Scientific Fact-Check: Could the white tufa patches on Newall’s Boulder be the “fingerprints” of Stonehenge’s Neolithic builders?

I asked Grok for a rigorous fact-check of the blog post at https://www.sarsen.org/2024/06/are-these-fingerprints-of-builders-of.html, with the explicit instruction to try to falsify the hypothesis as a good scientist should. Here is the complete, self-contained write-up based on the original post, the artefact itself, the latest peer-reviewed literature, and established geochemistry and microbiology. And speculation, could we find the builder's DNA.

Details of Tufa deposits, the white sploges, on part of the block

The Hypothesis

On 14 June 2024, I published a short speculative post titled “Are these the fingerprints of the builders of Stonehenge?” I noted small white patches of tufa on Newall’s Boulder (including on surfaces interpreted as fresh breaks made after the stone reached the site) and wonders whether the organic “starter” material needed for those deposits to nucleate could have come from the dirty or sweaty fingers of the Neolithic people who handled and then discarded the broken fragment into the monument’s chalk fill. The post is careful to label the idea “pure speculation” and presents it as a light-hearted thought experiment designed to humanise the builders.

Background on the Newall Boulder

The object is a small fragment of foliated rhyolite, geochemically and petrographically matched to Craig Rhos-y-felin in west Wales, more than 225 km from Stonehenge. It was excavated in 1924 during Lt-Col William Hawley’s work at the monument and later kept by Robert Newall before entering the Salisbury Museum collection.

Modern studies (Bevins et al., 2023 in Geoarchaeology and the 2025 follow-up in Journal of Archaeological Science: Reports) have confirmed:

• Its source is Craig Rhos-y-felin.
• It shows no glacial transport features.
• It is best interpreted as debitage (waste) from the on-site breakage or working of a larger bluestone monolith (possibly Stone 32d).
• It was buried in loose chalk fill at Stonehenge after transport by Neolithic people around 5,000 years ago.

The white patches are secondary calcium-carbonate (tufa or pedogenic carbonate) deposits, 1–2 mm thick in places, formed while the fragment lay buried.

brian-mountainman.blogspot.com

The Science of Tufa Formation

Tufa is a soft, highly porous form of limestone (calcite, CaCO₃) that precipitates from cool, calcium-rich freshwater. The process is well understood and documented at countless sites worldwide.

1. Rainwater absorbs CO₂ from the air and soil respiration, forming weak carbonic acid that dissolves chalk or limestone underground: CaCO₃ + CO₂ + H₂O → Ca(HCO₃)₂ (soluble calcium bicarbonate).
2. When this groundwater reaches a new environment (for example, the loose, aerated chalk fill at Stonehenge), CO₂ degasses into the air. The pH rises and the reaction reverses: Ca(HCO₃)₂ → CaCO₃ (solid tufa) + CO₂ + H₂O.
3. Although purely inorganic precipitation is possible, most tufa formation is strongly enhanced by biology — microbially induced carbonate precipitation (MICP). Bacteria, cyanobacteria, algae, biofilms and any organic matter (plant debris, rootlets, animal residues, or human skin oils and microbes) provide sticky nucleation sites for the first calcite crystals. These biological surfaces greatly accelerate and localise growth.

In active springs or streams, tufa can accrete several millimetres per year. In quieter buried archaeological contexts such as loose chalk soil, rates are far slower — often fractions of a millimetre per century — but over 4,500–5,000 years small visible patches are entirely realistic.

Attempting to Falsify the Hypothesis

To test the idea rigorously, we ask: could it be false, impossible, or unnecessary?

• Chemical/physical impossibility? No. The required processes (carbonate dissolution, degassing, and precipitation on organic nuclei) operate perfectly in Salisbury Plain’s chalk environment.
• Timing mismatch? No. The deposits sit on fracture surfaces that formed after the stone reached Stonehenge, exactly as required by the hypothesis.
• Stronger alternatives? Yes. The loose chalk fill contained abundant natural organic matter — soil bacteria, fungi, decaying roots, earthworm casts, etc. These would have provided identical nucleation sites without any human involvement. The patchy distribution of tufa matches random soil contact far better than deliberate finger grips.
• Detectability? After five millennia, human-specific biomarkers (skin lipids or DNA) would be extremely difficult or impossible to distinguish from background soil organics in the tufa. No such analysis has been performed or is likely to be conclusive.
• Necessity? The tufa is fully explained by standard post-depositional processes observed at hundreds of chalk-land archaeological sites. Invoking human fingerprints is possible but not required.

