<link rel="stylesheet" id="wp-block-library-css" href="https://c0.wp.com/c/5.8.2/wp-includes/css/dist/block-library/style.min.css" type="text/css" media="all">
loader image

What would disprove evolution? A well-dated species being found far beyond its time. Like if we found a fossil dinosaur gnawing on a modern human. Creationists claim to have found such a thing in the form of 10 million-year-old stone tools. These “artefacts” – technically called eoliths – would thus be millions of years older than any species which could have made them. A real problem for our understanding of human evolution.

This point has been argued by several people for – literally – centuries. The first eoliths were found in the Victorian period. However, the recent advocates for eoliths come from Answers in Genesis. This multi-million dollar creationist organisation has published a 34-page scientific paper on the subject. It claims to prove eoliths are ancient stone tools, turning the chronology fo evolution upside down.

(hat tip to Belac for pointing out this article to me)

Creationist claims

The creationist paper is long. It’s 34 pages of describing eoliths found around the turn of the century. These rocks were discovered when our understanding of the stone age was in its infancy. These stones came from ancient layers, looked crudely made, and fit in with all the stereotypes Victorians had about early man. Like coming from Europe, where obviously humans evolved. Because Europe is the coolest place, and Victorians were oh so racist.

However, a growing body of evidence suggested humans evolved in Africa. Combined with an increasing understanding of how rocks could wind up looking like tools through natural means (the results termed “geofacts”) led to eoliths being abandoned. Oh, plus a better understanding of how old they were. Most come from Miocene sediments, several million years older than the Pliocene strata where the oldest tools are found. They were then locked away in museum collections, an old quirk of a bygone, naive age. Like Piltdown man.

However, the Answers in Genesis paper contends this dismissal came too early. It examines several dozen eoliths (out of a collection of thousands) from Europe. Armed with a better understanding of what a stone tool looks like, the creationist began comparing the two. It turns out he was able to find accepted stone tools that looked like all these eoliths. Surely then, eoliths must be real tools?

A figure from the creationist paper, comparing a Miocene eolith (left) to an accepted stone age tool (right)

A figure from the creationist paper, comparing a Miocene eolith (left) to an accepted stone age tool (right)

There’s not much more nuance to it than that. The entire paper is essentially anecdotes, finding eolith rocks that look like man-made rocks. There’s no large-scale survey or analysis. It’s all a series of photos.

Photos that do show some striking similarities. Might Answers in Genesis be onto something? To examine this, we first have to examine how we identify a stone tool.

Spotting stone tools

The problem with studying stone tools stems from how our ancestors made them. They used a variety of techniques, some of which were incredibly complex. But they all boil down to one basic method: hit a rock until it breaks into the shape you want. Thus, anything else which results in a rock being hit can create a similar trace. From tumbling down a mountain, being trampled by an animal, to washing down a river, so-called “geofacts” can be a pain in the neck for archaeologists.

Fortunately, most stone tools can be easily differentiated from geofacts. There are various tell-tale signs that help tell the two apart. Such natural damage, for example, rarely produces more than a handful of “hits” to the rock. Those hits are rarely patterned in a focused way and rarely produce symmetrical results. Perhaps the most telling sign is how often a finished “tool” is repeated. The random pattern of nature means two geofacts rarely look alike. Finding artefacts in the process of being made or used is another good way of spotting a tool.

The object which perhaps most fulfils the criteria, being the most obvious tools, is the famous handaxe. The evidence of repeated, focused striking producing symmetrical “cookie-cutter” tools is clearly human.

A classic handaxe

A classic handaxe

Unfortunately for us, our ancestors weren’t always trying to create these recognisable, symmetrical tools. Sometimes they just needed a sharp blade, so broke a sharp fragment of a rock. These simple flakes begin to look a bit more similar to these geofacts.

Statistical eoliths

Even when examining ancient stone tools that lack the complexity of a handaxe, the same sort of approach for identification still works. Was a rock hit more times than could be explained naturally? Are there examples of tools in varying stages of production/use? Is the same form repeated more often than chance would dictate?

