The ERC BANG project aims to uncover the untold traces of West Africa’s last confirmed language isolate, Bangime, and the genetic affiliations of its speakers, the Bangande. Bangime is currently spoken by around 1,500 people living in seven small villages at the edge of a valley in central-eastern Mali. These villages are perched on and among boulders at the foot of the Bandiagara Escarpment. The Bangande live alongside Dogon communities, whose languages may constitute an early branch of the Niger-Congo language phylum that otherwise dominates much of sub-Saharan Africa. The origins of both the Dogon and the Bangande in the Cliffs remain unknown. In the valley below, Bozo-speaking farmers and Fulbe herders contribute to a complex web of linguistic and cultural diversity.
Although Bangime shows no genetic or linguistic ties to Dogon languages, the Bangande identify as Dogon and describe their language as Dogon. The ERC BANG project seeks to unravel these paradoxes through an interdisciplinary approach that combines linguistics, genetics, archaeology, ethnography and history. We employ computational methods to analyse large datasets across these disciplines.
Preliminary linguistic results suggest that Bangime is not a Dogon language, despite the speakers’ self-identification, but that Bangime has been shaped by contact with multiple distinct Dogon languages over time. Our analyses also indicate that the Dogon languages themselves may represent an early branch of Niger-Congo, and their unusual syntactic patterns may help reconstruct features of the proto-Niger-Congo system. Innovative approaches to loanword detection — including across semantic domains — point to Bangime as a language heavily influenced by Dogon, while retaining a substrate possibly linked to Mande. This points to a previously unrecognised depth of linguistic and cultural diversity in the Bandiagara region, now being explored further through historical records and oral traditions shared by West African griots.
Genetic analyses, in collaboration with international partners, are expected to complement the linguistic findings and provide additional insight into the complex population history of the Bangande and their neighbours. The results will be shared in open-access formats for both researchers and community members, contributing to a better understanding of the hidden diversity of the origins of West Africa’s populations.