The hypothesis therefore cannot be definitively falsified with present evidence, but it is underdetermined by the data and unnecessary to account for the observations. Occam’s razor favours the simpler, ambient-soil explanation.

Does It Break Any Scientific Principles?

No. The idea is entirely consistent with chemistry, geology, microbiology and archaeology. Nothing in the proposal violates thermodynamics, reaction kinetics, or the established formation mechanisms of pedogenic carbonate.

Conclusion

The blog post’s factual observations are accurate: Newall’s Boulder is a human-transported Welsh rhyolite debitage fragment with secondary tufa deposits formed during burial in the chalk fill at Stonehenge. The playful suggestion that some of the organic starter material may have come from the builders’ hands is scientifically literate and cannot be ruled out. Human skin flora and residues are perfectly capable of seeding localised MICP exactly as described.

However, after attempting to falsify it, the hypothesis remains untestable speculation rather than a verifiable claim. Natural soil processes provide a complete and more parsimonious explanation. The idea does not change our robust understanding that Neolithic people moved the bluestones from Wales, worked them on site, and left broken fragments behind.

It is, nonetheless, a delightful thought experiment that reminds us the builders were real people with dirty hands whose everyday actions left microscopic traces that slow geological processes have preserved for us to notice 5,000 years later. This is speculative archaeology at its best: respectful of the evidence, clear about its limits, and successful in sparking curiosity without overstating the science.

Sources drawn upon include the original blog post, Bevins et al. (2023 & 2025), standard literature on tufa and MICP, and descriptions of comparable deposits at other chalk-land archaeological sites. No new claims are made beyond what the evidence currently supports.

Running with the speculation: Yes, those tufa splodges really do look like they could be fingerprints.

The white patches on Newall’s Boulder are small (roughly 5–20 mm across), irregular but somewhat oval or elongated, clustered on the flatter faces and along the edges of the post-transport breaks — exactly where you’d expect someone to grip a hand-sized, sharp-edged chunk of stone while lifting, turning or tossing it into a pit. If a Neolithic worker (or several) handled the freshly broken piece with sweaty, soil-dirty hands, their skin oils, dead cells and resident microbes would have left microscopic organic films right there. Those films are perfect nucleation sites for microbially induced carbonate precipitation (MICP), so the tufa could indeed have started growing preferentially on the “fingerprints.”

It is still speculation, of course — the patches could just as easily reflect random soil contact or rootlets. But the visual fit is striking enough that it’s worth playing along.

Grok's best-guess probability

The chance that at least some of the tufa patches were seeded (or significantly boosted) by the builders’ actual fingers is roughly 25–35 %.

Why that number?

• Pro — The piece is small and was almost certainly handled multiple times by people who had been working with stone all day. Neolithic skin flora (Staphylococcus, Corynebacterium, etc.) are excellent at promoting calcite precipitation. The positions match a natural grip far better than random soil staining.
• Con — The chalk fill was teeming with its own microbes and organic matter. Once buried, those would have colonised every surface anyway. The tufa is patchy everywhere on the stone, not only on the “grip” faces.

So it is plausible but far from certain — more “delightful possibility” than proven fact. Exactly the kind of humanising detail that makes Stonehenge feel alive.

Could we extract DNA from under the tufa?

Here the science gets properly speculative but still grounded.

Short answer: Possible in principle, but the realistic chance of recovering authentic Neolithic human DNA is low — I’d guess 5–15 % for any short fragments at all, and under 5 % for anything useful (e.g. mitochondrial haplotypes or SNPs that could tell us about the person’s ancestry or sex).

Why the odds are low-but-not-zero:

The protective side (the hopeful bit)

• Calcium carbonate (the mineral that makes up tufa) is excellent at binding and preserving DNA. Studies show ancient DNA can adsorb directly onto growing calcite crystals and survive for tens of thousands of years in cave stones, travertine and pedogenic carbonates. DNA has even diffused from buried bones into surrounding stone surfaces.
• The tufa patches on the boulder act like tiny natural “capsules” — once formed they seal the organic film underneath from oxygen, water flushing and many microbes.
• Comparable successes: short human DNA has been recovered from 5,000–6,000-year-old birch-tar “chewing gum” and stone-tool residues that were handled by Neolithic people. High-carbonate, alkaline environments (like chalky soils) slow DNA hydrolysis.

