Savage et al. and Mehr, Krasnow, Bryant, & Hagen argue that collective music-making (e.g., group singing and dancing) is part of our evolved human nature because of its adaptive function over human evolution, either because of its role in social bonding (Savage et al., target article) or in signaling coalition strength to other groups (Mehr et al., target article).

The social bonding theory, in particular, has old roots (e.g., Dunbar, Reference Dunbar, Oller and Griebel2004; Roederer, Reference Roederer1984), yet both theories face an important, unaddressed challenge: Namely, the existence of significant variation in the extent to which traditional, small-scale societies engage in collective music-making. Simply put, the traditional practices of some small-scale societies include little to no collective music-making, whereas in other indigenous cultures collective music-making is prominent. Documenting and understanding this variation is crucial for collectivist theories of music's origins, because unless this variation can be explained, such theories remain incomplete.

Consider the Tsimane of the Bolivian Amazon, a population of forager-horticulturalists, with whom one of us (CvR) has worked since 2005. Ethnographic research indicates very little collective music-making in Tsimane society. Between 2002 and 2007, structured observations (n ≃ 80,000) of hundreds of people in different villages included only nine instances of collective music-making – all pairs of young children singing together in the context of play (Tsimane Health and Life History Project, unpublished). Traditionally, Tsimane music-making was largely solo, including singing by shamans or other older adults whose songs conveyed traditional knowledge, reinforced cultural norms, and propitiated ancestors and the guardian spirits of forest animals (Huanca, Reference Huanca2006; Godoy, personal communication). Missionary influence has contributed to the demise of traditional song, and led to establishment of churches in several villages. Where CvR has observed collective singing among adults, it occurred in the context of a church service, and Tsimane participation in collective singing was fairly reluctant, timid, and uncoordinated. Little experience with collective music-making may help explain unusual features of Tsimane music perception, such as no esthetic preferences related to consonance and dissonance, and a lack of octave equivalence (Jacoby et al., Reference Jacoby, Undurraga, McPherson, Valdés, Ossandón and McDermott2019; McDermott, Schultz, Undurraga, & Godoy, Reference McDermott, Schultz, Undurraga and Godoy2016; Thompson, Sun, & Fritz, Reference Thompson, Sun, Fritz, Rentfrow and Levitin2019; Zatorre, Reference Zatorre2016).

The Tsimane are not unique in having little collective musical behavior. Coordinated group music (vocal and instrumental) is largely absent in traditional cultures in parts of Siberia, including among the Tuvans and Yakuts, both of whom engage in animal husbandry (Levin, Reference Levin2006; Nikolsky, Alekseyev, Alekseev, & Dyakonova, Reference Nikolsky, Alekseyev, Alekseev and Dyakonova2020). Our informal canvassing of a few ethnographers who have carried out years of fieldwork with hunter-gatherers revealed a relative absence of collective music-making in the traditional practices of several such groups. These include the northern Aché of Paraguay and the Agta of the Philippines, both of which are egalitarian cultures (Kim Hill, Bion Griffin, and Thomas Headland, personal communication), and the Ayoreo people of Bolivia and Paraguay (Lucas Bessire, personal communication). We strongly suspect that these societies are far from a complete list of those where traditional music is largely performed solo.

Compiling a list of such small-scale, traditional societies and understanding what drives their tendency toward solo music have implications for the frequency and importance of collective music-making over human evolution. We offer three hypotheses to explain cultural variation in collective music-making:

1. (1) Music-making is less likely to be collective where there is less collective action in general. In particular, a relative absence of inter-group conflict may predict less collective, synchronous singing and dancing, given less demand for signaling group cohesion (Mehr et al., target article) and lower returns to social bonding and identity fusion (Savage et al., target article). The Tsimane may be an example: historical documents and ethnographic accounts suggest a relative lack of inter-group coalitional violence over the past several centuries (Godoy, Reference Godoy2015). Furthermore, collective action in productive activities outside of the extended family is infrequent in Tsimane society (von Rueden, Gurven, Kaplan, & Stieglitz, Reference von Rueden, Gurven, Kaplan and Stieglitz2014). It may be that inter-group conflict or exchange fosters the genesis and spread of collective music making via cultural evolutionary dynamics, because collective music-making facilitates the competitive success of groups.

2. (2) Music-making is less likely to be collective where it is a principal means of conveying expert knowledge or individual accomplishment. In such cases, individuals may be more reluctant to produce music or be discouraged from doing so unless their status (e.g., shaman) justifies their music-making. Among the Ayoreo, who lack synchronous singing, men (and sometimes women) take turns singing at night to describe notable events, but singers are implicitly ranked on skill and “bad” singers meet with critical commentary from others (Bessire, personal communication, Reference Bessire2006, Reference Bessire2014).

3. (3) The adoption of certain musical styles can constrain subsequent adoption of collective music-making. For example, timbre-based music, common in Tuva and elsewhere in Siberia (Levin & Süzükei, Reference Levin, Süzükei, Dolan and Rehding2018; Nikolsky et al., Reference Nikolsky, Alekseyev, Alekseev and Dyakonova2020), is less conducive to collective, synchronous music-making than is pitch-based music. Timbre-based music instead enables highly individualized forms of expression that are used by Tuvans and Yakuts as a form of personal identification (Nikolsky et al., Reference Nikolsky, Alekseyev, Alekseev and Dyakonova2020).

In summary, collective music-making is far from ubiquitous across small-scale, traditional societies, and documenting and seeking to explain this variation should be part of any evolutionary theory which appeals to the adaptive value of collective musical behavior. We do not claim that collectivist theories of music evolution are necessarily wrong. It is possible that the musical capacities that enable solo music-making evolved largely in the context of collective music-making. However, it is also possible that collective musicality arose and spread via purely cultural dynamics and did not play a necessary or sufficient role in genetic selection of traits underlying human musicality. This is an important null hypothesis for gene-culture coevolutionary theories of collective musicality (e.g., Savage et al., target article; Patel, Reference Patel and Honing2018).