05.11.2024 - Paul Verdu

Interdisciplinary research between cultural and biological anthropology has fostered numerous discoveries about the history of our species and the evolutionary mechanisms at the root of human diversity worldwide. With the advances of human population genetics during the 20th century, a pioneer study by Cavalli-Sforza et al. (1988) highlighted striking homologies between human populations phylogenetic trees and language classifications; possibly substantiating the parallel between the formation of languages and the formation of species initially proposed by Darwin in “The Decent of Man” (1871). Since then, conceptual analogies between genetic and linguistic evolutions have been extensively explored by computational linguists and evolutionary biologists, revealing the links, or their absence, between human biological and cultural histories. Most of these studies rely on genetic and linguistic samples often collected entirely separately. Furthermore, they compare the genetic and linguistic histories of populations and languages reconstructed separately by each discipline at the macro-evolutionary scale. This makes it fundamentally challenging to formally compare genetic and linguistic inferred histories, and to elucidate the possible interactions, or the absence of interactions, between genetic and linguistic evolutions across and within populations. We formalized a novel interface between population genetics and a proposed “population linguistics”, rooted in evolutionary linguistics principles, and to build a general historical model describing individuals interacting both genetically and linguistically in an agent-based approach. We then developed a novel software allowing the massive simulation of joint genetic and linguistic variation centered on individuals, and deployed Approximate Bayesian Computation to infer the parameters of the historical genetic and linguistic models underlying observed genetic and linguistic data collected in the same individuals. We apply this novel paradigm and methodology to genomic and utterance data collected in Cabo Verde, in order to reconstruct the genetic and linguistic history of the peopling of each island in the archipelago since the 15th century.