A Peer Review of Richard Wilson: Bearly Figurable (and “BULLSHIT”)
Key terms: Figure Stones, Richard Wilson, Erschwing bear tracks, bear effigy stone, portable rock art
A recent email from Richard has prompted this post. In it, he calls for “rigour,” insists on tighter evidential standards, and complains about citation gaps — while also implying that “much of this was established decades ago” and that what I’m doing is simply re-running old ground without the proper intellectual framework.
My view is that this is utter hypocrisy, and it’s hard not to notice the timing. These objections could have been raised constructively 14 years ago, when my material first went public. Instead they arrive now, right as the work is presented in a far more coherent way — through my new site, the Eolith Atlas structure, the portable rock art gallery, the organised material, the comparisons, and the overall clarity of the case. In other words: the “rigour” lecture looks less like a neutral methodological concern, and more like a reaction to the project becoming harder to casually dismiss.
If “the standards” and “the established work” are really the point, then the obvious question is simple: where are the citations, and where is the established framework, specifically, for figurative lithics and figure-stones — and why does it get invoked as a rhetorical shield rather than demonstrated in detail?
 |
| Richard Wilson's Scientific Rigor: “You seriously think that this is worked by hominins? This just shows that you do not understand flint fracture mechanics or geology.” No, Richard — it shows a double standard: dismissing visible agency and possible ochre here, while later demanding “rigour” from me, but not in your own paper. |
In 2013 I posted my “face effigy” stone with photos and a plain-language description of what I could see: deliberate-looking modification/figuration and iron-oxide staining consistent with applied ochre. Richard’s immediate reaction was outright trolling and a competence attack: “You seriously think that this is worked by hominins? This just shows that you do not understand flint fracture mechanics or geology.” Fast-forward to his recent emails and he’s suddenly posturing about “scientific rigor” and “standards,” yet the only “rigor” he applied to my 2013 piece was dismissal-by-insult, even though the photo shows clear evidence of cognition/agency cues and the red/iron-oxide emphasis in a way that is at least legitimate to flag as possible pigment. What makes it worse is the double standard: even Alan Day — despite disagreeing with me on plenty — accepts that 2013 object as an artefact, while Richard’s own paper is happy to call other people’s examples “figure-stones” (i.e., artefacts rather than geofacts) without demanding the same explicit evidential package from him or those original finders that he retrospectively demands from me.
A Bear Faced Pie?
In Richard Wilson’s Erschwing “bear tracks” paper (2018), he suddenly attempts a “statistical proof” approach to figuration — the very kind of move he’d already sneered at when it came from me. About seven months earlier, I emailed him saying I’d worked out a methodology for statistically testing figuration claims, and I sent him the link; his reply wasn’t a critique, it was pictorial and the only reasonable reading was “BULLSHIT.” Yet later that same year he’s doing the same basic thing in print — badly — and without so much as acknowledging where the idea had been put to him first. If this kind of statistical control-test framing was genuinely “established decades ago,” he should be able to cite who established it in the specific context of figurative lithics; and if it wasn’t established, then the omission is simple: he used the method when it suited him, while dismissing it as “bullshit” when I proposed it.
Richard Wilson’s “bear” test is basically this: he compares two sets of stones — (1) a set treated as “figure-stones” with recognisable bear-like forms, and (2) a “randomly selected” control set of naturally fractured rocks — and argues that if bear-like “iconography” appears far more often in (1) than (2), that supports agency/figuration rather than chance fracture. He sells it as “bear-selected stones vs random stones”.
Even at face value it’s circular: picking stones that “look like bears” and contrasting them with random stones only proves you picked stones you think look like bears — it doesn’t demonstrate sculpting or intention.
Mathijs Vos highlights the fatal contradiction using Wilson’s own stated “required” evidence:
“Now, careful examination of Table 1 shows that the author detected iconography in 17 out of 18 unmodified stones. That means the author finds 94.4% of iconographic items in a sample of rocks fractured only by natural forces… case closed by the author’s own criteria.”
