Introducing Mattheson: A Proof-of-Concept App for Researching Affective Accountability
I have recently begun developing Mattheson, an experimental app intended to support my research into “affective accountability” in musical performance. Mattheson is currently a proof of concept, and both the software and the experimental method are still in the early stages of development and testing. I am sharing this initial demonstration to test the basic idea, identify technical and methodological problems, and gather feedback from friends, colleagues, musicians, researchers, and interested listeners.
The app is my own developing venture, bringing together my artistic research, work as a performer, and interest in creating digital tools for the study of historical performance and audience reception. In the future, I hope to use Mattheson as the principal platform for organizing the audience-laboratory components of several artistic-research projects.
What is affective accountability?
I use the term “affective accountability” to describe the responsibility performers have toward their audiences and themselves when attempting to create an affective experience. Historically informed performance is often considered in relation to historical evidence and the performer’s interpretative intentions. Affective accountability introduces a further question: what does the listener actually perceive or experience? This does not mean that an audience should decide whether an interpretation is correct. Listener responses are not votes on artistic quality or objective measurements of musical meaning. They can, however, provide useful evidence about the relationship between intention and reception. If I intend a passage to communicate tenderness, pride, fear, sublimity, or a combination of passions, it is valuable to ask whether listeners perceive anything resembling that intention.
The aim is to place reception within the artistic research process. Historical evidence informs an interpretation; the performer develops and tests expressive strategies; the instrument and performance situation shape the result; and audience responses contribute to further reflection and experimentation.
What does the Mattheson app do?
In its current proof-of-concept form, Mattheson allows listeners to respond to a performance while it is taking place. Instead of asking someone to describe an entire performance afterward, the app records responses at particular moments in the musical timeline.
Participants watch or listen to a recording and tap the passion that most closely corresponds to what they are experiencing at that instant. A quick tap records an immediate response. Holding the button for slightly longer allows the participant to refine the intensity, from mild to strong. If the available categories do not adequately describe the experience, a free-text field can be used instead. Participation is voluntary and anonymous, and nothing identifying the participant is recorded.
The responses can then be displayed alongside the recording. Each dot represents a response at a particular moment, while its size indicates the reported intensity. Responses from different listeners can be compared with one another and with the passions that the performer intended to communicate. At present, these functions are approximate and experimental. The app is being shared to determine whether this basic method is understandable and usable, and whether the resulting information might become meaningful within a larger artistic-research process.
The first experiment: Beethoven’s Op. 26
For this first demonstration, I have selected an excerpt from Beethoven’s Piano Sonata No. 12 in A-flat major, Op. 26.
The experiment uses twelve categories: cheerfulness, melancholy, wonder, sublimity, love, hatred, sympathy, pride, anger, gratitude, desire, and fear. This vocabulary is adapted from Thomas Brown’s writing on the emotions from 1820 and places the experiment in dialogue with historical attempts to classify and understand the passions.
Such taxonomies are never neutral or universal. Historical writers divided and named the passions in different ways, and their terminology may not correspond exactly to how we understand our own experiences today. Part of the experiment is therefore to discover how usable—or unusable—this particular historical vocabulary might be for present-day listeners.
Before publishing the demonstration, I marked the passions that I believed I was attempting to communicate throughout the excerpt. These “intended passions” appear as shaded areas on the results graph. These markings are necessarily approximate. I entered them according to how I personally understood and felt the changing character of the passage. They are not definitive answers about what Beethoven’s music means, nor do they claim to represent the only historically plausible interpretation. Several passions may overlap, change rapidly, or resist precise classification.
The comparison between my intended passions and the listeners’ responses is therefore a starting point for enquiry. Where do intention and reception coincide? Where do they diverge? Do several listeners respond similarly at the same moment? Are some expressive intentions communicated more consistently than others? What might these results suggest about timing, articulation, dynamics, gesture, rhetoric, or the character of the historical instrument?
Take part in the experiment
You can participate anonymously here:
Participate in the Beethoven Op. 26 experiment
Begin the recording and tap whichever passion best describes what you feel as the music unfolds. There are no correct answers: I am interested in your immediate experience of the performance.
You can explore the developing results here:
The results page displays responses across the musical timeline. Larger dots represent greater reported intensity, overlapping responses become visually darker, and the shaded bands show my approximate intended passions. The interface also offers an initial comparison between intention and audience response.
The next stage
Mattheson remains a proof-of-concept app rather than a finished or validated research tool. This demonstration is intended to test the software, the response method, the historical vocabulary, and the way the resulting information is presented.
I would therefore be very grateful for feedback about any part of the experience: the instructions, the emotional categories, the response controls, the graphs, the usability of the interface, or the broader research concept. Reports of technical problems are equally welcome.
I hope to develop Mattheson further through a series of experiments and eventually use it as an important part of upcoming artistic projects. It could support live and recorded audience laboratories, comparisons between contrasting interpretations, and repeated cycles in which audience responses contribute to the continuing development of a performance. For now, this first Beethoven experiment offers an opportunity to test the central idea, discover its limitations, and begin considering what this approach might contribute to the study and practice of musical expression.