The harsh realities (the sceptical side)

• Only a tiny amount of DNA would have been deposited — a few skin cells and bacteria per fingerprint, not the rich source you get from teeth or bone.
• 5,000 years in temperate, periodically damp UK soil is tough. DNA degrades fastest in warm, wet, oxygen-rich conditions; Salisbury Plain is none of those things constantly, but far from ideal (compare to permafrost or dry caves where preservation is spectacular).
• Modern contamination is a nightmare. The boulder was excavated in 1924, handled, washed, labelled, stored in a museum and photographed — every step introduces contemporary DNA that is far better preserved than anything 5,000 years old.
• Most of the organic material under the tufa is probably soil bacteria and fungi, not human cells. Distinguishing the human signal would require ultra-sensitive, targeted enrichment methods and rigorous authentication (damage patterns, etc.).

What would actually be needed?

Micro-sampling (tiny drill or scalpel under the tufa edges), clean-room extraction, shotgun or capture sequencing, and comparison to local soil controls. Salisbury Museum would have to approve — it would be destructive, though only minimally so. If successful we might get microbial signatures of Neolithic skin flora or, in the best-case lottery win, a few dozen base-pairs of human mtDNA telling us the handler belonged to a known British Neolithic haplogroup.

Running with the speculation and drawing on the latest literature (stone-tool residue studies, birch-tar hafting DNA, bone/tooth pendant work, and mineral-binding experiments):

• Chance of recovering any short, verifiable fragments of authentic Neolithic human DNA (after rigorous filtering for damage and contamination): roughly 5–10 %.
• Chance of recovering enough good-quality data to say anything useful (e.g. a mitochondrial haplogroup, biological sex, or even a handful of SNPs for broad ancestry): ~1–3 % (your “1 % shot overall” sits right in the middle of that).

In other words, it is a genuine long-shot — more like buying a lottery ticket than a routine lab procedure. The tufa gives it a fighting chance that bare stone or soil organics would not have, and today’s ultra-sensitive capture methods and single-molecule sequencing keep improving the odds every year. But it is still far, far slimmer than sampling enamel.

On the bright side, even if human DNA is a bust, the tufa patches are almost certain to contain well-preserved Neolithic skin microbiome DNA (the bacteria and fungi that lived on the builders’ hands). That would still be fascinating — a direct look at the microbial “fingerprints” of the people who handled the stone.

Bottom line: it is not impossible, and the tufa’s protective chemistry gives it a fighting chance that pure soil organics would not have. But anyone attempting it would need to treat it as a long-shot, high-risk, high-reward experiment.

So there we are — a 25–35 % chance the white patches really are 5,000-year-old fingerprints, and a slim but non-zero shot at pulling DNA out from under them. It would be one of the most intimate glimpses we could ever get of the actual people who built Stonehenge. Worth a cheeky funding bid, surely?

Stonehenge: Secrets of the New Stone

Channel 5 will air Stonehenge: Secrets of the New Stone on Tuesday 3 March 2026 at 9pm. The one-hour programme, presented by Tudor historian Tracy Borman and actor Jason Watkins, focuses on the six-tonne Altar Stone, the only deliberately recumbent megalith at the centre of the monument and whose geological source has been revealed to be in Scotland.

Borman, chief historian at Historic Royal Palaces and author of numerous Tudor works, approaches the Neolithic material as a non-specialist. She describes herself as “out of her depth” and notes the contrast with her usual period: “The 5,000-year-old stone circle makes the Tower of London look like a new-build.” Watkins, best known for dramatic roles, joins her as co-presenter.

The documentary follows their investigation, including laboratory visits with Professor Richard Bevins at Aberystwyth University, practical trials recreating possible megalith transport techniques at the Ancient Technology Centre in Dorset, and a trip to Skara Brae in Orkney. Borman reflects on encountering evidence of long-distance connections and sophisticated Neolithic communities for the first time as a Tudor specialist.

The programme is produced by Lion Television and will be available internationally as Stonehenge: The Final Mystery Revealed.

Battling Bullshit with Bayes and Brandolini

We can give fringe views far too much credence by applying false equivalence rather than proper Bayesian reasoning. We should assign prior probabilities to hypotheses, update them with new evidence to reach posterior probabilities, and let strong data drive the probability of evidentially unsupported ideas close to negligible levels. Refusing to do so inverts the hierarchy of evidence and grants nonsense unearned legitimacy.

