Interviewed and edited by Helen Wu
Ève Poudrier is an assistant professor of music theory at UBC School of Music. Her recent projects include the Polyrhythm Project, a collaborative work focusing on the human experience of complex musical rhythms, how people perceive and understand “polyrhythm,” the superposition of rhythms that are not simple manifestations of the same meter. The project approaches the development of notated Western classical music from a geographically and historically informed perspective, integrating listener feedback to explore the relationship between musical structure and human perception. Poudrier combines close score analysis, computer-assisted musicology, and behavioral experimentation in her research. The research platform and tools were realized in collaboration with Craig Stuart Sapp (CCARH/PHI, Stanford University).
Bryan Bell is a fourth-year PhD student in music theory at UBC School of Music, research assistant, and collaborator for the Polyrhythm Project. His primary area of research is rhythm and meter, employing digital corpus methods alongside empirical approaches to analyze large datasets of musical scores, recordings, and performance data, as well as conducting experiments and observations to understand how music is created, perceived, and used.
What is the traditional method in music research?
Ève Poudrier: The traditional method for music theorists is generally based on prior knowledge and what they hear. It’s score study, very similar to close reading in the older literary scholar tradition. They come up with some kind of explanation for a compositional practice, and these observations and explanations are mostly supported by the scholar’s own intuitions.
How did digital scholarship come into play in your research, especially in the context of your polyrhythm project?
Poudrier: I was working with the music of American composer Elliott Carter (1908-2012) for my dissertation. His music has a lot of polyrhythms with an ungrounded feeling. In my literature search, I found Louis-Marc Suter’s study on polyrhythm which featured 719 samples from 450 works of western classical music from 1900 to 1950. Part of empirical musicology is to look at a large set of materials, analyze them and extract information which might answer the questions researchers might have about style. Suter published many examples to show the development of polyrhythmic technique within each composer’s output, with 20 composers in total, but it was before personal computers became common.
For me, it was a large set of data that I could examine polyrhythm with. However, it’s quite a lot, and you wouldn’t want to prepare all the materials by hand. I was interested in using computational methods for that, encoding the music in a language that a computer can work with.
[Vdeio] Ève explaining how computational methodology is essential to her polyrhythm research
How did computational methodology enhance your music research?
Bryan Bell: Like Poudrier said, one factor is the size of data. Using computers or digital methods expands the amount and the complexity of the measures that you’re doing research with. For Poudrier’s corpus, which has over 700 samples in it, it would take hours to go through by hand, especially given the complexity of the data. Even though it’s also time consuming to implement the analysis into the computer, once you have it, the computer will apply it to an arbitrarily large piece of data, which is still much faster.
There is also the public sharing aspect. Today, the default for sharing with people is online, through platforms like Open Science Framework or GitHub. It’s common for digital musicology scholars to launch their own websites and have interactive projects online like Poudrier’s, where more people can access and interact with the music samples.
What specific digital tools or software did you use for this project?
Poudrier: The Polyrhythm Project website connects with Verovio Humdrum Viewer (VHV), an online tool that allows researchers to view, edit, analyze music in Humdrum files (sheet music encoded as structured text data) and turn it back into standard-looking sheet music. You can see what the encoding looks like on the left side – this is what the computer reads; on the right side is the human notation of the music.
Verovio Humdrum Viewer (VHV) interface.
Bryan mentioned sharing your research with other people. In what way can people benefit from research using digital methodologies?
Poudrier: I want to prepare corpora that are not just for my own use, but potentially for other people as well. For example, the composers can use the tool for their own compositions by studying other composers’ polyrhythmic works; there might be educators who might use the website for educational purposes.
The project involved a lot of metadata. It’s not just the name of the composer, the year it was composed, but detailed information about the works that could be of interest for musicians and scholars. Users can search by instrumentation, length, time signature, or the year of composition, premiere or publication.
There is also a timeline map, which shows where in the world these pieces were first heard. Besides premieres in North America and Europe, there is one in South America and another in Australia. Some people might be curious: what work in this corpus was first heard in South America? They click on that link and see that it was Aaron Copland, an American composer, premiering his piano sonata in Buenos Aires in Argentina.
Navigating the search tool.
Timeline map showing detailed information about the premiere of each piece in the corpus.
Is this kind of computational methodology commonly used in music research?
Poudrier: When I was doing my own PhD, computational methods were not quite accepted as a fundamental methodology for music theorists. Nowadays there is some understanding of music cognition and computational musicology, which is what Bryan and I do. But it’s still hard for our work to be accepted as true music theory studies because our primary area of research is interdisciplinary: it overlaps with computer science, music psychology, and music engineering. Now it’s a quite bustling field though, especially in Europe, where they do Music Information Retrieval, which includes computers extracting data from music.
Bell: There is also some regionalism to the discipline. Things like grant funding are easier to secure in Europe than in North America. Some governments are more hesitant to fund humanities-style research in general, even if it is digital.
What does digital scholarship mean to you as a music researcher?
Poudrier: With digital scholarship, we now can share the research project with a larger public, whose interest and use of what I contribute might be beyond my specific research question. It also means collaboration with people in different areas, getting to something that none of us individually could have produced. The Polyrhythm Project involves a lot of people: digital specialists who curate this encoding/computational language, scholars experienced with both music cognition and computational analysis, and research assistants, now collaborators, like Bryan. There are a lot of students at different stages of their studies; some of them had no prior experience or knowledge about encoding, and they learnt it with us.
Bell: I went into music theory because I was interested in certain questions about music. When I started out on a traditional path, I became a bit disappointed with the answers I could get. So I pivoted into digital and empirical methods, not for the sake of digital scholarship but for the types of answers that it can provide to the research questions that I’m interested in, such as why music is structured the way that it’s structured, why it sounds the way it does, and how that relates to the experiences that people have listening to it. For me, digital scholarship is more satisfying in terms of the potential that it has and the answers that it offers.
Dr. Poudrier, what’s an upcoming project for you?
Poudrier: I’m starting a project on Francophone folk songs, which have not been much examined from the perspective of musical structure. The interesting part is that, in developing an expertise in digital scholarship through the polyrhythm project, there are all kinds of new topics and new questions that become possible to explore. I would never have embarked on the Francophone folk songs project had I not had the experience of working on the polyrhythm project. The experience of doing digital scholarship has opened a lot of possibilities for me.