As you might begin to see, the difference between these sorts of tools and geofacts becomes more quantitative rather than qualitative. Whether or not a rock was hit doesn’t tell you much about whether it is a stone tool or not. But the quantity of smacks, their distribution, the number of tools in the assemblage, and so forth. That’s where information starts becoming useful.

However, it also means that some of these features will crop up on geofacts occasionally. It’s the luck of the draw. When you’re examining a handful of tools from a site, the result might be ambiguous. But when you zoom out and see the whole site, trends become apparent. And it’s those trends that can allow you to reliably identify if the artefacts from a site are man made.

The following example from Kirmington, England, highlights this quite well. A site was found with ambiguous artefacts that might be geofacts. Each rock earned a score based on how many man-made characteristics it had. This was then compared to the score for flakes made of the same material collected from non-archaeological sites nearby. As the chart below shows, there is an overlap. There are some ambiguous results. But when you consider the whole formation, the artefacts earn a higher score.


The graph I just described, making this label pointless

The same approach has been tried for an ambiguous site in Poland, called Kończyce Wielkie. There the results showed the opposite, with the suspected artefacts scoring much lower than known artefacts from other Polish sites. However, it was a similar score to natural rocks found at Mokzeszow.

Another useless label on a previously described graph

Another useless label on a previously described graph.

Cargo-cult creationists

So, if you’re really curious about whether or eoliths are actual artefacts, the trick is to examine the site as a whole. Zoom out, do a complete statistical analysis and compare it to known artefacts and “control rocks” from natural sources.

When you realise the sort of work that needs to be done to examine eoliths, the creationist paper becomes laughably awful. Not a single site is given a statistical treatment. In fact, there’s not even any qualitative data presented. It’s all a case of trying to find known artefacts that look like the eoliths. As the above examples show, that would be possible and yet prove nothing. Only a complete statistical analysis would yield any useful data.

In essence, the creationist paper is just a bit of cargo-cult science. Bad research dressed up in scientific clothes to give it the appearance of credibility. It has figures, sub-headings, scientific terms, and more. But without the key statistics, it all means nothing.

Fortunately for people genuinely curious about the status of eoliths, someone has done these sorts of large-scale comparisons. A researcher took a classic eolith assemblage from Kent and subjected it to the same battery of tests discussed above. The rocks were scored based on how many tool attributes they had. This score was compared to control artefacts and natural samples.

Eoliths from Kent compared to natural samples and known artefacts

Eoliths from Kent compared to natural samples and known artefacts

The results clearly show that eoliths fall within the range expected for naturally occurring rocks. They aren’t ancient tools, and if Answers in Genesis had done even the most basic statistical tests, they would have realised that. But they aren’t good scientists so they didn’t. Instead, they have 33 pages of pseudo-science waffle stuck up on their website. An eternal reminder of how they only have the appearance of science, but not an ounce of the credibility the discipline actually has.


Peacock, E., 1991. Distinguishing between artifacts and geofacts: a test case from eastern England. Journal of Field Archaeology, 18(3), pp.345-361.

Wiśniewski, A., Badura, J., Salamon, T. and Lewandowski, J., 2014. The alleged Early Palaeolithic artefacts are in reality geofacts: a revision of the site of Kończyce Wielkie 4 in the Moravian Gate, South Poland. Journal of Archaeological Science, 52, pp.189-203.

Related posts

Categories: Creationism


Natural Historian · 17th October 2016 at 7:14 pm

Very interesting. That AiG paper had completely escaped my attention. Had I known about it I would have certainly have used it in my response to Terry Mortenson who tried to argue that those artifacts in Lybia were not real artifacts but were natural products – geofacts. Amazing that he went to so much trouble to suggest that rocks that were so much more obviously crafted by human hands were just geofacts when his own organization had published this paper in which they try hard to make geofacts into human-made artifacts. But it is never really surprising to find creationists contradicting themselves.

    Adam Benton · 18th October 2016 at 12:38 am

    I hadn’t considered the implications for that issue. It’s certainly rather telling how much variation one can encounter within creationist beliefs, depending on what they’re trying to prove. It reminds me of your review of their opinions of Homo naledi.

