Musical structures and cardiac arrhythmia
Senior CNRS researcher and pianist, Elaine Chew joined the Music Representations team at the research lab STMS in the spring of 2019. She conducts and leads scientific research on the mathematical representation and computational analysis of musical structures, as applied to music analysis, cognition, expressivity, composition/improv, and electrocardiographic traces of arrhythmia. We met with her for this interview.
Elaine Chew, you are a winner of the European Research Council'sAdvanced Grant 2018 call to lead the COSMOS (Computational Shaping and Modeling of Musical Structures) project. You have chosen the CNRS and joined the Musical Representations team of the UMR Sciences and Technologies of Music and Sound (STMS) at Ircam to conduct this research. Can you explain this choice to us?
The joint research lab STMS at Ircam is a historic and internationally renowned center for music computing research that is deeply rooted in musical practice. It is home to the world’s largest concentration of full-time researchers working on all aspects of music, with high quality training programs for junior researchers and for composers. There are not many places in the world where music research can be carried out in such a favorable environment. STMS brings together the CNRS, Ircam, Sorbonne University, and the French Ministry of Culture. Of these, the CNRS is host to the largest number of ERC grantees, with the excellent infrastructure and experience to support ERC research.
My own connection to Ircam dates back to 2002 when Michel Fingerhut, then Director of the Ircam Médiathèque and illustrious organizer of an early International Conference on Music Information Retrieval (ISMIR), contacted me at the University of Southern California (USC) in Los Angeles. Together with my colleagues Shrikanth Narayanan and Jay C.-C. Kuo, we had just won only the second grant awarded by the National Science Foundation (NSF) for MIR research, and Michel wrote to extend an invitation for us to attend ISMIR. It was around that time that he introduced me to Gérard Assayag, Head of the Music Representations Team. Since then, many collaborations and exchanges have taken place between my research labs at USC, and later Queen Mary University of London (QMUL), and the Music Representations Team at Ircam, including invited presentations and concerts and joint organizing of the international workshop Mathemusical Conversations at the National University of Singapore. Most recently, I was a member of Ircam’s Scientific Council and had the opportunity to stay up to date with the STMS lab’s latest developments.
Given the research context and professional history, it was natural to choose to base the ERC AdG project COSMOS at the CNRS and STMS.
COSMOS aims to study musical structures created during performances but also during cardiac arrhythmia sequences. How do you connect musical structures and cardiac arrhythmia?
Cardiac arrhythmia is a medical condition caused by problems with the electrical wiring of the heart. This results in abnormal heart rhythms, which can be very musical. There are many rhythmic and temporal similarities between cardiac and music signals. For example, atrial fibrillation (AF), a common arrhythmia that afflicts over 8.8 million people over the age of 55 in the EU, has an irregular rhythm that can be likened to free jazz or some contemporary classical music. However, because it is not common for the fields of music and cardiology to cross, the methods of representing information in music and heart signals, and analytical approaches, have developed entirely separately.
When I was a graduate student and being coached in chamber music by violinist Lynn Chang at MIT, he used to speak of the importance of the “divine pulse”—we were playing music with a pulse. Rhythm, pulse, timing are some of the most visceral aspects of music performance, some might even say superseding pitch, harmony, and melody. Over the centuries musicians, who are the purveyors of time, have developed numerous ways to represent and talk about rhythm, broadly defined. With the advent of computers and digital technologies, music scientists have likewise created tools and analytical techniques to quantify properties of, and variations in, musical time and frequency, which are especially important in performance studies.
How musicians conceptualize and organize time can potentially benefit the field of e-cardiology. For example, arrhythmias are broadly categorized by rate—tachycardias are fast and bradycardias are slow—regularity—tachycardias are fast and regular, fibrillations are fast and irregular—and source—atrial, ventricular, or junctional. No self-respecting musician would be satisfied describing music as fast, regular, and primarily in the treble registers. To be able to characterize arrhythmic behaviors beyond the existing crude classifications using musical descriptors can potentially help stratify arrhythmias into finer subgroups that relate better to symptoms and treatment outcomes.
You are a researcher but also a virtuoso pianist. How do you reconcile your research activity with that of a musician?
To me there is no intellectual conflict being a researcher and a musician—being a musician makes me a better researcher, and a researcher a better musician—the only constraint is time—time to do the research as well as practice—and perhaps overcoming preconceived notions of what a researcher is (i.e. not a very good musician) and what a musician is (i.e. more often an intuitive rather than an analytical thinker, notable exceptions being Pierre Boulez, Leonard Bernstein, etc.).
Elaine Chew at the official opening of the Kunsthall Stavanger Museum of Modern Art in Norway, 2013
The act of performing carries with it a wealth of embodied musical knowledge that goes well beyond the physical action of playing an instrument. Great performers are also incisive thinkers—able to take apart the music and put it together in a way that makes sense—and articulate communicators—capable of deciding how best to convey new ways of hearing to the listener and to execute it on their instrument. This knowledge is often unexplained and relegated to the realm of intuition. Much of my research activities seeks to uncover and make concrete this hidden knowledge.
Your participation on March 12, 2019 in the "Women in Mathematics Days" organized by the London Mathematical Society seems to testify to your interest in the place of women in research. What is your position on the subject?
There are three senior female researchers at the STMS, myself included, which is a little under 10% of the research staff. The STMS lab itself is headed by a female Director, Brigitte d’Andrea-Novel, who is Professor at the Sorbonne University and École des Mines, a control theory specialist, and director of the Musique aux Mines festival. Worldwide, recognition for women in Science and Engineering (WiSE) as well as for the importance of female composers and conductors are gaining traction. In addition, female-male parity and equal treatment of women are currently central issues at the CNRS as evidenced, amongst other things, by an open letter from the President and CEO Antoine Petit to all female staff in July of this year, encouraging eligible candidates to apply for competitions, promotions, and management positions, so the situation is expected to improve.
Lack of female peers while studying as the sole or only one of two women in high school computer science, undergraduate mathematical sciences, and PhD operations research courses did not bother me as much as when I was working professionally as a female researcher and witnessed firsthand how women are disadvantaged and held back from advancement by simple things such as being assigned ‘housekeeping’ work rather than more visible ‘glamour’ work. I could not help but be interested, perhaps self-interested, in the prospects of women in research.
As a beginning Assistant Professor, I benefited from enlightened leaders and generous WiSE programs that sought to level the playing field. I have also seen how the presence of female (and other kinds of minority) role models makes a difference in changing the research landscape by challenging stereotypes and perceptions of leadership. It has been gratifying to see that many of my former students and postdocs (both female and male) have gone on to become dynamic WiSE champions, founding and running the Women in MIR (WiMIR) mentoring programs and workshops. They have accomplished far more than I could ever have done.
For an organization to function optimally in a competitive environment, it must be able to benefit fully from the skills and strengths of both women and men. Research has shown that women in the boardroom improve strategic decision making and business performance. Ensuring that all voices are represented is not only the job of women; it is important that both men and women work together to make sure that this happens.
Photo: Elaine Chew au Women in Mathematics Days, 2019