Brandolini’s law, the “bullshit asymmetry principle”, compounds the problem. Producing a confident but unsupported claim is quick and easy; refuting it properly requires time, expertise, and careful explanation. Weak ideas can therefore saturate public discussion faster than they can be dismantled. The Dunning–Kruger effect adds a further distortion: people with limited domain knowledge often lack the background needed to recognise the limits of their understanding, making them resistant to updating beliefs even in the face of contrary evidence. This dynamic is particularly visible in alternative archaeology.

Academic publishing is polite and cautious; traditionally high costs filtered low-quality material. Online publishing has removed those barriers, allowing superficially “scientific” nonsense to spread freely. Academia must raise its game: use robust public language, rebut directly and quickly, and call falsified ideas exactly what they are. Politeness is no substitute for rigour.

Refusal to debate is sometimes valid, but it carries risks if unexplained. Silence may be interpreted as uncertainty or evasiveness. The goal is not to engage endlessly with committed proponents, but to inform the broader audience. Brief, evidence-based correction is often more effective than performative debate, which can create the illusion of a live controversy where little exists.

Open-mindedness is not refusing to close questions. As Walter Kotschnig warned in 1939, "don’t keep your mind so open that your brains fall out". It is letting strong evidence close them so better ones can be asked. When data drives a hypothesis near zero, we must say plainly: “Tested. Probability now vanishingly small. Move on.”

Review of Clarke et al. (2026) ‘Altar to Attic to Analysis: Geochemical Authentication of a Rediscovered Victorian Thin Section of Stonehenge’s Altar Stone’

Anthony J.I. Clarke, Christopher L. Kirkland, Arthur de Oliveira Vicentini, Lisa Brown, Altar to Attic to Analysis: Geochemical Authentication of a Rediscovered Victorian Thin Section of Stonehenge’s Altar stone, Journal of Archaeological Science: Reports, Volume 70, 2026, 105619, ISSN 2352-409X, https://doi.org/10.1016/j.jasrep.2026.105619. (https://www.sciencedirect.com/science/article/pii/S2352409X26000544)

Clarke et al. (2026), published in Journal of Archaeological Science: Reports, examine thin section S45 from the William Cunnington III collection (1876–1881), rediscovered in 2021 at the Wiltshire Museum. The study applies automated mineralogy (TIMA SEM-EDS) and laser-ablation ICP-MS U-Pb dating of zircon and apatite to assess the section’s authenticity as material from Stonehenge’s Altar Stone and to constrain its geological provenance.

The modal mineralogy of S45 comprises quartz (53.9 vol. %), calcite cement (19.2 vol. %), plagioclase (12.5 vol. %), white mica, chlorite, and trace heavy minerals (rutile, chromite, zircon, apatite), with fabric and phase abundances that align closely with previously verified Altar Stone fragments such as 2010 K 240 and MS-3. Apatite U-Pb analyses define two isochrons yielding lower-intercept ages of 1043 ± 29 Ma and 449 ± 24 Ma. Zircon data are affected by common-Pb contamination from residual Canada balsam resin; the authors address this by subdividing time-resolved ablation signals into short (typically 3–6 s) integrations and performing unanchored lower-intercept regressions on a grain-by-grain basis. Twelve regressions meet the stated acceptance criteria, producing dates between 389 Ma and 1850 Ma with main density peaks at approximately 435 Ma and 1021 Ma. These age populations, together with the mineral assemblage, are consistent with derivation from the Upper Old Red Sandstone of the Orcadian Basin in northeast Scotland.

The paper adds to knowledge of the Altar Stone by authenticating a historic thin section without requiring new sampling of the 6-tonne megalith, thereby supporting material preservation while reinforcing the Scottish provenance established in earlier work. It also presents a data-reduction procedure for extracting U-Pb information from resin-contaminated legacy thin sections.

Limitations arise from the small size of the analysed chip (~250 mm² surface area, ~20 % of a standard thin section), which restricted the dataset to 55 zircon analyses and only 12 acceptable ages. Conventional concordant ages could not be obtained, and the sub-set integration approach, although functional in this case, rests on bespoke acceptance thresholds (minimum three integrations, initial ²⁰⁷Pb/²⁰⁶Pb ≥ 0.837, MSWD 0.1–2.0, ≤10 % age uncertainty) that have not been validated against uncontaminated reference materials. The overall results remain confirmatory of prior mineralogical and isotopic studies rather than introducing new interpretive elements. Minor overdispersion in the apatite regressions is noted but attributed to natural variation in closure temperatures.