    No doubt some part of this range of opinions stems from an inability to evaluate and (potentially) reject evidence; beyond “does it agree with our interpretation of the Bible”.

Ashley Haworth-roberts · 17th October 2016 at 7:42 pm

I believe this is the article you are mainly critiquing (the link takes me to a post flagging around four AiG papers/blogs):
From the abstract, the main YEC argument appears to be that there is a lack of ‘reasonable scientific support’ that these were geofacts not eoliths. A sort of deep time and evolution ‘aren’t scientific’ claim (but I have not studied the whole paper – and largely skimmed your blog post also – so there may be slightly more to it). I note that you point to a lack of appropriate statistical analysis in the AiG paper.
Chance versus design – and the YECs are claiming ‘design’ once again.

    Ashley Haworth-roberts · 17th October 2016 at 7:44 pm

    ‘Chance’ being somewhat shorthand for geological processes.

    Adam Benton · 18th October 2016 at 12:39 am

    If you follow the link under “34 page scientific paper” that should take you straight to the AiG article. Is that not working for you?

      Ashley Haworth-roberts · 18th October 2016 at 6:04 pm

      My error. I was clicking on the link at “(hat tip to Belac for pointing out this article to me)”.

Belac · 23rd October 2016 at 5:53 am

Hat-tip received.
Also, any thoughts on this article? I feel as though there are a few exaggerations and half-truths riddled within.

    Adam Benton · 24th October 2016 at 3:15 pm

    This might be worth tackling in another post at some point. It tries to make the point that fraud is common, yet only provides 4 solid examples. A failure rate of 4 out of the literally trillions of artefacts and fossils associated with human evolution hardly seems like there’s a “common” problem. One of those examples is eoliths, which as discussed here is hardly a case of fraud.

    Beyond that it’s mostly poisoning the well and character assassination through cherry picked quotes; the majority of which come from a handful of sources; often decades old by this point.

Anonymous · 2nd June 2017 at 9:12 am

Well I put up a strong argument for Eoliths being Genuine flint tools on my blog http://eoliths.blogspot.co.uk/

    Adam Benton · 2nd June 2017 at 2:04 pm

    As I explain in the article, many diagnostic features of stone tools can occur naturally. The key distinction is the relative frequency of such marks compared to a control sample. I see no such work in your argument.

Brett Martin · 6th June 2017 at 12:35 am

The frequency of marks is an invalid argument, the test is fatally flawed, here is why:

1. Natural stones have no flake removal scars at all, if there is flake removal how can we know it to be natural, the natural stones were not observed to be naturally ‘worked’ in the first place. So the samples used for comparison are an invalid selection from the get go.

2. A flint tool can be created with one blow, frequency of blows/marks is irelevent.

3. There are no genuine control conditions used in the test, only what the tester has perceived to be genuine tools, or not genuine, or natural formations so its subjective, not factual.

4. The study does not account for time, and differing patinas as my study does, this is crucial for determining genuine tools from the outset.

5. The Chosen ‘natural’ samples could have been worked figure stones, and he would of been none the wiser, or even debotage.

If we look at the comparison pictures you have used and run a simple probability analysis according to my study at http://eoliths.blogspot.com/ , just examining the parallel removal of flakes we can see its statistically incredibly unlikely, in order of trillions. Add the sharpening features and proximal blade sharpening flake removal, again we have a massive likely hood that its statistical very unlikely, probably in the order of billions to one. If the stone has an even patina, we cannot tell from the picture, but lets assume it does, then all flake removal would of happened in one proximal time period, and none since, again according to my study another multiplier, so trillions, times billions , times billions. Likely hood of being natural…. about zero.

    Brett Martin · 6th June 2017 at 12:40 am

    Both stones were also created using the levallios process, this out right proves it to be a genuine worked flint tool.

    Adam Benton · 7th June 2017 at 3:01 pm

    You talk about statistical probability, but I didn’t see any calculations in your blog. Did I miss something?

