What is the function of music, how did it originate and evolve, and how has it changed us over time? The proposed mechanisms underlying the evolution of musicality in both Savage et al. and Mehr et al. are supported with cogent arguments and convincing evidence. Both discuss important parallels between human music and animal models of vocal learning and coordination. Complementing this discourse, we suggest additional insights drawn from birdsong neuroethology and evolution that support the broad ideas presented in these articles.

Birdsong has evolved under selection pressures associated with multiple functions, including sexual selection, within-species interactions, and species recognition. Songbird behaviors may provide insights into neural mechanisms of coordination between individuals during music production that are currently difficult to study in humans. For example, duetting is a relatively rare phenomenon in birds, involving two individuals producing a temporally coordinated song consisting of overlapping or alternating phrases (Hall, Reference Hall2009). Recently, researchers have simultaneously recorded vocalizations and electrophysiological signals in duetting pairs of birds, finding correlated neural activity between paired individuals. During a duet, individuals modulated their vocalization tempo to fit into the silences between their partner's vocalizations, just as humans can use auditory feedback to match an external tempo. Interestingly, these birds exhibited premotor neural activity both while singing their own portions of the song and during their partner's parts of the song (Hoffmann et al., Reference Hoffmann, Trost, Voigt, Leitner, Lemazina and Sagunsky2019). A similar phenomenon is observed in zebra finches producing single-note calls to establish social contact: Individuals dynamically modulate the tempo of their calls to fall in the silences between their partner's calls and, again, inhibition of call production is associated with premotor neural activity that anticipates a partner's vocalization (Benichov & Vallentin, Reference Benichov and Vallentin2020; Benichov et al., Reference Benichov, Benezra, Vallentin, Globerson, Long and Tchernichovski2016). In other words, remaining silent to accommodate a vocal partner is an active neural process mediated by both auditory and social stimuli.

Songbirds also provide an example of evolution of auditory preferences based on culturally transmitted (i.e., learned) signals. In nature, juvenile songbirds are exposed to many species' songs. To be recognized as the correct species and successfully attract a mate, birds must identify and selectively learn conspecific songs. We posit that this might occur similarly to how, according to Mehr et al., human babies attune to rhythmic vocalizations. In songbirds, these relatively coarse song-selection filters are thought to be innate, because studied species of songbirds react differently to conspecific vocalizations while still in the egg or nest (Colombelli-Négrel et al., Reference Colombelli-Négrel, Hauber, Robertson, Sulloway, Hoi, Griggio and Kleindorfer2012; Hudson, Creanza, & Shizuka, Reference Hudson, Creanza and Shizuka2020; Hudson & Shizuka, Reference Hudson and Shizuka2017) and selectively learn conspecific song without prior exposure to it (Colombelli-Négrel et al., Reference Colombelli-Négrel, Hauber, Robertson, Sulloway, Hoi, Griggio and Kleindorfer2012; Marler & Peters, Reference Marler and Peters1977; Soha & Marler, Reference Soha and Marler2000). Selectivity in song learning varies widely between species and is based on species-specific song features, which can be sound properties of individual syllables (e.g., timbre or pitch modulation), or the pattern of these syllables. For example, white-crowned sparrows attune to a pure-tone whistle at the beginning of songs, whereas swamp sparrows, which sing a trilled song of multiple frequency sweeps, selectively learn conspecific syllables regardless of temporal organization (Marler & Peters, Reference Marler and Peters1977; Soha & Marler, Reference Soha and Marler2000). In contrast, zebra finches raised by Bengalese finches learn Bengalese finch syllables but transpose these syllables to match a typical zebra finch song temporal pattern (Araki, Bandi, & Yazaki-Sugiyama, Reference Araki, Bandi and Yazaki-Sugiyama2016). These findings suggest that, for zebra finches, there is an innate template for rhythm of syllable production. Understanding how the brains of young songbirds selectively attune to certain auditory patterns may shed light on how human brains attune to rhythm.

In addition to their diversity of innate auditory preferences, songbirds exhibit wide variation in the duration of their song-learning window. We recently modeled and analyzed the evolution of songs alongside the duration of song learning; we found that selection for more elaborate songs can drive the evolution of the capacity to learn throughout life (Creanza, Fogarty, & Feldman, Reference Creanza, Fogarty and Feldman2016; Robinson, Snyder, & Creanza, Reference Robinson, Snyder and Creanza2019). We propose that this evolutionary paradigm in songbirds – that selection on a learned trait can drive evolution of the brain – provides a possible example of the phenomenon depicted in Savage et al. (Fig. 2, left panel): Musical features can act as an intermediary between social functions and their neurobiological underpinnings.

Savage et al. describe musicality as a “cognitive toolkit.” How might the framing of musicality as a set of tools affect our understanding of its evolution? Our lab modeled the evolution of birdsong features as culturally transmitted functional traits, similar to tools, wherein learners aim to imitate proficient tutors (Hudson & Creanza, Reference Hudson and Creanza2021). Like other fitness-altering cultural traits, functional signals based on rhythmicity or pitch modulation could have gradually become more complex if learners preferentially choose tutors with complex signals. Over time, the cultural development of functional signals could elevate the minimum cognitive baseline to recognize and reproduce these signals, thereby influencing brain evolution to favor attention to and learning capacity for these acoustic features. In this context, elements of musicality might have been under selection for purposes other than the umbrella explanation of “social bonding.” Savage et al. describe the neural synchronization between auditory and motor brain regions during rhythm perception to explain the origins of dance, but only briefly mention other functions of coordinated behavior. Could rhythmic movement have functioned as a fitness-enhancing tool? Rhythmicity allows for synchronization of actions between individuals and for individuals to accurately predict the actions of others. It is thus conceivable that the development of rhythmicity would have facilitated a large repertoire of coordinated behaviors that could have impacted group survival.

Finally, both target articles discuss the hypothesis that musicality evolved through sexual selection, concluding that it is inadequate to explain the evolution of musicality. However, this hypothesis is framed from an intraspecific mate selection perspective, where females choose males with the most attractive musical displays. Studying the evolution of birdsong and its role in species recognition suggests another perspective: in our evolutionary past, could musicality have served an interspecific function, mediating the interactions between the ancestors of Homo sapiens and other hominin lineages? Although musicality appears to be uniquely human among extant species, Mehr et al. conjecture that the basic elements of musicality are ancestral to all primates – just as song is to all songbirds. Did musicality contribute to species recognition when our ancestors formed groups or selected mates, perhaps before the emergence of language? We are unable to know how much musical predisposition we shared with our evolutionary cousins – those we interbred with, and those we didn't. However, considering songbirds as a model system suggests that the evolutionary implications of musicality need not be limited to interactions within our own species.