Linguistic diversity is disappearing (almost) as fast as biological diversity: very soon, we may no longer be able to know the sounds and structures of many of the world's languages. As for sounds, phonetic documentation is thus essential to preserve this disappearing heritage, and to learn more about our linguistic and cognitive evolution. This is particularly crucial in those areas of the world where language density is higher, such as the Democratic Republic of the Congo (henceforth, DRC). Therefore, the first focus of my research is to document the sounds of the “vanishing voices” of the last hunter-gatherers of the Lower Kasai Region of the DRC. These groups, commonly known as “Batwa” or “Pygmy”, live in one of the least well-surveyed regions of the planet, where language loss is particularly fast. Historically, the so-called “Bantu Expansion”, i.e. the southward and eastward spread of Bantu-speaking groups over large parts of sub-Saharan Africa, resulted in the disappearance of pre-existing languages in Central Africa. “Pygmies”, in particular, while often retaining their genetic and cultural identity, have shifted from their ancestral languages to those of the newcomers. Hence, the second focus of my research: the contrastive study of “Batwa” and “non-Batwa” Bantu language varieties, and of their (socio-)phonetic contrast, will contribute to a better historical appreciation of the striking genetics/linguistics mismatch we witness in the tropical rainforest.