      Brett Martin - Eoliths · 7th June 2017 at 6:34 pm

      You should know that any numbers I came out with would only be guide lines, but I do mention pairs of X,Y,Z coordinates in producing a linear flake removal, if we use 360 degrees increments for all values, I see 4 linear flake removal scars 3 on the front and the whole back being the 4th, there we have our statistics for probability, we get the first one free, so x,y,z and x,y,z times x,y,z and x,y,z times x,y,z and x,y,z we can see that he linear flake removals all have a relationship to each other.

      Brett Martin - Eoliths · 7th June 2017 at 6:40 pm

      You should really have other worries besides the numbers, just about everything you have learned about supposed progression from ape to man has been proved false 😛

      Brett Martin · 7th June 2017 at 10:33 pm

      That’s besides the point, but my response is awaiting your moderation.

      Brett Martin · 9th June 2017 at 5:43 pm

      Why didn’t you allow my reply? did I hit a nerve, are you stuck in denial about the falsehoods of your *science* heroes? Or was it that I proved this post to be utter garbage?

      To clarify probability, maths statistics and debunking that eolith to be a natural formation I have written a new post on my blog here -> http://eoliths.blogspot.co.uk/2017/06/eoliths-are-natural-not-on-your-nelly.html

      This should help you to understand probability and what my points were and also prove that the artifact in question, found outside mainstream time lines for human evolution is in fact a flint tool like any other.

        Adam Benton · 14th June 2017 at 4:48 pm

        That whole post revolves around the concept that random parallel lines are extremely rare. But do we know that the damage your talking about is truly random? At least, in the same mathematical sense as the equations that you cite. After all, rocks may be exposed to repeated similar forces like freeze-thaw cycles, gravity, alluvial forces, and so forth. Whilst hardly planned, such interactions aren’t quite random either.

Brett Martin - Eoliths · 14th June 2017 at 10:02 pm

I get what your saying, although it is very weak. Did you notice how the eolith resembles an alligator or croc head side profile? This is exactly what my study into ancient tools shows, head profile shapes.

Eoliths are thought to be crudely shaped, this misconception comes from a lack of understanding of ancient tool topology, if archeologists started looking for the right thing they would realize there is evidence of peoples well into the dinosaur times. And these peoples created stunning figurative art in stone, a record of fauna in any stratified layer.

    Adam Benton · 15th June 2017 at 1:32 pm

    Yeah, I mean none of the factors I listed could ever produce parallel lines. I can’t see any head. It’s a common archaeological convention to draw pictures of tools. This allows you to highlight the key features of it. Perhaps you should do that with your rock to pick out the head better.

      Brett Martin - Eoliths · 15th June 2017 at 3:21 pm

      You didn’t mention layers of sediment? which I assume is the cause of those in the picture in the comment I’m replying to.

        Brett Martin - Eoliths · 15th June 2017 at 3:38 pm

        It’s not my rock in your post here that resembles an alligator or croc head profile, just to make it perfectly clear its the pic here in your post, the comparison one, second drawing, rotate it CCW 90 degrees, and it is extremely easy to compare with a face profile of a croc or alligator.

        Adam Benton · 16th June 2017 at 2:12 pm


        Adam Benton · 16th June 2017 at 2:13 pm

        I think that rock was produced by freeze-thaw cycles, although its in Japan so a bit of the information is beyond my grasp

Brett Martin - Eoliths · 27th June 2017 at 4:18 pm

I really don’t understand this accusation of pareidolia, the image is present, it can be seen by any who chose to look, shaping and tiny flake removal suggest it was deliberately created to form the representation of a ‘crocodile’ head profile. Many flint tools have these same figurative descriptions, and multiples of. If you truly believe the image seen is just imagined, I wonder how you manage in every day life? When crossing the road, do you say to your self, that’s not a bus, its pareidolia? When reading a book or looking at road signs to you have the propensity to ignore the information your eyes are supplying to you brain?

    Adam Benton · 28th June 2017 at 12:53 pm

    The image you’re talking about came from a page trying to prove eoliths are real tools and even they didn’t spot this crocodile face. Nor have any of the hundreds of other people who have read either this article or that original post. Clearly, these images aren’t as present as you think they are.

Leave your filthy monkey comments here.

This site uses Akismet to reduce spam. Learn how your comment data is processed.