Although not a novel hypothesis, Savage et al.'s music and social bonding (MSB) is an impressive synthesis of evidence suggesting social bonding is the primary adaptive function for the origin and evolution of musicality in humans. This intriguing idea has seen rapid development in recent years, both in terms of theoretical refinement and increasing evidence across human and nonhuman animals (Oesch, Reference Oesch2019). Putting aside relative strengths, three aspects, relevant to both articles, seem worthy of elaboration.

First, the analogy with vision borders on tautology. In opening, Savage et al. curiously state: “Turning to music, ‘social bonding’ provides an umbrella explanation analogous to ‘vision is for seeing’.” Indeed, vision is for seeing; but testing an adaptive explanation for a trait is to determine what kind of vision is best adapted for a particular organism within a specific ecological niche. For example, colour vision seems to have specifically evolved in howler monkeys in foraging for ripe fruit and leaves, whereas convergently evolved in bower birds for conspecific identification and assessing mate quality during courtship (for a review, see Gerl & Morris, Reference Gerl and Morris2008). Similarly, invoking social bonding as an “umbrella explanation” for the function of musicality blurs the contexts and functions in which musicality exists in human and nonhuman animals. For example, recent evidence suggests song can facilitate both social bonding (Oesch, Reference Oesch2019; Whittingham, Kirkconnell, & Ratcliffe, Reference Whittingham, Kirkconnell and Ratcliffe1997) and courtship in various bird species (Catchpole & Slater, Reference Catchpole and Slater2008), whereas similar effects are observed in humans (Oesch, Reference Oesch, Shackelford and Weekes-Shackelford2020). Although Savage et al. do acknowledge it as “unlikely that a single ‘main’ evolutionary function for complex, multi-component abilities like language or music exists,” we believe an ideal analysis would, when space allows, simultaneously: (1) avoid circular reasoning, (2) invoke substantive data from both human and nonhuman animals, and (3) be based on a distinction between the ancestral, primary, or direct function of a trait or adaptation like musicality and its derived evolutionary function(s) (Oesch, Reference Oesch2019, Reference Oesch, Shackelford and Weekes-Shackelford2020). Fortunately, evolutionary theory offers clear criteria for distinguishing direct from derived functions: shared ancestry and phylogenetic history in the former case, and unshared with a common ancestor in the latter (Ridley, Reference Ridley2003). Moreover, in addition to social bonding, it seems likely that there were several other selective factors and derived functions at play, which Savage et al. at least implicitly acknowledge.

Second, is credible signalling truly disjunct from the social-bonding proposal? Ultimately, cooperation and even signalling are arguably part of the family of hypotheses epitomized by MSB. As such, the target articles appear closer to one another than to hypotheses following Pinker's “cheesecake” analogy. If Mehr et al.'s main thesis is that music-making is a proxy measure for signalled group coherence, then we fail to see how this is not a special case of MSB, albeit that signalling focuses on between-groups and social-bonding on within-group relations. In their group of friends example, surely something is required to explain why they are already closely-bonded enough to walk together in a dangerous neighbourhood at night: manifestations of musicality could themselves plausibly have contributed to the bonding being signalled. In fact, much of the evidence presented by Mehr et al. support a social-bonding interpretation at least as well as a signalling one – if, in fact, the dichotomy is warranted.

Third, although social bonding hypotheses provide a plausible seed for musicality, accounting for specific features of music necessitates deeper examination of the by-products of the gene-culture feedback loop, as played out many iterations later. The application of consonance versus dissonance in music is an interesting case in point. Musicality may well be what allows for the affective value of dissonance, just like – as Savage et al. note – metre aids synchronized performance. Indeed, dissonance's role in music, as in animal vocalizations, is arguably to calibrate the arousal “dial” (Fitch, Neubauer, & Herzel, Reference Fitch, Neubauer and Herzel2002). Because of its deep biological substrate in the harmonic series of vertebrates' affective vocalizations (Bowling, Hoeschele, Kamraan, & Fitch, Reference Bowling, Hoeschele, Kamraan and Fitch2017; Bowling & Purves, Reference Bowling and Purves2015), sensory dissonance imparts strong affect that music aptly builds upon (Bravo et al., Reference Bravo, Cross, Hopkins, Gonzalez, Docampo, Bruno and Stamatakis2019) – once again, within a biology-culture loop. We feel the logic of such feedback loops – not lost on Savage, Mehr and colleagues – is emblematic for the apparent (mechanistic, rather than logical) circularity intrinsic to theories of cultural or biological evolution, and indeed ultimately intrinsic to any systematic account of connected events whereby past effects become future causes, as seen in history.

In Western tonal music, for example, although by no means rare, dissonance tends to be used as a “spice”; one that, in the language of predictive coding, “provides its own reward prediction error,” as Savage et al. remark. Sensory and tonal dissonance (Johnson-Laird, Kang, & Leong, Reference Johnson-Laird, Kang and Leong2012) abounds in rock's distorted guitars, jazz's flatted-5th chords, European art music back to at least Monteverdi – and, plentifully, in the Second Viennese School's atonality. That ideas revolutionary at their time – such as Schoenberg's wish to “emancipate dissonance” and strive for total chromaticism (Harrison, Reference Harrison1996; Schoenberg, Reference Schoenberg2010) – only survived as an elitist niche, whereas music, including most art music, remained firm in its tonal structure, is evidence of the biologically-based invariance of universal features of music such as the cognitive framework of tonality (Mehr et al., Reference Mehr, Singh, Knox, Ketter, Pickens-Jones, Atwood and Glowacki2019). It seems unlikely Schoenberg's idea would have arisen from social-related pressures, but rather from pressures of cultural evolution (arguably, the demand for novelty).

The Western musical canon is anchored in salient features of cognition (e.g., Huron, Reference Huron2016). Although current evidence is mixed, humans and other species (Izumi, Reference Izumi2000; Sugimoto et al., Reference Sugimoto, Kobayashi, Nobuyoshi, Kiriyama, Takeshita, Nakamura and Hashiya2009; Watanabe, Uozumi, & Tanaka, Reference Watanabe, Uozumi and Tanaka2005) appear to prefer sounds with minimal roughness and/or abundant fusion (Masataka, Reference Masataka2006; Popescu et al., Reference Popescu, Neuser, Neuwirth, Bravo, Mende, Boneh, Moss and Rohrmeier2019; Trainor, Tsang, & Cheung, Reference Trainor, Tsang and Cheung2002). Thus, consonance preference appears not merely an arbitrary cultural invention – despite recent suggestions to the contrary (Bowling et al., Reference Bowling, Hoeschele, Kamraan and Fitch2017; McDermott, Schultz, Undurraga, & Godoy, Reference McDermott, Schultz, Undurraga and Godoy2016; McPherson et al., Reference McPherson, Dolan, Ossandon, Valdes, Undurraga, Jacoby, Godoy and McDermott2019) – but plausibly emerges from iterated cycles of gene-culture coevolution. In effect, Savage et al.'s MSB explains a biologically-evolved set of propensities upon which music builds, using universal principles of cognition such as prediction (Koelsch, Vuust, & Friston, Reference Koelsch, Vuust and Friston2019) and dendrophilia (Fitch, Reference Fitch2014). Once established, this basis can give rise to culturally-evolved features such as scales, some more pervasive across cultures than others (Fitch & Popescu, Reference Fitch and Popescu2019; Mehr et al., Reference Mehr, Singh, Knox, Ketter, Pickens-Jones, Atwood and Glowacki2019).