- what the source states
- Fóti argues that the spectroscopic workhorses of provenance archaeology are "essentially half-blind": ICP-MS and portable XRF cannot detect hydrogen, carbon, or oxygen, and calibration fills the gap with a natural-stone assumption ("there is no 'artificial stone' calibration on the device… Its dumb little internal table will confidently report that it's a natural stone"). His summary: "With this table, you're ready to debate anyone who insists that various spectrographic methods can determine everything about a material or rock—even pinpoint the quarry on a map. Such claims are deeply misguided." He expressly exempts other tools: "Of course, I know there's petrography, microscopy, acid testing, titration, and so on." Applied to Stonehenge, he reads the Nash et al. 2020 sarsen result as a statistical correlation, not an exact match, and proposes that crushed West Woods material transported "in buckets" and cast on site would carry the same metallic fingerprint, slightly shifted by additives. On European menhirs he notes attributed quarries "often 400–500 miles away, which is almost certainly a sign that this is a mistake."
- what we infer
- Endorsed as the case's live proponent counter-argument by editor Eugene (2026-08-22): quarry-matching and materials analysis are distinct methodologies run by different specialist communities and answer different questions, so "we found the quarry" conclusions can overextend into anti-casting claims the chemistry does not support. The editor's illustrative point — accepted provenance inferences can imply surprising logistics — is real at Stonehenge: the mainstream- accepted bluestone transport is ~230 km from the Preseli Hills. Accuracy guardrails, per the verified record: the Nash 2020 sarsen match is ~25 km (West Woods), and the geo verification batch found Stone-58's immobile-element signature directly matched West Woods — stronger than Fóti's "statistical correlation only" framing (see GEO-E009 and the SRC-FOTI-2024 reliability notes). And the critique's scope is bounded by Fóti's own concession: it cuts against trace-element provenance reasoning, not against the binder-aware petrography of GEO-E018–E021, which examines the monument stone itself.
- limitations
- Proponent primary text, not peer reviewed; quotes are from the project library's full-text extraction.
- Fóti's methods table contains errors: it lists NMR as unable to see C, O, or Si, yet solid-state ²⁹Si NMR is standard and is exactly what the pro-casting MacKenzie et al. 2011 study used. The H/C/O-blindness point is correct for XRF and ICP-MS, not for the whole table.
- The critique does not answer the direct petrographic record (GEO-E018–E021): those studies examine monument stone with binder-aware methods, and Jana 2007's positive control showed a real casting signature is detectable when present.
- Related catalog claims GEO-C507, GEO-C522, and GEO-C575 carry the same argument from the same author — one independence group (provenance-method-critique), not corroboration.