In balance, Clarke et al. (2026) demonstrate the continued utility of 19th-century thin-section collections for modern archaeometric questions, subject to the constraints of sample volume and preparation artefacts. The work contributes incremental but useful reinforcement to current models of Altar Stone provenance and to methodological options for heritage-science applications. 

Correcting Brian John's Silly Mistake

Last year Brian John published a paper in Archaeology in Wales Vol 63 arguing that “Carn Goedog on Mynydd Preseli Was Not the Site of a Bluestone Megalith Quarry" Unfortunately, his argument was built on false bedrock. The journal has published a correction.

Pearce, N.J.G., Bevins, R.E. and Ixer, R.A. (2026) ‘Carn Goedog is a 10m thick, northward dipping sill: a comment correcting geological errors in “Carn Goedog on Mynydd Preseli was not the site of a bluestone megalith quarry” by Brian John (2023)’, Archaeology in Wales, 64.

In the 2023 paper, John noted that dolerite crops out on the north face of Carn Goedog from roughly 305 m down to 235 m elevation — a vertical drop of about 70–75 metres. He declared: “If all of the rocks of Carn Goedog belong to the same sill, it must be at least 75 m thick.”

A sill is a flat, sheet-like body of igneous rock (here, spotted dolerite) intruded between older layers. Its true thickness is the shortest distance measured perpendicular to its top and bottom surfaces. It only the same as the vertical drop if the sill is perfectly horizontal. The Carn Goedog sill is not. It dips gently northward at about 23°.

This is the geological equivalent of looking at my soughdough heel and announcing that the inserted slice is 2 inches thick because the bread is 2 inches tall.

John used his 75 m “thickness” as the foundation for three key claims:

• the sill must be internally chemically differentiated (like much thicker sills elsewhere), • the geologists’ sampling was inadequate, • the geochemical provenancing to Carn Goedog was unreliable.

Once you correct the thickness to ~10 m and recognise it as a thin, simply dipping sheet the entire geological critique collapses like an old quarry face.

Cross-section from the highest part of Carn Goedog (approx. 300m AOD) to an elevation of about 215m with the position of the Carn Goedog sill marked (shaded red). Horizontal and vertical scales are the same. - From https://www.academia.edu/164795472/Carn_Goedog_reply  

This was not a subtle difference of interpretation — it was a basic first-year structural geology error. The 2025 correction paper by Pearce, Bevins & Ixer politely but firmly points this out with clear cross-sections, structure contours and LiDAR. They also note, with admirable restraint, that John misunderstands how analytical geochemistry works: no two samples from the same outcrop are ever perfectly identical because of natural heterogeneity and analytical precision.

John also asserted that the Carn Goedog sill forms one vast, kilometre-scale continuous outcrop stretching 3 km west and 2 km east across the Preseli ridge, implying the geologists had under-sampled a gigantic body. In reality, as Pearce et al. show the exposures are nothing more than isolated crags; the supposed “continuous sill” is an illusion created by joining up unrelated dots on the BGS map. Carn Goedog and Carn Breseb, for example, sit 75 m apart stratigraphically and belong to completely different geochemical groups. John’s claim that columnar jointing covers less than 10 % of the outcrop is equally wide of the mark — the entire 10 m sill is columnarly jointed, and the jumble of blocks on the slope is simply the weathered, disaggregated result. All of this fed his central complaint that the geologists’ sampling was “inadequate”; once you realise they had sampled the full 7–10 m thickness of each thin, northward-dipping sheet and analysed continuous vertical sections, that complaint evaporates too.

I couldn't resist calling it a "silly" mistake, but it is actually rather more than that, his whole paper is built on a fundamental geological misunderstanding, he needs to withdraw.

Stonehenge Glacial Transport Theory Takes Another Major Hit

 New Quaternary Newsletter paper dismantles the Ramson Cliff erratic as proof of high-level ice flow

In the February 2026 issue of the Quaternary Newsletter (Vol. 167), Tim Daw, Rob Ixer and Paul Madgett (one of the original 1974 discoverers) publish a detailed re-examination of the 700 kg altered epidiorite block at 80 m OD on Baggy Point, north Devon. New petrographic analysis of the original thin section shows the rock is fully compatible with Cornubian greenstones from the Dartmoor metamorphic aureole, lacking any diagnostic Scottish or Welsh minerals. The boulder was first recorded standing upright in pasture, has no pre-1969 map or aerial photo record, shows no beach abrasion, and is the only claimed high-level erratic on the south Bristol Channel coast. The authors conclude its evidential value for Irish Sea ice overriding the cliffs “should be reassessed” and that it “warrants consideration as potentially a manuport rather than unequivocally glacial.”

