Perception of Chord Sequences Modeled with Prediction by Partial Matching, Voice-Leading Distance, and Spectral Pitch-Class Similarity: A New Approach for Testing Individual Differences in Harmony Perception

Tools

Eitel, Matthew; Ruth, Nicolas; Harrison, Peter; Frieler, Klaus and Müllensiefen, Daniel. 2024. Perception of Chord Sequences Modeled with Prediction by Partial Matching, Voice-Leading Distance, and Spectral Pitch-Class Similarity: A New Approach for Testing Individual Differences in Harmony Perception. Music & Science, 7, ISSN 2059-2043 [Article]

Preview

Text
eitel-et-al-2024-perception-of-chord-sequences-modeled-with-prediction-by-partial-matching-voice-leading-distance-and.pdf - Published Version
Available under License Creative Commons Attribution.
Download (865kB) | Preview

Official URL: https://doi.org/10.1177/20592043241257654

Abstract or Description

The perception of harmony has been the subject of many studies in the research literature, though little is known regarding how individuals vary in their ability to discriminate between different chord sequences. The aim of the current study was to construct an individual-differences test for the processing of harmonic information. A stimulus database of 5076 harmonic sequences was constructed and several harmonic features were computed from these stimulus items. Participants were tasked with selecting which chord differed between two similar four-chord sequences, and their response data were modeled with explanatory item response models using the computational harmonic features as predictors. The final model suggests that participants’ responses can be modeled using transitional probabilities between chords, voice-leading distance, and spectral pitch-class distance cues, with participant ability correlated to three subscales from Goldsmiths Musical Sophistication Index. The item response model was used to create an adaptive test of harmonic progression discrimination ability (HPT) and validated in a second study showing substantial correlations with other tests of musical perception ability, self-reported musical abilities, and a working memory task. The HPT is a new free and open-source tool for assessing individual differences in harmonic sequence discrimination. Initial data suggest this harmonic discrimination ability relies heavily on transitional probabilities within harmonic progressions.

Item Type:

Article

Identification Number (DOI):

https://doi.org/10.1177/20592043241257654

Additional Information:

Funding: The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was partially funded by InnovateUK Smart Grant 96040 awarded to SoundOut Ltd. and Daniel Müllensiefen at Goldsmiths, University of London.

Data Access Statement:

We are happy to share the harmony progression discrimination test as an open-access tool for researchers interested in exploring either individual differences in musical abilities or the cognitive underpinnings of harmony perception. An R implementation of the test is available on GitHub (https://github.com/NicolasRuth/HPT), and a demo can be accessed at https://shiny.gold-msi.org/longgold_demo. Due to issues related to the consent obtained from participants the data from this study cannot be made fully open access. However, all data can be made available upon request from the corresponding author in a fully anonymised format.

Keywords:

Chord sequences, discrimination, Goldsmiths musical sophistication index, harmony, implicit learning, individual differences, musical training, pitch class spectral similarity, ppm, voice leading

Departments, Centres and Research Units:

Psychology

Dates: