The date of the first settlement of the Americas remains a contentious subject. Previous claims for very early occupation at Pedra Furada in Brazil were not universally accepted (see Meltzeret al.1994). New work at the rockshelter of Boqueirão da Pedra Furada and at the nearby open-air site of Vale da Pedra Furada have however produced new evidence for human occupation extending back more than 20 000 years. The argument is supported by a series of14C and OSL dates, and by technical analysis of the stone tool assemblage. The authors conclude that the currently accepted narrative of human settlement in South America will have to be re-thought. The article is followed by a series of comments, rounded off by a reply from the authors.

No abstract

International audienc

Current archaeological paradigm proposes that the first peopling of the Americas does not exceed the Last Glacial Maximum period. In this context, the acceptance of the anthropogenic character of the earliest stone artefacts generally rests on the presence of projectile points considered no more as typocentric but as typognomonic, since it allows, by itself, to certify the human character of the other associated artefacts. In other words, without this presence, nothing is certain. Archaeological research at Piauí (Brazil) attests to a Pleistocene human presence between 41 and 14 cal kyr BP, without any record of lithic projectile points. Here, we report the discovery and interpretation of an unusual stone artefact in the Vale da Pedra Furada site, in a context dating back to 24 cal kyr BP. The knapping stigmata and macroscopic use-wear traces reveal a conception centred on the configuration of double bevels and the production in the same specimen of at least two successive artefacts with probably different functions. This piece unambiguously presents an anthropic character and reveals a technical novelty during the Pleistocene occupation of South America.