Perceptual fusion of musical notes by native Amazonians suggests universal representations of musical intervals

Music, a universal human phenomenon, exhibits cross-cultural regularities suggesting underlying biological constraints. One potential constraint arises from harmonicity—the presence of frequencies that are integer multiples of a fundamental frequency (f0)—a prominent feature in natural sounds. Harmonicity aids sound segregation, allowing listeners to distinguish concurrent sound sources. In Western music, consonant intervals (note pairs with f0s related by simple integer ratios) are perceived as pleasant and often exhibit spectral properties similar to a single harmonic sound. This has led to the hypothesis that harmonicity might influence musical structure and perception. However, consonance preferences vary cross-culturally. This study explores the potential universal perceptual mechanisms underlying the phenomenon of fusion (perceiving concurrent notes as a single sound), which has been linked to consonance in Western music. The researchers hypothesized that fusion, a consequence of harmonicity-based sound segregation, might be present cross-culturally, even in groups with limited exposure to Western harmony. The Tsimane', a native Amazonian group with minimal exposure to Western music and lacking consonance preferences, provided a unique population to test this hypothesis. Understanding the interplay between universal perceptual mechanisms and cultural influences on musical perception is crucial to explaining cross-cultural regularities and variations in musical systems.

The relationship between harmonicity and consonance in Western music has been extensively studied, with evidence suggesting a link between simple integer ratios of frequencies and perceived consonance. However, the extent to which harmonicity fully predicts consonance preferences remains unclear, due to complexities in auditory system representation of harmonicity and the influence of factors like harmonic position within the series and musical experience. Cross-cultural research on music perception is limited, raising questions about the universality of harmonicity's influence on musical behavior. Prior work with the Tsimane’ showed a lack of consonance preferences, despite some sensitivity to harmonicity, suggesting potential cultural influences on aesthetic judgments. The relationship between consonance and fusion, another consequence of harmonicity, is also debated. Existing research on fusion has largely focused on Western listeners, and the extent to which fusion reflects learned schemas or universal perceptual constraints remains unclear.

The study involved two participant groups: US non-musicians and Tsimane’ from Bolivia. Both groups participated in fusion experiments where they judged whether stimuli (concurrent notes separated by various intervals) sounded like one or two sounds. The intervals included both consonant and dissonant intervals, alternating in size. The study used both synthetic tones (with just intonation and equal temperament tunings) and sung notes to assess the robustness of the findings across different acoustic stimuli and tuning systems. Control experiments assessed task comprehension by using concurrent talkers and harmonic/inharmonic sung vowels. Separate preference experiments used the same musical intervals to rate the pleasantness of each sound. The same experiments were conducted online with Western participants to increase sample size and reliability. Statistical analysis included non-parametric tests (Wilcoxon signed-rank, permutation tests) for fusion data and parametric tests (t-tests, ANOVAs) for preference data. Split-half reliability was assessed for individual differences in fusion and preference.

Both Westerners and Tsimane’ participants showed a higher degree of fusion for consonant intervals compared to dissonant intervals, irrespective of the tuning system used with synthetic notes. This suggests a universal perceptual mechanism underlying fusion based on harmonicity. However, consonance preferences, as measured by pleasantness ratings, differed significantly between the groups. Westerners showed a clear preference for consonant intervals, while Tsimane’ did not. The pattern of fusion across intervals was similar across groups, with the octave showing the strongest fusion. Importantly, in Westerners, the pattern of fusion did not perfectly mirror the pattern of consonance preferences at the level of individual intervals, suggesting that consonance involves factors beyond mere fusion. Individual differences in fusion did not predict consonance preferences in Westerners, even with a larger online sample. The online study corroborated the findings from in-person experiments and revealed a correlation between average fusion and consonance across intervals, but with notable dissociations in the case of the octave and fifth, indicating that consonance is influenced by factors beyond simple fusion.

The findings support the hypothesis that basic perceptual mechanisms for sound segregation, relying on harmonicity, are cross-culturally consistent. The similarity in fusion judgments between the two groups, despite their vastly different musical backgrounds, suggests a universal basis for the phenomenon. However, the divergence in consonance preferences highlights the role of cultural experience in shaping aesthetic responses. The lack of correlation between fusion and consonance preferences at the individual level in Westerners further strengthens this conclusion, indicating that consonance likely involves other factors, possibly cultural learning and familiarity. While harmonicity-based sound segregation might influence cross-cultural regularities in musical systems, the results show it does not directly determine aesthetic associations, which are shaped by cultural factors.

This study demonstrates cross-cultural consistency in the perceptual phenomenon of fusion for consonant musical intervals, pointing to universal perceptual mechanisms related to harmonicity. However, aesthetic preferences for consonance are strongly influenced by cultural experience. The dissociation between fusion and consonance in Westerners emphasizes the multifaceted nature of musical perception. Future research should investigate other aspects of harmonicity, such as melodic intervals, and further explore the influence of cultural factors on musical experience.

The study focused on a limited set of musical intervals, and future research might consider a broader range of intervals and musical contexts. The limited sample size of the Tsimane’ participants might restrict the generalizability of findings to other Amazonian groups. While control experiments addressed task comprehension, individual differences in attention and task understanding could still influence results. The online experiment, while increasing sample size, might have introduced biases due to participant self-selection and online environment.