Correcting the Corrector: The Newall Boulder is Still Not Glacial

Brian John's latest self-published critique (preprint deposited on ResearchGate, January 2026), titled The Newall boulder at Stonehenge: correcting the “corrections”, sees him lurching forth once more like Monty Python's indefatigable Black Knight – "Tis but a flesh wound!" – swinging away despite the accumulating evidence. Written in a style that repeatedly invokes "the present author" with the regal detachment of a Victorian memoir (one is almost tempted to dub him "Princess Present Author" for the monarchical flair), it is a 13-page exercise in persistence, but persistence does not equal persuasion. His bombastic use of "refute", boldly declaring in the abstract that "The contents of the paper are therefore refuted", is equally baseless: a self-published assertion without decisive counter-evidence might contest, challenge, or question, but only an egotist would claim outright refutation when even he concludes the debate remains "scientifically disputed". It recycles familiar claims while misrepresenting our peer-reviewed paper (Bevins et al., 2025, Journal of Archaeological Science: Reports 66: 105303), which presents new petrographic, automated SEM-EDS mineral mapping, portable XRF geochemical data, and contextual analysis to reaffirm the boulder's origin at Craig Rhos-y-Felin and dismiss glacial transport.

The piece opens with the now-familiar falsehood: our manuscript was supposedly "rejected by the Journal of Quaternary Science" before acceptance elsewhere. As previously noted, no such rejection occurred; it was submitted directly to JAS:Reports, the appropriate geoarchaeological venue, and passed standard peer review. This invention sets the tone for the rest – assertions dressed as fact, selective omissions, and a heavy reliance on blog posts as "evidence".

Here are the main points addressed systematically:

1. Provenance to Craig Rhos-y-Felin (Sections 3–4)

John claims "no convincing evidence" for a Rhos-y-Felin source, citing insufficient sampling density, no identical matches, reliance on old museum fragments, and potential mislabelling. He insists we must demonstrate that all nearby foliated rhyolite outcrops are "substantially different".

Response: Our study builds on 15+ years of cumulative work (e.g., Bevins et al. series from 2011 onward), incorporating new analyses of previously unsampled areas on the boulder. Key immobile elements (Zr, Nb, Th), mineral assemblages (foliation, spherulites, quartz phenocrysts), and textural features show strong convergence with Rhos-y-Felin rhyolite Group C. Provenancing in igneous terrains uses probabilistic multi-method matches, not identical twins across every metre – a standard approach the John ignores (cf. comparable studies on dolerites or the Altar Stone). Museum fragments are contextualised with fresh field sampling; mislabelling risks are acknowledged but do not invalidate the overall dataset. The demand for exhaustive exclusion of every nearby outcrop is unrealistic and not required in geoarchaeology.

2. Morphology and "Glacial" Features (Sections 5–6)

His Fig. 1 photo is annotated with labels such as "Fresh rock exposed in fracture scars", "Eroded facet", "Slickenside surface with quartz crystals and lineations", and "Weathered 'top' surface". He insists these are "strong indicators of glacial transport", apparently supported by a poll of 11/12 unnamed "senior geomorphologists".

Response: These features – variable facets, edge rounding, localised lineations, and fresh scars – are consistent with natural jointing in the rhyolite, prolonged surface weathering exploiting discontinuities, and human-induced breakage/percussion (e.g., small scars near the tip matching debitage from monolith shaping). Subglacial abrasion produces more uniform striae, polish, and faceting; we see none of that. The "expert poll" (non-blind, photo-based, unpublished) lacks methodological rigour – photo interpretation of geomorphology is notoriously subjective. Our high-resolution imaging and comparison to Rhos-y-Felin debitage/monolith fragments support breakage from Stone 32d (the stump), not ice transport. Senior geomorphologist input on our team (e.g., Scourse) finds no compelling glacial signature.

3. Glacial Transport and Ice Limits (Section 6)

BJ argues Bristol Channel erratics prove inland reach to Salisbury Plain, dismisses ice-rafting claims, and invokes pre-Devensian (Anglian/Wolstonian) speculation despite "abundant evidence".

Response: Far-travelled erratics in the Bristol Channel relate to coastal/rafting or limited Devensian advances – not inland chalkland penetration. BRITICE-CHRONO models and recent work (Scourse, 2024) constrain LGM limits well west/north; no ground-truthed glacial deposits (till, striated bedrock, classic erratics) exist on Salisbury Plain. Pre-Devensian ice lacks supporting stratigraphy or dating here. The boulder's context (Neolithic layers, fresh fractures) fits human debitage far better.

4. Broader Bluestone Assemblage (Sections 7–9)

John then claims 46+ lithologies prove erratic scatter, accuses bias (ignoring hammerstones/etc.), and insists most are unmodified erratics/"rubbish stones".

Response: Our focus is on monoliths and associated debitage (the monument's core). Variability is constrained to ~12–15 Preseli lithologies with multi-source origins – not random glacial randomness. "Rubbish" reflects selection/breakage processes. Claims of 12–15 "quarries" misrepresent; we identify specific extraction sites (Rhos-y-Felin, Carn Goedog) supported by excavation (Parker Pearson et al., 2019+), while the John's blog critiques lack new fieldwork.

5. Local Opportunistic Use in Wales (Section 9)

He further cites Bedd yr Afanc as evidence of local, opportunistic stone use without significance.

Response: This supports our view – Preseli stones were abundant locally and used where convenient. It does not undermine targeted selection/transport for Stonehenge.

6. Press and "Pseudo-Science" (Final sections)

The shy and retiring Brian John criticises the Aberystwyth release/media as "over-sold". We agree media can sensationalise; our paper is measured, data-led, and corrects prior errors (especially in John's 2024 piece). It adds substantial new analyses – hardly "little new information".

In summary, the Newall boulder is rhyolite debitage from Craig Rhos-y-Felin, transported by Neolithic people – not a glacial erratic. The data overwhelmingly support human transport; the glacial hypothesis lacks any credible foundation, relying on special pleading to ignore the vast moraine of counter-evidence and an unsavoury dismissal of Neolithic ingenuity and persistence simply because they are from a previous time.

Full references and data are in Bevins et al. (2025).