A Bayesian analysis of the four key lines of evidence (petrography, upright position when found, unabraded texture, and lack of early documentation) starts from a neutral-to-glacial prior and updates sequentially. The combined likelihood ratios drive the posterior probability of pure glacial emplacement at 80 m OD to less than 0.3 % — vanishingly unlikely. Academics must phrase conclusions cautiously, but the data speak clearly: this boulder almost certainly required human transport to its present position.

Brian John has repeatedly used the boulder as key support for his glacial-transport theory for the Stonehenge bluestones. Since 2015 he has described it as “proof that ice-sheet override of Baggy Point” occurred, citing its height, supposed Scottish origin, and rough texture as incompatible with shoreline rafting (see his detailed 2015 posts here: https://brian-mountainman.blogspot.com/2015/01/the-erratics-at-baggy-point-croyde-and.html and https://brian-mountainman.blogspot.com/2015/01/baggy-point-erratic-at-80m-altitude.html). He repeated the claim in 2023 and featured it in his September 2024 table of “high-level erratics” to debunk the “myth of shoreline erratics”.

Another one bites the dust. The Ramson Cliff erratic — long presented as one of the strongest pieces of evidence for high-level Irish Sea ice reaching elevations comparable to those needed for bluestone transport to Salisbury Plain — has now lost its glacial credentials. The glacial-transport model for Stonehenge has just lost a central pillar.

Reference:
Daw, T., Ixer, R., Madgett, T., 2026.
A review of the Ramson Cliff erratic: evidence of high-level ice flow?
Quaternary Newsletter, Vol 167, p13
https://doi.org/10.64926/qn.20517

Review of Daw, Ixer & Madgett (2026): "A Review of the Ramson Cliff Erratic: Evidence of High-Level Ice Flow?"

Daw, T., Ixer, R., Madgett, T., 2026.
A review of the Ramson Cliff erratic: evidence of high-level ice flow?
Quaternary Newsletter, Vol 167, p13

This article, published in the Quaternary Newsletter (Vol. 167, February 2026, pp. 13–19), offers a concise yet multi-disciplinary re-evaluation of a single large boulder — the so-called Ramson Cliff erratic — long cited as key evidence for high-level (c. 80 m OD) Irish Sea ice incursion onto the north Devon coast. The authors (Tim Daw, Rob Ixer, and Paul Madgett, the latter being one of the original 1974 discoverers) combine fresh petrographic analysis, archival research, and regional Quaternary context to argue that the boulder’s evidential value for Pleistocene ice-sheet dynamics should be substantially downgraded. The piece is timely, appearing shortly after Bennett et al. (2024) reaffirmed the glacial interpretation in their review of Devon’s Quaternary geology, and it engages with the wider debate on possible glacial transport of Stonehenge bluestones (John, 2024).

Summary of Content and Argument

The boulder — a ~700 kg, roughly 0.25 m³ block of altered epidiorite/greenstone — was first formally reported in 1969 as standing upright in pasture on the crest of Baggy Point (SS 4356 4070). It was initially identified as indistinguishable from Scottish Highland epidiorites and therefore interpreted as a far-travelled glacial erratic emplaced by ice at high elevation. The authors present:

• A new transmitted-light description of the original thin section (prepared by the Soil Survey in the 1970s and previously examined by R.J. Merriman). The rock is an amphibolitised microgabbro with distinctive secondary green amphibole knots in chlorite; crucially, it lacks epidote, analcime, calcite, and other minerals diagnostic of Welsh or Scottish sources.
• Historical and cartographic review: no clear pre-1969 records on maps, Tithe maps, Ordnance Survey sheets, or 1940s Luftwaffe aerial photographs; the stone was moved to the field edge for ploughing in the early 1970s and now lies adjacent to the South West Coast Path.
• Morphological notes: angular, rough-surfaced, unabraded — inconsistent with prolonged high-energy beach transport.
• Regional context: all other confirmed erratics on the north Devon coast lie below ~30 m OD; high-level examples elsewhere (Lundy, Shebbear, Ilfracombe–Berrynarbor) are either contested, local, or non-glacial (e.g., sarsen).

