In eighteen ninety-one, a pastoralist named Joseph Bradshaw rode into a remote Kimberley gorge and found painted figures on the sandstone walls so sophisticated he refused to believe Aboriginal Australians had made them. He was wrong. But the mystery he stumbled into has never been fully resolved. The Gwion figures depict slender human forms wearing conical headdresses, arm bands, tassel belts, and ceremonial gear, painted with a precision that researchers have struggled to explain for over a century. Strands of hair are rendered at a thickness of one to two millimetres. No corrections or changes were made during painting. An imprint of a feather found at one site suggests the artists used quills rather than fingers or brushes. In twenty twenty, a study published in Science Advances confirmed the main tradition was painted between twelve thousand seven hundred and eleven thousand five hundred years ago, during the final upheaval of the last ice age. One motif dated to seventeen thousand years and nobody knows why it stands alone. The organisms that should have destroyed the art are now protecting it. And somewhere in the Kimberley right now there is a shelter nobody has looked inside yet.
In 1891, Joseph Bradshaw found painted figures in a Kimberley gorge unlike anything seen before.
The cattlemen recorded delicate human shapes wearing tassels and headdresses, painted in
colours he compared to ancient Egypt. Aboriginal guides later told a stranger story. A blackbird
bled on the stone and painted spirits that still watch over the cliffs. But for more than a century
scientists have fought over who made these images, how old they really are and whether
they prove a lost chapter in Australia's human past. And in the last few years mud wasp nests
changed almost everything we thought we knew. If you want to keep discussing ancient Australian
history beyond what fits in a video, join the Ancient Australia School community. It is free,
spots are limited and the link is below. Joseph Bradshaw was not an archaeologist.
He was a Melbourne pastoralist looking for grazing land in one of the most remote corners of the
continent. In 1891 he pushed up the Roe River in the Kimberley deep into country that had barely
been mapped by Europeans. He was searching for suitable ground for cattle, not ancient mysteries.
But what he found on a sandstone escarpment stopped him cold. The paintings were unlike
anything recorded before. George Gray, the colonial explorer who would later become South
Australian Governor, had seen and documented rock art in the Kimberley back in 1838. But Gray's
recordings showed the Wajina style, large pale spirit faces with dark eyes and no mouths. Bradshaw's
discovery was completely different. The figures were slender, active and dressed in what looked
like ceremonial gear. He sketched them carefully, noting pale blues, yellows and browns that reminded
him of ancient Egyptian art. When he presented his findings to the Royal Geographical Society in
Victoria, he knew he had found something unique. But the world was not ready to listen. For decades
the art vanished from scientific attention. American archaeologist Daniel Sutherland Davidson
looked at Bradshaw's sketches in 1936 and dismissed them. He argued that Bradshaw's
encounter was too brief, that the drawings lacked any Aboriginal interpretation and that they might
simply be inaccurate. Davidson claimed the sketches carried a Eurocentric bias and without
the original paintings to check, nobody could prove him wrong. The original mural sat undisturbed in
the gorge, slowly fading, while backburning from modern land management began to damage what the
It was not until 1938 that a proper expedition arrived. Anthropologist A.P. Elkin led a team
from the Frobenius Institute into the Kimberley with the specific goal of studying this mysterious
art. They brought Aboriginal guides who recognised the figures immediately. But unlike the Wanjina
paintings, which the guides treated with reverence, these images drew a different reaction.
When pressed, one guide told a creation story that had been passed down for generations.
He spoke of Kujon, a black bird who struck his bill against the stones until it bled.
With that blood, he painted human-shaped figures on the rock. The guide said these were not animals,
they were spirits, and they were still watching. Here is the part most documentaries skip.
Before we chase dates and debates, we need to look at what Bradshaw actually saw,
because these are not crude scratches made by fingers dipped in ochre.
The Kujon figures are painted with a precision that should not exist in ancient Australia.
Researchers have identified four distinct styles across the Kimberley. The earliest and most
detailed are the tassel figures. Large and ornate, these figures are identified by tassels hanging
from their arms and waists. They wear conical headdresses, armbands, and they carry boomerangs.
Then come the sash figures, similar but sturdier, with three-pointed sashes across their bodies
and bags attached to their belts. The elegant action figures are completely different.
You will almost always see them running, kneeling, or hunting with multi-barbed spears.
