Not a Fossil After All: Why San Artists Weren't Painting Extinct Animals

I once studied South African San rock art and came to appreciate the weird and wonderful images painted on otherwise dusty cave walls. Some images were seemingly ordinary eland, rhebuck, and other animals. Others depicted clapping women and dancing men. Sometimes people were painted with hooves, or eland were seen walking into cracks; some people had winglike projections others were surrounded by bees and fish. But can you understand such paintings without first truly understanding the painters’ world?

Each image holds its own special meaning, a meaning only really understandable when one goes deep into the San ethnographic literature. Because without it I would interpret the eland as simple eland, and the hooved man as some monster.

In recent decades, a trend among some palaeontologists suggests that certain southern African groups, like the San and Basotho, acted much like western scientists. It is argued that they found ancient fossilised bones, recognised them as extinct animals, and reconstructed them or wove them into their myths (a concept called geomythology) and rock art.

Based on these claims, two sets of rock paintings have been interpreted respectively as a monstrous dinosaur and a creature related to mammals called a dicynodont. 

David Witelson, a rock art researcher at Oxford, argues that neither image depicts what is being claimed. While it is possible some individual San may have once stumbled upon a fossil and recognised it as something no longer around, no ethnographic accounts mention fossils, dinosaurs, or dicynodonts. And when held up against what we know about these people and what the images purportedly show, the narrative starts to crumble.

Rock Art And Geomythology

Southern African rock art is rather lucky because, unlike ancient rock art in Europe or the weathered art of Indonesia, it has a detailed ethnography that records the beliefs, culture, and worldviews of the people who painted these images. In some rare cases, there are even accounts of interpretations of the art  or art copies by the indigenous peoples themselves. Some of the most extensive material was recorded by Wilhelm Bleek and Lucy Lloyd not just in English but in the indigenous language of the ǀXam San.

Much of what we know about southern African rock art comes from these accounts. But 'geomythology', the study of earth-related myths from non-Western cultures through a Western lens, looks at local geology and the fossil record to interpret how those people thought about earth phenomena.

One problem with geomythology is that it is largely unprovable, hard to date, and prone to confirmation bias. In the two rock-art cases Witelson examines, one core problem is a logical fallacy: the researchers assumed that if a San or Basotho person noticed fossils, they must have thought of them exactly like modern Western palaeontologists. For this reason, they concluded these people were 'indigenous palaeontologists.'

The argument also assumes that non-Western people were simply using primitive means to try and explain nature. In this way of thinking, a Western palaeontologist knows it is a dinosaur, but an indigenous person might see an ancient extinct monster from their myths.

But this assumption suggests two completely different worldviews would independently arrive at the same interpretations and conclusions. But where is the evidence?

If I see a perfect circle with some odd shapes painted on a wall, and say it depicts planet Earth, how do you prove me wrong? And how do you prove an ancient or indigenous person, be they San, Polynesian or even a Neanderthal, would come to the same conclusion?

One way is to examine their myths, stories, and culture, but what if you turn up empty-handed? Surely your assumption is wrong; well, a geomythologist may argue that something can still be true, simply because it hasn't been disproven.

Now that you have a relatively good idea of geomythology and how it interprets things, let's look at the two cases and see whether the geomythological arguments have evidence supporting their interpretations.

The Living Fire

The first case examines a case made by palaeontologist Paul Ellenberger. Ellenberger and colleagues argued that ‘prehistoric Bushmen’ noticed ancient dinosaur bones scattered across the landscape and were impressed by the giant footprints these beasts left behind. In response, they sang and told stories about 'monsters' that made trackways and explained their existence through a mythical giant creature called ǁKhwai-hemm, the "All-devourer" which held a central place in ǀXam San folklore.

Ellenberger and colleagues further equated the San's ǁKhwai-hemm with the large, swallowing Basotho beast 'Kholumolumo'. Despite some similarities, they are not equivalent.

Ellenberger's belief that the San thought of dinosaurs and incorporated them into their folklore stemmed from an image he saw at Mokhali Cave in Lesotho as a 12-year-old boy in 1930. He copied three black bird-like figures and a red three-lobed shape; he thought that red shape was a dinosaur footprint, and the bird-like creatures were the San's attempt to reconstruct the extinct creature, basically reconstructing ǁKhwai-hemm.

But that assumption has some problems. First, the three birds have long since vanished, if they ever existed, making them hard to verify, and the drawings of them lack the three-toed feet one would expect. Finally, later researchers note the red shape was painted using a technique more similar to that of Basotho tradition, so the birds and foot may not even have been made by the same people. That said, what about ǁKhwai-hemm?

In Bleek and Lloyd's collection of ǀXam manuscripts, they show that ǁKhwai-hemm was not an animal or beast, but a personified veld fire. He has no bones, but he does have a fiery flame-like licking tongue and ‘swallows up’ vegetation as he burns it. That doesn't make him a beast; after all, in English we also describe 'licking flames' and fires that 'angrily consume' and 'rage' over the land. That doesn't make a fire a person with anger issues and a penchant for eating anything in his way.

Furthermore, unlike what Ellenberger said, ǁKhwai-hemm is not the central figure of San mythology; ǀKaggen, the Mantis, is, and the myth in which ǁKhwai-hemm appears centres on Mantis getting his comeuppance for his antisocial behaviour. ǁKhwai-hemm is dangerous, solitary, all-consuming veld fire, not a dinosaur.