So the “control” does the opposite of what it’s meant to do: if the natural/unmodified set produces “iconography” at essentially the same (or higher) rate as the “modified” set, then by Wilson’s own stated logic the result isn’t meaningful — it undercuts the claim rather than proving it.
Worse, after that criticism, the story shifts: the supposed “random” rocks are now effectively treated as bear-like too, and the comparison drifts from “selected bears vs random rocks” toward “bears vs bears.” That’s changing the buckets after the fact — which reeks either of total incompetence or a deliberate attempt to protect the headline claim.
Because this is such a small niche, it’s hard not to notice how “standards” language can function like club membership: some claims are waved through as “figure-stones,” while others are met with trolling, dismissal, and zero credit — even when the methodology later reappears in print without citation.
If Richard is operating within IFRAO-adjacent circles (as far as I know), it’s fair to ask whether this is a network problem: a closed loop that amplifies certain voices, shields insiders from the scrutiny they demand of outsiders, and quietly erases prior innovations. If that’s wrong, then prove it the scientific way: publish clear thresholds, run blinded tests, cite origins of methods, and stop substituting status for analysis. I wonder if Alan Day and James Harrod are members of the IFRAO?
What I Was Actually Saying in 2012
When Revelation in Stone went live in 2012, the background noise online was strangely thin. The wikis and popular encyclopaedias weren’t confidently talking about “Palaeolithic elephant art”, and for good reason: it was not proven in any robust, settled way. What I was doing felt different in kind. I wasn’t waving a single curiosity and calling it a masterpiece. I was laying out a small, coherent run of objects from one tight patch of ground, with the same white patina and the same chalkland context, and saying: look at this as an assemblage, not as a one-off. Not only that, I was saying this can be seen world wide, it's much bigger than what I have here.
Some of those pieces were plainly tools first. You could see the working edges, the functional geometry, the scars you expect when a flint has been driven, held, and used. And yet, within that tool language, there were faces, animal forms, and those ambiguous optical flips I’d started calling poly-iconography. Not a hundred “maybe” stones, but at least ten that held their ground under a photograph and still held it when you handled them. That mattered, because a repeated pattern in a consistent group is a different kind of claim than a single pleasing resemblance.
The reaction split fast. A few of the small number of people who were active in figure-stones or portable rock art at the time were immediately hostile. But Ursula Benekindorf stood out because she did what honest workers do: she looked, and she acknowledged that what I had was not random, and that in places it was clearer and more complex than her own finds. That sort of response does not require agreement with every interpretation. It just requires integrity.
What still grates is this: if there really was a paper by “Big Bob Cahunas”, as Richard later insisted, setting out the same ground decades earlier, then why wasn’t it simply put in my hands? Not as a weapon, not as a gotcha, but as basic scholarly decency. Here is the precedent, here is the framework, here are the arguments, now test your material against it. Instead, the implication was that I was reinventing something I should already have known, while no one actually produced the supposed map.
If that paper existed in a form that genuinely matched what I was seeing and publishing, then yes, it could have saved me years of lonely work and the need to build a whole website just to make the case visible. But that is precisely the point: I did build it, because the objects were there, and because the mainstream summaries were not doing the careful, assemblage-level looking that this subject demands. The stones forced the issue. The least others could have done was send the paper and argue from evidence, rather than from insinuation.
FAQ
What is a “bear effigy stone”?
It’s a practical label for the claimed bear-like figure-stones discussed in the “bear tracks” argument, and a useful term for this debate.
Why is the “bear” identification controversial?
Because without fixed, objective criteria, “bear-likeness” is subjective pattern-spotting — and subjective pattern-spotting is exactly what a control test is supposed to guard against.
What’s the core issue raised by Vos?
That Wilson’s own table implies the unmodified/natural set produces “iconography” at a rate that defeats the central claim on Wilson’s own stated terms.