No other Kujon figures have ever been painted over or defaced by later artists.
Which suggests they occupied a special place in the tradition. Finally, the most recent style
is the clothes peg figures. That style is simpler, painted in red pigment, and shown in aggressive
poses. One panel even shows a battle with opponents lined up in ranks, facing each other.
The detail is what shocks you. Image processing has revealed that the artist painted the outlines
first, then filled them in. Some faces are painted with anatomically correct features,
detailed enough to be considered portraits. Strands of hair are painted at a thickness of
just one or two millimeters. That level of control suggests the artist may have used feather quills
instead of brushes or fingers. An imprint of a feather found at one site supports this idea.
And here is what makes it stranger. No evidence has been found of corrections or changes during
painting. These artists planned their compositions and executed them in a single attempt. For decades,
Western observers misread the figures. Many paintings showed what researchers called paunches,
suggesting pot-bellied bodies. But recent work has revealed something the original viewers missed.
The figures are painted as if they are facing into the rock itself. That means what looked like a
belly was actually the back. The so-called paunches were gluteal structures, not stomachs. An entire
generation of research was built on a simple misreading of perspective. The distribution of
these paintings is as distinctive as the art itself. While Wanjina art is restricted to specific
sites and painted mainly on ceilings, Gwion art appears on vertical surfaces high up in escarpments.
Rock shelters here are shallow with small overhangs and irregular rocky floors that were not suitable
for living in. That means these were not domestic scenes painted where people slept.
They were something else entirely. Territorial markers, ritual signals or messages left at
locations that mattered in their world. Gwion art appears across the sandstone regions of Western
and Central Kimberley and isolated examples have been found as far east as the Napier Ranges.
Someone or some group was moving across vast distances and leaving the same distinctive
signature behind. This raises a bigger question. These paintings are sophisticated and we need to
know exactly how long they have been here. For years the answer seemed settled. Most sources
claimed the Gwion tradition stretched back 17,000 years or more. That estimate came from a 1997 study
that used optically stimulated luminescence on quartz particles inside a mud wasp nest
overlying a Gwion painting. The method measures when sand grains were last exposed to sunlight
and the results suggested the nest and so the painting beneath it was at least 16,400 years old.
That number stuck. It became the baseline that textbooks repeated and documentaries cited.
But there was a problem. Rock art pigment usually contains no carbon which means
standard radiocarbon dating does not work on the paint itself. The surfaces are exposed to intense
weathering and nobody knows exactly what techniques the artists used thousands of years ago.
So archaeologists needed another way to pin down the age. They found it in the most unlikely place
mud wasp nests. In 2020 a team led by Damien Finch at the University of Melbourne published a
breakthrough study in science advances. Working with the permission and assistance of the Balangara
and Dambamangari traditional owners they collected samples from over 108 rock art sites across the
Kimberley. Instead of dating the pigment they dated the mud wasp nests built on top of the paintings
and beneath them. A nest built over a painting gives a maximum age for the art underneath.
A nest underneath gives a minimum age. By comparing both you can bracket when the painting
was actually made. The team reported 101 radiocarbon dates. Results were remarkably consistent.
Most of the Gwion figures were painted between 12,700 and 11,500 years ago.
That is at least 5,000 years later than everyone had assumed. One motif however
dated to approximately 17,000 years standing alone as an outlier and those numbers were
about to start a very different kind of fight. Their work also revealed that wasps had been
building nests in these shelters quasi continuously for over 20,000 years which
means the method can be applied to other archaeological features across the region.
This places the main Gwion tradition at a very specific moment in history. 12,000 years ago
the last ice age was ending. Sea levels were rising rapidly swallowing the northern coastline
and displacing clans from land that had been dry for millennia. The Kimberley landscape was
transforming and right in the middle of that upheaval someone was painting these extraordinary
figures on sandstone walls. But here is where it gets strange because while the science was
catching up another story was running in the opposite direction. In the late 1970s an amateur
archaeologist named Graham Walsh began working in the Kimberley. Over three decades until his death
in 2007 he recorded one and a half million rock art images and located over 1,500 new sites.
