PUBLIC RESEARCH PREVIEW Research in progress · uncertainties and source access are recorded throughout.

SITE-019 · 2026-10-28

Vale da Pedra Furada

C7γ: 27,615–24,071 cal BP · 80% modelled phase range

South AmericaBrazilearly human activitydatingstone tools
Working research. Assessments and source-access limitations are retained from the research record. This entry has not completed its public-release review.

Begin with object 255660

This open-air site is distinct from Boqueirão da Pedra Furada, the nearby rock shelter. The detailed argument here concerns FUMDHAM 255660, a sandstone plate measuring up to 21.0 × 18.5 × 2.9 cm, recovered flat in C7γ-a. Open its object and dating ledger. [S66 · Boëda et al. (2021)]

The authors infer five working stages from overlapping removals, bevels, notches and differing edge wear. This is an argument about a sequence of actions, rather than resemblance to a familiar tool. A possible hafting area and hoe/shovel comparison do not establish a handle, exact function or agriculture. Matching production flakes were not reported for the notching operation. [S66 · Boëda et al. (2021)]

What “24,000 years” actually means

No clock was applied directly to the stone’s manufacture. Charcoal and sediment constrain its stratigraphic context. The supplement gives 27,615–24,071 cal BP as the C7γ phase time range at 80%. The main text rounds this to 27,600–24,000. Its 25.0 and 24.5 ka beginning/end MAP values are different model summaries, not uncertainty bounds. [S66 · Boëda et al. (2021); S78 · Boëda et al. (2021)]

The focused ledger preserves six C7 charcoal entries, four C7 OSL sediment samples and two underlying C8 charcoal entries. Single-grain and multi-grain results for the same sample remain linked. Published calibrated segments are not collapsed across gaps or substituted by the phase model. [S78 · Boëda et al. (2021)]

Objections must match the evidence

A regional critique questions geological context, raw-material selection, artifact-selection criteria, capuchin activity and the adequacy of trace studies. Its abstract does not document an independent reconstruction of object 255660. Our comparison therefore asks what those processes could explain on this particular object, rather than treating a regional objection as its automatic explanation. [S79 · Gómez Coutouly (2021)]

The ledger also flags sample-count, layer-label and calibration-table discrepancies. These need reconciliation before a full chronology reproduction; a reporting inconsistency and a challenge to human manufacture are different issues. [S78 · Boëda et al. (2021)]

Keep deeper claims separate

The paper reports much older activity in deeper levels, around 41,000 years ago. That claim needs its own object-level audit. A strong C7 argument cannot establish every deeper stone’s human origin. Old and revised layer names cannot be joined solely by matching labels; C10′ additionally comes from a separate sector without direct stratigraphic contact. [S66 · Boëda et al. (2021); S78 · Boëda et al. (2021)]

Inspect the C7 samples and model · Compare early South America · Locate on the atlas