What about the songs and stories that, according to Ellenberger, speak of ancient trackmakers or fossils and bones linked to ǁKhwai-hemm? No such songs or stories exist.

In fact, even some modern palaeontologists have raised doubt about Ellenberger's interpretation.

A Smear That Looked Like Tusks

You may be familiar with this one, as it spread like wildfire across the media. A study a few years ago claimed a San artist depicted a tusked rain-animal inspired by fossils of extinct dicynodonts. Dicynodonts were ancient relatives of mammals, though they weren't mammals themselves, that lived hundreds of millions of years ago. Two tusks jutting down from their upper jaws are among their most recognisable features. Their fossils are rather abundant in South Africa, so when palaeontologist Julien Benoit came across a 19th-century copy of rock art showing a tusked animal and comments about rain-animals from a San person, he interpreted the painting as indicating a local geomyth about a huge, now-vanished creature.

This interpretation has several problems. First of all, Stow's rock art copies are notoriously unreliable; he often changed the colours, positions and details, left out figures, and in one case he was even accused of forging a copy.

Second, the San woman named 'Kou'ke, who was shown several copies, made comments on what they showed. However, her words are not recorded in her language; instead, they are translated and filtered through Stow's own geological understanding.

Let me give you an example. In the picture below, we have two snippets of text (not the whole): one in ǀXam, the other in English, transcribed and translated by Dorothea Bleek (daughter of Wilhelm Bleek) when she showed a copy of rock art from Sehonghong Shelter to a San man named Diä!kwain in 1974:

That doesn't make much sense without context, but at least here we have the original ǀXam, so we do not need to rely solely on an English translation. But if it were to be 'translated' only into normal English, without any ǀXam to compare back to, it would lose its original meaning. That is basically what happened to 'Kou'ke's words; the original was never recorded, only Stow's interpretation of them.

So Stow's words may not be the most reliable. Regardless, he reports that 'Kou'ke describe vanished giants: a great serpent, a horned water-creature, and one she specifically named the 'master of water.' Rain-animals are well established in many San accounts and a well-understood category of San belief. And when she says they lived in ancient times, she likely refers to the San concept of mythical Primal Time, though Witelson argues a more probable explanation is that these rain-animals belonged to San rainmaking rituals and vanished when that tradition broke down. After all, 'Kou'ke said that her people had hunted them.

Each animal she explained, in some form or another, did not refer to fossils or ancient evolutionary extinction, but to animals, and most importantly, she never mentions tusks at all.

Even if she had, remember Stow's copies are unreliable, and Witelson's visit to the shelter shows those tusks were never tusks at all. The supposed tusks are actually an irregular, black smear of paint possibly added later beneath the animal's jaw. It appears on only one side of its face rather than symmetrically on both sides, unlike genuine tusk depictions in other San rock art images.

Even its shape and placement don't match a dicynodont's tusks, which usually project from the upper jaw, not from its chin. Nor is it like the tusks of a warthog, which the San often depict (one on each side).

Witelson also points out that the "tusks" only look greyer than the black head, a feature Benoit took as typical of San tusk painting, because an oily film still coats the head while the smear has weathered; an old archival slide shows they were once the same black. The ‘tusks’ are a single curved brush stroke rather than a deliberate pair of tusks.

If anything, it looks more like an ill-maintained moustache, placed on a spotted slug. And since you can't prove me wrong, maybe that's what it is.

No, no, just kidding.

So what is it then: a rain-animal, but not one inspired by dicynodont fossils.

But Witelson stresses a bigger logical problem: even if this one painting did show something tusk-like, myths and traditions do not hinge on one person's interpretation; they work at a cultural level and are shared across communities. San mythology contains no evidence of extinct tusked creatures, and no other rock art example shows similar dicynodont imagery. Benoit's claim rests on a singular, ambiguous painting, rather than a documented cultural pattern.

So, does that mean it is impossible that the San ever took note of fossils or thought of extinct animals? No. But possibility is not evidence; unless someone can provide a San story about ancient bones, dinosaur trackways, or repeated imagery of tusked dicynodonts, the honest answer is we have no reason to think they did.

Two of the more well-known cases of geomythology do not hold up against the known San myths and beliefs. One is likely a mistaken veld fire that even later palaeontologists, including Benoit, think may not hold up to scrutiny. The other is a supposedly 'tusked' dicynodont-inspired animal that doesn't have tusks.

But why would it matter if we interpreted it falsely? Because science is about telling the story as accurately as the evidence allows, and once a claim built on too little evidence is accepted, it spreads, gets repeated, erasing the original meaning intended by the people who painted it. Reinterpreting it to fit our worldview simply replaces theirs.

Had you heard about the dicynodont rock art claims before, and what's your take on geomythology?

Let me know what you think, and if you'd like to support this blog and the person writing it, you can chip in once, monthly, or yearly. Only if you'd like to.

References:

Hollmann, J. (2022). Customs and Beliefs of the| Xam. NYU Press

Witelson, D. (2026). Arguments and evidence in rock art research: an assessment of two geomythological cases. Southern African Humanities, 39.