Review: Measuring Polyphony: Digital Encodings of Late Medieval Music

Transcription is hard. Complicated or unfamiliar music makes it harder. In his book on music editing, James Grier reminds us that the edited or transcribed musical text is “not so much a tool, leading to higher ends, as an active, critical participant in those ends, fostering further critical study and the ultimate goal, one hopes, of all types of musical endeavour, the animation of the music in performance.”1 These are high stakes, especially when the pieces at hand are motets from the end of the medieval era, written in mensural notation. The website Measuring Polyphony uses the most recent technology in digital transcription to “animate” upward of sixty of these motets as found in the pages of several well-known thirteenth- and fourteenth-century manuscripts. This project is more than simply a digital update to the more traditional format of the printed edition; each piece is linked to a high-resolution image of its appearance in the original manuscript, and an audio playback (in MIDI) is available for listening while watching the notes in the various voice parts turn red as they sound. These visual and auditory features lift these motets out of their traditional identity as academic notation exercises and imbue them with life, appealing to the “musician” within the “musicologist.” As a medievalist, my view is that we badly need this kind of revitalization, especially in the current academic climate, in which our field flirts increasingly with extinction. We must use all the technology now available to us to fill the experiential gaps left by traditional print editions in order to cultivate a genuine appreciation of the richness of the ars nova repertoire in the next generation of scholars. As Margaret Bent warns, “without that symbiotic relationship of shared expectations, notated music is like seaweed out of water; as time passes, it becomes divorced from what originally gave it vitality and beauty.”2 Measuring Polyphony is a helpful, hopeful little pool that will benefit the musicological community by growing and deepening in the future.Digital humanities methodologies notwithstanding, it is still easy for students of thirteenth- and fourteenth-century mensural notation to get sidetracked by the initial tasks of learning to recognize the various Franconian note types, memorizing the rules for their interpretation, and then testing their understanding by transcribing each voice part in modern notation, hoping against hope that they all end at the same time. With these priorities, is there time to really think about what these pieces might have sounded like at their first performances, or what color the ink on the manuscript page was, or whether the audiences were pleased with the new sounds? Craig Wright describes his experience as a student in a “Notation of Polyphonic Music” course in the 1960s in this way:The transcriptions and links on Measuring Polyphony address this question and welcome other researchers into that process.From a theoretical standpoint, late medieval motets are a perfect fit for the methods and assumptions of twenty-first-century digital transcription. The logic informing their compositional style is well documented. Anonymous IV's account of the Notre-Dame school and Franco of Cologne's Ars cantus mensurabilis are still key sources for most undergraduate survey courses, and upper-year students in notation seminars regularly come across fourteenth-century treatises such as Johannes de Muris's Notitia artis musicae and Jacques de Liège's Speculum musicae.4 Several other thirteenth- and fourteenth-century books could augment this list further.5 Such rigorous documentation in the fourteenth century of a new way of conceiving of music that uses time as a guiding principle is a witness to the deliberate proliferation of this new music across western Europe and its academic associations with mathematics and geometry in the scholastic quadrivium. Expressing and measuring time in and through music was a radically new way of composing and conceiving of the demands and implications of harmony, let alone examining what a piece of music actually is in the first place. Measuring Polyphony celebrates, with this same kind of attention and exactitude, the complexity and ingenuity of late medieval harmony; it is the latest in a centuries-long chain of musical projects that take these fascinating motets as opportunities to look deeply into what they represent in terms of musical possibilities, assumptions, and expectations.In practical terms, Measuring Polyphony offers the user MEI (Music Encoding Initiative) transcriptions of more than sixty pieces found in three fourteenth-century manuscripts.6 The transcriptions can be viewed, downloaded, or printed, in either modern notation or using a digitized form of the original notation through Verovio, an open-source, online library for engraving MEI.7 The pieces can also be heard through the browser by activating an audio playback feature of Verovio. These pieces include twenty-four ars antiqua and ars nova motets from the Montpellier Codex, eighteen motets from the Roman de Fauvel, and twenty-two from the Ivrea Codex.8 The texts for the motets in the Roman de Fauvel and the Montpellier source were transcribed from images of the originals, and those for the motets in the Ivrea Codex were taken from an edition.9The website itself is divided into two main pages: “About,” in which the entire project is explained, and the names of researchers, developers, and contributors are listed; and “Browse,” in which the motets are listed by composer and by the opening text of each voice part. Someone less familiar with the repertoire might appreciate the addition of a column indicating the manuscript source for each motet, though this information is included with the individual transcriptions. The motet list under “Browse” (see Figure 1) links to the individual transcriptions, an example of which is shown in Figure 2. Here, the transcription is displayed digitally using the Verovio engraver to represent the original notation on screen. If the user would rather see it in modern notation, the red “Switch score to modern notation” command in the upper right-hand corner will toggle the view. To the left of this switch is the option to play the transcription on MIDI as the notation scrolls by. Beneath the transcription, the user can click to download the MEI code for the transcription (which requires an environment in which to run XML code) or a printable version of the transcription itself in either modern or mensural notation. Appropriate metadata for the piece is also given, including its source, concordances, and notes on its transcription.Karen Desmond of Brandeis University is both Project Director and visionary for Measuring Polyphony and was forthcoming in answering my questions about the development of and future plans for the project. She was quick to credit the technical expertise at McGill University, where Martha E. Thomae developed the scripts necessary to transform transcription in modern notation into the MEI-encoded mensural notation. Several graduate students and postdocs from both Brandeis and McGill are listed on the site as having provided editorial