Researchers have developed a groundbreaking model to explain how humans perceive harmony in music, drawing on a novel approach called spectral knowledge representation. This model, based on the oscillatory behaviour of the brain, has the potential to redefine our understanding of musical perception by providing a mathematical framework for analysing pitch, chords, and key distances. The study, published in the Journal of Cognition, integrates findings from cognitive neuroscience and music theory, and its results align closely with empirical data on how humans experience harmony.
The research, conducted as part of a broader programme aiming to link psychological phenomena with physiological processes, uses the resonance space model to represent sounds as combinations of oscillatory functions called resonances. These resonances mimic the brain’s response to sound, creating a framework to study musical perception. Key to this approach is the understanding that resonances are the brain’s way of encoding sensory experiences, and that they can be recalled through sympathetic resonance, much like how memories are triggered.
In particular, the model successfully simulates the perception of musical features such as tone and chord affinity. By mapping the brain’s perception of harmony onto a geometric structure called resonance space, the researchers show how categories of musical elements, such as keys and chords, can be represented as regions in this space. The study proposes that harmonic structures, including the well-known major and minor keys, can be mapped to specific resonances, thus explaining their relationship in cognitive terms.
One of the most significant findings is the model’s ability to explain key affinity – the degree to which a tone fits within the context of a key. The model demonstrated a strong correlation with established empirical profiles, particularly those of Krumhansl and Kessler. These researchers had previously shown that certain tones are perceived as more consonant or dissonant depending on the key in which they are heard. The new model mirrors this understanding by using cosine similarity between harmonic spectra to measure how well a given tone aligns with a key. The results show that the model is competitive with other leading theories in the field, achieving a correlation coefficient of 0.95 with empirical data.
Further work will explore how this mathematical model can be applied to other auditory phenomena, including rhythm and metre, which may also be explained through the same resonance-based framework. This is particularly relevant for understanding how the brain processes sequences of musical events over time, an area the researchers plan to investigate by extending the time-based elements of their model.
The implications of this research are vast, offering a unified approach to understanding music perception. By grounding their work in both cognitive science and mathematical principles, the researchers hope to build a more comprehensive theory that can account for the complex ways in which humans perceive and experience music. This model opens up new avenues for exploring how the brain encodes, processes, and recalls auditory information, potentially influencing the development of AI systems that mimic human auditory cognition.
