Large-scale iterated singing experiments reveal oral transmission mechanisms underlying music evolution

Simple frequency ratios naturally make precisely perceived melodies

Almost all human music is built on discrete scales of pitch.1 Culturally prominent scales, such as the diatonic major scale of Western music, make use of the simple frequency ratios 2:1, 3:2, and 4:3 between notes.2 It is generally believed that these ratios were chosen to optimize the consonance of simultaneous notes.3,4,5,6,7 Alternatively, or in addition, it is conceivable that these ratios are intrinsically advantageous for the perceptual encoding of melodies.8,9,10,11,12 Here, we provide behavioral support for this hypothesis. In three experiments, young Western adults had to detect pitch anomalies ("sour notes") in partly random pure-tone melodies based on various musical scales, including novel ones. The task did not require any musical knowledge. Most importantly, the listeners were extensively trained in order to saturate familiarity with the scales: for a given scale and listener, more than 2,000 (up to 5,280) trials were run. Practice largely improved performance. This occurred even for the diatonic major scale, suggesting that performance in our task was not biased by previous musical enculturation.13,14 Frequency ratio simplicity also favored performance. Crucially, its benefit was not smaller in the final test sessions, when performance for each scale was presumably optimal and no longer improvable by practice, than in the initial test sessions. Thus, frequency ratio simplicity appeared to be intrinsically advantageous, rather than advantageous merely due to familiarity. The naturalness of melodic intervals defined by simple frequency ratios is likely to have contributed to the cultural selection of musical scales.

Iterative learning experiments can help elucidate music's origins

Anglada-Tort et al. Current Biology, 33, 1472-1486.e12, (2023) conducted a large-scale iterative learning study with cross-cultural human participants to understand how musical structure emerges. Together with archaeological, developmental, historical cross-cultural music data, and cross-species studies we can begin to elucidate the origins of music.

Large language models predict human sensory judgments across six modalities

Determining the extent to which the perceptual world can be recovered from language is a longstanding problem in philosophy and cognitive science. We show that state-of-the-art large language models can unlock new insights into this problem by providing a lower bound on the amount of perceptual information that can be extracted from language. Specifically, we elicit pairwise similarity judgments from GPT models across six psychophysical datasets. We show that the judgments are significantly correlated with human data across all domains, recovering well-known representations like the color wheel and pitch spiral. Surprisingly, we find that a model (GPT-4) co-trained on vision and language does not necessarily lead to improvements specific to the visual modality, and provides highly correlated predictions with human data irrespective of whether direct visual input is provided or purely textual descriptors. To study the impact of specific languages, we also apply the models to a multilingual color-naming task. We find that GPT-4 replicates cross-linguistic variation in English and Russian illuminating the interaction of language and perception.

Commonality and variation in mental representations of music revealed by a cross-cultural comparison of rhythm priors in 15 countries

Music is present in every known society but varies from place to place. What, if anything, is universal to music cognition? We measured a signature of mental representations of rhythm in 39 participant groups in 15 countries, spanning urban societies and Indigenous populations. Listeners reproduced random 'seed' rhythms; their reproductions were fed back as the stimulus (as in the game of 'telephone'), such that their biases (the prior) could be estimated from the distribution of reproductions. Every tested group showed a sparse prior with peaks at integer-ratio rhythms. However, the importance of different integer ratios varied across groups, often reflecting local musical practices. Our results suggest a common feature of music cognition: discrete rhythm 'categories' at small-integer ratios. These discrete representations plausibly stabilize musical systems in the face of cultural transmission but interact with culture-specific traditions to yield the diversity that is evident when mental representations are probed across many cultures.

Perceptual (but not acoustic) features predict singing voice preferences

Why do we prefer some singers to others? We investigated how much singing voice preferences can be traced back to objective features of the stimuli. To do so, we asked participants to rate short excerpts of singing performances in terms of how much they liked them as well as in terms of 10 perceptual attributes (e.g.: pitch accuracy, tempo, breathiness). We modeled liking ratings based on these perceptual ratings, as well as based on acoustic features and low-level features derived from Music Information Retrieval (MIR). Mean liking ratings for each stimulus were highly correlated between Experiments 1 (online, US-based participants) and 2 (in the lab, German participants), suggesting a role for attributes of the stimuli in grounding average preferences. We show that acoustic and MIR features barely explain any variance in liking ratings; in contrast, perceptual features of the voices achieved around 43% of prediction. Inter-rater agreement in liking and perceptual ratings was low, indicating substantial (and unsurprising) individual differences in participants' preferences and perception of the stimuli. Our results indicate that singing voice preferences are not grounded in acoustic attributes of the voices per se, but in how these features are perceptually interpreted by listeners.

Group singing is globally dominant and associated with social context

Music is an interactive technology associated with religious and communal activities and was suggested to have evolved as a participatory activity supporting social bonding. In post-industrial societies, however, music's communal role was eclipsed by its relatively passive consumption by audiences disconnected from performers. It was suggested that as societies became larger and more differentiated, music became less participatory and more focused on solo singing. Here, we consider the prevalence of group singing and its relationship to social organization through the analysis of two global song corpora: 5776 coded audio recordings from 1024 societies, and 4709 coded ethnographic texts from 60 societies. In both corpora, we find that group singing is more common than solo singing, and that it is more likely in some social contexts (e.g. religious rituals, dance) than in others (e.g. healing, infant care). In contrast, relationships between group singing and social structure (community size or social differentiation) were not consistent within or between corpora. While we cannot exclude the possibility of sampling bias leading to systematic under-sampling of solo singing, our results from two large global corpora of different data types provide support for the interactive nature of music and its complex relationship with sociality.

Music cognition: Can you carry over that tune?

Almost universally, music uses scales consisting of a small number of notes. Could this increase the fitness of melodies for oral transmission? By reproducing the process online, a new study reveals how cognition, sound and culture may interact to shape music.