His database was unprecedented but his conclusions were controversial. Walsh proposed that the Gwion
art predated the Pleistocene and he implied that it might not have been created by Aboriginal people
at all. The theory gained traction because the art looks so different from every other tradition in
Australia. It was too detailed, too advanced and too visually distinct. To some observers it felt like
it must have come from somewhere else. The idea found its way into popular books, documentary
pictures and even political talking points. In 2015 a senator publicly suggested that people
other than Aboriginal Australians may have occupied the continent in the past. He pointed to
anthropological disagreements as evidence but the fact that a politician repeats a theory
does not make it true. Here is what the science actually says. A team of ARC supported archaeologists
from the University of Melbourne and the University of Western Australia has spent years working
directly with traditional owners to date and document the Kimberley sequence. Professor Peter Veth,
one of the project's chief investigators, put it plainly. He said the Gwion figures are clearly of
Aboriginal origin and they form part of a long tradition of signalling places of importance
within the wider landscape. There is no archaeological evidence supporting a pre-Aboriginal
population in Australia. None. The theory was built on Eurocentric assumptions about what
Aboriginal art should look like combined with the romantic appeal of a lost civilization.
The idea keeps coming back because the Gwion figures really are an anomaly. They appear
suddenly in the record after the glacial maximum with techniques and visual conventions that do not
match what came before or what came after. Their style is so distinct that even mainstream
archaeologists struggle to explain how it emerged so fully formed. Make no mistake, the mystery is
real. It is just not the mystery some people wanted it to be. Strip away the speculation
and here is what is left. The people who always knew this art best never stopped telling its story.
The names vary by region and language. The Ngarrindjan people call the figures Gwion Gwion.
In the Napier Ranges and around Broome Bay they are known as Jiro Jiro or Kiro Kiro.
Researcher David Welch traced these words back to a single root, Kujon, the name of the blackbird
from the creation story first recorded in 1938. These are not different traditions.
They are regional accents of the same story. The story itself is specific and strange. Kujon,
the bird, struck his bill against the stones until it bled. He used that blood to paint
human-shaped figures on the rock. He painted no animals, only spirits. Aboriginal guides in 1938
told the Frobenius Expedition that these were depictions of bush spirits or Dimi. Similar
stories were recorded by Ian Crawford in 1969 and Patricia Vinicum in the 1980s.
The continuity is hard to dismiss and the connection to living culture is even stronger.
Aboriginal people in the Kimberley still perform dances that feature apparel matching the paintings.
Headdresses, boomerangs and string. The Gwion figures are not just ancient images on a wall.
They are referenced in contemporary art produced for sale in the region today. Artists like Kevin
Wehner paint Gwion Gwion, works that connect modern Aboriginal identity to this deep tradition.
The art never stopped. It just changed form. The paintings are also territorial or ritual markers.
Professor Veth's team describes them as signals placed at locations that mattered deeply in their
world. They were not made for casual viewing. They were made for people who understood the
language of the symbols, the stories of the spirits and the meaning of the places where
they were painted. The mystery is not who painted them. It is what they were trying to say.
12,000 years ago, as the last ice age ended, the Kimberley was in flux. The sea was rising,
the coastline was shifting and the environment that people had known for generations
was changing rapidly. It was against this backdrop that the Gwion tradition spread across the region.
The paintings were not isolated experiments. They were part of a widespread, socially complex
response to massive environmental upheaval. The art itself tells us these people were not
struggling to survive. These communities had time, skill and social organisation to produce
work this refined. Trade networks stretched across hundreds of kilometres. Complex rituals
required specific costumes and gear and the artists painted every piece with astonishing accuracy.
Some of the paintings have survived because of an unexpected partnership. Many Gwion figures
are colonised by bacteria and fungi, including a black fungus that forms a symbiotic relationship
with the red pigment. The organisms that should have destroyed the art are now protecting it.
Nature itself has become the conservator, but much remains unknown. The 2020 dating study
answered one big question and opened several smaller ones. One motif dates to 17,000 years,
while the rest cluster at 12,000 and nobody knows why. We do not know what happened to
the tradition after the closed peg period and we have not even found all the shelters
that might hold more paintings. Every field season in the Kimberley adds new sites. The Gwion figures
are not a closed case. They are a window into a world we barely understand, painted by people
who saw their landscape transformed and left their mark on the stone anyway.
Somewhere in the Kimberley, right now, there is a shelter nobody has looked inside yet
and it might change everything.