Global relationships between musical, linguistic, and genetic diversity

Abstract

Music and language are both universal but diverse cultural traits shaped by cultural and biological evolution. However, there is disagreement on the relationships between music, language, and human history. Some argue that musical and linguistic similarities trace ancient migrations of people and their cultures, while others argue that they primarily reflect more recent contact between neighboring societies and local micro-evolution independent of population migration. Previous direct comparisons of musical, genetic, and linguistic diversity were restricted to small regional samples that gave conflicting results. Here, we analyze global patterns of diversity from newly public global databases containing over 5,000 traditional songs with standardized “Cantometric” codings and genomic profiles from over 4,000 individuals. We directly compare musical, linguistic, and genetic diversity for a subset of 152 matched societies (represented by 1,054 songs, genomic profiles of 1,719 individuals, and 152 languages). For both genes and music, differences within groups are greater than those between groups, but musical differences between groups are approximately three times greater than genetic differences. Song style and basic vocabulary both show relatively weak relationships with each other and with genetic distance and geographic proximity, in contrast to the much stronger relationships found between genes and geography. Thus, to our surprise, our findings suggest that music and language are weak proxies for human migrations.

The Global Jukebox: A public database of performing arts and culture

“…Lomax's correlations between song style and social complexity have been disputed [14][15][16][17][18][31][32][33][34][35][36]61], but our bivariate and PCA reanalyses provide evidence suggesting that his original correlations are a) reproducible, and b) not clear artifacts of geographic or linguistic autocorrelation. We do not believe, nor is there any evidence that Lomax believed, that social factors directly produce effects upon music.…”

“…Unfortunately, the original analyses cannot be precisely reproduced, as it is not possible to reconstruct the specific datasets and procedures that were used for previous analyses performed during the mid-20th century before data/code-sharing capabilities were widely available. For reanalyses of Cantometric data and analyses of similar data, see [ 19 , 21 , 22 , 52 , 53 , 55 – 61 ].…”

“…(cf. [ 61 ]). However, the probability that certain social and musical traits and configurations will consistently cooccur raises questions about what drives aesthetic preferences.…”

“…[62] Understandably, readers may be skeptical about any attempt to establish predictive or causal relationships between society and music. Whether such correlations reflect causal relationships, autocorrelation due to shared ancestry, or something else, has been greatly debated [14][15][16][17][18][31][32][33][34][35][36]61]. Here we attempt to see whether the basic correlations are a) reproducible, and b) artifacts of geographic or linguistic autocorrelation.…”

“…For example, are aesthetic preferences embodied in vocal representations of emotions and / or physical states that may arise under certain persistent conditions? Different evolutionary theories propose several possible mechanistic pathways that could be tested [61], such as patterns of interaction [69], roles of social bonding [70], signaling [71], and psychosocial effects [16]. Our database can generate new hypotheses with preliminary evidence for causal and other kinds of relationships; these could then be tested by modeling archival information and data from the field.…”