The authors list four plausible emplacement mechanisms and conclude that a Cornubian (Devon–Cornwall) provenance is “a realistic alternative” to a northern glacial source. They stop short of declaring the boulder a definite manuport but stress that its status as unequivocal evidence of high-level ice flow is no longer tenable. Wider implications for the glacial-limit model in south-west England and for bluestone-transport hypotheses are noted.

Strengths

• Petrographic rigour: Rob Ixer’s detailed mineralogical account is the article’s strongest element. The absence of key northern indicators and presence of features compatible with altered metabasics in the Dartmoor metamorphic aureole (Meldon–Sourton–Belstone area) directly challenges the original Scottish identification. This is the first modern re-examination and sets a high standard for erratic provenancing.
• Interdisciplinary approach: Integration of geology, history, and landscape archaeology (standing-stone possibility, D-Day training disturbances, field names such as “Mearlands” and “Long Stone”) is effective and transparent.
• Cautious tone and self-criticism: The authors (including the original finder) acknowledge that absence of early records is not proof of recent placement, that provenance remains “inconclusive” at outcrop scale, and that hybrid scenarios (ice-rafted to foreshore then human transport) cannot be ruled out. This intellectual honesty strengthens the piece.
• Broader relevance: Explicit linkage to Stonehenge debates and the “myth of shoreline erratics” (John, 2024) situates a local curiosity within national Quaternary discourse.

Limitations

• Provenance is suggestive rather than definitive; no new samples, geochemical data, or precise outcrop match is provided. The Dartmoor aureole hypothesis is plausible but untested.
• Archival “negative evidence” (no pre-1969 mentions) is handled carefully but remains inherently weak.
• No new fieldwork (e.g., excavation around the original find-spot or detailed comparison with foreshore erratics) is reported; reliance on 1974 thin section and 2008 photographs limits fresh morphological data.
• As a newsletter article rather than a full journal paper, depth is necessarily constrained; quantitative abrasion or fabric analysis is absent.

Overall, this is a high-quality, evidence-led contribution typical of the Quaternary Newsletter’s role in fostering debate. It exemplifies “re-assessment science” — returning to a long-accepted outlier with modern techniques and finding it less anomalous than supposed. The piece will be essential reading for anyone citing the boulder in future syntheses of British-Irish Ice Sheet limits or megalith transport.

Visual context of the boulder (for readers unfamiliar with the site):

brian-mountainman.blogspot.com

Stonehenge and the Ice Age: The Erratics at Baggy Point, Croyde and Saunton (1)

(The photograph shows the characteristic rough, angular block in its high-level grassland setting on Baggy Point — note the absence of beach rounding.)

Bayesian Analysis: Likelihood of Manuport versus Glacially Placed Emplacement

The query asks for the probability that the boulder is a manuport (human-transported and placed at its current ~80 m OD location) versus glacially placed (emplaced at that elevation by natural glacial processes — either direct ice-sheet deposition or ice-rafting to high relative sea level). I use Bayesian reasoning because it makes prior assumptions explicit, shows how each piece of evidence updates belief, and is fully traceable for non-statisticians. No specialist software is required; everything is simple multiplication of probabilities or odds.

Step 1: Bayes’ Theorem in Plain English

• Prior probability: What we believe before looking at the new evidence in this article (based on the long-standing glacial interpretation in the literature).
• Likelihood: How probable each piece of evidence is if a hypothesis is true.
• Posterior probability: Updated belief after the evidence (what we should believe now).
• Formula (odds form, easiest for two hypotheses): Posterior odds = Prior odds × Likelihood ratio (for each piece of evidence in turn). Odds = P(H1) / P(H2). We update sequentially, treating each major line of evidence as independent.

Step 2: Define the Two Hypotheses Clearly

• H_G (Glacially placed): The boulder reached ~80 m OD purely by glacial processes (Irish Sea Ice Sheet or floating ice during high relative sea level). This is the traditional claim.
• H_M (Manuport): Humans transported and placed the boulder at its find-spot (possibly as a boundary/standing stone, or dragged up from the shore). Hybrids (glacial to foreshore + human lift) are possible but counted under H_M for “glacially placed” vs “manuport” framing.

