Following an audition with artistic director Neville Creed on the 24th of June, 2019, Courtney will join the London Philharmonic Choir as a full member. Courtney passed the audition for the full choir and the semi-chorus, and will have her first performance with the second sopranos in early September at the BBC Proms. The LPC will sing in the European premiere Prom 66: John Luther Adams’s In the Name of the Earth at 11:00 Sunday 8th September, 2019.
Courtney will be a speaker at a workshop-symposium in cross-cultural research methods in music and emotion, presented by the The Music and Science Lab of Durham University’s Department of Music.
The Workshop-Symposium on Research Methods in Music and Emotion will address methodology and research practices in empirical work with the aim to enhance music emotion research by adopting practices from anthropology and ethnomusicology. The one-day symposium will take place on the 14th of September, 2019 at St.Chad’s College, Durham University.
Title: The Listening Participant and The Living Instrument: A Thematic Analysis of Why Listeners Actively Annotate Emotion Changes Through Live Performance
Author: Courtney N. Reed, Simin Yang, Elaine Chew, Mathieu Barthet
Abstract: The most widely used tool in music emotion research (MER) is the listening test; participants listen to a given excerpt while reflecting on its intended emotion or their own induced emotion. Existing research using listening tests has determined why musical features, such as harmonic progression, dynamic changes, timbre, and tempo are associated with different emotions and emotional changes; however, isolating these features within musical context is a widely recognised challenge in MER.
Alternative to focusing on individual features, an empirical study was conducted to determine the high-level reasoning behind why participants feel compelled to make emotion-related annotations. In a traditional listening test setting, 21 participants with varying musical backgrounds annotated varying emotion points through a recording of a live classical trio; after, they were asked to reflect on why they made each annotation.
Through Thematic Analysis of these open-ended reflections, we have underpinned the logic behind these annotation choices and identified seven key themes in participants’ reflections that cause them to actively make annotations about the music’s perceived emotion: Instrument Entities, Expectation & Violation, Discrete Features, Instrumentation, Musical Structures, Performer Expression, and Visual Cues.
Instrument Entities in particular presents a novel view of the listener perspective through interaction with “living instruments.” In describing musical change, participants were found to use language and metaphor which mirrors that used to describe perceived emotion in human interaction; instruments were personified and viewed as inherently having emotions and wanting to communicate them to the audience through their sound.
This focus on high-level themes and the core communication relationships that drive participant engagement proposes a more feature-driven methodology for study design, modelling of behaviour, and analysis in MER. Listening tests can thus be developed with this aim to collect more annotations from participants and therefore gather larger, more relevant data for future MER applications.
The following research will be presented at the seventh Seminar on Cognitively Based Music Informatics (CogMIR) 2019 at the Tow Center for the Performing Arts, Brooklyn College (Brooklyn, NY, USA) on the 8th of August, 2019.
Ms. Simin Yang (first author) will give an oral presentation of work done in collaboration with Ms. Courtney N. Reed, supervised by Professor Elaine Chew and Dr. Mathieu Barthet.
Title: Listener-informed Features for Time-varying Emotion Perception in Live Music Performance
Authors: Simin Yang, Courtney N. Reed, Elaine Chew, Mathieu Barthet
Abstract: The purpose of this study was to gain a deeper understanding of the factors that influence time-varying emotion perception in music performance. A two-stage study was launched to investigate the reasons behind why listeners produce different emotion annotations in a complex classical music piece. In an initial experiment, we collected time-varying emotion ratings (in valence and arousal dimensions) from listeners during live classical trio performance; in a follow-up experiment, we further explored the reasoning behind listeners’ time-stamped emotion ratings through the re-evaluation of seven pre-selected music segments of various agreement levels informed from the initial study. 15 and 21 participants with varying degrees of music training were involved in the initial and follow-up experiments, respectively. Thematic analysis of the time-stamped annotation explanations revealed themes pertaining primarily to commonly discussed musical features including dynamic change, tempo, and melody as the main factors influencing emotion perception. The analysis also uncovered prominent listener attention to the personification of instrument entities and their interactions, melody onset and repetition, performance embellishments, and visual cues, which are less discussed in previous music informatics research. With recent advances in MIR subtasks, such as automatic instrument recognition, playing technique detection, and source separation, we propose potential benefits from the incorporation of these listener-informed features into MIR-based audio content analysis.