The primary purpose of studying music's evolution is to understand the selective pressures that caused different musical traits to evolve. When studying the evolution of a given musical trait, the first question to ask is whether the trait can be parsimoniously explained as a byproduct of non-musical adaptive functions. An example might be ears: Ears are indisputably useful for appreciating music, yet in all likelihood the reason that we possess ears is for general auditory perception rather than music listening. If no byproduct explanation is forthcoming, the way forward is clear for an alternative hypothesis explaining the musical trait in terms of a particular adaptive function of music.

Although the two target articles by Mehr et al. and Savage et al. ultimately come to contradictory explanations for music's evolution, their approaches share certain commonalities. First, both articles minimize the sense in which musical traits may be explained as byproducts of non-musical adaptive functions. Second, both articles present unified theories purporting to explain the evolution of musicality in terms of music's own adaptive functions.

Both articles provide substantive positive arguments for their respective theories, making it clear that both social bonding and credible signalling are plausible adaptive functions of music-making with broad cross-cultural and historical relevance. However, we would push back against the claims that (a) byproduct explanations have little relevance for music evolution, and that (b) the articles' theories provide complete unified accounts of music evolution.

Savage et al. address byproduct explanations towards the end of their article, writing that “whether music is a domain-specific evolutionary adaptation for social bonding, as opposed to a byproduct of the evolution of other adaptations, is open to debate.” However, they conclude against this possibility, writing that the cross-cultural universality of music and language implies that music and language fulfil independent functions. This argument holds little water; if one believed that music were a byproduct of language, then music's universality would simply be a byproduct of language's universality.

Mehr et al.'s arguments against byproduct explanations are numerous but weak. Music's ancient nature, cross-cultural universality, early developmental manifestation, and complex language-like syntax could surely be explained by music being the byproduct of ancient, cross-culturally universal, and early manifesting capacities such as language cognition. The convergent evolution of music-like behaviours in other species could likewise be the consequence of music being a byproduct of other adaptive traits; for example, Patel (Reference Patel2014) argues that cross-species rhythmic capacities may be a byproduct of vocal learning abilities. Finally, much of music's neural specialization might be explained by neural plasticity (e.g., Münte, Altenmüller, & Jäncke, Reference Münte, Altenmüller and Jäncke2002).

Meanwhile, having minimized the role of byproduct explanations, Savage et al. conclude that their theory of music-induced social bonding provides a “unified” and “comprehensive theory” that “synthesize[s] and extend[s] previous proposals into a new, parsimonious framework.” The implication is that there is no space for alternative explanations for music's evolution, such as credible signalling or sexual selection; however, the authors do not present any arguments against these alternative explanations. Section 6 rather muddies the waters, with credible signalling and sexual selection being described as both “complementary” and “competing” with respect to the authors' own hypothesis. It seems as if the authors want to claim that their theory provides a complete account of music evolution, but simultaneously want to avoid the obligation of disproving competing theories.

In contrast, Mehr et al. do systematically attempt to disprove competing explanations of music evolution, including the byproduct hypothesis as discussed above. Their critique of Darwin's sexual selection theory is particularly convincing, but their critique of Savage et al.'s social-bonding hypothesis is problematic. The authors interpret social bonding as altruism, and make the important observation that music-induced altruism would be counteracted by free-rider mutations. However, social bonding does not have to be altruistic in nature. If music-making has the byproduct of enhancing my ability to physically coordinate and nonverbally communicate with my fellow tribespeople, with the consequence that I'm less likely to die on our next hunting exhibition, then my participation is not altruistic, because I benefit myself as well as the group. The authors also claim that music offers no clear benefits for inducing social bonding over language; this might be the case if social bonding is operationalized as altruism, but it is not true if we consider more complex aspects of social bonding, such as developing joint coordination and communication skills, for which music seems to be particularly well adapted (see e.g., Cross & Morley, Reference Cross, Morley, Malloch and Trevarthen2008).

We are left wondering why both articles are so keen to downplay byproduct explanations for musicality, and to promote their own theories as complete unitary explanations for musicality. On the one hand, Occam's razor tells us to prefer simple theories, and as music researchers we have a vested interest in discounting byproduct explanations that might be seen to downgrade music's evolutionary importance. On the other hand, a broader biological perspective provides countless examples of evolved traits that were shaped by a variety of selection pressures, some with only tangential relevance to the trait's ultimate function. For example, spider silk initially evolved to line burrows, but in most extant spider species webs are formed for other functions such as prey trapping, pheromone release, water capture, and dispersal parachuting (Duffey, Reference Duffey and Selden1998; Nentwig & Heimer, Reference Nentwig and Heimer1987; Vollrath & Edmonds, Reference Vollrath and Edmonds1989). It seems quite plausible that music's evolution could have similarly complex explanations.