Step 3: Choose Transparent Priors

Historical literature (pre-2026) largely accepted the glacial interpretation (Madgett & Madgett 1974; Bennett et al. 2024). However, the boulder is unique as a high-level example and contradicts the consensus low-level limit (<30 m OD). A reasonable prior reflecting this mixed picture: P(H_G) = 0.60 (60 % — the “default” glacial view). P(H_M) = 0.40 (40 %). Prior odds (H_G : H_M) = 60 : 40 = 1.5 : 1.

(You can start with 50 : 50 if you prefer neutrality; the final conclusion is robust to reasonable changes.)

Step 4: Key Pieces of Evidence and Likelihoods

I use four independent lines drawn directly from the article. For each I assign P(E | H) on a 0–1 scale, grounded in the text and regional geology. These are informed judgements, not arbitrary — the article itself supplies the reasoning.

1. Petrography (Cornubian-compatible, incompatible with Scottish/Welsh sources)
   • If H_G true (northern ice): very unlikely — Irish Sea ice brings northern rocks; Cornubian source would require extraordinary local reworking. → P(E1 | H_G) = 0.10.
   • If H_M true (human move): likely — altered greenstones abound south of Baggy Point (Dartmoor aureole etc.). → P(E1 | H_M) = 0.70. Likelihood ratio (LR1) = 0.10 / 0.70 = 0.143.
2. Upright/standing position when found in 1969
   • H_G: ice does not erect boulders. → P(E2 | H_G) = 0.05.
   • H_M: humans commonly erect stones (boundary markers, standing stones). → P(E2 | H_M) = 0.60. LR2 = 0.05 / 0.60 = 0.083.
3. Rough, unabraded surface texture (no beach rounding)
   • H_G (direct ice or ice-rafted high): possible, but if it spent time on the high-energy foreshore it should show abrasion. Article notes similarity only to freshly exposed Head-deposit erratics. → P(E3 | H_G) = 0.25.
   • H_M: expected for a fresh inland or short-distance move. → P(E3 | H_M) = 0.80. LR3 = 0.25 / 0.80 = 0.312.
4. No pre-1969 cartographic/photographic record + uniqueness as high-level erratic
   • H_G: possible (small stones often unmapped), but weakens the “long-known” glacial claim. Regional absence of other high erratics makes this one anomalous. → P(E4 | H_G) = 0.30.
   • H_M: expected if placed relatively recently or overlooked as a field stone. → P(E4 | H_M) = 0.65. LR4 = 0.30 / 0.65 ≈ 0.462.

Step 5: Sequential Updating (Full Workings)

Start with prior odds 1.5 : 1 (H_G : H_M).

• After E1 (petrography): 1.5 × 0.143 = 0.2145 : 1
• After E2 (upright): 0.2145 × 0.083 ≈ 0.0178 : 1
• After E3 (texture): 0.0178 × 0.312 ≈ 0.00555 : 1
• After E4 (records/uniqueness): 0.00555 × 0.462 ≈ 0.00256 : 1

Posterior odds ≈ 0.00256 : 1 → P(H_G) ≈ 0.26 %, P(H_M) ≈ 99.74 %.

Even if you start with a stronger glacial prior (say 80 : 20 or 4 : 1), the combined likelihood ratios (~0.002) still drive the posterior below 1 % for pure glacial emplacement at high level.

Step 6: Sensitivity and Interpretation

• The result is robust: petrography and upright position are the strongest “knock-out” factors.
• A hybrid scenario (glacial to foreshore + short manuport) would have higher probability than pure H_G but still falls under “involves human placement” for the query’s framing.
• Uncertainties: likelihoods are subjective but explicitly tied to the article; different readers might adjust them ±0.1–0.2 without changing the conclusion that H_M is overwhelmingly more probable.

Conclusion of analysis: On the evidence presented, the probability that the Ramson Cliff boulder is a manuport (or at minimum required human transport to its current position) is >99 %. Pure glacial emplacement at 80 m OD is extremely unlikely (<1 %). This does not disprove glaciation in north Devon, but it removes one of the few cited high-level “smoking guns” and supports the article’s call for reassessment.

The piece is a model of careful Quaternary scholarship and the Bayesian update illustrates why single anomalous boulders should be treated with caution in ice-sheet reconstructions. Further work — new geochemical provenancing, test-pitting, or OSL dating of surrounding soil — could refine these probabilities further.

(Independent analysis by Grok in response to a neutral prompt)