and recomposition of the model’s points of imitation (e.g., Quereau 1982); there have been few analyses focusing specifically on the treatment of homorhythmic models (see Crook 1994, 191ff.). This paper investigates this overlooked topic by presenting highlights from the analyses of two imitation masses by Orlando di Lasso, based on Claudin de Sermisy’s “Il me suffit” and Pierre Sandrin’s “Doulce Memoire.”

Lasso’s imitation technique in his chanson-masses has been described as at times “perfunctory” (e.g., Reese 1954, 697), presumably due to the frequent literal quotation of entire phrases from the model and the alternation of such passages with seemingly unrelated, newly-composed material. In this paper, I argue that far from being a time-saving technique, these homorhythmic blocks of quoted polyphony, or block quotations, as I refer to them, allow various degrees of preservation of the model’s original polyphonic fabric, thus various degrees of literal quotation, and as such offer a rich palette of compositional uses and opportunities.

I define four types of block quotation, namely simple block, inverted block, revoiced block and reharmonized block quotation. While simple block quotation preserves the model’s original polyphonic fabric relatively unchanged, the other block quotation types vary this fabric through the inversion of voices or the extensive revoicing or reharmonization of the block’s constituent sonorities and soggetti. These different types of block quotation can underly new, seemingly freely-composed passages, create varied repetitions, or serve as motives. Furthermore, their placement and interaction with quoted soggetti can create various formal narratives across individual mass movements or their inner sections.

I reflect on the ways and extent to which CRIM Intervals can help detect block quotations. I show that tracking the progression of harmonic sonorities proves problematic due to melodic alterations in the block’s constituent soggetti. Despite these alterations, normally one of these soggetti is preserved exactly; this soggetto can be found using the Melodic n-gram corpus tool and its association with a block quotation can then be manually assessed.

This study will help us reassess Lasso’s chanson-masses and provides a new understanding of his imitation technique.