Conference Agenda

Amongst Luvale (and related) communities in Zambia, drummers tune their instruments by applying ulezo (tuning paste) and heat. Heating a drum (either in the sun or by fire) raises the pitch while ulezo lowers it. In addition to adjusting pitch, this two-step process dramatically impacts timbre (Ferreira 2013; Toulson 2021). However, timbre is a complex set of auditory attributes that work in tandem (Eerola et al., 2009; Juslin and Laukka 2003; McAdams 2017, 2019). Different dimensions of timbre could be salient in different contexts and through a combination of statistical acoustic analysis, cultural contextualization, and ethnographic methods, we attempt to determine which of these attributes are relevant to listener perception in drum tuning. In this project, we attempt to trace the timbral effects of tuning in Luvale drumming through attention to both the acoustic and perceptual domains of timbre (Fales 2002). A data set of individual drum tones was obtained from field recordings of multiple dance troupes at various stages of the tuning process. We begin by generating acoustic descriptors for these sounds in the revised Timbre Toolbox (Kazazis et al., 2021) which measure temporal, spectral, and spectrotemporal attributes. We then employ statistical analysis to reveal which acoustic descriptors significantly change throughout the tuning process. Finally, we map these acoustic attributes onto emic vocabulary used by Luvale musicians when tuning and describing the timbre of drums. We demonstrate how the process of tuning changes certain sonic properties which, in turn, influences the perception of various timbral qualia. In essence, this exploratory study translates the acoustic domain into the perceptual domain. Analysis of interviews, concept maps, performed action, and lived experience reveals both the timbral goals of tuning and the ways in which drum timbre is conceptualized amongst Luvale musicians. It adds to the literature of timbre semantics with a novel cultural case study, wide consideration of timbral qualia, blending of quantitative and qualitative data, inclusion of local descriptors, and interdisciplinary methodology.

According to Robert Gjerdingen, “charting the exact musical knowledge of modern listeners would be a daunting task. Musics from popular culture, musics from the near and distant past, and musics from various ethnic regions coexist today in a vast commercial marketplace of sound” (1996, 380-381). Music informatics researchers have attempted to address this issue by developing large-scale corpora that represent musical data in machine-readable symbolic and audio formats (e.g., the Million Song data set; Bertin-Mahieux et al. 2011). Nevertheless, the size, scope, and format of these projects often require extensive computational training (e.g., Dang et al. 2012). As a result, music corpora often eschew the kinds of musical engagements favored by scholars in music theory and (ethno)musicology.

To address these issues, the first part of this paper presents the Music Informatics for Radio Across the GlobE MetaCorpus (MIRAGE-MC), a data set consisting of music meta data for one million songs monitored across 10,000 Internet radio stations around the globe during the months October-January 2022-2023. For each song, MIRAGE-MC includes meta data pertaining to the radio station (e.g., city, country), the encoding stream (e.g., name, genre, url), and the song itself (e.g., artist name, song title), along with links to audio recordings on streaming services like Spotify and YouTube. MIRAGE-MC also features an online dashboard that allows researchers with no computing background to develop their own corpora by selecting and exporting data for further analysis and study.

To demonstrate the potential of MIRAGE-MC for the music theory community, the second part of this paper examines the song list from Best FM 104.1 in Johor Bahru, Malaysia, which features primarily Malaysian and Indonesian artists and both western (pop, rock, hip-hop) and southeast Asian (dangdut, kronkong, campursari) popular music genres, thereby offering several striking examples of the “compromise between western formal influence and local materials” (Moretti 2000). The paper concludes by considering the methodological challenges imposed by Internet radio, a music-listening format that places diversity center stage.

Pitch centricity is commonly conceived as the focus upon a particular pitch class above others in a given musical passage. This paper refines that conception by explicitly repositioning pitch centricity not as a property of music itself, but rather as a product of the perceptual processing of music. Such an approach privileges the varied perspectives created by individual listeners’ modes of engagements with particular musical contexts—rather than saying “this music is centered on A,” this posture emphasizes the advantages of saying “I/you/we hear this music centering on A.”

Locating pitch centricity in a given musical experience allows for observing and classifying the degrees and types of listeners’ engagement with this critical perceptual component. Any musical encounter might involve conscious attention to centricity, subconscious acknowledgement of centricity as a background factor providing context to other musical phenomena, or the decision (conscious or not) that centricity doesn’t take hold. Because this model stations centricity in the perceptual process rather than in the music itself, centricity’s role ebbs and flows according to the idiosyncrasies of a musical stimulus, the kind of attention given it by the listener, and the acculturation that colors the listener’s parsing of it.

This paper posits four modes of engagement with pitch centricity that are reflexively applied by listeners. These modes are not pristinely discrete from one another, and listeners move freely among them—with significant perceptual consequences. Further, any of these modes might emerge in any musical context, though particular musical features catalyze otherwise “neutral” listeners into favoring particular modes. The capturing of pitch centricity’s dynamism in this four-mode model illuminates its impact on the listening/processive experience.