Coevolution of languages and genes on the island of Sumba, eastern Indonesia

Abstract

a geographic scale, which is too coarse to permit any refined inference about the dynamics of language change in these contact zones. Information at a finer scale is essential to characterize the nature of contact relationships and infer mechanisms of linguistic and genetic transformation over recent temporal and fine spatial scales. Toward this goal, we examine linguistic and genetic variation in a contact zone on the Indonesian island of Sumba. Broader regional studies support the initial settlement of Southeast Asia/Oceania by foraging societies by 40,000 to 45,000 BP (13). Languages of the geographically expansive Austronesian family occupy much of the Indonesian archipelago, except in far eastern Indonesia, where diverse and unrelated Papuan languages dominate. Recent syntheses place the Neolithic transition, considered to mark the arrival of Austronesian colonists in the vicinity of Sumba, at between 4,000 and 3,500 years ago (14, 15). At that time, small numbers of farmers speaking an Austronesian language likely came into contact with an indigenous population of foragers speaking aboriginal languages. Several circumstances favor Sumba as a site to investigate the relationship between population incursion and language change. Sumba is remote and culturally conservative, the last island in the archipelago where the majority adhered to a tribal or pagan religion at the close of the 20th century. Today, nearly all Sumbanese live in traditional farming villages composed of patrilocal clans. Contact between villages is limited, and population density is low. Perhaps the most telling indicator of the extent of contact between villages is the large number of languages now spoken on the island despite its small size (220 ⫻ 75 km2). In this report, we reconstruct Sumbanese language relationships using a 200-word Swadish list, and we examine Y chromosome SNP and short tandem repeat (STR) diversity in a sample of Sumbanese villages. We propose a model of language–gene coevolution to explain the striking associations we observe among linguistic, Author contributions: J.S.L. designed research; J.S.L., S.S.D., B.H., T.M.K., and H.S. performed research; S.S.D., B.M.G., B.H., T.M.K., P.N., J.W.S., and J.C.W. analyzed data; and J.S.L., M.P.C., B.H., P.N., J.C.W., and M.F.H. wrote the paper. The authors declare no conflict of interest. This article is a PNAS Direct Submission. Abbreviations: PAn, Proto-Austronesian; PS, Proto-Sumba; STR, short tandem repeat.