or user-interface (UI)/design assistance. Laurent Pugin (RISM, Switzerland) provided the Verovio customization.The project is built on Desmond's transcriptions of these motets into Sibelius in 2015. Each piece took between forty-five minutes and two hours to type into the program (or scan from an edition and then manually correct). At that time, Desmond's research interests centred on the notation itself, so the transcriptions were ultimately checked against images of the original and double-checked whenever ficta appeared. Since Sibelius does not support queries or analysis of the transcription, Desmond turned to music analytical software, under development at the time, to help to identify aspects of her Sibelius transcriptions that would have seemed “new” (ars nova, after all) to a fourteenth-century audience. By working closely with the MEI community, she developed a schema for MEI that would address the challenges of mensural notation, using the fourteenth-century rules. Fortunately, most of these rules are clearly set out in the sources like 700-year-old “ReadMe” files.Turning a manuscript score into a Sibelius file and then into MEI holds our assumptions about musical syntax and grammar up to the light. In the thirteenth and fourteenth centuries, mensural notation for the first time expressed rhythm at a level as fundamental as pitch, adding a new layer of complexity to the relationships between voice parts. There were as yet no standards concerning what might be allowed and what was not, and the variations in complexity are sometimes downright intimidating. For example, there are currently seven motets that are impossible to display using Verovio on account of their use of mixed mensurations mid-piece. It will be the MEI community itself, made up of researchers who use MEI and driven by a management board and technical team, that addresses this problem, using it as a test case for improving its overall adaptability to other notation challenges. The summer of 2019 will see development of an input tool for ars antiqua and ars nova notation, allowing users to contribute their own transcriptions online and listen to them as they go. Desmond envisions musicologists locating an image on DIAMM and transcribing the note shapes they see (longs, breves, semibreves, etc.) in a separate input tool window on their screen. She successfully piloted the site in a recent university “Early Notations” course and saw both the need and the capacity to develop the tool in order to increase the site's usefulness. Such an input tool would certainly increase the number of pieces in the collection—on Desmond's wish list are the one hundred motets in the Bamberg Codex—helping musicologists to gain a broader and more substantial overview of the “new style.” It is important, however, that the input tool should not presuppose extensive knowledge of Sibelius or XML; users should be able to simply transcribe what they see. Perhaps the audio playback could also be made to play only select voices at a time, in the event that someone would like to test their skill by playing or singing against it. Similarly, if two competing versions of a piece exist, the playback systems could allow the researcher to make instantaneous aural comparisons, as opposed to laborious visual ones.With the technical challenges for encoding and displaying mensural notation now well in hand, the next step will be: Analysis. One of the most powerful ways in which computers can aid musicology is in comparative study on a scale impossible for individual researchers, allowing us to skip over the initial step of repetitive manual work and moving the starting line for research right up to the edge of notational terra incognita. With even a roughly sketched treasure map of this new land in hand, we will get a better start.The Measuring Polyphony project takes a progressive stand on “open source” as a methodology and philosophy. In addition to the free downloads of all the editions in both notation types, the code and the Sibelius transcriptions are also available on the open-source version of GitHub, a development platform used all over the world to host code and facilitate projects in a way that is open to all.10 This kind of generous scholarship goes hand in hand with Desmond's resolve to make the results of this project available before they are perfect, preferring to solicit help and advice from the scholarly community rather than present a “finished product”—the idea of which, especially in the world of editions, is suspect at the best of times.The astonishing (and daily increasing) number of high-quality manuscript images now available on the Internet ought to constitute nothing but good news for musicological scholarship, but their value seems to be determined by the collective imagination of the researchers who use them. Are these images to be understood as simply electronic facsimiles, now more widely and cheaply available than their hard-copy predecessors? Or are there ways of working with these images that will further scholarship itself? The answer depends on the kind of attention we pay to the music and how intensely we scrutinize the way it is represented on the page. The particular layouts and the tiny horizontal and vertical distributions of notes and text tell us about the way this music was understood, just as any glossing or additional comments written in at a later date give us an indication of its life over time. Measuring Polyphony has grown with the various phases of online image policies over the past five years, building in the ability to refer to online images from IIIF (the International Image Interoperability Framework, a standardized method of providing images and metadata on the Internet),11 as well as from individual library and archive websites.As it stands, Measuring Polyphony is a great pedagogical resource that can be used to introduce the next generation to the intricacies of mensural notation and the complex and astounding music that it encodes. The ability to see each motet in both a representation of the original notation and modern notation puts to rest questions about note values and time signatures in a quick and clear way—something that previous editions have seemed at great pains to do, with mostly underwhelming results. Of course, “digital natives” expect to use the digital images and metadata to power through the spadework of transcribing, to accelerate the business of comparing repertoires for musical and exegetical similarities, and to explore how the manipulation of tempos and tuning systems can be used to render this music more vibrant and alive than formerly.As image-document analysis software becomes more able to deal with all the usual challenges (such as warping of the page, faded ink colors, and marginalia) it will perhaps automatically decipher, identify, and classify mensural notation with only minimal help from a human musicologist to train it on the Franconian rules. Ideally, there would be a way of coupling the input tool, currently under development, with the classroom as well as our colleagues, so that the number of pieces available on the site can continue to grow and inspire our understanding of our musical history.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.